ANNALS OF THE f r \^ 1-=^ (^ VOLUME 59 1972 1 ^ -^ 3- - ^J^^'=\ Missouri Botanical Garden c \ Published by the Missouri Botanical Garden Press St. Louis, Missouri 63110 CONTENTS Mosses Re elf earn, Jr. 1 Disjunctions in Plants: A Symposium Otto T. Solbrig 105 Moiphology and Phy togeography : The Classical Approach to the Study of Disjunctions Carroll E. Wood, Jr 107 Approaches to Disjunct Populations: The Contribution of Palynology Donald R. Whitehead 125 Genetical Features of Ferns as Contrasted to Seed Plants Edward J. Klekoivski, Jr 138 Chemosystematic Data: Their Use in the Study of Disjunctions B. L. Turner 152 Disjunctive Distributions in the Lichen-Forming Fungi William Louis Culberson 165 Disjunctions in Bryophytes W. B. Schofield ir H. A. Crum 174 Disjunctions in Homosporous Vascular Plants W. H. Wagner, Jr 203 The Floristic Disjunctions between the "Monte" in Argentina and the ^C Sonoran Desert" in Mexico and the United States Otto T. Solbrig . . 218 Species Disjunctions in Larrea: Evidence from Morphology, Cytogenetics, Phenolic Compounds, and Seed Albumins J, H, Hunziker, K A. Palacios, Amalia G. de Valesi & Lidia Poggio 224 Plant Species Disjunctions: A Summary Peter IL Raven 234 Notes on Panamanian Trees and Shmbs Collected in 1971 by L. R. Holdridge and others John D. Dwyer 247 Solanaceae Studies II: Typification of Subdivisions of Solamim W, G. D'Arcy 262 Nev^ Taxa and Recombinations in Lopezia (Onagraceae) Uzi Plitmann, Peter H. Raven, ir D. £. Breedlove 279 The Comparative Morphology of the Cochlospennaceae. III. The Flower and Pollen Richard C. Keating 282 New World Juglandaceae, III. A New Perspective of the Tropical Members with Winged Fruits Donald E. Stone ' 297 I Rejoice in What He Affirmed 323 Edgar Anderson, 1897-1969 John /. Finan 325 The Publications of Edgar Anderson Erna R. Eisendrath 346 Student Days with Edgar Anderson or How I Came to Study Sunflowers Charles B. Heisefy Jr , . 362 Edgar Anderson; Recollections of a Long Friendship G. Ledyard Stebbins 373 382 Hybridization, Evolution^ and Systematics Duncan M. Porter 380 Cytological Studies of Natural Intergeneric Hybrids and their Parental Species in the Moss Genera Astomum and Weissia Lewis E. Anderson ir Betty E. Lemmon Cytological Evidence of Introgression Between Drosophila Species in Panama Sarah Bedichek Pipkin Levels of Confidence in the Analysis of Hybridization in Plants L, B. Gottlieb 417 435 Hybridization, Taxonomy and Avian Evolution Lester L. Short 447 Variation and Speciation in the Genus Hudsonia Judith Troop Skog ir Norton H. Nickerson 454 C-4 and C-3 Carboxylation Characteristics in the Genus Zy gophyllum (Zygophyllaceae) R. Kent Crookston ir Dale N. Moss 465 New Combinations in Compositae R. P. Wunderlin 471 NOTES Capparis hypoleuca Presl: A Synonym of Solanum schlechtendalianum Walp. Hugh H. litis 474 The Chromosome Number of Utricularia denticulata Benjamin Katsuhiko Kondo 474 VOLUME 59 NUMBER 1 ■ _ * CONTENTS ■\.- V , - Mosses of the Interior Highlands of North America Paul L, Redfearn, Jr 1 VOLUME 59 1972 NUMBER 1 ANNALS OF THE Missouri Botanical Garden The Annals contains papers, primarily in systematic botany, contiibuted from the Missouri Botanical Garden and the Department of Biology of Washington University. Papers originating outside the Garden or University will also be accepted. For information on preparation of manuscripts, see the inside back cover. i The Annals appears three times a year, and three numbers, totaling about 400 pages, constitute a volume. For information concerning subscriptions, see the back cover. Matters regarding exchange of publications are handled by the Library, Missouri Botanical Garden, 2315 Tower Grove Avenue, St, Louis, Missouri 63110. Editorl\l Committee Marshall R. Crosby, Editor Missouri Botanical Garden Sheri G. DaviSj Assistant to the Editor Missouri Botanical Garden Duncan M. Porter, Editor, Flora of Panama Missouri Botanical Garden ir Washington University John D. Dwyer Missouri Botanical Garden 6- Sf . Louis University Joan W. Nowicke Missouri Botanical Garden ir Washington University Published by the Missouri Botanical Garden Press, St. Louis, Missouri 63110 © Missouri Botanical Garden 1972 ALLEN PRESS, INC. ''";;;'o LAWRENCE, KANSAS VOLUME 59 1972 NUMBER 1 ANNALS OF THE Missouri Botanical Garden MOSSES OF THE INTERIOR HIGHLANDS OF NORTH AMERICA Paul L. Redfearn, Jr.^ Abstract A manual of the mosses of the Interior Highlands of North America that include diagnostic keys to the families, genera, and species. The Interior Highlands of North America includes several diverse physio- graphic regions (Figure 1) that generally coincide with the boundaries of the oak-hickory forest region as outlined by Braun ( 1950 ) . The northern portion of this region includes two extensive plateaus, the Salem and the Springfield. Up- land forest associations include both oak-hickory forests and sugar maple-white oak forests. Along north slopes and in narrow ravines more mesic foresjts, such as the sugar maple-red oak association, may be present. Along flood plains mature forests are often dominated by sugar maple and bittemut hickory. Local relief is frequently rugged and consequently extensive rock outcrops and bluffs, com- posed chiefly of limestone and dolomite, are common. On the thin soils of the summits of ridges and drier south-facing slopes, cedar glades and rocky barrens are often encountered. Springs are also a common feature of these plateaus. Some exposures of sandstone may be encountered, particularly along the westem border of the Springfield Plateau adjacent to the Prairie Plains and in the eastern part of the Salem Plateau. Within the Salem Plateau there is a residual upland area of rugged relief, the St. Francois Mountains, where crystalline rocks, mainly felsite and granite, are extensively exposed. In such areas open, rocky barrens commonly develop, particularly on the steeper slopes near and on the summits of the ridges. Oak-hickory forests are very well developed in this region. To the south of the northern plateau region lie the Boston Mountains or r Boston Plateau, a peneplained area dissected by numerous narrow, deep ravines, particularly along the Buffalo River drainage system. Sandstone is commonly exposed at the higher elevations, while shale and limestone occur at the lower elevations. On the ridges oak-pine or oak-hickory forests are usually well devel- oped. Narrow ravines and north-facing slopes are frequently occupied by rich, ^ Department of Life Sciences, Southwest Missouri State College, Springfield, Missouri 65802. Anx. Missouri Bot. Card, 59: 1-103. 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 FiGlJRE ; regions of the Interior Highlands of North America. Inset Highlands in Eastern North America, Boundaries approximate. mesophytic forests that often contain a large percentage of beech and umbrella magnolia. ■ r The southern border of the Interior Highlands is represented by the Ouachita Mountains. The eastern region of the mountains in Arkansas is characterized by a series of low hills separated by narrow, rocky valleys where shale, sandstone, quartzite, and chert are exposed. Pine-oak forests are common, although on deeper soils, oaks and hickories may completely replace pine. The western region of the Ouachita Mountains lies in western Arkansas and southeastern Oklahoma. The valleys separating the mountains are relatively high and broad. Mesophytic forests are common in the ravines which indent the lower and middle slopes. Extensive erosion has exposed shale, sandstone, chert, limestone, and norvaculite. The Arkansas Valley separates the Boston and Ouachita Mountains, This valley, while of generally low relief, contains numerous synclinal mountains that have tlie appearance of steep-sided buttes. Most of the slopes are covered with a mesophytic forest- The rocks most commonly exposed are sandstone and shale. On the benches and flat summits, drainage is often poor enough to produce small 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 3 Sphagnum bogs. Along the smaller streams of the valley the sandy soils are occupied by forests that include both pine and sweetgum. The Ozark Hills of southern Illinois contain two basic types of escarpments (Voight & Mohlenbrock, 1964): Those composed of sandstone and extending in an east-west direction, and those composed of limestone running in a north-south direction paralleling the Mississippi River. The forests near the summits of ridges are xerophytic, oak-hickory type, while the forests of ravines and lower slopes are of the mesophytic type. Along ledges and ravines, extensive exposures of sandstone or limestone are common and support a diverse rock ledge vegetation. In the development of the diagnostic keys and the arrangement of the fami- lies no originality is claimed. Previously published treatises, particularly by Grout (1928-1940) and Welch (1957), have been particularly useful although extensive modification was often required. Recent monographs and revisions have consulted Nomenclature has been updated so as to be consistent with not only tlie most recent monographs and revisions, but also in general accord with "A list of the mosses of North America" by Crum, Steere, and Anderson (1965) and "Index Muscorum" (Wijk et al, 1959-1969). The arrangement of the families follows the sequence given by Brotherus (1924-1925). Descriptions are limited to the families and genera and have been synthesized from onlv the taxa occurring within the Interior Highlands. Consequently, these ions in t sense ^j . — — ^j - ., w Synonymy is not meant to be complete, but is included only for taxa where recent changes may result in some confusion when compared th earlier works prim an collected during this study or examined in herbaria of the University of Missouri, the University of Arkansas, the Missouri Botanical Garden, and the Field Museum of Natural History. Literature reports, particulariy by Gier (1955a) for Missouri, have also been considered. The commonness of each species refers to its distri- bution in expected habitats (Gleason & Cronquist, 1964) and is, of course, a judgement based upon my own field experience. The terms used to express commonness — rare, imcommon, common, and very common — must be considered r purely subjective. Support for the field and herbarium studies was made possible not only by my research grants from the Society of Sigma Xi and the National Science Founda- tion (NSF Grants G-9059, GB-67, and GB-4095), Their support is gratefully acknowledged. specimens it Missouri State College (SMS). All collection numbers cited are mine unless otherwise indicated. One final caution on the use of the keys. There is an old adage that "keys are made by people who don t need them for people who can't use them.*' There certain amount of truth dyn 4 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 being keyed out and that a particular collection may not possess all the stages that make identification relatively sure. Therefore, one should develop the habit, particularly if experience with mosses is limited, of comparing carefully each collection identified with its description and with a herbarium specimen that has been named by someone more experienced with mosses. If such collections are not available, any number of bryologists, including the author, are usually more than willing to examine unidentified or identified collections, provided the speci- mens are supplied in good condition and with adequate collection data. Ideally, such data should include the type of substrate (soil type, rock type, or species of tree and location on tree) upon which the moss was growing, general habitat of the species, exact location, date of collection, and collector. You should send enough material so that the person identifying the collection may retain a duplicate for his records. To aid in the use of the keys, a glossary of the more common bryological terms is included in Appendix I. An index to families and genera follows in Appendix II. Key to the Mosses of the Interior Highlands of North America^ 1. Leaves composed of a single layer of tvvo kinds of cells, narrow linear chlorophyl- lose-cells forming the meshes of a network enclosing large rhomboidal hyaline cells Sphagnaceae (1) 1. Leaves not as above 2 2. Plants large, erect, dendroid (tree-like) in growth form 3 2, Plants not dendroid - 4 3, Leaves irregularly serrate near apex Thamnobryum (29) 3. Leaves serrate, often coarsely sOj but not irregularly serrate Climacium (25) 4. Leaves composed almost entirely of costa made up of 2-3 layers of hyaline cells whose walls are perforated with circular pores and enclose a central layer of narrow chlorophyllose cells; plants in cushion-like tufts, whitish green _. Leucobryum (8) 4. Plants not as above 5 5. Plants acrocarpous, generally but not always erect in habit 6 5. Plants pleurocarpous, generally but not always creeping in habit 12 6. Leaves in two ranks or if in three ranks, leaves of some branches complanate Section A 6, Leaves in three ranks, not appearing complanate 7 7. Leaves ecostate Section B 7. Leaves costate 8 8. Leaves with lamellae on upper surface of costa Polytrichaceae (43) 8. Leaves without lamellae 9 9. Gemmae bearing structures conspicuous Section C 9, Gemmae bearing structures not conspicuous; gemmae, if present, on leaves or in axils of leaves 10 10. Leaves with hyaline apices or excurrent costae Section D 10. Leaves not as above 11 11. Leaves papillose or mammillose Section E 11. Leaves smooth Section F 12. Leaves costate . 13 12. Leaves ecostate _. Section G 13. Costa short and double (may be absent in some leaves) Section H 13. Costa reaching the middle of the leaf or beyond 14 14. Leaves not papillose - 14. Leaves papillose Section I Section J 2 Nmuber in parenthesis is the number of the family or the family in which the taxon may be found. _i rf_ J>;Cri*— l-Li^-*^L"^--^^"- ^ ^■'* ■ T I- ^ ~ri I -iz 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 5 Section A Plants acrocarpous: leaves distichous or in three ranks and complanate. 1. Leaves distichous _„ 2 1. Leaves only appearing distichous, in three ranlcs and complanate Mnium (17) 2. Leaves split at base and clasping stem Fissidens (2) 2. Leaves not split at base Bryoxiphium (5) Section B form not complanate, ecostate. 1. Leaf cells smooth Venturiella sinensis (21) 1. Leaf cells papillose - Hedwigm ciliata (26) Section C ranlced 1. Leaves ovate to ovate-lanceolate; many sterile stems bearing a gemmae-bearing cup composed of small leaves; peristome single, composed of 4 teeth „ Tetraphis pellucida (15) 1. Leaves long-lanceolate; some stems bearing pseudopodia that are naked or bearing minute leaf-like gemmae in a cluster at upper end; peristome double, teeth 16 Aulacomnium (18) Section D Plants acrocarpous; leaves three-ranked, costate, apices or excurrent costa hyaline; gemmae not present on special structures. 1. Plants dark green, blackish green, or golden green in color 2 1. Plants light green to whitish green in color - Bryum argenteum (16) 2. Leaves mostly lanceolate-subulate; alar cells inflated Campylopus introflexus (7) 2. Leaves not as above 3 3. Leaves obtuse, emarginate to truncate at apex; cells papillose; basal cells hyaline, rectangular to long — 5 3. Leaves not possessing the above combination of characters 4 4. Costa subpercurrent to shortly excurrent; septate brood bodies absent .. Grimmia (12) 4. Costa ending below apex; septate brood bodies present _. Orthotrichum diaphanum (23) 5. Plants small, 2-5 mm high; awn smooth Desmatodon plinthobius (11) 5. Plants larger, 5-15 mm high; awn smooth or serrate; if plants small, growing on trees ._ Tortilla (11) Section E Plants acrocarpous; leaves three ranked, costate, apices or excurrent costa not hyaline, cells papillose or mammillose; gemmae not present on special structures. 1. Plants minute, usually growing in wet, sandy sites Ephemerum (13) 2 1. Plants small to large 2. Leaves strongly serrate, bordered from base to apex by hyaline, elongated cells; the inner cells of sheathing base, thin-walled and hyaline -— Syrrhopodon texanus (9) 2. Leaves not as above . 3 3. Hyaline cells extending up the margin of the leaf base further than along the costa, forming a V; leaves oblong-lanceolate to broadly lanceolate-acuminate .. 4 3. Hyaline cells absent along the leaf base, or if present, only faintiy showing a V- shaped arrangement; leaves narrowly lanceolate to linear-lanceolate 5 4. Margins of leaf entire; basal hyaUne cells extending to the costa „ TorteUa humilis (11) 4. Margins of leaf minutely serrate near apex; basal hyaline cells not extending to the costa Pleurochaete squarrosa (11) 5. Leaves bordered with bands of 2-3 rows of light-colored cells Desmatodon (11) 5. Leaves not so bordered — 6 6, Papillae formed by projecting end walls of cells, evident at least on lower cells if not on upper cells Bartramiaceae ( 19 ) 6. Papillae centered over the lumen of all cells . — 7 g ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 rent; alar cells usually inflated 8. Leaves not as above 7. Leaves with a hyaline or yellowish basal sheath; margins strongly serrate from sheathing base to apex Timmia ( 20 ) 7, Leaves not as above , 8 8. Apices of leaves channelled, acuminate or acute; costa percurrent to short-excur- Dicranum ( 7 ) __. 9 9. Costa excurrent 10 9. Costa percurrent or ending below apex 15 10. Costa excurrent as a short, yellow or pellucid, rounded mucro; cells strongly papillose on back; leaves oblong to lingulate from ovate base, apices usually obtuse - - Barbuh ( 11 ) 10. Leaves not as above — - H 11. Costa excurrent into a long awn 12 11. Costa not excurrent as a long awn, leaves often ending in a small, short, sharp, nearly hyaline to hyaline point; margins plane or strongly involute above POTTIACEAE (11) 12. Awn smooth or nearly so 12. Awn spinose or sharply serrate 13 14 13. Leaves obovate, hngulate or panduriform; apices rounded or truncate; propagula often present at apex of stem, in axils of upper leaves, or on upper surface of leaves ToHuUl ( U ) 13. Leaves ovate to oblong-lanceolate; apices acuminate; propagula absent Phascum (11) 14. Plants small; leaves nearly linear, subulate-lanceolate; apices mosfly aciuninate or subulate Bruchia (7) 14, Plants moderately large; leaves oblong-spatulate; apices obtuse or truncate __ Tortula (11) 15. Costa ending below the apex 16 15. Costa ending in the apex or excurrent 18 16. Leaves lingulate; apices obtuse; capsule oblique-pointed and immersed in hair- like perichaetial leaves - _— Diphyscium (42) 16. Leaves and capsule not as above 17 17. Leaves elongate-ovate; apices usually obtuse and apiculate; margins coarsely toothed in upper V2-% . Aulacomnium (18) 17. Leaves lanceolate to lingulate; apices subacute, acuminate, or obtuse Pottiaceae (11) 18. Leaves lingulate, subspatulate; apices obtuse; margins plane and coarsely serrate above; propagula present in axils of upper leaves; plants usually growing on rocks in or along streams Hyophila involuta (11) 18. Plants not as above - 19 19. Leaves lanceolate to ovate-lanceolate or oblong-lanceolate - 20 19. Leaves oblong-lingulate to lingulate Desmatodon (11) 20. Costa widi elongated cells on upper surface; capsules exserted Barhula (11) 20. Costa without elongated cells on upper surface; capsules immersed to emergent Orthotrichaceae (23) Section F Plants acrocarpous; leaves three-ranked, costate, apices or excurrent costae not hyaline, cells smooth; gemmae not present on special structures. 1. Leaves distinctly bordered, at least in part, by 1 to several rows of narrow, hyaline cells „_„ 1. Leaves not distinctly bordered by narrow, hyaline cells 16 2 2. Alar cells clearly differentiated, usually enlarged and inflated; leaves broadly to narrowly lanceolate 2. Leaves not as above DlCRANACEAE (7) 3 4 3. Plants dark green, brownish green, or blackish green 3. Plants green, light green, whitish green, or reddish green 7 4. Plants minute, leaves subulate, growing on limestone -„ . Seligeria (6) 4. Plants not as above _. 5 5. Plants with primary stems prostrate and bearing numerous erect, crowded, and densely foliate branches Drummondia ( 23 ) 5. Plants not as above 6 ^-iri^-w- cJ ...rf^^t '-t -m t' 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 7 6. Leaves crisped when dry, linear from a lanceolate base; margins plane, entire or serrate above; apices obtuse to almost blunt . Ptychomitrium (22) 6. Leaves usually imbricate when dry, ovate-lanceolate to linear-lanceolate; margins plane to narrowly or strongly recurved; apices muticous to subobtuse ._ Grimmia (12) 7. Leaves crispate when dry _._ 8 7. Leaves not crispate when dry 9 8. Plants very small, stems usually less than 1 mm long; calyptra mitrate; seta arcuate 1 Campylostelium saxicdlum (22) 8. Plants larger, stems 2-10 mm long; calyptra cucullate; seta not distinctly arcuate - DiCRANACEAE (7) 9, Leaves secund, frequently falcate - Dicranaceae (7) 9. Leaves not secund or falcate 10 10. Plants small, stems usually averaging less than 4 mm in height 11 10. Plants larger, stems usually averaging greater than 5 mm in height 12 11. Capsules immersed, globular, dehiscing at equatorial line ApJianorhcgma (14) 11. Capsules not as above 15 12. Plants densely tufted; leaves oval, ovate, ovate-lanceolate to narrowly lanceolate „_. 13 12, Plants loosely gregarious; leaves oblong-lanceolate, obovate-lanceolate to obovate - FUNARIACEAE (14) 13. Median leaf-cells thick-walled, quadrate to hexagonal or roundish-quadrate, 6-10 ^ in diameter ___ 14 13. Median leaf-cells thin- or thick-walled, prosenchymatous, rhomboidal-hexagonal to linear, 14-100 ^ in longest direction - Bryaceae (16) 14. Margins entire; median leaf-cells rounded quadrate Didymodon trifarms (11) 14. Margins denticulate to serrate near apex; median leaf-cells quadrate, hexagonal- quadrate or irregular Ceratodon purpureus (4) 15. Capsules ovoid, cylindric, or pyriform -___ 19 15. Capsules sphaerical, indehiscent and immersed 18 16. Plants with large terminal rosette of leaves Rhodobryum roseum (16) 16, Plants not as above - 17 17. Median leaf-cells parenchymatous, mostly hexagonal or rounded _ Mnium ( 17 ) 17. Median leaf-cells prosenchymatous, rhomboidal, rhomboid-hexagonal to linear ... Bryaceae (16) 18. Leaves lanceolate-subulate; spores few {ca. 24), large, 145-174 fi in diameter, angular or tetrahedral, deep yellow or orange Archidium ohioense (3) 18. Leaves broadly ovate; spores smaller, 40-50 fi in diameter . Acaulon muticum van rufescens (11) 20 19. Capsules ovoid or cyHndric 19. Capsules pyriform , Bruchia (7) 20. Plants minute; capsules inoperculate . Ephemerum (13) 20. Plants larger; capsules operculate - DrrmciiACEAE (4) Section G Plants pleurocarpous; leaves tliree-ranked, ecostate, 1. Plants large and normally aquatic Fontinalis (24) 1. Plants small or large, not aquatic 2. Leaves papillose 2. Leaves smooth 2 3 4 3. Plants slender; leaves ovate to ovate-lanceolate; alar cells abruptly quadrate to short rectangular Schwetschkeopsis fabronia (32) 3. Plants more or less robust; leaves ovate-elliptical; alar cells roundish-quadrate; sec- ondary stems sparingly branched, branches curved outward and upward from sub- stratum Leucodon (28) „ 5 4. Branches and stems complanate-foliate at least in part 4. Branches and stems not at all complanate-foliate 6 5, Leaves oblong-ovate, soft, thin, Hght-green; median leaf cells 50-60 /i wide , - Hookeria acutifolia ( 30 ) 5. Leaves ovate-lanceolate, lingulate, or cultriform, green to yellow-green or brownish green; leaf cells 10 fi or less wide Neckeraceae (29) g ANNALS [Vol. 59 6, Secondao' stems not to sparingly branched, branches curved outward and upward from substratum; leaves ovate to ovate-lanceolate; alar cells numerous, roundish- quadrate 6. Secondary stems and leaves not as above Leucodon (28) 7 7. Leaves squarrose-spreading, triangular-cordate - Campylium (35) 7. Leaves not both squarrose-spreading and triangular-cordate — 8 8. Plants very small; leaves to 0.6 mm long; alar cells not inflated Flatydictya ( 35 ) 8. Plants larger; leaves to 2.5 mm long; alar cells inflated Sematophyllaceae (39) Section H Plants pleurocarpous; leaves three-ranked, costa short and double. 1. Leaves papillose 14 1. Leaves not papillose 2 2. Branches complanate-foliate 3 2. Branches not complanate-foliate ^-^ 5 3, Leaves falcate-secund Hypnaceae (40) 3. Leaves not falcate-secund 4 4. Nimierous quadrate alar cells extending 8-10 rows up margin Entodon (37) 4. Alar cells not as above Plagiotheciaceae (38) 5. Foliated stems and branches julaceous 6 5, Foliated stems and branches not julaceous 10 6. Stems conspicuously red; leaves broadly ovate-elliptic; alar cells enlarged and somewhat inflated - Pleurozium ( 37 ) 6. Plants not as above 7 7. Secondary stems curved outward and upward from substratimi; apices of leaves abruptly short-acuminate; alar cells numerous, roimdish quadrate Leucodon (28) 7. Plants not as above 8 8. Quadrate alar cells numerous, extending 10-20 rows along the margin __ Entodon (37) 8. Alar cells not as above 9 9. Alar cells composed of 3-8 conspicuously large, inflated, hyaline or colored cells Sematophyllum (39) 9. Alar cells not as above Plagiotheciaceae (38) 10. Tips of shoot frequently crowded with bud-like axillary gemmae; one inflated cell at extreme basal angle of leaf Platygyrium ( 40 ) 10. Plants not as above 11 11. Branches erect; leaves ovate to oblong-ovate, abruptly and broadly short-acuminate, concave; margins entire; alar cells numerous and quadrate or transversely elongate and extending upward % length of margin HomomalUum adnatum (40) 11. Plants not as above 12 12- Secondary stems numerous, abundantly subpinnately branched, often bending outward and upward from substratimi Forsstroemia ( 27 ) 12. Secondary stems not as above 13 13. Alar cells distinctly enlarged Sematophyllaceae (39) 13. Alar cells numerous, quadrate Pylaisiella (40) 14. Plants light to bluish green _ Myurella sibirica (31) 14. Plants green to dark green Pterigynandrum filiforme (37) Section I Plants pleurocarpous; leaves three-ranked the leaf or beyond, cells not papillose. L Plants dendroid from creeping stems 1. Plants not dendroid 2 3 2. Leaves irregularly serrate near apex Thamnobryum (29) 2. Leaves serrate, often coarsely so, but not irregularly serrate Climacium (25) 3. Plants aquatic to subaquatic; leaves carinate-conduphcate Fontinalaceae (24) 3. Plants not aquatic; if aquatic, not carinate-conduplicate - 4 4. Plants glossy, golden yellow to brownish yellow; branches turgid, terete-foliate, julaceous; branch leaves very concave, ovate, acumination of rounded apices twisted V2 turn Bryoandersonia (36) 4. Plants not as above 5 . ..litrbJi '-'■ — --^-^^ 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS g 5, Plants complanate-foliate; leaves oblong-cultriform, rounded obtuse ._ Neckeraceae (29) 5. Plants not as above 1. 6 6. Plants possessing conspicuous paraphyllia Cratoneuron (35) 6, Paraphyllia absent or inconspicuous 7 7. Branches and stems falcate-secund; leaves with distinct, hyaline, enlarged and inflated alar cells that form distinct auricles Drepanocladus ( 35 ) 7. Plants not as above . 8 8. Leaves with a distinct border of elongated, thick-walled cells Sciaromium (35) 8. Leaves not as above 9 9. Leaves squarrose to erect-spreading _ C ampylium (35) 9. Leaves not as above — 10. Plants small and slender, in soft, thin mats; quadrate alar cells filling entire base and extending 12-20 rows up margin; margins entire to laciniate-dentate or serrate- dentate 10, Plants not as above 10 Fabronia (32) --.- 11 IL Plants slender; quadrate basal cells numerous, distinctly different in shape from median leaf -cells 12 11. Plants not both slender and with numerous quadrate basal cells that differ markedly from median leaf cells 13 12. Leaves distinctly serrate; costa of branch leaves usually reaching middle of leaf or beyond ___„ _ Brachytheciaceae (36) 12, Leaves entire to obscurely serrulate Clasmatodon (32) 13, Branch leaves ovate-lanceolate; median leaf-cells rhombic-hexagonal; alar cells few, thin-walled, quadrate to rectangular; capsule erect, strongly contracted beneath mouth when dry Anacamptodon (32) 14 13. Plants not as above 14. Median leaf -cells short, rounded-ovate to isodiametric; if long, alar cells very numerous, rounded-quadrate to transversely oblong and capsule on a short seta Cryphaeceae (27) 14. Median leaf-cells long; alar cells not as above 15 15. Leaves ovate-lanceolate or ovate, apices long and slenderly acuminate, usually not plicate; margins entire to slightly serrulate; costa reaching middle of leaf to percur- rent or excurrent; median leaf -cells oblong-hexagonal, rhomboidal, to linear, 5-15 : 1; capsule curved and inclined, strongly contracted beneath mouth when dry Amblystegiaceae (35) 15. Leaves ovate-lanceolate, ovate, ovate-cordate to triangular-cordate, often plicate; margins serrate to serrulate; costa usually ending near middle of leaf in a dorsal spine; median leaf-cells linear, linear-flexuose, oblong-rhomboidal to elongate-rhom- boidal, 5-12: 1; capsule erect to inclined, symmetrical to asymmetrical, not strongly contracted beneath mouth when dry Brachytheceacae (36) Section J Plants pleurocarpous; leaves three-ranked, costa reaching middle of leaf or beyond, cells papillose. 1. Costa long and double Rhytidiadelphus triqxietrus (41) 1, Costa single - 2 2. Plants robust; leaves secund and strongly rugose; papillae dorsal, extending from angle of cell and strongly directed forward Rhytidium rugosum (41) 2, Plants not as above 3 3. Leaves papillose because of thickened cell angles 4 3. Papillae developed over lumen of cells 5 4, Plants slender, julaceous; cells strongly papillose Pterygynandrum filiforme (37) 4. Plants not as above; leaves of stem loosely imbricate Bryhnia (36) 5. Plants glaucous-green, branches jidaceous Theliaceae (31) 5. Plants not as above 6 6. Apex of leaf hyaline; costa ending near middle of leaf „ Lindbergia hrachyptera (33) 6. Apex of leaf not hyaline; if hyaline, costa ending near apex „ 7 7. Paraphyllia inconspicuous or entirely absent 8 10 THE [Vol. 59 7. Paraphyllia numerous, often branching THmniACEAE (34) 8. Capsules usually numerous, immersed in perichaetial leaves; secondary stems sparingly branched; costa protruding on dorsal side of leaf Cryphaea ( 27 ) 8. Plants not as above 9. Leaf bases not clasping the stem 9 Leskea (33) 9. Leaf bases subclasping to clasping Thuk)iaceae (34) 1. Sphagnaceae 1. Sphagnum L. The following treatment is adapted from a monograph by Andrews (1913). Plants primarily of bogs, wet rock ledges and sandy creek banks, erect, whitish green to bright green, often tinged with purple, pink, red, or brown, extensively branching in fascicles disposed spirally about stem, near apex branches often densely crowded; stems without central strand, central portion composed of thin-walled parenchymatous cells, passing gradually into thick-walled, pigmented prosenchymatous cells, these latter cells enclosed by one or more layers of thin- walled parenchymatous cortical cells that may be dimorphic and porose, and have spiral fibril-bands reinforcing walls inwardly; leaves of stems and branches of two kinds of cells, narrow, linear chlorophvUose-cells forming the meshes of network uiw ophyt smooth absent. a. Cortical cells of stems and branches reinforced by spiral fibril-bands b a. Cortical cells of stems and branches without spiral fibril-bands b, Chlorophyllose-cells of branch leaves usually isosceles-triangular in section \^ S palustre b. Chlorophyllose-cells of branch leaves equilateral-triangular in section 2. S. imbricatum c. Cortical cells of branches uniform, each with a pore at upper end; chlorophyllose- cells of branch leaves included 3. s. compactum c. Cortical cells of branches dimorphic, the large retort-cells with a neck and a pore in the axils of leaves; chlorophyllose-cells of branch leaves not completely included d d. Chlorophyllose-cells of branch leaves exposed exclusively or more broadly on outer surface d, Chlorophyllose-cells of branch leaves equally exposed on both surfaces or ex- posed exclusively or more broadly on irmer surface _ f e. Cortical cells of stem small and thick-walled 4^ g^ recutvum e. Cortical cells of stem large, thin-walled, in 1-3 layers __ . 5. S. cuspidatum f . Chlorophyllose-cells of branch leaves equally exposed on both surfaces g^ g^ subsecundum f. Chlorophyllose-cells of branch leaves exposed exclusively or more broadly on inner surface ^ g. The free convexity on die outer surface of hyaline cells of stem leaves mostly less than Va their diameter 7a. Sphagnum capillaceum var. capillaceum g. The free convexity on the outer surface of hyaline cells of stem leaves usually more than Ml their diameter ____ 7b. s. capillaceum var, tenerum 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS H 1. Sphagnum palustre L. Rare, Boston and Ouachita Mtns., Salem and Springfield Plateaus; on sandy soil, edges of streams and creeks. L r 2. Sphagnum imhricatum Hornsch. ex Russ. var. imbricatum Rare, Ozark Hills; around seepage areas of springs. 4 3. Sphagnum compactum Lam, & DC. Mtns.. Ozark Hills, eastern bluffs 4. Sphagnum recurvum P. Beauv. a. Branch leaves undulate . - :.- 4a. S. recurvum var. recurvum a. Branch leaves hardly undulate because of reduced size - 4b. S. recurvum var. tenue r 4a. Sphagnum recurvum var. recurvum Uncommon, Ouachita Mtns.; on sandy moist soil at bases of bluffs. 4b. Sphagnum recurvum var, tenue Klinggr. Rare, westem Salem Plateau; on moist sandy soil along narrow gully. 5. Sphagnum cuspidatum Ehrh. ex Hoffm. var. cuspidatum Mtns beneath creeks ( Waller 1967 ) . 6. Sphagnum subsecundum Nees ex Strum ■ Mtns.. Salem Plateau; on soil, edffes of creeks streams 7si. Sphagnum capillaceum (Weiss) Schrank var. capillaceum Reported from Boston Mtns., in Franklin County, Arkansas, by Wittlake (1950). 7b . Sphagnum capillaceum var. tenerum (Sull. & Lesq. ex SuU.) Crum (1971) Rare, Ozark Hills, eastern Salem Plateau; on moist shaded sandy soil of and 2. FiSSIDENTACEAE Members of this First, leaves are arranged in two vertical rows in a single plane ( distichous ) and, second, each leaf consists of a "boat-Hke" portion ( the vaginant or sheathing laminae ) , which clasps the stem, a dorsal (lower) lamina, and a terminal (apical) lamina. Both laminae this Only the J2 ANNALS OF THE MISSOURI BOTANICAL GARDEN £Vol. 59 1. Fissidens Hedw. a. Leaves rigid, opaque, several cell layers in thickness 15. F. grandifrons a. Plants not as above b b. Plants aquatic; apical lamina 2-3 times as long as die vaginant lamina 16. F. fontanus c. c. b. Plants not aquatic; apical lamina as long as the vaginant lamina or shorter; i£ longer, not 2-3 times — c Leaves obtuse, entire 4. F. obtusifolius Leaves acute to acuminate - d d. Leaves distinctly bordered with a row of elongate cells that extends to the apex or subapical region of the apical and dorsal laminae ___ e d. Leaves not distinctly bordered; if bordered, the border confined to the vaginant lamina or border not composed of elongated cells g e. Leaves widely spreading, forming an angle of 45° or more with the stem 3. F. viridulus e. Leaves not widely spreading, forming an angle of 30-45° with the stem f f. Leaf cells 6-10.0 /a in diameter, obscure; leaves narrowly lanceolate to oblong lanceolate; apex acute to acuminate - 1. F. minutulus f. Leaf cells 10-13 fi in diameter, pellucid; leaves lanceolate to ligulate; apex acute to obtuse 2. F. pusillus g. Leaves bordered by elongated cells, border confined to the vaginant laminae, at times indistinct, absent or confined to perichaetial leaves, often intramarginal; if absent plants less than 5 mm tall^ h g. Leaves not bordered by elongated cells, plants 5 mm or more tall k h. Leaf cell papillose i h. Leaf cells smooth or bulging j i. Costa ending 2-several cells below apex .-. 8. F. garberi i. Costa percurrent to excurrent 7. F. ravenelii j. Leaf margins entire except near apex 5, F. exiguus j. Leaf margins sub-entire to irregularly crenate; margins of the vaginant laminae strongly crenate-dentate 6. F. exilis k. Leaves coarsely and irregularly crenulate-serrate 1 k. Leaves finely and evenly crenulate by projecting cell angles — _ n 1. Costa ending several cells below apex and covered with low mammillose cells 14. F. subhasilaris 1. Costa percurrent and not covered with mammillose cells m m. Leaf cells obscure, enlarged and elevated tumid cells appearing singly or in groups in surface view of apical lamina; leaf cells 7-9 ^ in diameter; border of paler or thicker-walled cells distinct; leaves crowded on stem, rolled inward from tips when dry 1 1. F. cristatus m. Leaf cells clear, tumid cells absent or few; leaf cells 13-16 fi in diameter; border of differentiated cells often indistinct; leaves distant on stem, curving underneath stem and with upper % wrinkled and distorted when dry 13, F. adianthoides n, Costa ending several cells below apex; sporophyte terminal 10. F. osmundioides n. Costa percurrent to shortly excurrent; sporophyte lateral o o. Costa stout, usually filling apiculus, short-excurrent; cells of apical lamina conically papillose or mammillose — - 9. F. taxifolius o. Costa slender, not filHng apiculus, percurrent; cells of apical lamina plmripapillose 10. F. bushii 1. Fissidens minutulus Sull. var. minutulus Common throughout Interior Highlands; often abundant on moist, shaded rocks. The bordered groups (Sections Bryoidium and Semilimbidium) of this genus are not well understood. Many workers, notably Crum (1964), consider not only Fissidens minutulus but F. viridulus var. viridulus and F. exiguus to be forms of ^ Occasionally smaller forms of Fissidens taxifolius or F. osmundioides may key out here. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 23 F, bryoides Hedw. Other workers do not even recognize F. pusilhis. Grout (1936) considers it a synonym of F. minutulus, and Wijk and Margadant ( 1960) consider it a subspecies of F. minutuJus, However, this work follows the treatment by Pursell (1957) that recognizes not only F. minutulus^ but also F. pusillus and F. viridulus var. viridulus. Since F. exiguus also seems distinct in Ozark mate- rial, it is also retained. For the present, such an arrangement best expresses the situation. 2. Fissidens pusillus (Wils.) Mild. Fissidens minutulus Sull. subsp. pusillus (Wils.) Wijk & Marg. Uncommon, Boston Mtns., Prairie Plains, Salem and Springfield Plateaus; on shaded rocks that are usually in or near creeks and streams. This species is undoubtedly closely related to the previous one. In Fissidens minutulus^ leaves are narrowly lanceolate to oblong-lanceolate with an acute apex, and the cells of the apical lamina are small (6-10 fi in diameter) and consequently appear obscure. In contrast, the leaves of F. pusillus are broader, lanceolate to ligulate with an acute to obtuse apex; the cells of the apical lamina are relativelv large <' 10-13 u in diameter) and annear pellucid. 3. Fissidens viridulus (Sw.) Wahlenb. var. viridulus Fissidens bryoides Hedw. var. viridulus (Sw.) Kindb, Common throughout the Interior Highlands; on rocks and soil, particularly of moist and well shaded forests. In addition to the character mentioned in the key, this species may be recog- nized in the field by its unusually distinct margins, relatively robust size (stems 3-8 mm in length), and the lustrous dark to light green color of the plants. 4. Fissidens obtusifolius Wils. var. obtusifolius Common, Boston Mtns., Prairie Plains, Sale limestone or sandstone rocks in or along the ed2( Locally, this species may be very abundant. A species easily recognized in the field because of its obtuse leaves and the pale green, at times almost glaucous, color of the colonies. 5. Fissidens exiguus Sull. subsp. exiguus Mtns cks This species is of doubtful validity. As noted previously, Crum (1964 Fissidens 6. Fissidens exilis Hedw. Reported from northern Salem Plateau by Gier (1955a:34) from collections made by N. L. T. Nelson. The location of Nelson's collection is unknown. 14 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 7. Fissidens ravenelii Sull. Uncommon, Boston and Ouachita Mtns., Ozark Hills, Salem and Springfield Plateaus; restricted to moist shaded crevices of sandstone. Superficially this species resembles Fissidens garheri; the presence of a per- current to excurrent costa in F, ravenelii, however, is usually a reliable distin- guishing character. 8. Fissidens garheri Lesq. & James Common, Boston and Ouachita Mtns., Prairie Plains, Salem Plateau; on sandy dolomite and sandstone along streams and creeks. Occasionally plants occur deep within rock crevices. The presence of this species in the Interior Highlands was first noted by Ireland from a collection made in Cedar County, Missouri (see Redfeam, 1961). The border of the vaginant lamina is often difficult to demonstrate. Breen (1963:7) notes that the border is usually restricted to the uppermost pair of leaves (perichaetial). In the Interior Highlands, the separation of this species from Fissidens ravenelii is not always easy. However, in F. garheri the costa in most leaves is not percurrent and ceases 2-several cells below the apex. Florschiitz (1964) places both this species and Fissidens ravenelii in synon- ymy under F. guianensis Mont. Considering the difficulty encountered in sepa- rating F. ravenelii and F. garheri in the Interior Highlands, he may be correct, 9. Fissidens taxifolius Hedw. var. taxifolius Very common throughout the Interior Highlands; on shaded soil, rock ledges and crevices, bases of trees, and decaying wood. 10. Fissidens osmundioides Hedw. var. osmundioides r Uncommon, Boston and Ouachita Mtns., Ozark Hills, Salem Plateau; on shaded sandstone ledges and soil banks along creeks and streams. Sterile forms of this species may be confused with Fissidens taxifolius or poorly developed forms of F. cristatus. However, a costa which distinctly ends 2 to several cells below the apex will serve to separate it from F. taxifolius^ and the even and finely serrulate to crenulate margins will distinguish it from F. cristatus. 11. Fissidens cristatus Wils. ex Mitt. var. cristatus typ the Interior Highlands; occurring in nearly every rocks trees, decay in This species may be confused with the next, however the distinctions the kev have been demonstrated bv Anderson and T?rvan MQKfi^ fn T^^ reliable. 12. Fissidens hushii (Card, & Ther.) Card. & Ther. Uncommon, Boston and Ouachita Mtns., Ozark Hills, Salem and Springfield Plateaus; on shaded soil and decaying wood. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 15 13. Fissidens adianthoides Hedw. var. adianthoides Uncommon, Boston and Ouachita Mtns., Salem and Springfield Plateaus; on moist shaded soil along creeks and streams, occasionally on rocks along seepage areas. 14, Fissidens suhhasilaris Hedw. Common throughout the Interior Highlands; on moist soil, shaded rocks, bases and trunks of trees, and decaying wood. 15. Fissidens grandifrons Brid. var. grandifrons Uncommon, Boston Mtns., Salem Plateau, St. Francois Mtns.; on submerged stones, usually near the emergence of swiftly flowing springs. very this to be locally very abundant. County, M near the spring outlet are densely covered with this species, a condition not unusual at other large springs. 16. Fissidens fontanus (B. Pyl.) Steud. Fissidens debilis Schwaegr, Common throughout Interior Highlands; on stones, roots, and branches in streams and creeks. 3. Archidiaceae 1. Archidium Brid. Plants erect, small, terrestrial; leaves narrow, lanceolate to subulate; margins entire perichaetial leaves long-pointed, to 2.25 mm, 3-4 times as long as the capsule; capsule cleistocarpous, sessile, ovoid, thin walled and containing about 12-24 very large (120-160 fi in diameter), yellow to orange spores clearly visible through the thin wall of the capsule; oil bodies conspicuous in spores. Sterile plants of this genus are difficult, if not impossible, to separate from When fertile, the ovoid, sessile, cleistocarp lOres found in the mosses will at once st this other 1. Archidium ohioen^e Schimp. ex C. Miill. Rare, known only on the Springfield Plateau from Dade County ca. 1 mile W of Bona along Highway H, SW V^ Sect. 33, R. 25 W. open overlooked when sterile because and the tendency of plants to be half buried in sand. IQ ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 4. DiTRICHACEAE Plants dark green to yellow-green, erect, small to moderate in size, usually less than 2 cm tall, gregarious to scattered; leaves lanceolate to lanceolate-sub- ulate from a broader base, not distinctly secund, occasionally contorted when dry; costa percurrent to excurrent and occupying the larger part of the awn; leaf-cells smooth to faintly papillose; alar cells not inflated; capsule immersed to exserted on a long seta, spherical, ovoid, or cylindrical, cleistocarpous or- dehiscing regularly; peristome, when present, of 16 teeth split to base or nearly so into 32 filiform divisions. Three genera of this family are represented in the Interior Highlands. When sterile, identification is difficult and dubious. Consequently the following key is based primarly upon sporophytic characters. This is not to say that generic identification of sterile material is not possible in some cases, but should only be undertaken after complete familiarity with fertile plants. a. Plants small, usually less than 2 mm high; capsule cleistocarpous; peristome absent 1. Pleuridium a. Plants usually larger; capsule dehiscing regularly; peristome present b b. Leaves ovate-lanceolate to narrowly-lanceolate, green to dark-green, often con- torted; margins conspicuously revolute, rarely to apex; upper leaf-cells quadrate to short-rectangular 2. Ceratodon b. Leaves lanceolate-subulate, light green to yellowish green, often slightly secund; median leaf-cells rectangular to linear 3. Ditrichum 1. Pleuridium Brid. Plants small, green to yellowish green; leaves erect-spreading, mucronate to subulate from an ovate to lanceolate base; margins entire to minutely serrate; costa sub-percurrent, to percurrent, to excurrent; cells mostly elongate-rectangu- lar, smooth; alar cells not enlarged or inflated; capsule erect on short seta, ovoid to subglobose, apiculate, immersed in perichaetial leaves; peristome absent. a. Broad base of perichaetial leaves short, scarcely reaching the base of capsule a. Broad base of perichaetial leaves longer, nearly reaching tfie end of the capsule -x- 1. P. subtdattiin 2. P. acuminatum 1. Pleuridium suhuhtum (Hedw.) Rabenh. Common, Arkansas Valley, Boston Mtns., Mississippi Valley, Ozarlc Hills, Salem and Springfield Plateaus; on soil of pastures, fields, and rocky ridges. Capsules present from early spring to summer* Because of the small size of this species, it is likely to be overlooked; how- ever, when fruiting it is easily identified by its yellow-green appearance, subulate leaves, and immersed, small, subglobose, cleistocarpous capsule. Pleuridium sub- uliitum is difficult to separate from P. acuminatum. In addition to the distinctions the kev to the sDCcies that (1956) reports cytological differences: P. $u P. acuminatum is a tetraploid ( n = 26 ) . 13) and t^ T- JL _ . 'J jj* ■ i"-JI-'^ _■ 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS yj 2. Pleuridium acumirmtum Lindb. Common, Prairie Plains, Salem Plateau; on moist soil of open forests, in fields and pastures. Like the preceding species, this one is frequendy overlooked, but when fruit- ing may be separated from P. subulatum by the characters noted above. 2. Ceratodon Brid. 1. Ceratodon purpureus (Hedw.) Brid. Caespitose plants, green when young, often becoming reddish when old; leaves ovate-lanceolate to narrowly lanceolate, twisted and contorted when dry; margins revolute from apex nearly to base, usually serrate above; median leaf- cells rectangular to square, weakly coUenchymatous; seta dark red to purplish, 1-2.5 cm long; capsule oblong, deeply grooved, more or less curved and unsym- metric; peristome with 16 bifid, dark red teeth. Common, Arkansas Valley, Boston and Ouachita Mtns., Ozark Hills, Prairie Plains, Salem and Springfield Plateaus; on various types of substrates — bases of tree, sandy soil, soil pockets of sandstone, dolomite, granite, and rhyolite boulders. In urban areas very common on cedar shingle roofs. This species is often anomalous particularly when sterile. Amateur and pro- fessional alike have been fooled by its many forms. However, the intersections of the coUenchymatous median leaf-cells are characteristic when viewed under low magnification {ca. 100 X) in that they usually appear as a row of small four-pointed stars due to refraction of light. 3. Ditrichum Hampe Small caespitose mosses; leaves loosely spreading to erect, green to light green or yellowish; costa broad, percurrent to excurrent; upper leaf-cells sub- quadrate to long rectangular, smooth; seta slender, long; capsule oblong, cylin- dric to ovoid, erect to inclined. a. Leaves long, linear-subulate, more or less erect to loosely spreading; awn slender, as long as to much longer than broad base; apex acute 1 b a. Leaves not long, linear-subulate, appressed; awn shorter than broad base; apex obtuse 2, D, lineare b. Seta red; upper leaves 2-3 mm long; costa in awn clearly distinct L D. pusillum b. Seta yellow; upper leaves often more or less secund, 3-5 mm long; costa in awn 3. D. pallidum indistinct 1. Ditrichum pusillum (Hedw.) Hampe Uncommon, Ozark Hills, Prairie Plains, Salem Plateau; on soil, especially along creeks and gullies. 2. Ditrichum lineare (Sw.) Lindb. Apparently rare, eastern Salem Plateau; in crevices of moist sandstone and rhyolite rocks. 18 ANNALS [Vol. 59 3. Ditrichum pallidum (Hedw.) Hampe Very common throughout Interior Highlands; on bare soil of woods and fields, occasionally in moist rock crevices. Sterile plants may be confused w^ith Dicranelh heteromulla; however in the leaves are 5-4 5. Bryoxu^hiaceae 1. Bryoxiphium Mitt Slender plants similar to Fissidens; leaves conduplicate, not clasping the stem at base; stem with a bulb-like base. Many authors place the genus Bryoxiphium in the Fissidentaceae. However, Love and Love (1953) have concluded from their rather exhaustive study of this genus that it should be placed in a separate family. 1. icum (Brid.) Mitt Mtns., Prairie Plains, eastern Salem plants may be abundant on moist, shaded vertical sandstone bluffs beneath overhanging ledges. 6. Seligeriaceae subulate brownish than 'ularged absent. the geria^ whose alar cells are undifferentiated. rocks growin 1. Seligeria B.S.G. a. Peristome lacking . 1, S. donniana a. renstome present _ r b. Awn short and thick, usually developing abruptly from the sheading base' £^^^ b. Awn not short and thick, usually slender, gradually tapering from a lanceolate base ,. 3 s ^^.^^^ 1. Seligeria donniana (Sm.) C. Miill. Sterile forms of this taxon cannot be adequately distinguished from the next taxon. Although not yet reported from the Interior Highlands, there is no phyto- geographic reason for its absence. Further field studies may yet reveal its presence in our area, 2. Seligeria calcarea (Hedw.) B.S.G. Uncommon, Boston Mtns., Salem and Springfield Plateaus; in crevices and poclcets of shaded calcareous rocks. Probably more common in the Interior 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS \Q Highlands than distribution records indicate as it is seldom collected carefully searched for. 3. Seligeria pusilla (Hedw.) B.S.G. eastern Salem Plateau in Franklin Countv, Missouri 1955fl: 38). 7. DiCRANACEAE Plants erect, variable in size, usually growing in mats or tufts, pale, to yellowish, to dark green; stems often branching, densely foliate and tomentose; leaves costate, variable in outline, straight, often curved, secund, sometimes cris- pate or subulate; leaf-cells smooth or papillose, variable in shape from rectangular near the base to shorter, often irregular or isodiametric near the apex; alar cells frequently clearly differentiated into distinct auricles; capsule short to long exserted, curved or erect, pyrifonn to cylindrical; peristome absent or of 16 teeth, each divided V2 or more into 2 slender divisions. Cytological evidence presented by Bryan (1956) indicates that the genera Bruchia and Trematodon should be included in this family as it was originally conceived by Brotherus (1924:172-214) and not in the Ditrichaceae. a. Plants small; leaves subulate to ovate-lanceolate, not secund; capsule obovoid to pyriform, inoperculate — 1. Bruchia a. Plants small to large; leaves narrowly to broadly lanceolate or subulate, often secund; capsule cylindrical, often curved, operculate b b. Capsule with a long, slender neck 2. Trematodon b. Capsule without a long, slender neck c. Alar cells conspicuously inflated or enlarged c. Alar cells not conspicuously inflated or enlarged - c f - d d. Leaves crisped when dry 4. Rhabdoweisia d. Leaves not crisped when dry ___ ..__ _ e e. Costa at least % the width of the leaf base; plants pale green; numerous rudimen- tary branchlets often present in the axils of upper leaves 7. Brothera e. Costa less than % the width of the leaf at base; plants darker green; rudimentary branchlets absent — 3. Dicranella f. Costa less than V2 the width of the leaf at base _ 5. Dicranum f. Costa V2-% the width of the leaf at base 6. Campylopus h Bruchia Schwaegr. Small gregarious plants of moist soil in open fields and pastures; leaves subulate to abruptly acuminate from an ovate to lanceolate base; seta shorter than to barely as long as perichaetial leaves; capsule pyriform, inoperculate, A difficult genus for which mature fruiting material is generally required for identification of the species. Although Bruchia donneJlii and B, texana have been reported from the Interior Highlands, they have been excluded for lack of reliable material to study. a. Leaves short acuminate, imbricate-appressed 2. B. hallii a. Leaves subulate-lanceolate, not imbricate-appressed _ „„ 1. B, flexuosa 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 1. Bruchia flexuosa (Sw. ex Schwaegr.) C. Miill Bruchia sullivantii Aust. lachita Mtns., Ozark Hills, and 5 pastures. Sporophytes maturing summer. 2. Bruchia liallii Aust. Rare, Springfield Plateau; on open sandy soil 2. Trematodon Michx. 1. Trematodon longicollis Michx. Plants gregarious to scattered, yellowish green in color; leaves more or less crisped when dry, gradually tapering from a broad, clasping base to a lanceolate or subulate apex; margins more or less revolute; costa percurrent; leaf-cells smooth, thin-walled, elongate hexagonal above, rectangular below; seta 2-3 cm long; capsule very distinctive, operculate, yellow to yellow-orange with a neck about twice as long as urn; operculum long rostrate. Uncommon, Boston Mtns., Ozark Hills, Springfield Plateau; on open sandy soil along creek and river bottoms and in open woods. 3. Dicranella (C, Miill.) Schimp, Plants small, gregarious or forming open to dense tufts or mats, bright green to yellowish green; leaves ovate-lanceolate to narrowly lanceolate from a sheath- ing base, often falcate secund; costa less than % the width of the leaf at base; differentiated; seta red to yellow, 5-30 curved, smooth to furrowed when dry 2-3 divisions. Key to Fruiting Plants a. Mature capsules erect, symmetric or nearly so, seta red-brown 1. D. rufescens a. Mature capsule inclined, iinsymmetric; if erect then distinctly furrowed when dry b b* Capsule plicate or furrowed when dry and empty; seta yellow 3. D. heteromalla b. Capsule smooth or only slightly furrowed when dry and empty; seta reddish 2. D. varia Key to Sterile Plants (Less Reliable) a. Margins of leaves recurved, entire =„„. 2. D. varia a. Margins of leaves plane, entire to serrulate b b. Margins of leaves entire except at apex; leaves erect-spreading to slightly secund; mostly of moist, open soil along streams and ditches 1, D, rufescens b. Margins of leaves serrulate nearly to base; leaves usually clearly falcate-secund; habitat various 3. D. heteromalla 1. Dicranella rufescens (With.) Schimp. Rare, Salem Plateau; on moist soil, chiefly bare of other plants. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 21 2. Dicranella varia (Hedw.) Schimp. Uncommon, Ozark Hills, Springfield and Salem Plateaus; on moist soil and rocks, chiefly along streams and ditches. 3. Dicranella heteromalla (Hedw.) Schimp. a. Capsule erect and symmetric, urn coarsely furrowed when dry and empty 3b. D, heteromalla var. orthocarpa a. Capsule inclined and asymmetric, capsule plicate when dry and empty 3a. D. heteromalla var. heteromalla 3a. Dicranella heteromalla var. heteromalla types soil or rock substrates. 3b. Dicranella heteromalla var. orthocarpa (Hedw.) Jaegr. & Sauerb Rare, Ozark Hills; on sandstone. 4. Rhabdoweisia B.S.G. 1. Rhahdoweisia denticulata (Brid.) B.S.G. m Plants small, tufted, yellowish to darlc green; leaves oblong to linear-lance- olate, crisped when dry, finely serrulate above, teeth often few and distant, apex acute; costa stout, ending just below apex; upper leaf-cells mostly round-quadrate. 2-5 teeth Mtns 5. Dicranum Hedw. Plants small to large, in loose to dense tufts, yellowish to dark green, usually radiculose below; leaves lanceolate to ovate-lanceolate, erect-spreading, crisped, or falcate-secund, narrowly costate with distinctly enlarged alar cells; basal leaf- cells rectangular to elongate, walls often pitted; upper leaf-cells rectangular, oblong-rhomboidal, or irregularly shaped, walls plain to pitted, often papillose on the lower surface; seta erect; capsule cylindrical and symmetrical to inclined, curved and asymmetiical; operculum long rostrate; peristome teeth red, single. the middle into 2-3 b a. Walls of lower leaf cells not pitted - a. Walls of lower leaf cells pitted - a b. Costa broad, Vs or more the width of the leaf base 3. D. fulvum h. Costa narrower, Vi-Vo tlie width of the leaf base c c. Leaves strongly crisped when dry, margins irregularly serrate to serrulate; flagelli- form branches absent 1- O- montanum c. Leaves crisped, erect-spreading to secund when dry, margins essentially entire; flagelliform branches usually present 2. D. flagellare d. Upper leaf-cells elongate, more or less pitted f d. Upper leaf-cells 1-2 : 1 e e. Leaves widest some distance above base; upper leaf-cells irregular in shape, highly mammillose lower surface 5. D. spunum 22 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 e. Leaves widest near the base; upper leaf cells more or less uniform in shape, smooth or slightly papillose on lower surface 4. D. sabuletorum L Leaves distinctly transversely undulate 6. D. polysetum f. Leaves not transversely undulate 7. D. scoparium 1. Dicranum montanum Hedw. Uncommon, Boston and Ouachita Mtns., Ozark Hills, Prairie Plains, Salem Plateau; on shaded dolomite, rliyolite, sandstone, and bases of trees. 2. Dicranum flagellare Hedw. Common, Boston and Ouachita Mtns., Ozark Hills, Salem Plateau, St. Fran- cois Mtns.: mostly on decavinff wood, occasionally on shaded sandstone. 3. Dicranum fulvum Hook. Common throughout the Interior Highlands; on vertical, shaded, non-calcare- ous rocks. 4. Dicranum sabuletorum Ren. & Card. Common throughout the Interior Highlands; on rocky soil, decaying wood, crevices and ledges of rocks. Often mixed with Dicranum scoparium, 5. Dicranum spurium Hedw. Uncommon, Boston and Ouachita Mtns., Ozark Hills, Salem Plateau, St. Francois Mtns.; on rock ledges, especially granite, sandstone and dolomite, occa- sionally on rocky, acid soil of ridges. 6. Dicranum polysetum Sw. Dicranum rugosum (Funclc) Hoffm. ex Brid. Dicranum undulatum Ehrh, ex Web & Mohr Rare, known only from the Springfield Plateau in Barry Co., Missouri, where it occurred on soil in a mixed oak-cedar forest along a ridge ca. V2 mile east of Roaring River State Park. 7. Dicranum scoparium Hedw. Very common throughout the Interior Highlands; on soil of ridges, wooded slopes, and rock ledges. Occasionally on decaying wood and bases of trees. 6. Campylopus Brid. Plants dark to yellowish green, in loose to compact tufts; leaves erect spread- ing to secund, narrowly lanceolate-subulate; costa often to % width of the leaf base, in cross section showing stereid cells above and below guide cells, upper surface cells larger than guide cells; seta 5-10 mm long; capsule ovoid, cylindric to pyriform elliptic; peristome of 16 teeth divided to near the middle into papil- lose divisions; apparently fruiting infrequently. a. Leaves, particularly older ones, ending in a hyaline hair point; narrow rectangular hyaline cells extending up the margin from base as in Tortella 2. C. introflexus a. Leaves without hyahne hair points; no hyaline cells extending up margin ._ 1. C. flexuoms 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS £3 1- Campylopus jlexuosus (Hedw.) Brid. Campylopus tallulensis Sull. & Lesq. Uncommon, Boston and Ouachita Mtns sandstone. 2. Campylopus introflexus (Hedw,) Brid. Uncommon, Boston and Ouachita Mtns.; on open flat (occasionally vertical) exposures of sandstone and novaculite. Often forming extensive cushions. Excluded species: Campylopus fragilis (Brid.) B. S, G. — The collection cited by Lowe (1919) is C. jlexuosus. 7, Brothera C. Mull 1. Brothera leana (Sull.) C. Miill. Common, Boston Mtns., Ozark Hills, Salem and Springfield Plateaus; usually in crevices of dolomite or sandstone of bluffs, particularly along the upper reaches of bluffs and cliffs. Excluded taxa: Dicranoweisia cirrata (Hedw.) Lindb. ex Mildc and D. crispula (Hedw.) Lindb. ex Milde — ^CoUections from Arkansas determined as these species are Weissia controversa. 8. Leucobkyaceae 1. Leucobryum Ilampe Plants in whitish to glaucous, dense, spongy, cushion-like tufts, often to one foot in diameter; stems erect, to 15 cm or more high; leaves thick, lanceolate, the base composed almost entirely of a broad costa composed of 2 or more layers of large, empty, hyaline cells surrounding a layer of small chlorophyUose-ccUs; seta erect; capsule erect and symmetric to inclined and asymmetric; peristome single, composed of 8 or 16 teeth, entire or split to the middle into two sections* a. Leaves in cross section near the base at the middle of the costa 4-5 layers of cells thick (2 layers of hyaline cells above and 1-2 layers below a single layer of chloro- phyllose cells); leaves 5-8 mm long 1. L. glaucum a. Leaves in cross section near the base at the middle of the costa 3 layers of cells thick ( 1 layer of hyaline cells above and below a single layer of chlorophyllose cells); leaves 2-4.5 mm long 2. L. albidum 1. Leucobryum glaucum (Hedw.) Angstr. in Fr. Very common throughout the Interior Highlands; on acid soil of ridges, on rock ledges, particularly granite, rhyolite, felsite, sandstone, and dolomite. Often mixed with Polytrichum and Dicranum. '.C--^. ^taf^Wc- II 24 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Plants growing under dry conditions, particularly along oak-hickory ridges, are difficult to separate from the next species. However, structural differences of the lower costa as well as examination of older stem leaves will usually prove the key characters to be reliable. 2. Leucobryum alhidum (Brid. ex P. Beauv.) Lindb. ledges. ock 9. Calymperaceae 1. Syrrhopodon texanus SuU. green or brown cushions, 1^ high; leaves crisped when dry, linear to lingu tips; marginal cells elongated, hyaline, and coarsely toothed; costa rough on both surfaces; upper leaf-cells papillose; basal cells hyaline, forming acute angles above; sporophytes unknown in material from the Interior Highlands. Rare, Boston Mtns., Ozark Hills, and eastern Salem Plateau; on vertical shaded sandstone and shaded soil. 10. Encalyptaceae 1. Encalypta procera Bruch Encalypta streptocarpa Hedw. {sensu Flowers). Plants in loose, brownish green tufts; leaves crisped when dry, broad, lingu- late; costa single; upper leaf-cells short, densely papillose; basal leaf-cells hyaline, oblong, smooth; filamentous brood bodies arising from leaf axils. Rare, Salem Plateau; on exposed soil of ridge. Known only from Ozark County, Missouri, from a NW facing cherty slope ca. 3 miles SE of Dora SW Y* angustatum. W. When dry 1 1 . Pottiace ae tufts to dark green to blackish in color; leaves variable in shape, frequently contorted when dry; costa strong, often long excurrent; basal leaf-cells usually thin walled and hyaline; upper leaf-cells small, isodiametric, thick-walled and more or less papillose; alar cells not differentiated; capsule usually erect, sessile to long exserted on a straight seta; peristome when present composed of 16 teeth, often divided and spirally twisted. a. Plants small, bulbiform; upper leaves broadly ovate; cells smooth; capsule immersed a. Plants not as above 13. Acaulon b b. hyaline cells of leaf-base extending up the margin further thanaiong'the costa formmg a V; leaves oblong-lanceolate to broadly lanceolate-acuminate b. Hyaline cells of leaf base not as above; if faintly showing V-shaped arrangement """" ot hyahne basal cells, then leaves narrowly lanceolate to linear-lanceolate d 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS £5 c. Hyaline cells of leaf base not reaching the costa 8. Pleurochaete c. Hyaline cells of leaf base reaching the costa 7. Tortella d. Margins of the upper half of leaf involute e d. Margins of the upper half of leaf plane or reflexed -^^ - f e. Peristome present, inserted at the mouth of the capsule; costa 60 /i or more wide at base; leaves linear-lanceolate from a broadly oblong to oblong-lanceolate base 6. Trichostomum e. Peristome absent; if present, inserted below the mouth of the capsule; costa 50 fi or less wide at base, if wider, then leaves narrowly oblong-lanceolate .- 2. Weissia f. Leaves erect-spreading, tips usually curled, very narrowly linear 6. Trichostomum f. Leaves not possessing this combination of characters g g. Leaves bordered by a band of 2-3 rows of slightly elongated, lighter colored cells 15. Desmatodon g. Leaves not bordered as above ^.. h h. Leaves strongly papillose with many of the papillae c-shaped i h. Leaves smooth; if papillose, papillae not c-shaped j i. Leaves without excurrent costa; if costa excurrent, upper hyaline basal cells with distinct trigones; plant small, usually less than 5 mm tall — - 15. Desmatodon i. Leaves with excurrent costa (may be short); upper hyaline basal cells without dis- tinct trigones; plant taller, 0.5 cm or more tall; if shorter, plants bearing propagula 16. Tortula j. Leaf margins in upper half reflexed; if plane, tlien leaves either torn and broken, or plants growing on trees K j. Leaf margins plane; leaves not torn or broken; plants not growing on trees s k. Costa v^dth a single dorsal stereid band 1 k, Costa with two stereid bands n L Costa ending below the apex, spurred above - 4. Husnotiella L Costa excurrent, not spurred above - m m. Leaves ovate to oblong-lanceolate; apex acuminate; growing on soil 14. Phascum m. Leaves obovate, Ungulate or spatulate, ovate-lanceolate to oblong; apices acute, obtuse, emarginate, truncate or blunt; plants on rocks and trees — 16. Tortula n. Leaf margins 2-3-stratose — -- -^ 10. Didtjmodon n. Leaf margins not bistratose o o. Upper leaf cells not to only weakly papillose p o. Upper leaf cells distinctly papillose Q p. Basal leaf-cells hardly different from upper leaf-cells 10. Didymodon p. Basal leaf-cells thin-walled, clearly different from upper-leaf cells 12. Barbula q. Peristome teeth twisted 12. Barbula q. Peristome teeth erect, not twisted ._.„„„ r r. Leaves entire throughout 10. Didymodon r. Leaves denticulate to strongly dentate at apex 11. Bryoerythrophyllum s. At least some of the leaves broadly lingulate -- y s- Leaves not broadly ligulate t t. Leaves toothed just above the base at the upper margin of the hyaline cells 5. Eucladium t. Leaves not as above u u. Costa ending somewhat below apex; plants growing on vertical calcareous rocks z u. Costa percurrent to excurrent; if ending below apex, plants growing on soil v V. Leaf apex obtuse 12. Barbula V. Leaf apex acute w. Leaves closely imbricate when dry w „ 12. Barbula w. Leaves crisped or curled when dry x x. Leaf cells pluripapillose; the cells of basal ^/4 of leaves rectangular, elongated, hyaline; plants yellowish green 6, Trichostomum X. Leaf cells bipapillose; the basal % of leaves with few and inconspicuously elon- gate cells; plants bluish green 1. Anoectangiu7n y. Plants with numerous polymorphous gemmae in axils of upper leaves 9. Hyophila y. Plants without gemmae . 17. Scopelophila z. Leaves bistratose in patches; archegonial inflorescences latcrial 1. Anoectangium z. Leaves not bistratose; archegonial inflorescences terminal 3. Gymnostomum 26 ANNALS [Vol. 59 1. Anoectangium Schwaegr. Plants leaves contorted when dry linear-lanceolate, anex obtuse ending below apex; upper leaf-cells papillose; basal leaf-cells hyaline and rec- theridial a. Leaves acute, bluish green 1, A. sendtnerianum a. Leaves obtuse, whitish 1. A. obtusifoUum 1. Anoectangium sendtnerianum B.S.G, Molendoa sendtneriana (B.S.G.) Limpr. Rare, Boston Mtns.; on moist calcareous rocks. This species may be confused with Trichostomum cylindricum, Iwatsuki and Sharp (1958) have observed that the latter taxon differs from A. sendtnerianum by having 3 or more papillae per cell and large, rectangular, elongated, and hyaline cells in the basal quarter of the leaf. 2. Anoectangium obtusifoUum (Broth. & Par. ex Card.) Grout Rare, Boston Mtns.; on sandstone. May grow mixed with Eucladium verti- cillatum. 2. Weissia Hedw. Small, densely tufted, yellowish to bluish green plants; leaves narrowly lance- olate, costate; margins strongly involute to plane above, erect-spreading when moist, crispate when dry; upper leaf-cells strongly papillose; capsule immersed or exserted; operculum functional to non-functional; peristome absent or of 16 teeth inserted below the mouth of the capsule, a. Capsules immersed in perichaetial leaves b a. Capsules exserted beyond perichaetial leaves b. Setae often clustered; leaf margins both plane and involute above - 2, W. ludoviciana b. Setae solitary; leaf margins strongly involute above 1. W, muehlenbergiana 0. Costa 70 fi or more in width at base of leaf 4. W, tortilis c. Costa less than 60 fi in width at base of leaf 3. W. controversa 1. Weissia muehlenbergiana (Sw.) Reese & Lemmon (1965). Astomum muehlenbergianum (Sw.) Grout Common diroughout the Interior Highlands; on open soil of fields, road banks, paths, and la^vns. Sporangia produced in late winter and early spring. 2. Weissia ludoviciana (Sull.) Reese & Lemmon (1965). Astomum ludovicianum ( Sull. ) Sull, Very rare, Arkansas River Valley; on open soiL 3. Weissia controversa Hedw, Weissia viridula Hedw. costate; cells very leaf margins entire, strongly involute above; capsule cylindrical, smooth or ■ ■■■ -^- ^ ^-^ - Y-~r--\- y k/ y .-zz^'S. -0-1 ■ J I 4. 1. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 27 ; plicate when dry, exserted on a yellow seta; operculum functional; peristome of 16 teeth. Two varieties are represented in the Interior Highlands and may be separated as follows: a. Seta 3-8 mm long _ 3a. W. cotdroversa var. controvcrsa a. Seta 10-15 mm long 3b. W. controvcrsa var, longiscta 3a. Welssia controversa var. controversa Very common throughout the Interior Highlands; on soil and rocks of open areas such as lawns, pastures, roadways. 3b» Weissia controversa var. longiseta (Lcsq, & James) Crum, Stccre & Anderson ( 1964 ) . Weissia viridula Hedw. var. atistralis Aust. Reported from eastern Salem Plateau in Franklin, St. Louis, and Ste. Gene- vieve Counties, Missouri, by Gier (1955(7:38). 4. Weissia tortilis (Schwaegr.) C. Miill. knowi Moore (1965) from Conway Co., eastern Boston Mtns Wlien sterile this species may be confused with Trichostomum jamaicensus. However, the linear-lanceolate leaves with oblong-ovate base easily separate this species from Weissia tortilis. May be a tetraploid of W. controversa (L. E. Anderson, personal communication). 3. Gymnostomum Nees, Hornsch. & Sturm Densely tufted mosses, usually growing on moist, vertical limestone, often forming dense cushions several centimeters deep, dull to bright green in color; leaves erect-spreading when moist, little contorted to suppressed, incurved, and contorted when dry, oblong- to linear-lanceolate, keeled, apex acutely obtuse to acute; costa strong, vanishing below apex; upi>er leaf-cells papillose; seta long exserted, erect; capsule erect, symmetric, oval to oblong, glossy at maturity; operculum long rostrate; peristome absent. a. One or both margins recurved below; leaves little contorted when dry; upper leaf- cells clear, 12-15 m wide; operculum usually remaining attached to exserted colu- mella 2. G. rccurvirostrtnn a. Leaf margins not recurved; leaves usually crispate when dry; upper leaf-cells ob- scure, 7-12 fi wide, densely papillose; operculum not persistent 1. G, aeruginosum 1. Gymnostomum aeruginosum Sm. Common, Boston Mtns., Ozark Hills, Prairie Plains, Salem and Springfield Plateaus; on vertical, moist limestone and sandstone especially at bases of bluffs. Confusion between this species and Gymnostomum calcareum Nccs & Hornsch. is, as noted by Crum and Anderson (1956), quite common in this country. These authors have come to the conclusion that nearly all specimens called G. calcareum in the United States are forms of G. aeruiLinosum. Only from Cahfomia have 28 GARDEN [Vol. 59 they seen specimens which they felt could essentially be assigned to G. calcareum. There is apparently a taxonomic problem involved here which can only be resolved by a careful world-wide monographic study. Until this is done a lands is assigned to G. aeruginosum. all the 2. Gymnostomum recurvirostrum Hedw. a. Papillae of leaf-cells comparatively few; stems not papillose 2a. G. recurvirostrum var, recurvirostrum a. Leaf-cells densely papillose; stems papillose 2b. G. recurvirostrum var. latifolium 2a. Gymnostomum recurvirostrum var. recurvirostrum Common, Boston Mtns., Salem and Springfield Plateaus; on moist, vertical limestone and dolomite. Occasionally on sandstone. latifolium (Zett.) Flow, ex Crum Gijmnostomum Uncommon, Boston Mtns., Salem Plateau; on limestone and occasionally on sandstone. 4. Husnotiella Card. 1. Husnotiella revoluta Card. var. palmeri (Card.) Bartr. Plants small, 4-6 mm tall, in dense, light green tufts; leaves erect spreading when moist, contorted when dry, linear-lingulate, acute to apiculate; margins entire, revolute above; costa strong, ending abruptly below the apex, wider and spurred on dorsal surface; leaf-cells small, 7-10 fi in diameter, obscure, rounded and more or less papillose. Rare, known only from a collection made by W. B. Drew (67) from the Salem Plateau at Alley Spring State Park, Shannon Co., Missouri, on soil of limestone ledge. May be confused with Didymodon tophaceus from which it may be distin- guished by the possession of a spurred costa and absence of strongly decurrent leaves. 5. Eucladium B.S.G. 1. Eucladium verticillatum (Brid.) B.S.G. tufts, 1-3 dry widely spreading when moist, linear-lanceolate to linear-subulate, acute; costa toothed basal cells; upper leaf-cells clear, papillose, rectangular to quadrate, thick-walled; lower leaf-cells hyaline and larger than upper cells. Common, Boston Mtns., Ozark Hills, Salem and Springfield Plateaus; on vertical to underside of overhanging limestone (rarely sandstone) subject to seepage. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS £9 6. Trichostomum Bruch Plants small to moderate in size, 5-30 mm tall, yellowish green to dark green, in loose to dense tufts; leaves linear-lanceolate, to 8 mm long, strongly contorted when dry, involute to nearly plane above; costa strong, usually cxcurrent; leaf- cells small and densely papillose; capsule cylindrical, exserted upon an erect seta 8-15 mm long; peristome teeth 16, inserted at the mouth of the capsule, a. Margin of upper leaf plane or involute only near apex b a. Margin of upper leaf strongly involute 1. T. jamaicense b. Leaves distinctly curled from tip when dry, very narrowly linear, plane to apex „ _„_ .___....__„_. 3. T. mollissimum b. Leaves twisted but not curled from tip when dry, linear-lanceolate to ligulate, often involute near apex 2. T, tenuirostre 1, Trichostomum jamaicense (Mitt.) Jaeg. & Sauerb. Weissia jamaicense (Mitt.) Grout Uncommon, Boston Mtns., Salem and Springfield Plateaus; on limestone, dolomite, felsite and sandstone. When sterile this species may be confused with Weissia controversa and W. tortilis. Generally plants of W. controversa are shorter, their leaves are tightly crispate, and the costa is less than 50 fi wide near the base. In contrast, plants of W. janmicensis are taller, and their leaves are more loosely crispate with the leaves curled mainly at the tips. The costa usually exceeds 65 /i in width near the base. 2. Trichostomum tenuirostre (Hook. & Tavl.) Lindb. Trichostomum cylindricum, (Bruch) C. Miill. Rare, Boston Mtns., Salem Plateau; on rocks. Reported from Taney County, Missouri, by Gier (1955fl: 38) and Missouri (no County given) by Crum and Anderson (1958). This latter paper should be consulted for an excellent discus- sion of the variability of this taxon. 3. Trichostomum mollissimum (Broth, ex Bartr.) Crum Tortella mollissima Broth, ex Bartr. Rare, SW-Salem Plateau in Douglas County, Missouri; on sandy, vertical shaded dolomite, growing with Myurella sihirica. status L. E. Anderson (personal County are T. tenuirostre. 7. Tortelh (C. Miill.) Limpr. 1. Tortella humilis (Hcdw.) Jcnn. Plants in erect, loose to dense yellowish green to green tufts; leaves crisped when dry, oblong-lanceolate, apex acute to obtuse; margins plane, entire, some- times undulate; costa shortly excurrcnt in a mucro-point; upper leaf-cells small, 30 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 densely papillose; lower leaf-cells hyaline, rectangular, extending obliquely higher up the margin than the costa, abruptly becoming small and chlorophyllose; capsule oblong, exserted on a seta about 1.5 cm long; peristome teeth 32, twisted 2-3 times. Very common throughout the Interior Highlands; on decaying wood, bases of trees, and soil; occasionally on rocks. Excluded species: Tortella tortiiosa ( Hedw, ) Limpr. — Reports from Taney County, Missouri, by Gier (1955^: 38) are based upon collection by H. S. Conard (Nov. 30, 1940) and Paul Rissens (Aug. 29, 1931) deposited in the Herbarium of the State Uni- versity of Iowa. Both of these collections have been examined and are Tortella humilis. 8. Pleurochaete Lindb. 1. Pleurochaete squarrosa (Brid.) Lindb. Plants in loose tufts, 1-10 cm tall, often decumbent, yellow-green in color; leaves flexuose when dry, squarrose when moist, lanceolate-acuminate from a wide, sheathing base; margins finely serrate, undulate; costa strong, excurrent to percurrent; upper leaf-cells quadrate, papillose; marginal basal cells rectangu- lar, hyaline for 5-7 cells in towards the costa below, decreasing to 1-2 cells above; median basal cells rectangular, colored; plants usually not fruiting. Common, Boston Mtns., Salem and Springfield Plateaus; on soil in glades when limestone or sandstone rocks are close to the surface. Usually forms very loose mat-like growths, especially beneath cedar trees. 9. Hyophila Brid. 1. Hyophila involuta (Hook.) Jaeg. & Sauerb, Hyophila tortula (Schwaegr.) Hampe Plants in dark green tufts; leaves inrolled and curled when dry. Ungulate, obtuse, nearly entire to coarsely serrate above; margins plane above, recurved below; costa strong, ending below apex to slightly excurrent; upper leaf-cells quadrate, slightly papillose to smooth; basal leaf-cells rectangular; polymorphic gemmae in axils of upper leaves; sporophytes absent. Common throughout the Interior Highlands; on shaded rocks in and along edges of creeks and streams. 10. Didymodon Hedw. lanceolate twisted to only incurved 3 obtuse: margins entire. 2-3 botli dorsal and ventral rows of stereids present in cross section; upper leaf-cells small, round, smooth or papillose, slightly elongated at base; capsule erect, cylin- 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 31 dric, exserted upon a short, 1-1.5 cm long, red seta; peristome teeth 16, variously divided, straight to slightly twisted. a. Margin of leaves 2-3 stratose in upper part 1. D. rigidtthis a. Margin of leaves unistratose, recurved in upper part b b. Leaves strongly decurrent; leaf-cells papillose 2, D, toptiaceus b. Leaves not strongly decurrent; leaf-cells smootli to only weakly papillose _ 3. D. trifarius 1. Didymodon rigidulus Hedw, Rare, known only from the southern Springfield Plateau from a collection by R. R. Ireland, Jr. {3520, US) from Roaring River State Park, Barry Co., Missouri; on limestone rock outcrop. 2. Didymodon tophaceus (Brid) Lisa Uncommon, Salem and Springfield Plateaus; on moist, shaded calcareous rocks, especially where seepage occurs, rarely on soil. 3. Didymodon trifarius (Hedw.) Rohl. r Reported from the Salem Plateau in Camden and Dallus Counties, Missouri, by Gier (1955a: 34). 11. Bryoerijthrophyllum Chen 1, Bryoerythrophyllum recurvirostrum (Hedw.) Chen Didyjnodon recurvirostris (Hedw.) Jenn. Plants in loose green tufts, tending to be reddish brown below; leaves erect- spreading when moist, crisped, flexuous, and curled when dry, lanceolate from a broad base; margins revolute, dentate (often obscurely so) at apex; costa strong, excurrent in a pellucid apiculus; upper cells quadrate, densely papillose; lower cells rectangular, h}'aline to reddish brown; capsule oblong-cylindric upon an erect, reddish seta; peristome teeth 16. Uncommon, Boston Mtns., Salem and Springfield Plateaus; usually on cal- careous rocks, occasionally on soil. 12. Barhula Hedw. loose tufts green to yellowish green; leaves linear-lanceolate to ovate-lingulate, contorted or crisped when dry; margins entire, plane or revolute; costa strong, ending just below the apex to excurrent; upper leaf-cells small, diick-walled, papillose; basal cells larger, thin-walled and smooth; propagula often present; capsule exserted, cylindric; peristome teeth divided into 32 thread-like segments, usually strongly twisted. a. Leaves oblong, lingulate or ligulate b a. Leaves lanceolate or subulate, tapering gradually from an ovate base . e b. Seta yellow; perichaetial leaves convolute 1. B. convoluta b. Seta red; perichaetial leaves not convolute c 32 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 c. Upper leaf-cells smooth to slightly papillose 4. B. ehrenbergii c. Upper leaf-cells densely papillose d d. Propagula present 3. B. cruegeri d. Propagula absent 2. B. unguiculata e. Cells of upper siuface of costa linear 5. B. fallax e. Cells of upper surface of costa quadrate . 6. B. acuta 1. Barbula convoluta Hedw, Uncommon, Boston Mtns., Salem Plateau; on shaded to open soil 2. Barbula unguiculata Hedw. Common, throughout the Interior Highlands; on calcareous soil and rock ledges, old fields, waste places, edges of creek and streams. 3. Barbula cruegeri Sond. ex C. Miill. Common, Ouachita Mtns., Ozark Hills, Prairie Plains, Salem and Springfield Plateaus; on moist, calcareous rocks and rock crevices. The status of this taxon in relation to Barbula unguiculata is confused espe- cially in the Interior Highlands where both taxa occur. Gier (1955a: 32) calls attention to this problem and suggests controlled experiments might help resolve this problem. 4. Barbula ehrenbergii (Lor.) Fleisch. Uncommon, Salem Plateau; on moist, shaded calcareous rocks. Plants are often encrusted with lime. 5. Barbula fallax Hedw. Uncommon, Salem and Springfield Plateaus; on soil of crevices and ledges of limestone. 6, Barbula acuta (Brid.) Brid. Two varieties are encountered in the Interior Highlands and may be sepa- rated as follows: a. Plants in short dense tufts 6a. B. acuta var. acuta a. Plants in loose, elongate tufts 6b. B. acuta var. bescherellei 6a. Barbula acuta var. acuta Gier (1955fl: 32). the eastem Salem Plateau in St. Louis County, M Jaeg Rare, Springfield Plateau; on flat exposed Hmestone along dry creek bed in small ravine ca. 4 miles south of Springfield, on Mo. Hwy. 160. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 33 13. Acaulon C. Miill. 1. Acaulon muticum (Hedw.) C. Miill. var. rufescens (Jaeg. & Sauerb.) Crum ( 1969 ) . Acaulon rufescens Jaeg. & Sauerb. Plants sub-globose, bulbiform, small, less than 2 mm tall, yellowish green; leaves broadly ovate, concave; margins plane below, sharply serrate and reflexed above; costa strong, percurrent to exeurrent; lower cells smooth, hyaline, rectan- gular, to 100 fi long; upper cells smooth, smaller, rhomboid-hexagonal; capsule erect, immersed, globose; spores smooth, 40-50 fi in diameter. Rare, Boston Mtns., Ozark Hills, eastern Salem Plateau, Prairie Plains; on sandy soil. 14. Phascum Hedw. 1. Phascum cuspiclatum Hedw. m oblong-lanceolate, acuminate; margins entire and revolute in the middle; costa and quadrate to hexagonal, papillose; capsule ovoid-globose, immersed to slightly emergent, cleistocarpous. the as follows: a. Costa exeurrent la. P. cuspidatum var. cuspidatum a. Costa long exeurrent lb. P. cuspidatum var. americanum la. Phascum cuspidatum var. cuspidatum Uncommon Prairie Plains, Springfield Plateau; on soil, often sandy, of old fields, pastures, and waste places. lb. Phascum cuspidatum var. americanum Ren. & Card, ex Ren. & Ther. Uncommon, Salem and Springfield Plateaus; habitats similar to var. cus- pidatum. 15. Desmatodon Brid. nail, usually less than 6 mm tall, in compact to loose, ye tufts; leaves closely appressed, often contorted when dry obtuse arly hair-point; upper cells irregularly quadrate, densely papillose with many c- shaped papillae; lower cells rectangular to elongate-hexagonal, smooth, pale; capsule oblong to cylindric, exserted upon an erect seta; peristome teeth 16, cleft to base, at times poorly developed. a. Leaves bordered by a band of lighter-colored cells . 1, D. ported a. Leaves not bordered . b b. Upper leaves awned or hair-pointed 2. D, plinthobius b. Upper leaves not awned or hair-pointed 3. D, obtusifolius 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 1. Desmatodon ported James ex Aust. Uncommon, Boston Mtns., Ozark Hills, Salem and Springfield Plateaus; on dolomite and limestone. 2. Desmatodon plinthobius SuU. & Lesq. ex SuU. Common, Boston and Ouachita Mtns., Prairie Plains, Salem and Springfield Plateaus; on limestone, dolomite, and sandstone. 3. Desmatodon obtusifolius ( Schwaegr. ) Schimp. Common throughout the Interior Highlands; on sandy soil and sandstone, rarely dolomite and limestone. 16, Tortula Hedw, Plants small to large, in dense to loose mats or rocks, soil or trees; leaves oblanceolate, obovate, spatulate, elongate-ligulate to elongate-spatulate, often forming a crowded rosette near the stem apex, twisted and appressed when dry, widely spreading when moist, apex obtuse, rounded, truncate or emarginate; margins plane to revolute; costa strong, excurrent as an awn; upper cells small, strongly papillose, basal cells much larger, smooth, hyaline, often appearing sud- denly differentiated on each side of costa; brood-leaves (propagula) often present on surface of leaf or in the upper leaf axils. a, Costate brood-leaves found in axils of upper stem leaves; upper leaves often torn and broken in various ways; costa shortly excurrent 2. T. fragilis a. Plants not possessing the above combination of characteristics b b. Ecostate brood-leaves present in axils of upper leaves or on upper side of leaf costa . c b. Brood-leaves absent d c. Brood-leaves borne in axils of leaves at apex of stem 4. T. pagorum c. Brood-leaves borne on upper surface of leaves _ 3. T, papillosa d. Aw^n smooth 1. T. muralis d. Awn sharply serrate . 5, T. ruralis 1. Tortula muralis Hedw. Reported from the Ouachita Mtns., McCurtain Co., Oklahoma, by Sharp (1930: 49). 2. Tortula fragilis TayL dry Known 3. Tortula papillosa Wils, ex Spruce Rare, southern Springfield Plateau; on oak on other trees. 4. Tortula pagorum (Milde.) De Not. Common throughout the Interior Highlands: vertical calcareous and sandstone rocks. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 35 5. TortuJa ruralis (Hedw.) Gaertn,, Meyer & Schreb. Uncommon, Springfield Plateau; on open calcareous soil and rocks, especially along bluffs. 17. Scopelophila (Mitt.) Lindb. 1. Scopelophila ligulata (Spruce) Spruce Merceya ligulata (Spnice) Schimp. Plants in loose to compact tufts; leaves light to bright green in color above, often reddish to brown below; stems radiculose; leaves broadly spatulate to ligu- late, broadest near apex, blundy pointed on obtuse; margins entire and plane (may be slightly recurved below); costa ending in the apex; upper leaf-cells quadrate, oblong, to irregular; basal cells larger, thin- walled and often inflated, marginal basal cells narrower. Rare, Ouachita Mtns., Ozark Hills; on sandy soil. Considerable variation in morphology related to moisture has been reported by Noguchi (1956) and confirmed by Zander (1967). Both the loosely pulvinate, flaccid "hydric** form and the more densely pulvinate "montane*' form occur in the Interior Highlands. 12, Grimmiaceae 1. Grimmia Hedw. Plants in green to blackish or brownish green cushions, mats, or dense tufts, often appearing whitish because of hyaline leaf apices; leaves crowded, usually imbricate when dry, erect-ascending, spreading, or recurved when moist, lance- olate, often from an oblong to ovate base, plane, concave, carinate, or caniculate; margins plane or revolute; costa sub-percurrent to excurrent, flat or terete in cross section, often terminating in a hyaline, smooth to spinulose hair-point; upper leaf-cells rounded-quadrate, unistratose to bistratose, papillose; lower leaf-cells smooth, quadrate to rectangular; cell walls thin and smooth to thick and sinuose; capsule symmetric, immersed to exserted upon a straight to arcuate seta; peri- stome single, composed of 16 perforate or cleft teeth; calyptra cucuUate or mi- trate; with rare exceptions all taxa grow on rocks. A difficult genus in need of critical revision. The following key is based upon a key to North American species of Grimmia by Sayre (1952), a. Upper leaves (excluding perichaetial ) without hair-points .___ b a. Upper leaves with hair-points (may be inconspicuous or absent in some leaves) e b. Apex of some or all of the upper leaves obtuse or rounded 2. G. alpicola b. Apex of upper leaves acuminate or acute c. Leaves less than 1.5 mm long c. Leaves 1.5-3.0 mm long c 1. G. apocarpa d. Peristome teeth red, pointed; capsule ovate; most leaves ovate-lanceolate __ 2. G, alpicola d. Peristome teeth pale, truncated; capsule oblong; most leaves linear-lanceolate 1. G, apocarpa e. Margins of leaves inroUed in some part of the leaf; apex often dra\\'n out to a channeled acumen . f 36 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 1. e. Margin of leaves plane or erect, or one or both recurved; apex not channeled g f Seta curved; costa terete, hair-point spinose 11. G. olneyi f. Seta straight; costa mostly flat, hair-point nearly smooth 9. G. ovalis g. Leaves broader above the middle or oblong, abruptly acxmiinate; hair-point IV2 times the length of upper leaves 3. c. wrightii g. Leaves broader below the middle, gradually acuminate or acute; hair-point shorter h h. Leaf-cells mostly imistatose, large and clear 4. G. rauei h. Leaf-cells mostly bistratose, small and obscure i i. Costa flat; leaf apex usually acute; capsule exserted 5. G. laevigata Costa terete; apex usually acuminate; capsule immersed j j\ Leaves almost lingulate, ending in a long denticulate hair-point to % the length of leaf 10. G. puhinata }. Leaves lanceolate to ovate-lanceolate, hair-point shorter, less than % the length of leaf ^ k. Most mature leaves 2 mm long or longer 1 k. Most mature leaves 1-1.5 mm long n 1. Both margins equally and distinctly revolute m 1. Margins plane or only one revolute and the other slightly recurved below .. 8. G. arizonae m. Leaves narrowly acuminate from ovate base; costa forming about % of the leaf base; basal leaf-cells rectangular, thick-walled and sinuose; alar cells hyaline 7. G,' pilifera m. Leaves lanceolate, some acute; lower leaf-cells 2:1; alar cells not hyaline ._ 1. G. apocarpa n. Plants not filiform; alar cells not differentiated; most leaves with both margins revolute 1, q apocarpa n. Plants filiform; alar cells differentiated, usually hyaline; some or all leaves with one revolute margin and the other incurved 6. G. teretinervis 1. Grimmia apocarpa Hedw. Schistidium apocarpum (Hedw.) B.S.G. This is a highly polymorphic taxon, and there is a great deal of confusion regarding not only the status of subspecific taxa but with the next taxon. Varie- ties have been described that may be local or general environmental modifica- tions. Nyholm (1956: 145) has suggested what is perhaps the only answer to this perplexing group, cultivation experiments. Five varieties are known, reported from, or may occur in the Interior Highlands and can be separated by the fol- lowing differences, a. Upper leaves (excluding the perichaetial) not piliferous b a. Upper leaves piliferous _ _ h. Leaves under 1.5 mm long lb. G. apocarpa Vox" dupretU b. Leaves 1.5-3 mm long „ le. G. apocarpa var. ohscurimridis c. Mature leaves 2 mm long or longer Ic. G. apocarpa var. stricta Mature leaves 1-1.5 mm long j d. Plants dark green or black; capsule oblong _-_.. la. G. apocarpa var. apocarpa d. Plants bright green; capsule globose Id, G. apocarpa var. conferia la. Gnmrnia apocarpa var. apocarpa Common throughout the Interior Highlands; on dry, exposed rocks, usually c. typ lb. Grimmia apocarpa var. dupretii (Ther.) Sayre Rare, Springfield Plateau; on flat exposed limestone. Known only from Roaring River State Park, Barry Co., Missouri. This variety has recently been placed in the species Grimmia alpicola by Crum n971- 168) 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 37 Grimmia apocarpa var, stricta (Turn.) Hook. & Tayl. Grimmia apocarpa Hedw. var. gracilis (Schleich.) Web. & Mohr Reported from Shannon Co., Missouri, by Gier {1955a: 35), Id. Grimmia apocarpa var. conferta (Funck) Spreng. Uncommon, Boston Mtns.; on rocks. le. Grimmia apocarpa var. ohscuriviridis Crum Not known from Interior Highlands, but Sayre (personal communication) believes that it should be. 2. Grimmia alpicola Sw. ex Hedw. GriTJimia apocarpa var. alpicola (Hedw.) Hartm. Schistidium alpicola (Hedw.) Limpr. Common throughout the Interior Highlands; on exposed rocks, especially limestone, along edges of creeks and gullies. May be confused with the preceding taxa. However, when fertile they may be separated by the shape of the capsule. Grimmia alpicola has a more or less globular capsule, while the capsule of G. apocarpa is more distincdy ovoid to ellipsoid. 3. Grimmia wrightii (SulL) Aust. Reported from Missouri by Sayre (personal communication). 4. Grimmia raui Aust. Reported from Missouri by Sayre (personal communication). 5. Grimmia laevigata (Brid.) Brid. Very common throughout Interior Highlands; as a pioneer on exposed rocks Often locally abundant. May be recognized by its decurrent, flat, spinulose, hyaline apex. 6. Grimmia teretinervis Limpr. Rare, Springfield Plateau; on open limestone. Known only from Greene Co., Missouri, on bluffs above James River just below Lake Springfield Dam. 7. Grimmia pilifera P. Beauv. Common throughout the Interior Highlands; as a pioneer species on exposed to shaded rocks of all types, rarely on decaying wood. Often locally abundant. anzonae Sayre (personal communication) reports that this species occurs in Oklahoma therefore, be 38 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 9. Grimmia ovalis (Hedw.) Lindb Rare, Salem Plateau; on rocks. 10. Grimmia pulvinata (Hedw.) Sm, ex Sm. & Sowerby Uncommon, Ouachita Mtns., Prairie Plains, Springfield Plateau; on dry, open flat sandstone exposures and occasionally on brick walls. Apparently most fre- quent in the Prairie-border regions of the Interior Highlands. 11. Grimmia olneyi SulL Rare, St. Francois Mtns.; on non-calcareous rocks in exposed sunny places. Excluded taxa: Rhacomitrium aciculare (Hedw.) Brid. — ^The collection from McCurtain County by Sharp (1930) is Grimmia alpicola. 13. Ephemeraceae 1. Ephemerum Hampe r Small inconspicuous plants, stems almost absent, growing from persistent protonema; leaves costate or ecostate, lanceolate; margins strongly toothed to serrate; capsule cleistocarpous, ovoid, apiculate; calyptra campanulate, 0.2 mm or more in over-all length. The following treatment is based upon a study of this family by Bryan and Anderson (1957), a. Cells in the middle third of leaf lamina smooth, in distinct diagonal rows from costa to margin 3. E. cohaerens a. Cells of the leaf lamina papillose above, usually conspicuously so; if smooth, not in diagonal rows b b. Spines of upper leaves mostly recurved at an angle of 45** or more; leaves usually narrowly linear-lanceolate; lamina papillose or smooth above; leaf-cells at middle of leaf about 4-8 : 1 2. E, spinulosum b. Upper leaves almost entire to strongly toothed; most of the teeth extending at an angle of less than 45"*; lamina conspicuously papillose above; leaf -cells at middle of leaf less than 4 : 1 c c. Upper leaves lanceolate, without shoulder la. E. crassinervium var. crassinervium c. Upper leaves broadly lanceolate-subulate, with slight to prominent shoulder lb. E. crassinervium var. texantim crassinervium (Schw^aegr.) Hampe var. crassinervium eastern Known Jefferson Co., Missouri this that reach 1.5 mm in length and 0.2 mm in maximum v^idth, margins serrate in upper %, dense areolation in upper half, and a strong costa running from base to apex (Bryan & Anderson, 1957). merum crassinervium var. texanum (Grout) Bryan & Anderson, ted from Boston Mtns., Conway Co., Arkansas, bv Moore (1965' i9'72] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 39 2. Ephemerum spinulosum Bruch & Schimp. ex Schimp. Rare, western Springfield Plateau; on soil, edge of creek. Known only from vicinity of Horse Creek, 2 miles south of Olympia on County Rd. A, Cedar Co., Missouri, Irehnd 3273 (US). Plants of this taxon may be distinguished from the preceding taxa by the presence of unusually abundant and persistent protonema at the time leafy gametophytes mature (Bryan & Anderson, 1957). 3. Ephemerum cohaerens (Hedw.) Hampe Rare, Salem Plateau; on moist, shaded, sandy soil. Known only from vicinity W., Redf Jack 14. FUNARIACEAE Plants annual, sometimes biennial, in low, hght green, loose to clustered tufts, characteristically growing in open areas on bare soil or soil sparcely covered by other plants; stems erect; upper leaves closely clustered, concave, costate, often serrate above; leaf-cells large, parenchymatous, ca. 20-30 fi in v^dth, rhombic- to oblong-hexagonal to rectangular, lax, border weakly defined; capsules im- mersed to exserted on long setae, globose to pyriform, symmetiic to strongly inclined and asymmetric; peristome when present composed of an outer row of 16 reddish to brownish red teeth and an inner row of 16 yellowish, often rudi- mentary segments. a. Capsules immersed; peristome absent b a. Capsules emergent to exserted; peristome present or absent c b. Capsule functionally inoperculate; exothecial cells not strongly coUenchymatous _ i^ Fhyscomitriella b. Capsule operculate, dehiscing along the equatorial line; exothecial cells strongly coUenchymatous . 2. Aphagnorhegma c. Capsule imsymmetric and curved .„_ c. Capsule symmetric, straight d d. Seta short, about length of capsule; calyptra 4-angled . 4. Pyramidula d. Seta longer than capsule; calyptra symmetrically rostrate .___— 3. Physcomitrium 5. Funaria 1. Fhyscomitriella B.S.G. 1. Physcomitrielki patens (Hedw.) B.S.G, Aphanorhegma patens (Hedw.) Lindb. Small, inconspicuous plants with immersed, inoperculate capsules; leaves erect-spreading, lanceolate, costate; margins serrate; median leaf-cells oblong- 3-4 Matures commonly overlooked This taxon is closely related to the next from which it may be separated by with Physcomitrium pyrifi ymatous 40 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 2. Aphanorhe gma SuU. 1. Aphanorhe gmu serratum (Hook. /. & Wils. ex Drumm.) SuU. Similar to Physcomitriella patens except that capsule is dehiscent along equa- torial line and exothecial cells are strongly collenchymatous. Uncommon, Salem and Springfield Plateaus; on open to shaded moist soil. Plants with mature capsules may be found from late fall to early winter. 3. Physcomitrium (Brid.) Fiimr. 1. Physcomitrium pyriforme (Brid.) De Not Small, light green, tufted plants; leaves oblong-oblanceolate to oblong-lance- olate, contorted when dry, spreading when moist; margins serrate above; capsules erect, globose-pyriform when fresh, turbinate and constricted below mouth when dry, exserted on setae 5-15 mm long. Very common through the Interior Highlands; on soil in pastures, lawns, and waste places. Physcomitrium collenchymatum Gier has been reported from Clay County, Missouri, just north of the Missouri River (Gier, 19551?: 330). This taxon may be collected in the Interior Highlands and may be distinguished from P. pyri- forme by its smaller size and the extremely flared condition of the capsule when dry and empty (Crum & Anderson, 1964). 4. Pyramidula Brid. 1. Pyramidula tetragona (Brid.) Brid. Small, looselv tufted, short-stemmed the thwesterri Known from Muskogee Co., Oklahoma, from a collection reported by Little (1936: 12). 5. Funaria Hedw. Plants in loose to dense tufts, upper leaves often forming bud-like clusters; leaves entire or serrate; setae long; capsule with conspicuous neck, elongated- pyriform, strongly unsymmetric to arcuate; annulus present or absent; peristome in two rows of 16 teeth each, outer row reddish to brownish red and somewhat twisted to right, inner row yellowish, often rudimentary. a. Annulus present, conspicuous - b a. Annulus absent J b. Segments of inner peristome obtuse, less than V2 length of outer teeth; spores 20-30 M in diameter 2. F. flavicans b. Segments of inner peristome lanceolate, slenderly pointed, at least % length of outer teeth; spores 12-18 fi in diameter c c. Capsules horizontal to pendent, turgid, strongly arcuate la. F, hygrometrica var. hygrometrica 0. Capsule slender, merely inclined, less arcuate lb. F. hygrometrica var. calvescens d. Leaves entire, costa excurrent 3. F. americana d. Leaves serrate, costa ending belovi^ apex in most leaves 4, F. serrata 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 41 la. Funaria hygrometrica Hedw, var. hygrometrica Common throughout Interior Highlands; on bare soil of waste places, lime- stone glades, around areas of recently burned trash. lb. Funaria hygrometrica var. calvescens (Schwaegr.) Mont Rare, eastern Salem Plateau; on moist soil. 2. Funaria flavicans Michx. Common throughout Interior Highlands; on moist, shaded to exposed soil and limestone, piles of ashes, and on rocks along edges of creeks. 3. Funaria americana Lindb. Rare, Springfield Plateau; on bare soil 4. Funaria serrata Brid. Rare, Boston and Ouachita Mtns.; on moist bare soil 15, Tetraphidaceae 1. Tetr aphis pellucida Hedw. Plants small to medium in size, scattered to tufted, bright greeii to brownish green; stems erect, simple or sparingly branched; leaves ovate to ovate-lanceolate, apex acute; margins plane, entire; costa ending in apex or below; median leaf- ceUs isodiametric, rounded-hexagonal, incrassate, capsule erect, symmetrical, ovoid to cylindrical; operculum conic; peristome single, teeth 4; shoots frequently bear- ing terminal gemmiferous cups, formed of 4-5 broadly cordate bracts, containing lenticular gemmae. Common, Boston and Ouachita Mtns., Ozark Hills, Prairie Plains, Salem and Springfield Plateaus; on moist shaded sandstone, rarely on soil at bases of sand- stone boulders and bluffs. When sporophytes are present, a peristome composed of only four teeth is distinctive. Sterile forms may be confused with Mnium stellare; however, that species has serrate leaf margins. 16. Bryaceae Plants in loose to dense tufts, dark green, yellow-green to silvery, often tinged with red; leaves ovate-lanceolate, obovate-spatulate, to narrowly linear-setaceous, frequentlv larger and crowded into a rosette at the stem apex; margins smooth with to percurrent or excurrent; cells smooth, thin- to thick-walled, hexagonal, rhom- the with th teeth, often with a. Plants whitish-green to silvery in color b a. Plants green, yellow-green, to reddish green c 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 b. Plants whitish green, growing in moist habitats; leaves loosely imbricate, lance- olate to linear lanceolate 1. Tolilia b. Plants silvery, growing in drier habitats; leaves closely imbricate, broadly ovate to obovate 3. Bryum c. Erect shoots with terminal rosette of large, obovate-spatulate, serrate leaves; shoots rising from stolons 4. Rhodobryum c. Plants not as above _.„__ d d. Leaves very narrowly linear-setaceous; costa broad and occupying most of the terminal end of leaf 2. Leptobryum d. Leaves wider, ovate, ovate-lanceolate to obovate; costa narrower and not filling tfie upper part of the leaf e. Margins serrate in upper iy4 of leaf, if denticulate, upper leaf cells 8-12 : 1 and leaves clearly imbricate on stem; border of elongated cells not well developed; costa sub- percurrent to excurrent 1, Pohlia e. Margins entire to only denticulate in upper % of leaf; border of elongate cells often well developed; costa sub-percurrent to excurrent 3, Bryum 1. Pohlia Hedw. green lanceolate near leaf-cells rhomboidal-linear or rhombic-hexagonal, smooth. a. Leaves closely imbricate - a. Leaves distant on stem "b b. Elongate, twisted- vermiform gemmae present 1. P, annotina var. loeskei b. Gemmae absent 4. p. wahlenbergii c. Leaves denticulate near apex; upper leaf-cells 8-12 : 1; branches filiform „ 2. P. filiformis c. Leaves serrate near apex; upper leaf-cells shorter, 5-7:1; branches not filiform _. 3. P. nutans 1. Pohlia annotina (Hedw.) Lindb. var. loeskei Crum, Steere & Anderson Rare, Springfield Plateau; on moist shaded sandstone or sandy soil, along creeks and streams. The presence of elongated, twisted-vermiform gemmae distinguishes this taxon from other Pohlias in the Interior Highlands. 2. Pohlia filiformis (Dicks.) Andr. var. filiformis Rare, Springfield Plateau of western Missouri; on sandstone outcrops, near Cedar Springs in Cedar County (Ireland 3462) and at Lichen Glade in St. Clair County (Redfearn 27333). 3. Pohlia nutans (Hedw.) Lindb. var. nutans Very common throughout the Interior Highlands; a weedy species growing on rocky soil, rock ledges and crevices, and occasionally around bases of trees. Frequently mixed with other bryophytes. this s having three cilia have been given varietal status as Pohlia nutans (Jenn.) Tenn. by some authors. However, this oractice is not folk the two 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 43 4, Pohlia wahlenhergii (Web. & Mohr) Aiidr. Common throughout the Interior Highlands; on moist soil and roclcs, edges of creeks and streams. 2. Leptobryum (B.S.G.) Wils. 1. Leptobryum pyrijorme (Hedw.) Wils. Plants in loose to dense, yellow-green tufts; leaves narrowly linear-setaceous; costa broad; capsule cylindrical, inclined to pendent, with pronounced neck, exserted upon a long seta, light to dark brown, lustrous. Uncommon, Prairie Plains, Salem and Springfield Plateaus; on moist sub- strates, especially sandstone or sandy soil, occasionally on decaying logs or stumps near water. Sterile collections may be confused with Ditrichwn or Dicranella. The broad costa and long linear-rectangular to fusiform basal leaf-cells serve to distinguish L. pyriforme from these taxa. 3. Bryum Hedw. Plants loose to densely tufted, silvery, yellow to dark or reddish green in color; leaves ovate, obovate, to lanceolate, often distinctly bordered by elongated cells; margins entire to denticulate near apex, plane or recurved; costa ending below apex to strongly excurrent; upper leaf-cells short-rhomboidal to rhom- boidal-hexagonal; lower cells rectangular; capsule exserted on a short to long seta, cylindrical, clavate, pyriform to oval, neck often well differentiated, inclined to pendulous. a. Plants silvery in color 2. B. argenteum a. Plant not silvery in color b b. Leaves obtuse, rounded-obtuse, to sub-acute; costa not excun-ent c b. Leaves acute to acuminate; costa often excurrent e -.. d 9. B. mniatum — 12. B. tortifolium c. Leaves decurrent c. Leaves not decurrent d. Leaf cells thin-walled d. Leaf cells thick-walled, at least at comers 7. B. gemmiparum e. Plants small with regularly to irregularly sphaerical gemmae attached to rhizoids ._ f e. Gemmae not present on rhizoids g f. Rhizoids usually deep violet in color 11. B, ruderale f. Rhizoids pale yellowish to brownish, but not violet in color 8. B. klinggraefii g. Costa percurrent to slightly excurrent . h g. Costa long-excurrent or upper cells forming a cuspidate apex i h. Leaves to 2.5 mm* or more long, strongly bordered; plants often dark reddish purple in color __ 10. B, pseudotriquetrum h. Leaves 1 mm or less long, not distincdy bordered 3. B. bicolor i. Leaves obovate to obovate-spatulate, usually spirally twisted around stem „ 5. B. capillare u Leaves ovate to long-lanceolate^ not spirally twisted about stem j j* Inner peristome with rudimentary, imperfectly developed cilia k ], Inner peristome wdth well developed cilia k. Inner peristome adherent to outer peristome by transverse ridges 1. B. angustirete developed. forms 44 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 k. Inner peristome not adherent to outer peristome 13. B, uliginosum 1. Inflorescense synoicous; upper leaf-cells 3-5 : 1 6, B. creberrimum 1. Inflorescense dioicous; upper leaf-cells 6-7 : 1 4, B. caespiticium 1. Bryum angusfirete Kindb. ex Mac. Bryum pendulum (Homscli.) Schimp. Rare, Salem Plateau; on soil. Without capsules, separation of this taxon from other taxa of Bryum is not possible. However, when capsules are present, the lack of well developed cilia on the inner peristome and the adherence of the inner peristome to the outer by transverse ridges is distinctive. 2. Bryum argenteum Hedw. Two varieties may be encountered in the Interior Highlands and may be separated as follows: Costa ending below apex 2a. B. argenteum var. argenteum Costa percurrent to excurrent 2b. B. argenteum var. lanatum 2a. Bryum argenteum var, argenteum Common throughout the Interior Highlands; on soil of disturbed habitats, crevices of sidewalks, cedar shinele roofs, rarelv bases of trees or on rocks. 2b. Bryum argenteum var. lanatum (P, Beauv.) Hampe Reported from Greene and St. Louis Counties, Missouri, by Gier (1955a: 32). This variety is not recognized by many authors (Nyholm, 1958: 252). 3. Bryum bicolor Dicks. Uncommon, throughout the Interior Highlands; on soil. This species is quite small, usually less tiian 0,5 mm tall, gregarious, and frequently mixed with other mosses. Axillary bulbiform gemmae, when present, are distinctive. 4. Bryum cae^iticium Hedw, Uncommon, Salem and Springfield Plateaus; on sterile soil, occasionally on rocks. Sterile plants are difficult to identify. Various authors have attempted to distinguish this taxon on the basis of leaf border and shape of leaf cells. How- ever, these characters are notoriously variable in Bryum and may be misleading rather than definitive. 5. Bryum capillare Hedw. Common, throughout the Interior Highlands; on soil, wooded areas, rocks, ledges, pastures, and waste places. Occasionally abundant in crevices of vertical rock exposures. that are spirallv twisted about the stem serve this 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS AK 6. Bryum creherrimum Tayl. Bryum cuspidatum (B.S.G.) Schimp. Common, Salem and Springfield Plateaus; on open to shaded soil and rocks. Sterile collections are difficult to identify with certainty. When fertile, the synoicous inflorescence, lanceolate leaves with short (3-5:1) upper cells, and excurrent costa, and peristome with well developed cilia are distinctive for this taxon. 7. Bryum gemmiparum De Not. Uncommon, Boston Mtns., Salem and Springfield Plateaus; forming dense cushions on calcareous rocks subject to inundation or seepage. The calciphile nature of this species combined with its obtuse to sub-acute leaves serve to separate it from all taxa except Bryum tortifolium. The latter species may be distinguished by the presence of distinctly rounded leaf apices and thin-walled upper leaf-cells. 8. Bryum klinggraefii Schimp. in Klinggr. Apparendy rare, eastern Salem Plateau; on bare or disturbed soil, often asso- 4^ f %f at margins of ponds and reservoirs. patens Reported by Crundwell and Nyholm (1964: 615) from St. Louis County Missouri Physcomitrium rostellatum Kindb.) 9. Bryum miniatum Lesq, Mtns Known 13884). Johnson Shut-In State Park, Reynold County, Missouri (Redi The taxonomic status of this species is open to question. It is very similar CO 10. Bryum pseudotriquetrum (Hedw.) Gaertn., Meyer Bryum bimum Schreb. the subject trees. 11. Bryum ruderale Crundw. & Nyh. Apparently rare; mixed with Physcomitrium rostellatum and Bryum kling- graefii (see discussion under latter species), 12. Bryum tortifolium Funck ex Brid. Rare, Salem and Springfield Plateaus; on rocks in swiftly flowing creeks. This species can be confused with Bryum gemmiparum (see discussion under that species). 46 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 13. Bryum iiUginosiim (Brid.) B.S.G. Bryum cemum (Hedw.) B.S.G. Rare, eastern Salem Plateau; on rock>^ soil of wooded slopes. 4. Rhodobryum (Schimp.) Hampe 1. Rhodobryum roseum (Hedw.) Limpr. Large plants with erect branches arising from a stolon; leaves in a large terminal rosette, loosely imbricate when dry, v^dely spreading when moist, spatulate thick the Common throughout the rarely at the bases of ti*ees. 17. Mniaceae 1. Mnium Hedw. Plants light to dark green, loosely to densely tufted, sterile stems often prostrate; leaves complanate, ovate, rounded-ovate, obovate to ovate-lanceolate, entire leaf c inflorescence terminal, synoicous or dioicous; capsule ovate to cylindrical, hori- zontal to pendulous, exserted upon a long seta; peristome double. a. Leaves distinctly bordered by elongated cells b a. Leaves not distinctly bordered by elongated cells 1. M. stellate b. Margins of leaves serrate; if entire, bases not cuneate b. Margins of leaf entire, bases cuneate 8. M. punctatum c. Marginal teeth double j c. Marginal teeth single g d. Leaves narrowly elliptic-ovate to ovate-lanceolate; costa toothed on back near apex, not reaching apex of leaves 2. M. hornum d. Leaves ovate lanceolate; costa not toothed on back, reaching apex in most leaves _ f f. Leaf-cells averaging 25 ^ in diameter, collenchymatous, tending to be arranged in longitudinal rows; synoicous or paroicous 4. jf. marginatum f. Leaf-cells averaging less than 20 m in diameter, walls uniformly thickened; dioicous 3^ ^ orthorrhynchum g. Leaves serrate nearly to base _ „ _ h g. Leaves serrate only in upper % or % ZZZIZZZ sTm. cuspTd^tum h. Teeth ot leaves 1-3-cells-long; apex obtuse or mucronate; dioicous 7. M. affine h. Teeth of leaves one-cell-long, rarely more; apex acute to acuminate; synoicous __6. M. medium 1. Mnium stellare Hedw. Mtns rocks The serrations along the border of the leaf are often obscure and a faint border of elongated cells may be occasionally noted. Sterile forms of Tetraphis pellucida may be confused with this taxon. However, T. pellucida has rounded leaf cells and entire leaf margins. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 47 2. Mnium hornum Hedw. Missouri, by Andrews (in Grout, 1933). e report from Jackson County, Illinois, by Redfeam (1966 the collection (Redfeam 18789) is Mnium marginatum. 3. Mnium orthorrliynchum Brid. Uncommon, Boston Mtns., Salem Plateau; on shaded moist limestone and sandstone exposures. Plants of this species are quite small and easily overlooked. Consequently, its distribution is probably wider than collection records indicate. This taxon may be confused with the next. When fertile, its dioicous condition serves to distinguish it from Mnium marginatum. Sterile plants are more difficult to sepa- rate. However, the leaf cells of M. marginatum are generally larger (<25 /t in diameter) and coUenchymatous, while the cells of M. orthorrhynchum are smaller (some do average <20 ^ in diameter) and thin-walled with little thickening at the comers. 4. Mnium marginatum (With.) Brid. ex P. Beauv. Mnium serratum Schrad. ex Brid. Common, Boston Mtns., Ozark Hills, Salem and Springfield Plateaus; on shaded, moist rocks, particularly in crevices and undersides of overhanging ledges, occasionally on soil or bases of shrubs or trees. ■ For distinctions from Mnium orthorrhynchum, see notes under that taxon. 5. Mnium cuspidatum Hedw. Very common throughout the Interior Highlands; on moist, shaded soil and rocks. An easily recognizable taxon with rounded, ovate to obovate leaves that are the 6. Mnium medium B.S.G. Reported from Conway County, Arkansas, by Moore (1965: 32). 7. Mnium affine Bland, ex Funck This taxon shows considerable variability and several varieties have been suggested. Of these, only the variety rugicum seems worthy of recognition in the Interior Highlands. a. Leaves decurrent; leaf c a. Leaves not decurrent; le .„ 7a. M. affine var. affine 7b. M. affine var. rtigtcum 7a. Mnium affine van affine Very common throughout the Interior Highlands; on shaded soil and rocks. May be confused only with Mnium medium from which it may be separated by the features noted in the key above. 48 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 7b. Mnium affine var. rugicum (Lour.) B.S.G. Mnium rugicum Laur. Rare, Salem and Springfield Plateaus; on shaded rocks and soil beside creeks and streams. The taxonomic validity of this taxon is questioned by many. However, it is distinct in its size and form and should be recognized, especially since there is not any evidence as to the basis of its remarkable differences. These differences may be due to environmntal or cytological factors, or both. Clearly, further study is needed. 8. Mnium punctatum Hedw. Common tlii'oughout the Interior Highlands; on rocks and decaying wood in moist, shaded areas, particularly in rock crevices and beside streams and gullies. 18. AULACOMNIACEAE 1. Aulacomnium Schwaegr. Plants in dense to loose tufts, pale to blackish green in color, often tomentose at base; leaves ovate to lanceolate, coarsely toothed to denticulate above; costa stout at base, ending below the apex; upper leaf-cells small (7-18 fi), coUenchy- matous with single papillae on both surfaces; lower leaf-cells larger, elongate, to 30 fi long, walls incrassate; fusiform brood-bodies often present on pseudo- podia; capsules on twisted seta, 1-1.5 cm long, suberect, longitudinally striate; operculum beaked; peristome double. a. Leaves ovate, coarsely serrate above 1. A. heterostichum a. Leaves lanceolate, denticulate above 2. A. palustre 1. Aulacomnium heterostichum (Hedw.) B.S.G. Common throughout the Interior Highlands; on moist, shaded soil banks and rock ledges. Often mixed with Mnium affine, M. cuspidatum, or Bartramia pomiformis. 2. Aulacomnium palustre (Hedw.) Schwaegr. Uncommon throughout the Interior Highlands; often very abundant on moist, dripping sandstone ledges and at bases of bluffs or on soil subject to a great deal of seepage. Not infrequently mixed with Sphagnum in bogs. 19. BAR'raAMIACEAE Plants yellowish green to bright green, in loose to dense tufts, densely tomen- tose below; leaves linear-lanceolate to ovate-lanceolate, costate; cells papillose by projection of end-walls; capsule subglobose to ovoid, furrowed when dry, inclined or cernuous, exserted upon a long seta; peristome double. a. Plants in large, soft tufts; leaves 5-6 mm long, linear- to elongate-lanceolate, sub- ulate, flexuous from an ovate, more or less clasping base; margins usually strongly- denticulate to sharply serrate above 1. Bartramia a. Plants in coarse, dense tufts; leaves 1.5-2 mm long, linear-lanceolate to ovate-lance- olate, not flexuous; serrate but not denticulate above 2. Philonotis ^^"^^I REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 49 1. Bartramia Hedw. 1. Bartramia pomiformis Hedw. Soft tufted mosses, bright green to bluish green; leaves 5-6 mm long, elongate- lanceolate, subulate, flexuous from an ovate, more or less clasping base; margins revolute to apex, strongly denticulate to sharply serrate above, bistratose distally; upper leaf-cells thick-walled and short-rectangular, papillose on both surfaces,' basal cells linear to oblong, smooth; capsule exserted on seta 1-2 cm long, brown dry Common throughout the Interior Highlands; on moist, shaded soil banks and rock ledges. Often mixed with Aulacomnium heterostichum. 2. Philonotis Brid. oc cous, branches whorled and subtending perigonia, stems densely tomentose triangular-lanceol margins ^^^ , ^ ^ - — — ^ — current to long-excurrent; upper leaf-cells oblong to linear, papillose at upper Of lower ends or both; basal leaf cells larger, smooth or papillose; autoicous or dioicous, perigonia gemmiform or discoid; seta long; capsule subglobose to ovoid, inclined to cemuous, reddish-brown, longitudinally furrowed; peristome double. a. Costa percurrent to slighdy excurrent; leaves distinctly papillose; apex acuminate to acute L a. Costa ending below apex; leaves weakly papillose; apex blunt or obtuse _„ 1. P. gracillima b. Papillae at lower end of cells b. Papillae at upper end of cells, occasionally at both ends c. Perigonia gemmiform; leaves narrowly lanceolate to triangular-lanceolate; leaf-cells linear to rectangular; basal cells not noticeably shorter and wider _'_„ 2. P. longiseta c. Perigonia discoid; leaves ovate-lanceolate to triangular-lanceolate; leaf-cells linear to oblong; basal cells distinctly larger and broader than median leaf-cells „ d d. Leaves ovate-lanceolate; median leaf-cells oblong to linear, weakly papillose papillae occasionally at both ends of cells, not forming noticeable parallel rows' basal cells tending to be smooth _....._ 4, p. muehlenhergii d. Leaves ovate-lanceolate to triangular-lanceolate; median leaf-cells mostly linear distinctly papillose and arranged in parallel rows; basal cells usually papillose _!. f e. Margins of leaves singly serrate, not recurved 6, P. caespitosa e. Margins of leaves doubly serrate, recurved above 7^ p^ fontana f. Costa percurrent or shortly excurrent; leaves ovate-lanceolate to triangular-lance- olate 3 p ^rchica r. Costa long and slenderly excurrent; leaves often slenderly lanceolate 5. P, capillaris 1* Philonotis gracillima Angstr. Rare, Ouachita Mtns., eastern Salem Plateau: on *;}ialp c^nA cor^^ef^^^ 2. Philonotis longiseta (Michx.) Britt var. longiseta Mtns shaded sandy soil, sandstone, dolomite, or shale, usually near flowing water. 50 THE [Vol. 59 3. Philonotis marchica (Hedw.) Brid. Common throughout the Interior Highlands; on rocks and soil, mostly cal- careous, along banks of gullies and streams. This taxon is difficult to distinguish from the next. Transitional forms may be encountered making assignment to either difficult if not impossible. Further research is needed to resolve the difficulties in this and related taxa. 4. Philonotis muehlenbergii (Schwaegr.) Brid. Uncommon throughout the Interior Highlands; on moist soil and rocks along gullies, creeks, and streams. Easily confused with Philonotis marchica. 5. Philonotis capillaris Lindb. ^ Rare, eastern Salem Plateau, known only from Pickle Springs in Ste. Gene- vieve County, Missouri; on shaded sandstone ledge. 6. Philonotis caespitosa Jur. var. caespitosa Reported from Conway County, Arkansas, by Moore (1965) 7. Philonotis fontana (Hedw.) Brid. var. fontana Common throughout the Interior Highlands; on moist soil and rock ledges along flowing water or standing surface water in glades. 20. Tim MI ACE AE 1. Timmia megapolitana Hedw. Plants in loose to dense tufts, green above, brown to yellow- green below, 3-5 cm tall; leaves crisped with inroUed margins when dry, spreading and margins plane when moist, 4-10 mm long and 1-1.5 mm wide, lanceolate to linear-lance- olate, apex gradually narrowing, base sheathing, concave to keeled; margins strongly toothed from just above the sheathing base to the apex with multicellular teeth; costa strong, papillose above, sub-percurrent to percurrent; leaf-cells rounded-hexagonal and coUenchymatous above, thin-waUed and linear in the sheathing base; autoicous; capsule erect to horizontal; seta red, 2-2.5 cm long; operculum rounded, apiculate; outer peristome teeth 16, yellowish, pellucid and slightly papillose below, longitudinally striate and often perforate above; inner peristome arising from a basal membrane about V2 length of outer teeth and ending in 64 cilia united into groups of 4; calyptra erect from seta behind the capsule. Uncommon, Boston Mtns., Ozark Hills, Salem Plateau; on moist, shaded cal- careous rocks and soil along banks of streams and gullies and on ledges. When young sporophytes are present, the peculiar position of the calyptra is diagnostic for this taxon. Sterile plants may be recognized by the rather large size of the plants and the long leaves that have inrolled margins and are crispate when dry, and are strongly serrate with multicellular teeth and have thin-walled linear cells in the sheathing base. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 51 21 . Erpodiaceae 1, Venturiella sinensis (Vent, ex Rabenh,) C. Miill. Small plants in creeping, flattened mats with terete, erect branches, 2-4 mm long; leaves ovate, imbricate-appressed, widely spreading when moist, dark green with hyaline, denticulate tips; upper leaf-cells hexagonal, smooth, quadrate at margins, elongated and becoming linear in hyaline acumination; seta short; cap- sule pale yellow, oval-cylindric, 1.5 mm long, immersed to emergent. Mtns Known near W 22. Ptychomitriaceae Small to minute plants in yellowish to brownish or blackish green tufts; leaves crispate when dry, erect-spreading when moist, linear-subulate to linear-lance- olate, not plicate; costa percurrent to sub-percurrent; upper cells smooth, rounded to quadrate; capsule cylindrical to ovoid, exserted; peristome teeth 16, variously split; calyptra mitrate to campanulate-mitrate. a. Leaves linear-subulate; seta arcuate when moist 1. C ampylostelium a. Leaves linear from a lanceolate base; seta straight when moist 2. Ptychomitritim 1. C ampylostelium B.S.G. 1. C ampylostelium saxicolum (Web. & Mohr) B.S.G. Plants minute, in loose, yellow-green tufts; leaves subulate from a narrow, ovate base, crisped when dry, keeled; margins entire; median cells green and rectangular; basal cells thin-walled, elongate-rectangular, hyaline; capsule ex- serted on an arcuate seta, cylindrical, sulcate with age; operculum long subulate. Rare, Boston Mtns.; on shaded, moist vertical sandstone. 2. Ptychomitrium Fiirnr. Plants in small, brownish to blackish green tufts; leaves crispate when dry, linear-lanceolate; margins entire to serrate in upper half; costa single, percurrent to sub-percurrent; seta short, 2-3 mm long, not arcuate; capsule ovoid; operculum subulate; calyptra mitrate to campanulate-mitrate. a. Leaves entire or nearly so; plants growing on rocks . b a. Leaves serrate in upper half; plants growing on trees : 3. P. dmmmondii b. Leaves 1.2-2.0 mm long; seta 2.0-2.5 mm long 1, P. incurvum b. Leaves 3.0-4.0 mm long; seta 4^ mm long 2. P. leibergii 1. Ptychomitrium incurvum (Schwaegr.) SuU. Common throughout the Interior Highlands; on exposed to shaded rocks. This species characteristically grows in pockets of boulders. Colonies are seldom large, although they may be quite frequent. 52 ANNALS [Vol. 59 2. Ptychomitrium leihergii Best Rare, southern Salem Plateau; on limestone in cedar glade. A distinctive spe- cies known only from a cedar glade above the west bank of Norfork Lake, ca. V4. mile west of Tecumseh, Sect. 16. T, 22, R. 12 W. this SDecies are remarkablv similar to the sinense J capsule. 3. Ptychomitrium drummondii (Wils.) SulL Uncommon, Ouachita Mtns., Ozark Hills, St. Francois Mtns., Springfield Plateau; on trees. This species may be more common than collection records indicate as sterile plants are likely to be mistaken for sterile species of Orthotrichum. 23. Orthotrichaceae Dark, tufted plants, growing on trees and rocks; stems erect or creeping; leaves crowded, appressed or crispate when moist, keeled above, lanceolate; margins entire; costa strong; upper leaf-cells small, irregularly rounded, papil- lose; capsule immersed to exserted, erect and symmetric, often plicate; peristome double; calyptra mitrate, hairy or smooth. a. Primary stems long and creeping .™__. .„„„_._. 3. Drummondia a. Plants erect, stems not creeping b b. Gemmae present in upper leaf axils b. Gemmae absent 4. Zygodon c c. Capsule long exserted 2. Ulota c. Capsule immersed to emergent .... 1. Orthotrichum 1. Orthotrichum Hedw, Small, dark to brownish green tufted plants; stems erect, branched above; leaves lanceolate to ovate-lanceolate; costa strong; upper leaf-cells small, irregu- larly rounded, papillose on both surfaces; basal cells rectangular, rhomboidal to linear; capsules immersed to emergent, often ribbed and strangulate when dry; stomata scattered and immersed in our species; outer peristome teeth 16, often united in pairs and reflexed when dry; inner peristome when present of 8-16 narrow segments; calyptra usually campanulate, covering most of the urn, plicate, hairy. A difficult genus where identification depends on the presence of old, mature capsules. The following key is based in part on the treatment of Orthotricha straminea by Crum and Anderson (1956) and on suggestions by Dr. Dale Vitt ( personal communication ) . a. Upper leaves with hyaline points a. Upper leaves without hyaline points b. Plants growing on rocks b. Plants growing on trees __ 7, O, diaphanum b c. Peristome papillose 2. O. strangulat c d um 19721 REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 53 c. Peristome striate 1. O. cupulatum e d. Exostome teeth recurved d, Exostome teeth reflexed „ f e. Exothecial cells poorly differentiated 4^ Q, pusiilum e. Exothecial cells clearly differentiated ^ ™Z" 5." o' pumilu7n f. Capsule constricted below mouth; exothecial cells different in color from rest of capsule g Q stelhtum I. Capsule not constricted below mouth 3. Q. ohioense 1. Orthotrichum cupulatum Brid. Rare, on rocks. Reported from McCurtain Co., Oklahoma, by Sharp (1930: 53). 2. Orthotrichum stranguhtum P, Beauv. Common, Boston Mtns., Prairie Plains, Salem and Springfield Plateaus; on open to shaded rocks. 3. Orthotrichum ohioense SuU. & Lesq. ex Aust. Mtns trunks with Plants often mixed with the next taxon. Capsules frequently strangulate 4. Orthotrichum pusiilum Mitt. Common throughout the Interior Highlands; on tree trunks, rarely on sand- stone. 5. Orthotrichum pumilum Sw, Uncommon, Prairie Plains, Salem and Springfield Plateaus; on tree trunks. 6, Orthotrichum stellatum Brid. Mtns tree trunks. 7. Orthotrichum diaphanum Brid. Mtns rocks 2. Ulota Mohr ex Web. 1. Ulota hutchinsiae (Sm, ex Sowerby) Hammar Ulota americana (P. Beauv.) Limpr. Erect, dark green tufted plants; leaves closely imbricated when dry, concave- carinate, lanceolate to ovate-lanceolate, obtuse to obtusely acute; margins re- curved; costa strong; upper leaf-cells thick-walled, circular to elliptic, papillose; capsule exserted, oblong-ovoid with long neck, 8-ribbcd; peristome teeth 16 m pau-s sandstone. Locally Ouachita Mtns.; on vertical shaded dolomite 54 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 3. Drummondia Hook, ex Drumm, 1. Drummondia prorepens (Hedw.) Britt Primary stems creeping with numerous erect, short, dark densely foliate branches; leaves closely appressed when dry, obtuse inrolled; upper leaf-cells small, rounded, thick-walled, smooth; capsule exserted, erect, on seta 2-3 mm long, symmetric, ovoid-globose; peristome teeth 16; calyp- tra conic, without hairs. Common throughout the Interior Highlands; on trees, rarely on shaded acidic rocks. r w 4. Ztjgodon Hook. & Tayl. 1. Ztjgodon apiculatus Redfearn (1967). Plants very small, 1-4 mm high, dark green, erect; leaves when dry erect, when moist widely spreading, crowded, keeled, oblong-lanceolate or oblong- ^ ovate, acute, rather abruptly short apiculate, terminal cells of apiculus little elon- gated, yellowish-brown or concolorous; margins plane or slightly wavy; costa conspicuous, ending below apex and papillose near tip; upper leaf-cells green, hexagonal, thin surfaces; gemmae abundant in 3-^ (^rarelv H) nnisfrratp pr11^tra cucullate. g2 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Very common throughout the Interior Highlands; on open to shaded dry rocky soil to moist, shaded rock ledges, and soil banks. Frequently very abun- dant with mats covering large areas. 6. Rhynchostegiella (B.S.G.) Limpr. 1. Rhynchostegiella compacta (C. Miill.) Loeske subsp. compacta AmblysteghtTTi compactum (C. Miill.) Aust. Plants small and slender in yellowish green, dense tufts; stems irregularly branching, radiculose, rhizoids often minutely scabrous; leaves erect-open, nar- rowly decurrent, ovate-lanceolate, ca. 1 mm long, lanceolate and smaller on branches, gradually and evenly acuminate; margins finely denticulate, teedi at base frequently recurved or double by the protruding angles of two adjacent cells; costa percurrent, frequently with rhizoids or delicate elongated gemmae present on back; leaf-cells linear-rhomboidal, thin- walled, 4-10:1; basal cells shortly rectangular to quadrate; seta 1-3 cm long; capsule oblong, erect or nearly so; symmetric, operculum conic-apiculate; peristome double. Uncommon, Ozark Hills and Salem Plateau; on shaded, moist limestone or sandstone, usually forming extensive loose mats. Robinson (1965: 318) does not agree with the placement of this species in the genus Rhynchostegiella^ but retains it in die genus Amblystegium. 7. Rhynchostegium B.S.G. {sensu Robinson, Bryologist 70:38. 1967) Plants with irregularly divided branches erect to ascending from prostrate stems, dark green to yellowish green, often slightly glossy; branch leaves loosely to closely imbricate, often dimorphic, sometimes slightly plicate, broadly triangu- lar to orbicular, apex acute in stem leaves, obtuse in branch leaves, leaf bases not to slightly decurrent; margins plane, serrate; costate from V2 to more than % the length; median cells 5-15 : 1, linear-oblong; apical cells distinctly shorter; capsules inclined to horizontal, ovoid to oblong-cylindric; operculum long-ros- trate; peristome double. a. Plants aquatic or subaquatic, robust, stems woody or wiry; seta smooth 3. R. riparioides a. Plants not aquatic, not robust, stems not woody or wiry; seta rough or smooth b h. Seta rough; branch leaves broadest Vs above base; stem leaves obtusely acute, not decurrent 2. R. hians h. Seta smooth; branch leaves broadest at base; stem leaves acuminate, slightly decurrent 1. R. pulchelhim 1. Rhynchostegium pulchellum (Hedw.) Robins. Eurhynchium pulchellum (Hedw.) Jenn. Two varieties are present in the Interior Highlands and may be separated as follows: a. Branch leaves loosely imbricate; branch ends attenuate _. la. R. pulchellum var. pulchellum a. Branch leaves closely imbricate; branches julaceous, ends blunt lb. R. pulchellum var. praecox 1972] REDFEARN--MOSSES OF THE INTERIOR HIGHLANDS §3 la. Rhynchostegium pulchellum var, pulchellum Common throughout the Interior Highlands; on shaded soil and rocks. Occa- sionally on tree trunks and bases. lb. Rhynchostegium pulchellum var. praecox (Hedw.) Dix. ^ Very common throughout the Interior Highlands; on open to shaded soil, rocks, and tree bases. Crum, Steere, and Anderson (1965: 421) consider this variety merely a form. 2, Rhynchostegium hians (Hedw.) Delogn. Eurhynchium hians (Hedw.) Sande Lac. Oxyrhynchium hians (Hedw.) Jenn. Common throughout the Interior Highlands; on moist soil, rocks, and logs. 3. Rhynchostegium riparioides (Hedw.) Card, in Tourret Eurhynchium rusciforme Milde. Oxyrhynchium riparioides (Hedw.) Jenn, Common, Boston Mtns., Salem and Springfield Plateaus; on wet or submerged rocks in stream and spring branches. Often very abundant 37. Entodontaceae The genera included here in this family are ihose included by Brotherus in the 2nd edition of Die natiirlichen Pflanzenfamilien (1925), a procedure also adopted by Crum, Steere, and Anderson (1965: 395-396). Plants slender to moderately large, in loose to dense, glossy, yellowish green to dark green mats, sparingly to subpinnately branched, prostrate to suberect; leaves loosely imbricate, julaceous, to strongly complanate-foliate, concave, ob- long-ovate, ovate, to obovate with acuminate, apiculate, to rounded or obtuse tips, gradually narrowed to base, sometimes slightly decurrent; margins plane to nar- rowly recurved near base and strongly infolded near apex, entire to denticulate above; costa absent or short and double; median leaf-cells smooth or papillose by projecting cell ends, linear to narrowly rhomboidal; apical cells shorter; basal cells numerous in several rows, quadrate to subrectangular; alar cells often enlarged and colored; capsule subcylindric, arcuate and cernuous to cylindric and erect; peristome double, teeth 16; propagula present in Pterigynandrum. a. Stems red; leaves strongly infolded near apex 3, Plcurozium a. Stems not red; leaves not strongly infolded near apex b b. Leaf cells smooth , 2. Entodon b. Leaf cells papillose - 3. Pterigynandrum 1. Pterigynandrum Hedw, 1. Pterigynandrum filiforme Hedw. var. filiforme Plants slender, in thin mats, branches irregularly spaced, often ascending and flagelliform, julaceous, with small, filiform to much branched paraphyllia; leaves obovate to ovate-oblong, obtusely pointed to shordy acuminate or acute, secund 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 or closely appressed when dry; margins slightly denticulate or entire above, reflexed below, slightly decurrent; costa absent or short and double; median leaf- cells papillose by projecting angles, short linear-vermicular to narrowly rhombic, 3-5 : 1: cells of aDex broader or shorter: basal cells irresularlv Quadrate to short- 2-8 2-3 Reported from Salem and Springfield Plateaus by Gier (1955a: 37). 2, Entodon C. Miill Plants in glossy yellow-green to dark green mats, stems julaceous to com- planate-foliate, subpinnately branched; leaves oblong-ovate, oblong-elliptical to oblong- or ovate-lanceolate, abruptly acute, short apiculate to narrowly acumi- nate, narrowed to insertion, concave; margins entire to denticulate at apex, often narrowly recurved below; costa absent to short and double; median leaf-cells smooth, linear-fusiform to linear-flexuose, 8-15 : 1, cells of apex shorter; basal cells numerous in several rows, quadrate and enlarged; seta red or yellow; cap- sules erect and cylindric. a. Leaves narrowly acuminate 1, E. brevisetus a. Leaves acute to apiculate b b. Branches julaceous, scarcely complanate-foliate 5. E. seductrix b. Branches complanate-foliate c c. Plants slender, branches 1 mm wide or less 3, E. compressus c. Plants moderately robust, branches more than 1 mm wide d d. Seta yellow d. Seta reddish brown __. 4, E. macropodus 2. E. cladorrhizans h Entodon brevisetus (Hook. & Wils. ex Wils.) Lindb. Reported from vicinity of St. Louis by Wilson (1842) 2. Entodon cladorrhizans (Hedw.) C. Miill. Common throughout the Interior Highlands; on shaded rocks, soil, tree bases, and logs. 3. Entodon compressus C. Miill. Entodon challengeri (Par.) Card. (See Mizushima, 1967.) Uncommon, Prairie Plains, Salem and Springfield Plateaus; on shaded rocks, decaying wood, and tree bases. The small size and complanate leaves slanting towards the substratum from the middle of the stem and thus resembling the roof of a house distinguish this species from all others in the genus. 4. Entodon macropodus (Hedw.) C. Miill. Entodon drummondii (B.S.G.) Jaeg. & Sauerb. Mtns rocks 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS §5 5. Entodon seductrix (Hedw.) C. Miill. var. seductrix Very common throughout the Interior Highlands; on moist shaded rocks, soil, logs, tree bases and trunks. Entodon seductrix var. minor Aust. ex Grout has been reported from the Ouachita Mtns. in Garland County, Arkansas, by Scully (1941), This is appar- ently a poorly segregated variety that is much smaller than var. seductrix. 3. Pleurozium Mitt. 1. Pleurozium schreberi (Brid.) Mitt. var. schreberi C oilier gonella schreberi (B.S.G. ) Grout Plants in large, loose, shining yellow-green mats, stems red, rather closely pinnate to irregularly branched, branches terete, loosely julaceous; leaves elliptic or oblong-ovate; margins entire or crcnulate-serrate above, strongly infolded above, reflexed below, apex rounded-obtuse; costa faint, short and double or lacking; median leaf-cells smooth, linear-flcxuose, 8-15 : 1; apical cells shorter; basal cells short, broader, incrassate and porose, often colored; alar cells enlarged, subrectangular and often colored, forming a triangular patch; capsule subcylin- dric, arcuate, and cemuous. Uncommon throughout the Interior Highlands; on moist, shaded rocks and soil, especially of stream banks. Often forming mats that cover extensive areas. A striking species that may be easily recognized by its large yellow-green loose mats and its red stem. 38. Plagiotheciaceae This treatment of this family essentially follows the recent revision of the genus Plagiothecium by Ireland (19696). Plants small to medium in size, irregularly to sparingly branched, prostrate; w pseudoparaphyllia present or absent; leaves erect-spreading, imbricate, com- planate, though sometimes julaceous, ovate, oblong-ovate, or triangular, straight or secund, acute to acuminate, rarely obtuse, bases not to strongly decurrent; margins plane or narrowly recurved, entire to serrate; costa short, double, or absent; median leaf-cells smooth or dorsally papillose, shortly elliptic, rhomboid- elongate, elongate-linear, to linear-flexuose; alar cells quadrate to rectangular, incrassate or thin walled and enlarged; capsule cylindric, mostly inclined and asymmetric, contracted below mouth when dry; operculum conic to long-rostrate; peristome double, teeth 16. , a. Leaves decurrent 1. Plagiotheci a. Leaves not decurrent tim b b. Pseudoparaphyllia present, leaf-like 3. Taxiphyllum b. Pseudoparaphyllia absent; if present, filamentous 2. Isopterygium gg ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 1. Plagiothecium B.S.G. Plants medium in size, dull to glossy green, in loose mats; pseudoparaphyllia absent; stems complanate-foliate or julaeeous, irregularly branched; leaves imbri- cate, oblong-ovate, ovate, acute or slenderly acuminate; margins plane or re- curved near apex, entire or serrate near apex; costa short and double; median leaf-cells smooth, linear to linear-rhomboidal; alar cells oval, quadrate to rec- tangular, forming a decurrent region auriculate oval to triangular in outline; brood bodies consisting of 2-7 cells often present; capsule erect to inclined, straight or arcuate; operculum rostrate. a. Plants usually julaceous with symmetric, concave leaves, decurrent portion of leaf composed of mostly rectangular cells in 1-5 vertical rows, triangular in outline 2. P. roeseanum a. Plants complanate-foliate with asymmetric, flat leaves, decurrent portion of leaf composed of many inflated sphaerical, oval, or elongate cells in 2—8 vertical rows 1. P. denticulatuin 1. Plagiothecium denticuhtum (Hedw.) B.S.G. Reported from Jackson Co., Illinois, in the Ozark Hills by Hatcher (1952) 2. Plagiothecium roeseanum B.S.G. throughout the Interior Highlands; on thin rocks 2. Isopterygium Mitt. Plants in flat, glossy, dark to light green to yellowish green mats; stems and branches usually complanate-foliate, sparingly and irregularly branched; pseudo- paraphyllia lacking or present; leaves imbricate to distant, erect-spreading to secund or squarrose, symmetric or asymmetric, not decurrent, ovate, ovate-lance- olate to oblong-lanceolate, acute but more often acuminate; margins plane or recurved, serrate to entire below or throughout; costa short and double or some- times lacking; median leaf-cells smooth, linear or linear-flexuose; apical cells often shorter and rhomboidal; alar cells not differentiated or differentiated and forming a small area of quadrate to rectangular cells; propagula usually present; capsule oblong or ovoid, straight or arcuate, inclined to cernuous; operculum conic to short-rostrate. a. Pseudoparaphyllia present, filamentous, of 1-2 rows of cells; alar cells often over 12 fi wide, marginal row usually quadrate or transversely elongate; asexual repro- ductive bodies with papillose cells „ 1. /. tenerum a. Pseudoparaphyllia absent; alar cells usually less than 12 fi wide ( except in I. muel- Jerianum)y marginal row rectangular, seldom quadrate; asexual reproductive bodies with smooth cells b b. Outer layer of stem cells large and thin-walled in cross-section; leaves usually abruptly short-acuminate, entire or minutely serrulate 2. I. muellerianutn b. Outer layer of stem cells small and thick-walled in cross section; leaves not abruptly short-acuminate; serrulate to strongly serrate in upper half c c. Leaves mostly cultriform and undulate; propagula twisted-vermiform with 1-5 teeth at apex 3. I, distichaceum c. Leaves rarely cultriform, mostly symmetric, seldom undulate; propagula resembling parent plant but smaller 4. I. eJegans 1972] REDFEARN — MOSSES OF THE INTERIOR HIGHLANDS §7 1. Isoptery gium tenerum (Sw.) Mitt. Isopterygium micans (Sw.) Kindb. Plagiothecium micans (Sw. ) Par. Plagiotheciutn micans var. fulvum (Hook. & Wils.) Par. Common throughout the Interior Highlands; on decaying logs and stumps, tree bases, sandy soil, less rarely on rocks. 2. Isopterygium muellerkinum (Schimp.) Jaeg. & Sauerb. Plagiothecium muellerianum Schimp. Rare, Boston Mtns., eastern Salem Plateau, Ozark Hills; on moist shaded noncalcareous rock crevices and ledges. 3. Isopterygium distichaceum (Mitt.) Jaeg. & Sauerb. Plagiothecium subfalcatum Aust. Rare, Boston Mtns., eastern Salem Plateau, Ozark Hills; on moist, shaded sandstone or humus, in crevices or on ledges. Closely related to and considered a variety of the next species by some authors. 4. Isopterygium elegans (Brid.) Lindb. Isopterygium borrerianum, (C. Miill. ) Lindb. Plagiothecium elegans (Brid.) Schimp. Common, Boston and Ouachita Mtns., Prairie Plains, Ozark Hills, and eastern Salem Plateau; on shaded, moist noncalcareous rocks, tree bases and humus. 3. Taxiphyllum Fleisch. Plants in glossy, dark to golden green mats; stems branched, often sparingly, complanate-foliate; pseudophyllia always present, large, foliate; leaves appressed- imbricate, close or distant, appearing distichous, symmetric to somewhat asym- metric, smooth or plicate, broadly ovate to ovate- or oblong-lanceolate, acuminate or abruptly narrowed to an acute, filiform, or rarely subobtuse apex; margins plane or narrowly recurved, serrulate to strongly serrate above; costa lacking or short and double; median leaf-cells smooth or papillose dorsally by projecting cell ends, linear-flexuose; apical cells shorter and rhomboidal; alar cells quadrate to short rectangular in 1- several rows with 1-8 in marginal row; capsule oblong or ovoid, cemuous, straight or arcuate; operculum obliquely rostrate. a. Leaves usually close, appressed-imbricate; median leaf-cells smooth .._,. 1. T. deplanatum a. Leaves distant, not appressed-imbricate; median leaf-cells often papillose ^- 2. T. taxirameum 1. Taxiphyllum deplanatum (C. Miill.) Fleisch. Isopterygium deplanatum (C. Miill.) Mitt. Plagiothecium deplanatum (C. Miill.) Spruce Common throughout the Interior Highland except for the Ouachita Mtns,; on shaded calcareous rocks and soil, occasionally on bases of trees and rotten logs. gg ANNALS [Vol. 59 2. Taxiphyllum taxirameum (Mitt.) Fleisch. Isopterygium geophilum ( Aust. ) Jaeg. & Sauerb. Isopterygium taxirameum (Mitt,) Jaeg. & Saeurb. Plagiothecium. geophilum (Aust.) Grout Taxiphyllum geophilum^ ( Aust. ) Fleisch. Very common throughout the Interior Highlands; on shaded rocks or soil. 39. Sematophyllaceae Plants slender to moderately large, creeping; branches prostrate to erect and cui-ved, irregular to subpinriate; paraphyllia absent or present; leaves erect- spreading to almost julaceus or complanate, often secund to falcate-secund, ovate- lanceolate to narrowly oblong-lanceolate, slenderly acuminate to short-acuminate; margins plane or revolute below, entire to serrate; costa short and double or absent; median leaf-cells smooth, linear-flexuose, long-rhomboidal or fusiform; alar cells quadrate to rectangular with 2-several large, pellucid, hyaline or col- ored cells conspicuous at basal angles; capsule erect and symmetric to more or less inclined and asymmetric; operculum rostrate; peristome double, teeth 16. a. Leaves distinctly serrate above - . 2. Brotherdla a. Leaves entire b b. Paraphyllia large and multiform 1. Callicladium b. Paraphyllia absent 3. SematophyUum L Callicladium Crum 1. Callicladium haldamanum (Grev.) Crum (1971) H eterophyllium haldanianum. (Grev.) Kindb. Plants in loose, irregularly branching, dark to brownish green mats; paraphyl- lia large and multiform; leaves loosely imbricate, ovate- to oblong-lanceolate, narrowed to a short acummation, concave; margins plane and entire; costa short and double or absent; median leaf-cells linear rhomboidal, 12-18:1; basal cells rectangular to quadrate, thin-walled, enlarged and inflated near extreme angles, forming distinct auricles; capsule suberect or inclined, somewhat curved. Rare, Ozark Hills, Prairie Plains, and eastern Salem Plateau; on shaded sandstone. 2, Brotherella Loeske ex Fleisch. 1. Brotherella tenuirostris (Bruch & Schimp. ex SulL) Broth. Plants in thin, glossy, green- to yellow-green mats; branches slender, creeping, pinnate or nearly so; leaves slightly curv^ed to secund with apices turned ventrally, loosely imbricate when dry, oblong-lanceolate and slenderly acuminate; margins reflexed below and usually sharply serrate above; costa absent or short and double; median leaf-cells broadly linear, 7-10 : 1; basal cells shorter, quadrate to subquadrate with angular cells inflated, hyaline or slightly colored; capsule erect and symmetric. Uncommon, Boston and Ouachita Mtns., Ozark Hills, Salem Plateau; on shaded moist acidic rocks, decaying logs and tree bases. 1972] REDFEAKN— MOSSES OF THE INTERIOR HIGHLANDS 89 Tliis inflated 3. Sematophylhim Mitt. Slender to moderately large plants; branches spreading to erect and curved upward, irregular or pinnate, green to yellowish green; leaves often secund, closely imbricate when dry, concave, oblong-lanceolate to oblong-ovate, gradually acuminate to short-acuminate or acute; margins reflexed to plane, entire; costa absent or short and double; median leaf-cells smooth, fusiform to linear-fusiform or linear-flexuose, 6-12 : 1; apical cells shorter; alar cells oblong to rectangular, often incrassate and colored with 2-6 inflated and enlarged cells at the basal angles; capsule oblong, erect and symmetric to curved and asymmetric, contracted below mouth when dry; operculum rostrate. a. Plants growing on bark of trees; capsule erect and symmetric 1. S. adnatum a. Plants growing on rocks; capsule inclined and asymmetric b b. Plants moderately large; leaves 1.2-1.8 mm long; median leaf-cells 8-12:1 3. S. marylandicum b. Plants slender; leaves averaging less than 1.2 mm long; median leaf cells 6-8 : 1 2. S. carolinianum 1. Sematophyllum adnatum (Michx.) Britt. n Uncommon, Boston and Ouachita Mtiis., Prairie Plains, Salem and Springfield Plateaus; on decaying logs and tree bases. 2. Sematophyllum carolinianum (C, Miill, ) Britt. var. carolinianum Common throughout the Interior Highlands; on moist shaded acidic rocks, occasionally on calcareous rocks. 3. Sematophyllum marylandicum (C. Miill.) Britt. Rare, Ouachita Mtns., Ozark Hills, eastern Salem Plateau; on moist shaded norvacuhte and sandstone. 40. Hypnaceae Plants usually prostrate, irregularly to regularly pinnate; leaves straight to curved, secund, or falcate-secund, concave, ovate-lanceolate to triangular, erect- spreading, apex acuminate, bases cordate, slightly oriculate or ovate; margins entire to serrate or dentate; costa short and double or absent; median leaf-cells smooth or rarely slightly papillose on back, narrow to elongate-elliptic or fusiform, 4-8 : 1; basal cells short- rectangular to quadrate, incrassate, sometimes slightly in- flated, often very numerous and extending up margin 8-10 rows, hyaline or colored; seta tall; capsule asymmetric and curved or symmetric and straight; oper- culum conic to rostrate; peristome double, teeth 16; gemmae rare except in Plutygyrium. a. Leaves sharply serrate from base to apex 5. Ctenidium a. Leaves entire or only slightly denticulate or serrulate, mostly in upper part b b. Most leaves clearly curved, sometimes falcate or secund 4. Ilypmirn b. Most leaves straight with only terminal branch leaves being curved or secund c 90 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 c. Terminal branch leaves secund; plants usually growing on tree trunks; capsule erect and symmetric 2, Pylaisiella c. Terminal branch leaves not secund; plants growing on rocks or tree trunks and logs; capsule straight or curved d d. Numerous small deciduous branchlets in clusters in upper leaf axils; leaf-margin narrowly recurved; median leaf-cells 8:1; capsule straight and symmetric; usually growing on trees or decaying wood 1. Platygyrinrn d. Small deciduous branchlets absent; leaf -margins plane; median leaf-cells 4-8 : 1; capsule curved and asymmetric; usually growing on rocks 3. Homomallium 1. Platygyrium B.S.G. 1. Platygyrium repens (Brid.) B.S.G. var. repens Plants glossy, usually dark green, in prostrate mats; leaves erect-spreading to loosely imbricate, oblong-ovate to oblong-lanceolate, acute to acuminate; margins entire, narrowly revolute; costa absent or short and double; median leaf-cells ^ smootb, elongate, elliptic to linear-rhomboidal 8:1; apical cells rhomboidal; basal cells quadrate, numerous, extending up margin of leaf for 8-10 rows; seta 1-2 cm long; capsule erect, ovate-cylindric, symmetric, gemmae, shaped like small decid- uous branchlets, abundant in upper leaf axils. Very common throughout the Interior Highlands; on decaying tree trunks, stumps, occasionally on acidic rocks. This species is not likely to be confused with other species, especially when gemmae are present, which is almost always the case. When it grows on rocks it may be confused with Homomallium adnatum, particularly if gemmae are few and not well developed. It differs from H. adnatum species in that its median leaf-cells are 8 : 1 and its capsule is erect or nearly erect and symmetrical. In H. adnatum the median leaf -cells are 4-8 : 1 and the capsule is inclined and asymmetrical. 2. Pylaisiella Kindb. Plants glossy, green to yellow-green, growing creeping over substrate, branch- ing pinnately; branches erect or ascending, strongly recurved at tips because of secund apical leaves; leaves ovate-lanceolate, long- acuminate, entire or nearly SO; margins plane or involute; costa absent or short and double; median leaf-cells linear-fusiform, 7-12 : 1, alar cells numerous, quadrate, extending up margin for up to 20 rows; seta 1-2 cm long; capsule erect and symmetrical, oval to oblong- cylindric; peristome double. a. Segments of inner peristome longer than and only partially adherent to teeth; quadrate alar cells numerous, extending to 15-20 rows along margin 1. P. selwynii a. Segments of inner peristome about same length as and wholly adherent to teeth; quadrate alar cells fewer, extending only to about 10 rows along margin 2. P. intricata 1. Pylaisiella selwynii (Kindb.) Crum, Steere & Anderson Pylaisia selwynii Kindb, Very common throughout the Interior Highlands; on tree trunks, decaying stumps and logs, rarely on sandstone. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS QX 2, Pyhisiella intricata (Hedw.) Grout Pylaisia intricata (Hedw.) Schimp. Reported from Prairie Plains and eastern Salem Plateau by Gier (1955a: 37) 3. Homomallium (Schimp.) Loeske 1. Homomallium adnatum (Hedw.) Broth. Plants in dark to light green, closely adherent patches with irregular, short, erect branches; leaves crowded, loosely appressed when dry, concave, ovate to oblong-ovate, abruptly short-acuminate; margins plane and entire; costa short and double or rarely single, often absent; median leaf-cells, smooth, short, linear- hexagonal to subrhomboidal, 4-8:1; alar cell quadrate, numerous, often extending up Vs length of leaf on margin; seta 0,5-1.0 cm long, purplish below; capsule oblong to oblong-cylindric, inclined and asymmetric; operculum conic to long- conic and apiculate; peristome double, teeth 16. Very common throughout the Interior Highlands; on boulders and small stones in shaded places, occasionally on tree bases. May be confused with Platygyrium repens. See notes under that species. 4. Hypnum Hedw. Plants prostrate, light yellow, green, to brownish green, loosely to closely pinnately branched; paraphyllia often present; leaves concave, ovate-lanceolate, decurrent, acuminate to a long, fine point, almost straight, secund, to falcate- secund, often slightly complanate; margins plane, smooth to denticulate; costa short and double; median leaf-cells linear-flexuose, smooth; alar cells distinct, small and quadrate to inflated and hyaline; capsule asymmetric, inclined to hori- zontal, oblong to cylindric; peristome double, teeth 16. a. Steins surrounded by thin-walled, hyaline cells; alar cells conspicuously inflated 5. H. lindbergii a. Stems surrounded by incrassate cells; alar cells not to only slightly inflated b b. Quadrate to short-rectangular alar cells numerous, 6-10 in the marginal row c b. Quadrate to short-rectangular alar cells few, 4-6 along margin, inflated cells often present, alar group distinct, often colored — d c. Leaf margins entire 2. H. cupressiforme var. resupinatum c. Leaf margins serrate 4. H. pallescens d. Leaves usually abruptly narrowed to a subcordate to cordate base; paraphyllia few; capsule curved, plicate when dry and empty . 3. H. curvifoliwn d. Leaves usually tapering to insertion; paraphyllia numerous, broad, often dliate; capsules suberect, not plicate when dry and empty _ 1, H, imponens 1. Hypnum imponens Hedw. Uncommon, Boston Mtns., Prairie Plains, Salem Plateau; on sandstone of shaded habitats. 2. Hypnum cupressiforme Hedw. var. resupinatum (Wils.) Schimp. Common, Boston and Ouachita Mtns.; on shaded, vertical norva sandstone rocks, sometimes on rock ledges. 92 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 3. Hypnum curvifolium Hedw. Common throughout the Interior Highlands; on shaded, rocky soil, roclc ledges, and logs. Often extremely abundant and forming extensive mats. 4. Hypnum paUescens (Hedw.) P. Beauv. (Includes Hypnum reptile Michx.) Uncommon, Salem Plateau; on shaded, decaying logs and sandstone. Also reported from McCurtain Co., Oklahoma, in the Ouachita Mtns., by Sharp (1930). This species may be confused with Brotherella tenuirostris (see remarks under that species). 5. Hypnum lindbergii Mitt. var. lindhergii Hypnum arcuatum Lindb. Hypnum patientiae Lindb. Very common throughout the Interior Highlands; on shaded to open rocks, soil, and decaying wood beside creeks and streams. Hypnum lindhergii var. demissum (Schimp.) Loeske has been reported from the Salem Plateau by Gier (1955^: 36). This seems to be a poorly distinguished variety separated from var. lindbergii by being in thin, light to yellow-green mats. 5. Ctenidium (Schimp.) Mitt. 1. Ctenidium moUuscum (Hedw.) Mitt. Hypnum molluscum Hedw, Plants regularly pinnate and plumose, golden-green to yellowish, to occasion- ally bright-green; leaves crowded, when dry falcate-secund, plicate, slenderly acuminate from an auriculate, cordate-triangular base; margins plane, serrate or dentate; costa short, double, or absent; median leaf-cells linear, 8-12 : 1, slightly papillose at upper ends of cells on back of leaf; alar cells shorter, quadrate- rectangular, not inflated or colored; seta smooth, 1-2 cm long, capsule ovate, curved, inclined to horizontal. F Common throughout the Interior Highlands; on shaded rocky soil, especially of steep slopes and ridges. Occasionally on tree bases. 41. Rhytidiaceae Large, robust, irregularly to sparcely divided plants in loose, intertangled, yellow- to gray-green mats; stems to 15 cm or more long with pseudoparaphyllia sparingly found in branch axils; leaves crowded on stem, imbricate, often strongly secund, to squarrose-spreading, ovate-lanceolate to triangularly cordate, slenderly acute to acuminate, plicate, often rugose; margins sharply dentate to denticulate and revolute; costa single or double; median leaf-cells linear, smooth or papillose; alar cells rounded-quadrate and numerous to few and hexagonal-oblong; sporo- phytes not known in material from Interior Highlands. a. Leaves secund and strongly rugose; costa single 1. Rhytidium a. Leaves spreading to squarrose, plicate, but not mgose; costa double 2. Rhytidiadclphus 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 93 1, Rhytidium (Sull.) Kindb. 1. Rhytidium rugosum (Hedw.) Lindb. Large, robust plants; leaves crowded-imbricate, ovate-lanceolate, strongly ru cells linear; basal cells rounded quadrate, numerous. Rare, central Salem Plateau; on open to shaded rocky soil near edges of high, north-facing bluffs along Jacks Fork River in Texas and Shannon Counties, Miss 2. Rhytidiadelphus (Lindb.) Wamst. 1. Rhytidiadelphus triquetrus (Hedw.) Warnst. var. triquetrus Large, robust plants; leaves widely spreading to squarrose, cordate-triangular, plicate; margins sharply dentate above, denticulate below; costa double; median leaf-cells linear, spinose-papillose on back by projecting cell angles; basal cells hexagonal-oblong, thick-walled and pitted. Rare, central Salem Plateau and Boston Mtns.; on open to shaded rocky soil along edges of high, north-facing bluffs or rarely on shaded soil, bases of bluffs. 42. DiPHYSCIACEAE 1. Diphyscium Mohr 1. Diphyscium foliosum (Hedw.) Mohr dry late, obtuse; costa broad and flattened, often indistinct, ending below apex; upper leaf-cells obscure, rounded-quadrate, strongly papillose on both surfaces; lower leaf-cells rectangular, hyaline and smooth; perichaetial leaves large, ovate- lanceolate, costa ending in a long setaeous point; capsule sessile, shorter than awns of perichaetial leaves, resembling a grain of wheat, golden brown when mature; peristome double, outer rudimentary or lacking, inner whitish, 16- plicate. Common throughout the Interior Highlands; on shaded, rocky soil, vertical rock exposures, and undersides of overhanging ledges, particularly sandstone. 43. POLYTRICHACEAE Plants erect, small to robust, growing in green to brownish or reddish green, open, short to tall tufts; leaves stiff, coarse, with or without a sheathing base, lamellae usually present on upper surface of costa; capsule erect to inclined, terete or angled, exserted; peristome of 32 or 64 short teeth attached by their tips to a thin membrane. a. Leaves with 0-8 rows of lamellae, bordered by a row of elongated cells, without a sheathing base; capsule terete, peristome of 32 teeth; calyptra smooth 1, Atrichum a. Leaves with 10-70 rows of lamellae, not bordered by a row of elongated cells. sheathing base present; capsule terete or angled, peristome of 32 or 64 teeth; calyptra hairy b b. Plants small, 2-8 mm tall, forming open, short tufts; leaves 3-4 mm long; capsule terete; peristome of 32 teeth . _« 2. Pogonatum 94 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 b. Plants large, 2-20 cm tall, usually forming dense tall tufts; leaves 2-12 mm long; capsule angled; peristome of 64 teeth 3. Polytrichum 1. Atrichum P. Beauv. Plants erect, small to robust, forming green to brownish green tufts; leaves ovate to lingulate or lanceolate, lamellae few ( 0-8 ) ; margins distinctly bordered by elongated cells; capsules erect to inclined, elongated, terete; peristome of 32 teeth; calyptra smooth. a. Leaves elliptical to obovate, lamellae 0-3 cells high 1. A. crispum a. Leaves lingulate to lanceolate, lamellae 2-14 cells high b b. Leaves 1-1.3 mm wide, lingulate, lamellae obscuring % or less of the leaf viddth at the midpoint of leaf length — 2. A. undulatum b. Leaves 0.5-1.0 mm viade, linear-lanceolate, lamellae obscuring %-% of the leaf width at the midpoint of leaf length 2. A. angustatum 1. Atrichum crispum (James) SuU. & Lesq. var. crispum Rare, Boston Mtns., Salem Plateau; on soil banks of ditches and gullies. 2. Atrichum undulatum ( Hedw. ) P. Beauv. This species is generally very robust with leaves that are strongly crispate when dry and markedly undulate when moist. Three varieties are represented in the Interior Highlands and may be separated as follows: a. Lamellae obscuring ^— % of the width of the leaf at midpoint of its length - 2b. A. undulatum var. minus a. Lamellae obscuring Yu-^ of the width of the leaf at midpoint of its length b b. Lamellae 6, 2-6 cells high; plants autoicous 2c. A. undulatum var. alte-crisfatum b. Lamellae 2-5, rarely 6, 6-8 cells high; plants monoicous 2a. A. undulatum var. undulatum 2a. Atrichum undulatum var, undulatum Common and locally abundant throughout the Interior Highlands; on shaded sandy or calcareous soils along gullies, ditches, and, especially, shaded ravines. 2b. Atrichum undulatum var. minus ( Hedw. ) Par, Uncommon, Boston and Ouachita Mtns., Prairie Plains, Salem and Springfield Plateaus; on shaded rocky soil in ravines. 2c. Atrichum undulatum var. alte-cristatum Ren. & Card. Reported from the Salem Plateau by Gier (1955a: 32). Frye this variety a poor one, not always distinct from A. undulatum although Ireland n969a: 363) considers it a distinct snenips 3. Atrichum angustatum (Brid.) B.S.G, var. angustatum Atrichum xanthopelma (C. Miill.) Jaeg. Atrichum macmillanii (Holz.) Frye Atrichum papillosum (Jenn.) Frye Very common throughout the Interior Highlands; on poor soil of ditches, gullies, lawns, and on bare soil and soil mounds in forests. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 95 Forms with papillose leaves are occasionally encountered. Such forms have been considered as a separate species, Atrichum xantliopelma, by Frye (1949). Thomas (1953) considers such papillose forms to be A. angustatum, a judgement followed in this treatment. 2. Pogonatum P. Beauv. Plants small, erect, scattered or forming open tufts; green to dark green; leaves lanceolate to lanceolate-subulate from a sheathing base, 3-4 mm long, lamellae 10-35; margins of leaves not bordered by elongated cells; capsules erect, terete; peristome of 32 teeth; calyptra hairy; protonema persistent. a. Leaves with 10-15 lamellae, serrate in upper half; plants usually scattered _ 1. P. pensihanicum a. Leaves with 20-70 lamellae, entire; plants usually forming dense tufts __ 2. P. brachyphyllum 1. Pogonatum pensilvanicum (Hedw.) Par. var, pensilvanicum Pogonatum brevicaule P. Beauv. Common throughout die Interior Highlands; on shaded, sandy soil banks. Plants usually scattered on a mat of dense, green protonema. Male plants are usually smaller than the female plants. 2, Pogonatum brachyphyllum (Michx.) P. Beauv. Common throughout the Interior Highlands; on shaded, sandy soil, especially along road cuts and ditches. Plants usually form dense mats. This species is easily recognized by its short, thick, entire leaves and is usually restricted to soil of weathered dolomite and sandstone. 3. Polytrichum Hedw. Plants usually 2-20 cm tall, forming open to dense tufts; leaves 2-15 mm long, rigid, lanceolate, linear-lanceolate, to linear-subulate; margins entire or toothed, plane or involute; lamellae numerous along upper surface; capsule angled, usually inclined; peristome of 64 teeth; calyptra densely hairy. a. Leaves entire b. Temiinal cells of lamellae flat, not appearing notched in cross section — 1. P. ohioense b. Terminal cells of lamellae notched in cross section 2. P. commune c. Awn of leaf reddish . 3. P. juniperinum c. Awn of leaf white or hyaline 4. P. piliferum 1. Polytrichum ohioense Ren. & Card. Very common throughout the Interior Highlands; on open to shaded rocky soil of ridges, ledges, rock crevices, and around the edges of glades. In this last area this species often forms dense carpets along the edges of flat rock exposures of both sandstone and limestone where it is interdispcrsed with equally well- developed colonies of Leucohryum glaucum, Philonotis fontana^ and Climacium americanum. 95 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 2. Polytrichum commune Hedw. a. Perichaetial lea\es distinctly longer than the upper stem leaves, forming an erect tuft about the seta 2b. P. comntune var. perigoniale a. Perichaetial leaves similar to the upper stem leaves, not forming an erect tuft about the seta 2a. P. commune var. commune 2a. Pohjtrichuvi commune var. commune Common throughout the Interior Highlands; on open soil around the bases of boulders, on rock ledges, or on soil pockets of large flat rock exposures, partic- ularly felsite and granite. Tufts are usually very tall, 10-20 cm. Small plants may be confused with P. ohioense^ but may be separated by the features noted in the key to species. E 2b, Polytrichum commune var. perigoniale (Michx.) B.S.G. r Apparently uncommon, Arkansas Valley, Boston Mtns., Salem and Springfield Plateaus; on soil of rocky woodlands, creek banks, and moist rock ledges. 3. Polytrichum juniperinum Hedw. var. juniperinum Very common throughout the Interior Highlands; on open, dry soil of oak- hickory and oak-pine forests along ridges, and on thin soil over flat rock exposures of sandstone, granite, and felsite. 4. Polytrichum piUferum Hedw. Rare, Prairie Plains and Springfield Plateau; on thin soil over sandstone, espe- cially along open bluffs. Often mixed with Polytrichum juniperinum. Appendix I Glossary AcROCABPOus: archegonia located at tip of branch, plants usually, but not always, of erect habit. Acumen: the slenderly tapering point of leaf. Acuminate: slender point of leaf that is gradually tapering. Alar cells: cells at the basal angles of leaves. Annulus: ring of specialized cells between rim of capsule and the operculum. AntheridiuM: multicellular male gametangium, Apiculate; leaf ending in a short, abrupt, sharp point. Apophysis: swelling of seta just beneath capsule. Appressed: leaves lying flat on an erect stem. Abchegonium: multicellular female gametangium. Arcuate: bent like a bow, Areolation: network formed by cell walls of leaf as seen in surface view. Aristate: with a bristle-shaped or awn-like point. Astomous: capsule not opening by a regularly dehiscent lid or operculimi. Attenuate: tapered or long drawn out leaves. Auriculate: ear-like lobes more or less developed at basal angles of leaves. Bicostate: leaves having two mid- ribs. Brood-bodies: vegetative reproductive cells; also called gemmae. Calcicolous: growing on calcareous substrates. Calyptra: cap or hood covering capsule, derived from upper portion of archegonium. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 97 Campantjlate: calyptras shaped like a bell. Canaliculate: leaves channelled in upper portion. Capsule: enlarged spore producing portion of sporophyte. Carinate: leaves folded sharply along midrib and thus forming a keel. Caulescent: possessing a stem. Cernuous: capsule drooping, nodding, or slightly inclined. Caespitose: growing in tufts or mats. Chlorophyllose : containing chloroplasts. Cilia; delicate hair-Kke structures between segments of inner peristome. CmciNATE: leaves coiled or bent into partial or complete circle, Clavate: club-shaped. Cleistocarpous: capsule opening irregularly. Collenchymatous: cell walls thickened at angles or corners. Complanate: leaves flattened or compressed along stem. Conduplicate : leaves folded lengthwise along costa. Convolute: leaves with rolled margins. Cordate: broadest portion of leaf near base. Cremate: teeth of leaf rounded or convex. Crenulate: leaf margins with small, rounded projections. Crispate: leaves curled or twisted, Cucullate: calyptra hood-like, usually split along one side. Cuneate: wedge shaped. CusproATE: tip of leaf with sharp or rigid point, Decurrent: lower margin of leaf extending down the stem below place of attachment. Dehiscent: splitting open. Deltoid: triangular. Dendroid: tree-like in growth form. Dentate: teeth of leaf extending outward. Denticulate: teeth of leaf minute. Dioicous: antheridia and archegonia on separate plants. Distichous: leaves arranged on stem in two vertical ranks. Dorsal: surface of leaf away from stem axis. EcosTATE; leaves without costa. Elliptic: oblong with rounded ends. Elongate: considerably longer than wide. Emarginate: leaf notched at apex. Emergent: capsule extending slightly beyond perichaetial leaves. Endostome: inner peristome. Entire: margin of leaf smooth. Erect- ascending: leaves almost erect, extending upwards or forward. Excurrent: costa extending beyond tip of leaf. ExostOMe: outer row of peristome. Exothecial: outer layer of cells of the capsule wall. Exserted: capsule extending beyond the perichaetial leaves. Falcate: leaves curved like a sickle. Falcate-secund: leaf curved like a sickle and all leaves extending in the same direction or turning to one side of the stem. Fasciculate: branches clustered, arranged more or less in bundles. Fibrils: spiral thickenings of the walls of hyaline cells in Sphagnum, Filament; thread-like structure. Fimbriate: fringed with ciha. Flaccid: soft in texture. Flagelliform : slender, string-like branchlets. Flexuous: irregularly wavy. Foliose: leaf -like. Fruit: referring to sporophyte. Fulvous: yellow, tawny, or dull brownish yellow. Fuscous: dull brown. Fusiform: spindle-shaped; narrowly oval with narrow, tapering ends. Gemma: cell or cluster of cells of gametophyte capable of reproducing the plant vegetatively. 98 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Gemmiform: bud-like. Geniculate: bent like a knee. Gibbous: capsule swollen more on one side than another. Glabrous: smooth, neither hairy or papillose. Glaucous: bluish-gray or bluish-white; whitened with a bloom. Globose: rounded like a ball, sphaerical. Gregarious: plants growing near each other or clustered. Guard cells: cells surrounding the stoma. Guide cells: large parenchymatous cells seen in cross-section of costa. Gymnostomous: failing to produce peristome. Habit: general appearance or aspect of a plant. Hatr-point: leaf tip which differs from rest of leaf in appearance. Heteroicous: more than one form of inflorescence in the same species. Homomallous: leaves or branches all pointing in same direction. Imbricate: leaves closely overlapping. Immersed: capsule hidden with perichaetial leaves, Incrassate: thickened cell-walls. Inflated: cells of alar region enlarged much beyond size of neighboring cells. Inflorescence: cluster of antheridia and/or archegonia. Insertion: attachment of leaves to stem. Involute: margins of leaves inrolled. Isodlvmetric; cells having equal diameter in all directions. Julaceous: smooth, slender, and cylindrical; leaves uniformly erect or appressed. Keel: sharp ridge formed by longitudinal folding of leaves. Laciniate: jagged or torn. Lamellae: longitudinal strips of tissue in the form of thin blades standing on edge and extending parallel with one another along the costa. Lamina: blade of leaf. Lanceolate: shaped like a point, broadest at base and tapering to a point. Leucocysts: cells without chlorophyll. Ligulate: strap-shaped. Linear: very narrow with nearly parallel margins. Linear-rhomboidal: narrow, somewhat elongate rhomboid. Lingulate: tongue-shaped, LUNiEN: cavity or space within a cell, Mammillate: operculum with a short point; or large, rounded papilla covering a cell. Median: pertaining to cells of leaf in the middle between the costa and margins, Mitrate; calyptra shaped like a peaked cap with margins undivided or wifli several equal clefts . MoNoicous: antheridia and archegonia on same plant. Mucro: very short, small, abrupt point, usually stout. Muticous: not pointed; awnless. Oblong: elliptic, obtuse at each end, with the length 3-4 times the width. Obovate: inversely ovate, narrowed towards the base and broadest at the distal end. Obtuse: blunt or rounded at the apex. Ochraceous : yellowish-brown. Operculum: lid covering the mouth of the capsule. Orbicular: almost circular. Oval: broadly elliptical. Ovate: egg shaped with broader end downward. Panduriform: shape of a fiddle. Papillae: minute, nipple-shaped, rounded, or acute projections. Paraphyllia: thread-like structures or minute leaf-like structures growing among the leaves. Paraphyses: jointed, hyaline hairs or minute threads among the antheridia and archegonia. Parenchymatous: cells joined witii broad ends, walls thin and intercellular spaces often conspicuous. Paroicous: antheridia and archegonia on same plant but not mixed. Patulous: leaves spreading from the stem at an angle of 46-90*. Pellucid: completely or partially transparent. Pendent: hanging down from its support. Percurrent: costa extending through the entire leaf, but not beyond. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 99 Perichaetial: leaves that enclose archegonia. Perigonal: leaves that enclose antheridia. Peristome; fringe of teeth and segments surrounding the mouth of the capsule. Peristome perfect; complete with teeth and segments. Peristome single: consisting of teeth only, PiLiFEROUS: with hair-like projections at tips of leaves. Pinnate: branches spreading in a feather-Hke fashion. Pleurocarpous: archegonia located on a short, lateral branch, plants usually of a creeping habit. Plicate; folded lengthwise into pleats and furrows. Plumose: regularly and closely pinnate. Porose: cell walls perforated by pores. Procumbent; trailing along the ground. Propagula: small structures capable of reproducing the plants vegetatively. Prosenchymatous: narrow cells with ends dove-tailed into each other. PsEUDOPODiUM: leafless, branch-like extension of gametophyte resembling a seta. Pulvinate: shaped like a cushion. Pyriform: pear-shaped. Quadrate: square, or nearly so, in surface view. Radiculose: covered with rhizoids, Reflexed: bent backwards. Reniform: kidney shaped. Reticulate: forming a net. Retort cells: enlarged, cuticular cells of branches of Sphagnum, Retuse: shallow notch in a rounded apex. Revolute; margins rolled backward. Rhizoids: simple or branched, reddish-brown, multicellular outgrowths of stems and some- times leaves. Rosette: terminal leaves so arranged as to resemble a rose. Rostellate: operculum having a short beak. Rostrate: operculum having a long beak. Rupestral: growing on rocks. Secund: leaves all turned to one side. Segments: projections of the inner row of peristome. Serrate: toothed. Serrulate: minutely serrate. Sessile: not stalked. Seta; stalk of sporophyte. Setaceous: tip of leaf bristle-like. Sinuate: margin deep-wavy, Sinuose: cells wavy. Spatxjlate: narrowly obovate and attenuate downwards. Spinose: with small spines. Squarrose: extending from stem at an angle greater than 45°. STEmo cells: small, thick-walled cells seen in cross-section of costa. Strumose: goiter-like on one side. Subulate: leaves with a very fine point, needle-like or awl-like. Sulcate: longitudinally grooved capsule. Synoicous: archegonia and antheridia mingled in same inflorescence. Teeth: projections of outer peristome. Terete: cylindrical, round in cross-section, Tomentose: covered with matted hairs. TmcoNEs: the thickening in the angles of the cells of leaves. Triquetrous: sharply three-angled. Tristichous: arranged in three rows or ranks. Trxjncate: ending abruptly. Tubulose: leaves with strongly inroUed margins that meet. Tumid: swollen, inflated, or distended. TtTRBiNATE: top-shapcd. Turgid: swollen or inflated. Uncinate: hook-shaped. 100 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Undulate: surface or margin wavy. Unipapillate : one papilla per cell. Vaginant: sheathing. Ventral: surface of leaf facing the stem when parallel with the axis Vermiform : worm -shaped. Whorled: branches arranged in a ring or circle about stem. Xeric: dry habitat. Appendix II Index to Family Numbers for Families and Genera The number after each entry is the number of the family or the family in which the le and small caps, accepted genera in genus may be found. Families italic, and genera cited only as synonyms in roman. Acaulon 11 AMBLYSTEGL4.CEAE 35 Amblystegiella 35 Amblystegium 35 Anacamptodon 32 Anoectangium 11 Anomodon 34 Aphanorhegma 14 Archxdiaceae 3 Archidium 3 Atrichum 43 AuLACOMmACEAE 18 Aulacomnium 18 Barbula 11 Bartramia 19 Bartramiaceae 19 Brachelyma 24 Brachytheciaceae 36 Brachythecium 36 Brothera 7 Brotherella 39 Bruchia 7 Bryaceae 16 Bryhnia 36 Bryoandersonia 36 Bryoerythrophyllum 11 Bryoxiphw.ceae 5 Bryoxiphixim 5 Bryum 16 Calymperaceae 9 Callicladinm 39 C oilier gonella 37 Campylium 35 Campylopus 7 Campylostelium 22 Ceratodon 4 Chamberlania 36 Cirriphyllum 36 Clasmatodon 32 Climaciaceae 25 Climacium 25 C ratoneuTon 35 Cryphaea 27 Cryphaeaceae 27 Ctenidium 40 Desmatodon 11 Vichelyma 24 Dicranaceae 7 Dicranella 7 Dicranum 7 Didymodon 11 Diphyscl\ceae 42 Diphyscium 42 Ditrichaceae 4 Ditrichum 4 Drepanocladus 35 Drunimondia 23 Encalypta 10 Encalyptaceae 10 Entodon 37 Entodontaceae 37 Ephemeraceae 13 Ephemerum 13 ERPODL4.CEAE 21 Eucladium 11 Eurhynchium 36 Fabronia 32 Fabroniaceae 32 Fissidens 2 Fissidentaceae 2 Fontinalaceae 24 Fontinalis 24 Forsstroemia 27 Funaria 14 Funariaceae 14 Grimmia 12 Grimmiaceae 12 Gymnostomum 11 Haplocladiuni 24 Haplohymenium 24 Hedwigia 26 Hedwigiaceae 26 Helodium 34 Herpetineuron 24 um Homalia 29 Homaliadelphtis 29 H omalotheciella 36 Homalothecium 36 H omomallium 40 Hookeria 30 HOOKERIACEAE 30 Husnotiella 11 Hygroamblystegium 35 Hyophila 11 Hypnaceae 40 Hypnum 40 Isopterygium 38 Leptobryum 16 Leptodictyum 35 Leptodon 27 Leskea 33 Leskeaceae 33 Leucobryaceae 8 Leucobryum 8 Leucodon 28 Leucodontaceae 28 Leucodontella 28 Lindbergia 33 Merceya 11 Mxiaceae 17 Mniutn 17 Molendoa 11 Myurella 31 Neckera 29 Neckeraceae 29 Orthotrichaceae 23 Orthotrichmn 23 Oxyrhynchium 36 Palamocladium 36 Phascum 11 Philonotis 19 Physcomitriella 14 Physcomitrium 14 Plagiotheciaceae 38 Plagiothecium 38 Platygyrium 40 Platydictya 35 Platylomella 35 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS 101 Pleuridium 4 PleuTochaete 11 Pleurozium 37 Pogonatum 43 Pohlia 16 POLYTRICHACEAE 43 Polytrichum 43 POTTIACEAE 11 Pterigynandrum 37 Ptychomitriaceae 22 Ptychomitrium 22 Pylaisia 40 Pylaisiella 40 Pyramidula 14 Rhabdoweisia 7 Rhacomitrium 12 Rhodohryum 16 Rhynchostegiella 36 BJiynchosteghim 36 Rhytidiaceae 41 Rhytidiadelphus 41 Rhytidiuni 41 Schistidium 12 Schtoetschkeopsis 32 Sciaromium 35 Scopelophila 11 Seligeria 6 Seligeriaceae 6 Sematophyllaceae 39 Sematophijllum 39 Sphagnaceae 1 Sphagnum 1 Syrrhopodon 9 Taxiphyllutn 38 Tetraphidaceae 15 Tetraphis 15 Thamnobryum 29 T/ie/m 31 Theliaceae 31 Thuidiaceae 34 Thuidium 34 TiAtMIACEAE 20 Timmia 20 Tortella 11 rortuia 11 Trematodon 7 Trichostomtim H (7/o^a 23 Venturiella 21 Wem/a 11 Zygodon 23 LlTERATtTRE CiTED Anderson, L. E. 1958. The mosses of North Carolina. VIII. Grimniiaceae to Orthotricha- ceae. Bryologist 61: 285-313. & Virginia S. Bryan. 1956. A cytotaxonomic investigation of Fisside7}s cristatus Wils. and F. adianthoides Hedw. in North America. Rev. Bryol. Lichenol. 25: 254-267. Andrews, A. L. 1913. Sphagnaceae. N. Amer. Fl. 15: 1-31. . 1957. Taxonomio notes. XIII. The genus Campylium, Bryologist 60: 127-135, Braun, E. Lucy. 1950. Deciduous Forests of Eastern North America. Philadelphia. Breen, Ruth S. 1963. Mosses of Florida. Gainesville. Brotherus, V. F. 1924-25. Musci (Laubmoose). In A. Engler & K. Prantl, "Die natiir- lichen Pflanzenfamilien." Ed. 2, Vol. 10-11. Leipzig. Bryan, Virginia S. 1956. Chromosomes and systematic position of the inoperculate mosses, Pleuridium and Bruchia. Amer. Jour. Bot. 43: 460-468. . 1957. Cytotaxonomic studies in the Ephemeraceae and Funariaceae. Bryologist 60: 103-126. & L. E. Anderson. 1957, The Ephemeraceae in North America. Bryologist 60: 67-102. Chamberlain, E. B, 1919. Anacamptodon splachnoides var. tayloriae in Missouri. Bryol- ogist 22: 16. Conard, H. S. 1959. Amblystegiiim. Bryologist 62: 96-104. Crum, H. a. 1964. Mosses of the Douglas Lake region of Michigan, Mich. Bot. 3: 3-63. 1966. A taxonomic account of the genus Thelia, Natl. Museum of Canada, Bull. 216. Contr. Bot. IV: 123-127. — . 1969. Nomenclatural notes on North American mosses. Bryologist 72: 240-246. — . 1971. Nomenclatural changes in the Musci. Bryologist 74: 165-174. — & L. E. Anderson. 1956. Taxonomic studies on North American mosses. L Gymnostomum in North America. Jour. Elisha Mitchell Sci. Soc. 72: 276-280, — , 1956. Taxonomic studies on North American mosses. IV- The "Orthotricha strami- nea* of eastern North America. Jour. Elisha Mitchell Sci. Soc. 72: 284-287. — , 1958. Taxonomic studies on North American mosses. VII. Leaf variation in Tricho- stomum cylindricum (Bruch.) C. M. Jour. Elisha Mitchell Sci. Soc. 74: 32-34. — -. 1964. Notes on Physcomitrium collenchyynattim. Bryologist 67: 350-355. — , W. C. Steere & L. E. Anderson. 1964. New names for some North American mosses. Bryologist 67: 162-164. , & . 1965. A list of the mosses of North America. Bryologist 68: 377^32. Crundvv^ll, a. C. & Elsa Nyholm. 1964. The European species of Bryum erythrocarpum complex. Trans. Brit. Bryol, Soc. 4: 597-637. Frye, T. C. 1949. Atrichum xanthopelma. Bryologist 52; 191-196. Florschutz, p. a. 1964. Flora of Suriname. Musci. Vol. VI, Part 1. Leiden. 102 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 GiER, L. J. 1955fl. Missouri bryophytes. Trans. Kansas Acad, Sci. 58: 24-49. . 1955fc. Physcomitrium collenchymatum. Trans. Kansas Acad. Sci. 58: 330-333. & Ellen Kennedy. 1955. Thelia lescurii — an environmental form. Trans. Kansas Acad. Sci. 58: 334-336. Gleason, H. a. & A. Cronquist. 1964. The Natural Geography of Plants. New York. Grout, A. J. 1928-1940. Moss Flora of North America North of Mexico. 3 vol. Hague, Stella M. 1934. Mosses from the Illinois Ozarks. Trans. Illinois State Acad. Sci. 27: 62-63. Hatcher, R. E. 1952. Some bryophytes of soufliem Illinois. Bryologist 55: 223-257. Ireland, R. R. 1969a. Taxonomic studies on the genus Atrichum in North America. Canad. Jour. Bot. 47: 353-368. 1969^. A taxonomic revision of the genus Plagiothecium for North America, north of Mexico. Publ. Bot. 1, Nat. Museums of Canada. Ottawa. IwATSUKi, Z. & A. J. Sharp. 1958. Molendoa sendtneriana in the United States. Bryologist 61: 356-359. Little, E. L. 1936. The bryophytes of Muskogee Co., Oklahoma. Bryologist 39: 8-16. LoA\^, R. L. 1919. Collecting in Arkansas. Bryologist 22: 14-15. Lox^E, A. & Doris Love. 1953. Studies on Bryoxiphimn. Bryologist 56: 73-94; 183-203. MizusHiMA, U. 1967. Entodon microthecius Broth, and E. compressus G. Miill. — critical notes on mosses (5). Jour. Jap. Bot. 42: 272-276. Moore, Jewel E. 1965. A study of the vegetation of Petit Jean Mountain in central Arkansas. Castanea 30: 1-37. NoGUCHi, A. 1956. On some mosses of Merceya^ with special reference to the variation and ecology. Kumamoto Jour. Sci., Ser. B. 2: 239-257. Nyholm, Elsa. 1954-69. Illustrated Moss Flora of Fennoscandia, II. Musci. Stockholm. Pursell, R. 1957. A taxonomic revision of North American Fissidens^ section Bryoidium, Doctoral Dissertation, Florida State University. Redfearn, p. L., Jr. 1961. Bryophytes of southwest Missouri. IV, Additions to the flora. Bryologist 64: 266-267. . 1966. Bryophytes of the Interior Highlands of North America. XI. Additions to the flora. Bryologist 69: 504-508. — . 1967. Bryophytes from the Interior Highlands. XII. A new species of Zygodon, Bryologist 70: 333-335. Reese, W. D. & Betty E. Lemmon. 1965. A natural hybrid between Weissia and Astomuin and notes on the nomenclature of the North American species of Astomum, Bryologist 68: 277-283. Robinson, H. 1962. Generic revision of North American Brachytheciaceae. Bryologist 65: 73-146. . 1965. A small collection of bryophytes from Kashmir. Bryologist 68: 313-320. . 1967. PreUminary studies on the bryophytes of Colombia. Bryologist 70: 1-61. Sayre, Geneva, 1952. Key to the species of Grimmia in North America. Bryologist 55: 251-259. Scully, F, J. 1941. The mosses of Hot Springs National Park and vicinity. Bryologist 44: 125-128, Sharp, A. J. 1930. Bryophytes of southeastern Oklahoma. I. A preliminary list with notes. Bryologist 33: 45-55. Thomas, T. 1953. Atrichum xanthopelma. Bios 24: 147-149. VoiGT, J, W. & R, H. MoHLENBROCK. 1964. Plant Communities of Southern Illinois. Carbondale. Waller, W. T., & J. C. Bass. 1967. Sphagnum cuspidatum new to Kansas and Missouri. Bryologist 70: 440. Welch, Winona H. 1957. Mosses of Indiana. IndianapoUs. . 1960. A monograph of the Fontinalaceae. The Hague. WijK, R. VAN der & W. D. Margardant. 1960. New combinations in mosses. V. Taxon 9: 189-191. . & P. A. Florschutz. 1959-69. Index Muscorum. Vol. 1-5. Utrecht. Wn.soN, W. 1842. Description of four new species of mosses, discovered in Louisiana by the late Mr. Thomas Drummond. Jour. Bot. (Hooker) 4: 419-422. Wittlake, E. B. 1950. The bryophytes of Spy Rock hollow. Proc. Arkansas Acad. Sci. 3: 39-40. 1972] REDFEARN— MOSSES OF THE INTERIOR HIGHLANDS ]^Q3 Note Added in Proof Monte G. Manuel has discovered tliat a collection determined as Cryphaea glomerata from Pike County, Arkansas, {Redfearn 20214) is Cryphaea ravenelii Aust. Tliis latter species may be distinguished from C. glomerata by its rounded-obtuse leaves and single peristome. — PLR. ]^04 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Editor's Note The moss Bryoxiphium norvegicum (Brid.) Mitt, is a rare member of the flora of the Interior Highlands. Julian A. Steyermark first discovered Bryoxiph- ium in the Interior Highlands at Pickle Spring, Ste. Genevieve Co., Missouri, one of the State's most interesting botanical localities, in 1934. This locality, which fortunately remains largely undisturbed, harbors many other interesting plants, both cryptogams and phanerogams. Askell and Doris Love's "Studies on Bryoxiphium" (The Bryologist 56: 73-94; 183-203. 1953) presents a detailed investigation of the history, taxonomy, and distribution of the genus. Since Steyermark's discovery and the Loves* study, several additional localities for this species have been discovered in the Interior Highlands; these are summarized by Paul L. Redfearn, Jr. in "Mosses of the Interior Highlands of North America" in this issue of the Annals. The illustration of Bryoxiphium norvegicum on the cover banner is taken from W. S. Sullivant's plate in his "Contributions to the bryology and hepaticology of North America" (Mem. Amer. Acad. Arts 3: 57-66 + 5 pi 1848). SuUivant was the first to collect the moss in North America, in Ohio and Kentucky. Bryo- xiphium norvegicum has never been found in Norway, indeed it is not known from Europe at all. According to the Loves' "Studies," the Dane Axel Morch first collected it — in Iceland in 1820. Morch sent his moss collections to the German bryologist C. F. Homschuch, who apparently determined the specimen as Weissia volcanica P. Beauv., a moss of Madagascar and Reunion Island. Hom- schuch seems to have wondered if the Icelandic moss could be conspecific with the one known from off eastern Africa and sent some material to Augustin Desvaux in Paris for further study. Desvaux realized that Morch's collection had been misdetermined and sent the specimen on to S. E. Bridel. Although Desvaux was correct in his determination, he noted on the specimen that it had been col- lected in Norway, Bridel published the name Phyllogonium norvegicum in his Bryologia Universa (1827) based on this collection, and Mitten {Musci Austro- Americani^ 1869) made the combination Bryoxiphium norvegicum^ the name by which it is known today. — Editor. vo issues of the Annals of the Missouri Botanical Garden 99^266, and Vol. 58, no. 3, dd. 267-369, were oublished on 21 J PREPARATION OF MANUSCRIPT The is a charge of $25 per printed page to help defray costs of pubHcation. Authors are asked to follow the suggestions below in order to expedite editing and publication. If an author feels that his manuscript presents special problems, he should write the editor concerning the best way to handle these before submitting the manuscript. Manuscripts must be typewritten on one side of substantial weight paper, 8% X H in. The manuscript should have wide margins and be double spaced throughout, including tfie abstract, footnotes, legends, tables, lists of specimens, and the bibliography. Tables should be typed separately and placed at the end of the text. Authors should indicate in the margins the approximate places for illustrations and tables. Submission of the original and one carbon or xerographic copy of the manuscript is desirable, and the author should also retain a copy of tiie final, typed draft. Acknowledgements to granting agencies, herbaria, illustrators, and technical assistants may be conveniently placed as a footnote on page one. The authors full mailing address should appear as a second footnote. An abstract must accompany each paper other than "Notes.** The abstract should succinctly summarize the findings and conclusions of the paper and should be completely comprehensible itself. A brief Latin diagnosis for each new taxon is preferred to a complete Latin description. A complete description should be given in English. The citation of specimens should be concise. Geographic names are put in order of de- creasing political magnitude. Only the barest essential data concerning each specific locality should be given. Collectors are cited by family name and collection number. If there is no collection number, the year of collection should be given. Herbaria are designated according to the current edition of Index Herbariorum. Abbreviations should be checked for consistency and to make sure they are unambiguous. Periods are used after all abbreviations except metric measures, compass directions, and her- barium designations. All illustrative material should be mounted on stiff cardboard. If the originals are too large to be conveniently mailed, photographic reductions should be submitted. The maximum size of a printed illustration is 5 X 7% in., and therefore, the height of an illustration must not exceea about 1.5 its width. Figures are numbered consecutively, since they are not printed as "plates." Numbering must be done with a mechanical device or with dry-transfer lettering and never by hand. The amount of reduction should be noted on the back of each illustration, together witli the figure numbers, author's name and title of the paper. Photographs should be sharp, glossy prints. Numbering should be applied directly to the surface of the photograph. Several photographs may be assembled to form a composite block, and each photograph should be numbered separately. The individual photographs should be mounted with the interior edges flush. Line drawings are prepared with India ink and must never be placed in the same block widi photographs. Authors wishing to have original illustrations returned must notify the editor when proofs are returned. The bibliography should be compiled with care. The author should check to make sure each entry in the bibliography is referred to in the text and each reference to a paper in the text is entered in the bibliography. Citations of periodical literature should appear as follows: Author's last name, initial(s). Year. Full title of article. Name of journal. Volume: inclusive pagination. Names of journals are abbreviated according to Botanico-Periodicum-Huntianum, except Journal should be abbreviated Jour, and Xeitschrift, Zeitschr. Citations of books should appear as follows: Author's last name, initial(s). Year. Full title. Edition, if applicable. Place or publication. Autfior will receive page proof. Since changes in proof are expensive, authors should not make revisions at this stage. Authors will be charged for alterations other than corrections of printer's errors. Reprints may be purchased at cost plus postage. An order blank for reprints will be en- closed with proofs, A reprint price list is available from the editor. All manuscripts should be addressed to the Editor, Marshall R. Crosby, Missouri Botanical Garden, 2315 Tower Grove Avenue, St. Louis, Missouri 63110. Publications of the Missouri Botanical Garden ANNALS The Annals of the Missouri Botanical Garden contains con- tributions mainly in plant systematics. The Annals appears three times a year, and three numbers, totaling about 400 pages, constitute a volume. Many back issues are available, and out-of-print numbers will be reprinted as required. All subscriptions to the Annals must be placed through Allen Press, Inc., 1041 New Hampshire, Lawrence, Kansas 66044. Allen Press also handles all claims for missing num- bers and all orders for back issues. The prices below are strictly net; there is no discount to agents. Subscription price $16.00 per volume, domestic $18.00 per volume, foreign Back issues $20.00 per volume $ 7,00 per issue FLORA OF PANAMA The Flora of Panama treats the various families of indigenous and naturalized plants of the Republic of Panama and the Panama Canal Zone. Each family is dealt with in a critical manner with synonymy, descriptions, discussions, illustrations, and citation of representative Panamanian collections for each species. The Flora appears in the Annals of the Missouri Botanical Garden as contributions accumulate, but all parts are available separately through Allen Press, Inc., 1041 New Hampshire, Lawrence, Kansas 66044. V VOLUME 59 1972 NUMBER 2 Missouri CONTENTS Disjunctions in Plants: A Symposium Otto T. Solbrig - . . . 105 Morphology and Phytogeography; The Classical Approach to the Study of Disjunctions Carroll E. Wood, Jr 107 Approaches to Disjunct Populations: The Contribution of Palynology Donald K Whitehead 125 Genetical Features of Ferns as Contrasted to Seed Plants Edward J. Klekowski, Jr, . 138 Chemosystematic Data: Their Use in the Study of Disjunctions B. L. Turner . . . 152 Disjunctive Distributions in the Lichen-Forming Fungi William Louis Culberson , 165 Disjunctions in Bryophytes W, B. Schofield ir H. A. Crura 174 Disjunctions in Homosporous Vascular Plants W. H. Wagner, Jr 203 The Floristic Disjunctions between the "Monte" in Argentina and the "Sonoran Desert" in Mexico and the United States Otto T. Solbrig 218 Species Disjunctions in Larrea: Evidence from Morphology, Cytogenetics, Phenolic Compounds, and Seed Albumins /. H, Hunziker, R. A. Palacios, Amalia G. de Valesi & Lidia Poggio 224 Plant Species Disjunctions: A Summary Peter H. Raven 234 Notes on Panamanian Trees and Shrubs Collected in 1971 by L. R. Holdridge and others John D. Dwyer 247 Solanaceae Studies II; Typification of Subdivisions of Solanum W. G. D'Arcy . . 262 Xcw Taxa and Recombinations in Lopezia ( Onagraceae ) Uzi Plitniann, Peter H. Raven ir D, E. Breedlove 279 The Comparative Morphology of tlie Cochlospermaceae. III. The Flower and Pollen Richard C. Keating 282 New World Juglandaceae, III. A New Perspective of the Tropical Members with Winged Fruits Donald E. Stone 297 s. *v ^ i VOLUME 59 1972 NUMBER 2 ANNALS OF THE Missouri Botanical Garden The Annals contains papers, primarily in systematic botany, contributed from the Missouri Botanical Garden and the Department of Biology of Washington University. Papers originating outside the Garden or University wdll also be accepted. For information on preparation of manuscripts, see the inside back cover. The Annals appears three times a year, and three numbers, totaling about 400 pages, constitute a volume. For information concerning subscriptions^ see the back cover. Matters regarding exchange of publications are handled by the Library, Missouri Botanical Garden, 2315 Towner Grove Avenue, St. Louis, Missouri 63110. Editorial Committee Marshall R. Crosby, Editor Missouri Botanical Garden John D. Dwyer Missouri Botanical Garden 6- St, Louis University Peter Goldblatt Missouri Botanical Garden Published by the Missouri Botanical Garden Press, St Louis, Missouri 63110 © Missouri Botanical Garden 1973 ALLEN PRESS, INC. ''Tn'* LAWRENCE, KANSAS VOLUME 59 1972 NUMBER 2 ANNALS OF THE Missouri Botanical Garden DISJUNCTIONS IN PLANTS: A SYMPOSIUM Otto T. Solbrig^ The following pages present all but one of the papers presented at a sympo- sium held in Edmonton, Canada, during the joint meetings of the Canadian and American Botanical Societies. The symposium was jointly sponsored by the American Fern Society, The American Society of Plant Taxonomists, the Botan- ical Society of America and the Bryological Society, Dr. Warren H. Wagner and myself were charged by these associations with the task of organization. All species of plants, with the exception of a few relicts now restricted to one population, have disjunct distributions in the strictest sense of the word. How- ever, the botanist uses the term in a more restricted way. Range disjunctions refer usually to large discontinuities in the distribution of the populations of a species. There is no rigorous criterion however as to what magnitude of dis- continuity qualifies as a ''range disjunction," and consequently we see the term applied equally to relatively small discontinuities in the order of 100 miles, and large bipolar ones involving thousands of miles. The interruption in the range of a species can have two origins: (1) the range was once continuous and the intermediate populations have become ex- tinct, or (2) the range was never continuous and the disjunct populations have become established with the aid of some event that carried the propagules over the inhospitable area. To these two possible origins, recognized since the prob- lem of range disjunctions was identified, a third explanation has been added lately: the movement of continents that has physically moved apart populations that once were contiguous. This applies particularly to amphiatlantic disjunctions. The study of disjunctions of single species is fraught with all the problems involved in trying to reconstruct what was a unique event. When the disjunction is the result of extinction of intermediate populations of a once continuous range, fossil evidence if available can help solve the problem in an unmistakable way. Such is the case with the many disjunctions involving southeastern North America and southeastern Asia, an example of which is the genus Nyssa. W^cre once a ^ DepcOiiment of Biology and Gray Herbarium, Harv^ard University, 22 Divinity Ave,, hrirlfrp. Mass. 02138. Camb Ann. Missouiu Box. Gaud. 59: 105-106. 1972. IQQ ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 continuous range from southeastern Asia to the southeastern United States via the lands bordering the Bering Straits was proposed, we know now that the con- tinuity existed in the opposite direction, across Europe and Asia. In this respect the new evidence regarding continental drift has both simplified and complexed the problem: simplified in that the position of the land masses in the past can now be assessed using independently obtained, non-biological criteria; complexed because the relative position of continents throughout geological time in relation to each other and to the poles and equator has to be taken into account. Both ancient and modern amphiatlantic disjunctions can be understood when the wan- dering of the continents are taken into account, from the tree ferns of the upper Carboniferous, to many angiosperms on both sides of the Atlantic. Evidence for the establishment of disjunctions as a result of long-range dis- persal of propagules cannot be obtained directly, and therefore the study of such disjunctions is less satisfying intellectually. Given the spotty fossil record of the past, and its bias toward certain types of plants, it is not possible to state that absence of a fossil record in an area indicates that a plant has never grown there. What is needed is a set of additional criteria that will allow us to recognize un- mistakably a relict disjunction due to long range dispersal. To produce reliable criteria, if indeed it is ever possible, we need to make use of all the tools that modern biology has to offer. The objective of the symposium was to gather to- gether a number of researchers who were working on these problems in order to discuss these problems. The symposium — and the papers here presented — was divided into two ses- sions. A morning session aimed at assessing the possible contributions of various techniques and approaches: moiphology and phytogeography (C. Wood); paly- nology (D. Whitehead); genetics and cytology (E. Klekowski) and chemistry (B. Turner). The afternoon session was devoted to assessing the biological pecu- liarities of diverse phyletic groups: Lichens (W. Culberson); bryophytes (W. Schofield and H. Crum); ferns (W. Wagner) and angiosperms (O, T. Solbrig, J. H. Hunziker, and T. Mosquin). The meeting was summarized by P. Raven. The presentations and the discussion showed that a great deal of progress has been made towards setting up criteria to assess disjunctions; that the new dis- ciplines have new insights to offer; and that different phyletic groups have dif- ferent probabilities of long-range dispersal and establishment. It also showed however that more work is needed and that it is an interesting and rewarding area of research, * ^^ In closing I would like to thank Dr. Soper and Dr. Taylor for chairing the sessions; all the contributors for their work; and our hosts, the Canadian Botanical Societies and the University of Alberta, for their hospitality. Finally, I would like to thank Dr. Peter Raven and the Missouri Botanical Garden that made pubHca- tion of this work possible. MORPHOLOGY AND PHYTOGEOGRAPHY: THE CLASSICAL APPROACH TO THE STUDY OF DISJUNCTIONS Carroll E, Wood, Jr.^ Disjunctions are evident in the distribution of almost every species of plant, for it is obvious that no species is composed of completely continuous populations. Discontinuities, however, range from small to very large. Although formerly widely distributed, the giant redwoods or big trees {Sequoiadenclron giganteum) now occur in groves (populations) of varying size scattered over a distance of 250 miles along the western slopes of the Sierra Nevada of California. The gaps between groves may be insignificant or up to 50 miles wide, but disjunctions of this size do not draw particular attention, except as showing that the intervening habitats presumably are now unfavorable for the establishment and growth of Sequoiadendron. If, however, a grove of big trees were to be found in some part of the world as remote from California as the Himalayas of Nepal or the Smoky Mountains of North Carolina and Tennessee, the find might be more exciting than the discovery of living Metasequoia in central China. To the inquisitive mind such a large gap in distribution would demand explanation, and, unsurprisingly, disjunctions of this size are the ones that have most intrigued biologists. A disjunction in the distribution of a single species or genus, even a very wide one as in the hypothetical example above, may make one wonder, but when numerous species in the flora of a given region show the same general disjunction, the curious scientist who realizes that there is a common pattern can hardly resist searching for the circumstances behind it. Patterns of disjunction between eastern North America and eastern Asia, between Europe and North America, between eastern and western North America, and between North and South America are among those that have proved to be especially intriguing (at least to biologists of the continents involved). Almost from Linnaeus onward, botanists have sought and advanced explanations as diverse as double creation of species, long-distance dispersal, disruption (by a variety of factors) of formerly continuous ranges, drift- ing of continents, and the either conscious or unintentional activities of man, A great deal of effort has gone into the study of disjunct distributions, and in all of the published material that has resulted, it is evident that comparative mor- phology is basic. Morphology is basic to the study of disjunctions (and to all of phytogeography, for that matter), because it is basic to taxonomy. Without a taxonomic frame of reference for each disjunct taxon comparisons of almost any sort become meaningless. As taxonomic ideas about a taxon change, so must con- clusions based upon the earlier taxonomy. The new data brought into taxonomy by biosystematics have not upset the fundamental taxonomic framework built on morphology but have generally confirmed its soundness. The new data have, moreover, served to broaden the parameters of taxonomy, at the same time ^ Department of Biology' and Arnold Arboretum, Harvard University, 22 Divinity Avenue, Cambridge, Massachusetts 02138, Ann. Missouri Bot. Card. 59: 107-124. 1972. 2Q3 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Strengthening its framework and providing new c]ues that help to untangle knotty evolutionary problems. Similarly, in the study of disjunctions new techniques are providing new clues tliat comparative morphology alone cannot. Modern bota- nists are using data from geology, palaeobotany, ecology, palynology, cytology, genetics, physiology, and chemistry in their search for further evidence that will help to explain the fascinating patterns of disjunction in the distribution of plants. Before considering various aspects of morphology and taxonomy involved in study ems eastern American plants: Tertiary relict disjunctions; North American disjunctions; and North American-South American disjunctions. Tertiary Relict Disjunctions (Maps 1-11). — This pattern of disjunction is an extension of the spectacular and long-known eastern North American-eastern Asian floristic relationship (Maps 1-3, 8) that was first commented on by Linnaeus (1750; cf, Graham, 1966), later recognized independently by Thunberg (1784) and Castiglioni (1790; c/. Li, 1955), but really brought to the attention of American and European scientists by Asa Gray through a series of papers pub- lished between 1840 and 1878 (Gray, 1840-1878). In the most important paper of the series Gray (1859) showed that there are numerous similarities between the floras of eastern North America and Japan, similarities so marked in many instances that it appeared to him that the same species occurred in both regions. Surprisingly, similarities between the floras of Japan and western North America were fewer, and many of the "identical species" common to the eastern parts of both continents were missing. To explain the presence of the same species in both Japan and eastern America Gray advanced an hypothesis involving migration and interchange of species of Asia and America across the region of the Bering Strait followed by disruption of the continuous ranges by the Pleistocene glaciations, ideas that have played a fundamental role in the development of plant geography. Subsequent discoveries, especially in China, have revealed even stronger floristic ties between eastern and southeastern Asia and eastern North America. At present, the two areas are known to share at least seven pairs of closely related genera, 62 genera that now occur nowhere else, and at least 24 more widely distributed genera that have closely related species or groups of species in the two regions (Maps 1^, 8). It has also become evident that most of Gray's (1859) original "identical species" are morphologically distinguishable in the two areas, and, in most instances, are quite distinct (c/. Li, 1952), with the exception of circum- boreal species and a few others. There seem to be only relatively few identical species that are disjunct between the two areas {e,g. Tipularia discolor, Cypripe- clium arietinum, both of which may possibly be examples of long-distance dis- persal, and Alnus maritima [A. japonica] with its truly amazing disjunctions be- tween Asia, Oklahoma, and Delaware), although there are a number of species that are treated by various authors as distinct at the varietal, subspecific, or spe- cific level (e.g. Adlumia fungosa, Penthorum sedoides, Gaultheria hispidula, Phryma leptostachya) . Beyond this bicentric distribution there are also at least ten genera that have 1972] WOOD— CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS 109 \^ sect. Cymbostemon 26 species IILICIUM {niiciaceae) t. Illiclum -| T"'y~~ 12 species sect, nudum 1 species sect. C^bostemon 3 species DESMODIUH sect. PODQCARPIUM (Legumi nosae) _ ^ CALYCANTHUS (Calycanthaceae) Calycanthus p' sinensis ' In i C. occidental is /^Mcy PERIDERIDIA (Umbel li ferae) x P. neurophylla 12 species Maps 1-6.— 1-4 1. Gdsemium (after Ornduff, 1970).— 2. Piem.—S. Illicium (after Smith, 1947). — 4. De^njo- J/wm section Podocarpium (after Isley, 1951). — 5-6. Two examples of taxa restricted to east- ern Asia, western North America, and eastern North America. — 5. Caltjcanthus (largely after Nicely, 1965).— 6. Perideridia (after Chuang & Constance, 1969). Maps 7-8. Two relict Tertiary genera. — 7, Extant species of Liquidambar (courtesy of A, L. Bogle; L. stijraciflua after Little, 1971). — 8. NyssUy hatched areas, extant species; dots, Tertiary occurrences (after Eyde, 1963, and Eyde & Barghoom, 1963). > CO O •-3 2 O t3d n > < o U3 ^" I »Ti..» var. mexicana I ^ W. idahoensis Calothyrsus californica FAGUS (FAGACEAE) *-■* *-•• ■ ■• • ■• '■--• ■ "• .vv \'v F. grandi folia var, grandi foil a WALDSTEINIA (Rosaceae) , III' W. fragarioides W. lobata AESCULUS (Hippocastanaceae) sect. Pa V 1 a A. glabra^ ^ ST octandra. S7 sylvatica 5Z Pav^a sect. Parryanae A. Parryi sect. Macrothyrsus A^ parviflora F. sylvatica } S /f M_m F. oriental is * 4 \ W. ternata -Y subsp. trifoli 1 C: ^ d^ sect. Aesculus i j5L ^lippocastanum Maps 9-11. Tliree genera with relict Tertiary distribution. — 9. Tralau, 1962; Little, 1971).— 10. Extant species of Walchteinia (after tions Aescuhis and Calothyrsus (after Hardin, 1957, 1960). F. crenata F. japonica -5 speci *• «■*«■ »* «/d«*** •>»«■*• «a*t«t* «•*■*. !*•• *d«i*«« ****** • ^»»* **•««•* n > IT" > > n H O H Hi o o C/5 c! O 212 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 related taxa in eastern North America, eastern Asia, and southeastern Europe- Asia Minor {e.g. Liquidambar [Map 7], Fagiis [Map 9], Carpinus, Epigaea) and 30 genera that have disjunct related taxa in eastern North America, western North America, and eastern Asia. At least a dozen genera (e.g. Erythronium^ Ostrya, Hepatica, Platanus, Waldsteinia [Map 10], Cercis, and Aescuhis [Map 11] occur only in all four areas, and at least six more wide-spread genera have related species distributed in this patteiTi, In all, at least 146 genera of eastern North America (13 per cent of the indigenous genera of seed plants) are involved; almost half of these (67 genera) are woody plants. Over the years paleobotanical findings have confirmed the widespread occur- rence of many of these same genera, mainly the woody ones, in Eurasia and North America in broad areas from which they are now missing, and it is clear that the extant representatives of the same widely distributed woody genera are the survivors of the gradual climatic deterioration, volcanism, orogenic move- ments, and the glaciations that followed. Wolfe (1969) has pointed out that the representatives of some lineages that were present in the Tertiary in Pacific Northwestern America were able to adapt to changing climatic conditions (especially the switch from wet summer to dry summer) and now survive in different associations, others became extinct, and others survive in what are essentially relict habitats (especially in the Klamath Range of northern California and southern Oregon). The same undoubtedly was true in other regions of the Northern Hemisphere. The four disjunct areas, though, are the present more or less mesic refugia for many of the suwiving descendants of formerly more Avidespread genera. The largest, and ecologically most com- plex, of the relict areas are eastern Asia and eastern North America; those of western North America and Europe-Asia Minor are smaller, and many genera that formerly occurred in them have disappeared (e.g. Ulmiis and Liquidambar [Map 7] from western North America; Tstiga from Europe; Liriodendron, Mag- nolia, and Nyssa [Map 8] from both). The general pattern of formerly wide boreal distributions that were fragmented by orogenic movements, gradual climatic cooling, volcanism, and the Pleistocene glaciations seems to be well established, but much is unknown. The majority of the plants involved are as yet unknown as fossils (but palynologists might well look routinely for their pollen grains, even though many are not wind pollinated). The taxonomic interrelationships in many of the genera need much more study; it is not certain that all of the disjunctions are of the same age; their areas of sur- vival during the Pleistocene are for the most part unknown; and the accumulating -> Maps 12-14. Western America. — 12. Agoseris aurantiaca var. aurantiaca, dots, localities mapped by Femald (1925) for A. gaspensis and A. gracilens; circles, additional collections mapped by Quentin Jones; see text. — 13. Lesquerella arctica, black areas and dots, localities mapped by Fernald (1925); circles, additional localities mapped by Rollins and Shaw (1973); stations in northeastern Greenland and Siberia not shown. — 14. Astragalus ahoriginum, dots and outlined area as mapped by Femald (1925); circles, localities mapped by Bameby (1964), three Alaslcan sta- tions not shown. 1972] WOOD— CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS 113 11^ ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 proof of continental drift introduces new spatial relationships that must be con- sidered in connection with the migrations of plants. Eastern North American-Western North American Disjuncts. — The now famil- iar pattern of species of the high arctic and of the cordillcra of western North America disjunct in the glaciated area of northeastern North America (see Maps 12-14, 21 ) has received much attention, largely as a result of Fernald's stimulating hypothesis (1925-1935) that western and arctic plants were able to survive in unglaciated nunatak areas around Lake Superior, on the Gaspe Peninsula, on the tablelands of the Long Range of Newfoundland, and in the Torngat Mountains of Labrador, and that the surviving populations of these "old species" were unable to spread from these areas. "This failure of the plants of the unglaciated spots to extend their ranges into closely adjacent areas which, upon the melting of the Labrador sheet, became open territory ready for invasion, is interpreted as a further evidence of the antiquity of these plants; at the close of the Pleistocene they were already too old and conservative to pioneer, although they are able to linger as localized relics in their special undisturbed crannies and pockets" (Fernald, 1925: 243). He provided a list of some 295 species and varieties that either show this disjunction or are endemics (about 80 species or varieties) re- lated to species of western North America and showed that these occur in areas that were thought by many geologists of the time to have escaped glaciation. However, subsequent work has shown that all of the areas involved were indeed glaciated, and further field work has filled in some of the disjunctions in places that vmdoubtedly were glaciated (of. Maps 12, 14, 21). Monographic work has changed the taxonomic status of many of the supposed endemics, suggesting a shorter period of isolation. The work of a number of ecologists, too, has shown that the arctic plants are restricted to high altitudes or to river canyons or other places where their requirements for cold are met, and that many of the other relicts or endemics are plants of unstable habitats (cliffs, talus slopes, serpentine barrens ) where they can grow but are removed from competition with the domi- nant eastern boreal plants (cf. Rune, 1954). In this last respect they apparently are similar to the endemic species of ultramafic soils studied by Kruckeberg ( 1951, 1967, 1969) and to the shale-barren endemics of Virginia and West Virginia studied by Piatt (1951). In reviewing the disjuncts from northwestern North America, Schofield (1969) summarizes, "The most plausible explanation of their disjunction is that the east- ern representatives are remnants of a more widespread flora of the past, possibly of pre-Pleistocene arctic-alpine distribution in North America. The Pleistocene glaciations can be assumed to have eliminated the north-central portion of the range, but since habitats were available in north-eastern and western North Amer- ica the species survived, probably south of the glacial boundary, but possibly in nunataks or coastal refuges, moving to their present sites following retreat of the ice sheet but being eliminated from their Pleistocene refugium by the encroaching vegetation and by a succession toward more niesophytic temperate vegetation. Evidence for nunataks and refugia in eastern North America has been disputed/' Much less attention has been focused on East- West disjuncts to the south of 19721 WOOD— CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS 115 a disjnnction to west- Maps 15-17. Some taxa of unglaciated eastern North America with em North America.— 15. Dicentra cucullaria (after Stem, 1961).— 16. Rhyncho.spora capi- tellata (from Wood, 1971). — 17. Xerophyllum (from Wood, 1971). 216 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 the maximum extent of the Pleistocene ice, although some vicarious pairs of species (e.g. Piniis strobiis-P. monticoJa, Thuja occidentalis-T . plicata, Dirca palustris-D. occidentalism CaJycanthtis floridus-C, occidentalis [Map 5], Xero- phyllum asphodeloides-X. tenax [Map 17]) have often been cited. Wood (1971) has sorted out over 150 genera with taxa showing disjunct relationships between the part of the Appalachian mountain system south of the area of glaciation and western North America. This group includes some 50 genera with one or more species that appear to be disjunctly distributed (e.g. Dulichium arundinaceum [Map 23], RyncJiospora capitelhta [Map 16], R. glohularis^ Ranunculus pusillus, Sibara virginica, Dicentra cucullaria [Map 15], Viola canadensis vars. canadensis and corymbosa^ and Trichostema brachiatum) and 112 genera with related taxa of the same ranT< on both sides of the continent. Among the latter group are 20 genera that are restricted (or nearly so) to North America and 46 with distribu- tions that suggest the Arcto-Tertiary relict pattern {cf. Maps 5-6, 10). In all, at least 158 genera ( 14 per cent of the indigenous genera of seed plants of eastern North America) have disjunctions between the southern Appalachians and west- em North America either within the same taxon or between related taxa. Most of the taxa involved have not been studied in great detail, and the dis- tributional history of most is unknown. It seems likely that the disjunctions may be of different ages, since various taxonomic ranks are involved. Some fit the relict Tertiary pattern, others probably are the result of the disruption of formerly more continuous ranges by the Pleistocene glaciations, and still others may possi- bly be examples of long-distance dispersal. In some instances there may have been several disruptions, such as those postulated by Russell (1956) to account for the East- West disjunction of Viola macloskeyi subsp. pallens and the occurrence of both subsp. pallens and subsp. macloskeyi in the West. X^Tiatever the causes of disruption, one of the key issues is the question of how much of a vegetational shift southward was brought about by the Pleistocene glaciations {cf. Whitehead, this symposium). Amphitropical Disjunctions between North and South America. — Monog- raphers and floristic workers over a period of many years have pointed out ex- amples of disjunctions from both boreal and temperate North America to corre- sponding climatic parts of South America, mostly on the western side of both continents (Maps 18-22, 24) and disjunctions from the Sonoran and Chihuahuan desert regions of North America to deserts of northern Argentina and northern Chile. In summarizing a symposium on the amphitropical relationships in the Quart Biol. 38: the catalog to date of the herbaceous examples and thoroughly discussed the possible ways in which such disjunctions could have come about. The amphitropical dis- junctions fall into three groups: "bipolar or high-latitude, with about 30 species; temperate, with about 130 species; and desert, with a substantial number." In the herbaceous plants of all three groups (the first two are entirely herbaceous), the disjunction involves either the same species or closely related subspecific taxa in both areas. In the bioolar examoles. the northern tavon is nl^n pithrr pirmmboreal 1972] WOOD — CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS 117 QSMORHIZA (Umbel! i ferae) AGOSERIS (Compositae) HICROSERIS sect. HICROSERIS ( Compos itae) QSHQRHIZA (Umbelli ferae) PROS ERA (Droseraceae) Maps 18-22. Some taxa with an amphitropical disjunction between North and South America. — 18. Osmorhiza occidentalis and O. glabrata, a species-pair (after Constance & Shan, 1948). — 19. Agoseris heterophylla and A. coronopifolia, another species-pair (after Jones, 1954, and Chambers, 1963). — 20. Microseris section Microseris, with five species in western North America and M. pygmaea in South America (after Chambers, 1955, 1963). — 21. Osmorhiza depauperata (after Constance & Shan, 1948). — 22. Drosera hrevifolia (in part after Wynne, 1944). in distribution or (in a few cases) belongs to a circumboreal group. Of the ex- amples of temperate disjuncts brought together by Raven, in 111 instances the North American taxon is a species of western North America, in 13 is wide-spread in North America, and in 6 is a species of eastern North America. The desert disjuncts include both woody plants (with no species common to the two areas and with few common genera) and herbaceous ones (with either the same or closely related species in both areas). Some of the amphitropical disjuncts in all three categories have been studied in considerable detail, others have been rec- ognized from herbarium materials on purely morphological grounds. In summary, "Pertinent general considerations are that: (1) the North and South American populations are closely related; (2) the plants are almost without exception self- compatible and often autogamous; (3) they constitute an unbalanced assemblage entirely unrepresentative of the two extratropical areas; (4) they grow almost exclusively in open communities, not in woodland or scrub associations; (5) there are no corresponding cases among terrestrial vertebrates and very few among the insects; and (6) the floras of the two areas have been distinct since at least the middle Cretaceous and are still very different at present. The only explanation that accounts for all of these facts seems to be that at least the great majority of the plants reached their disjunct areas by long-distance dispersal relatively re- 118 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 23-24. Two genera formerly distributed in the Eastern but now restricted to the Western Hemisphere. — 23. Dulichium, the single species disjunct between east and west in North America; dots, Tertiary; circles, interglacial records (after Tralau, 1959); black, extant distribution, generalized (after Wood, 1971). — 24. Proserpinaca, P. palustris disjunct in Cen- tral America, Colombia, and Brazil; P. pectinata disjunct in Tennessee (both after Fassett, 1953); triangles. Tertiary occurrences (after Dorofeev, 1958). cently. For the bipolar species, the Pleistocene seems the most likely time of dis- persal; for the temperate species, the late Pliocene or Pleistocene; and for the thos the recent past. the nor har th . . . and are almost entirely herbaceous. The desert disjuncts, on the other often appear to have originated in the south . , . or to have diverged from a common tropical ancestor. Many of them are woody" (Raven, 1963). Throughout all three of these examples of general patterns of distribution it is study crucial taxonomic relationships and comparative morphology Gray's 'Identical species yy m Japan have already been mentioned, as has the status of some of Femald's Gaspe endemics that have been shown to be a part of the variation in more widespread western species. As one example of the latter, both Agoseris gaspensis and the later-described A. naskapensis Queb Femald 1972] WOOD— CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS HQ have been included by Quentin Jones (1954) in the western American A. aurantiaca var. aurantiaca (Map 12). Jones notes that both of the eastern dis- juncts fall well within the range of variability in the polymorphic var. aurantiaca and that if the source of the collection were unknown it would be impossible to refer it with any assurance either to the disjunct eastern populations or to the Rocky Mountain ones. This changed taxonomy not only merges A. gaspensis and A. gracilens with A. aurantiaca, but includes an eastern population (A. naskapensis) from an area that undoubtedly was glaciated, producing a rather different prob- lem in disjunction from that seen by Fernald (1925). Similarly, the enlargement of Perideridm (Umbclliferae) by Chuang and Constance (1989) to include the Asiatic PterygopJeurum neurophyllum produces a different phytogeographical problem. Ferideridia has been regarded as an en- tirely American genus with most of the species in the western United States and P. americana disjunct in the east-central United States (Map 6). However, Chuang and Constance showed that Pterygopleurum neurophyllum, which is dis- tributed in southern Japan and Korea, has corky fruit ribs like those of Perideridia hoioelln, has foliage closely similar to that of Perideridia gairdneri subsp. horealis, and, most decisively, has "exactly the type of polystelic tuberous root found in most species of Perideridia (except P. HowelUi and P. Kellogii), a feature un- known in Cenolophium, Ligusticum^ or Sium^" the other possible genera to which it might be related. Accordingly they included the species in Perideridia as P. neurophylla. "The confirmation of the close similarity and probable affinity of this Asiatic species to the taxa of Perideridia inhabiting Pacific Northwest America suggests that the genus may once have been considerably more widely distributed r in the northern hemisphere, perhaps before Pleistocene glaciation. At a different taxononiic level, disjunct relationships between the species of a genus may be seen to be very different when the taxonomic framework is over- hauled. The genus Aesculus (revised by Hardin, 1957, 1960) has an interrupted distribution in Asia, southeastern Europe, eastern North America, and western North America (Map 11). The four species of the Himalayas and China, how- ever, belong to a different section of the genus from the geographically adjacent A. turhinata of Japan, which, with A. hippocastanum, of the Balkan Peninsula, constitutes section Aesculus. The Himalayan and Chinese species find their closest relative in the North American A. californica, the five comprising section Calo- thyrsusl The remaining American species constitute three sections (two mono- typic) without disjunctions between their species. Osmorhiza (Umbclliferae), Amsonia (Apocynaceae), Gaultheria (Ericaceae), and Styrax (Styracaceae) all have representatives in eastern North America, as well as in the western part of the continent. However, in each genus the species of the eastern United States belong to a different alliance from their western congeners, and, in each instance, they are morphologically closer to species of eastern Asia than to the much nearer western ones (cf. Wood, 1971: 378). In Gaultheria, for example, the two eastern American representatives, G. procumhens and G. hispidula var. hispidula, both members of section Gaultheria, are respec- tively most closely related to G. pyroloides (of section Leucothoides) and G. yy ]90 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Voi.. 59 hispidula var. japonica, of Japan. Of the three quite different western American species, G. ovatifolia and G. Jiumifusa belong to section Amblyandra, and G. shallon belongs to section Brossaeopsis, which also includes two species of the West Indies and those of Mexico {cf. Shaw, 1940). Consequently, although the genus is of disjunct distribution in North America, the proper phytogeographic compari- son of the eastern species is not with those in western North America but with species of Asia. It is obvious, too, that as the state of knowledge of a flora changes, as it be- comes better known, and as floras of other areas become better known, new dis- coveries may greatly alter phytogeographical relationships. At the time Gray (1859) wrote his important paper on the relationships between the floras of Japan and North America, he had rather scant Japanese material and little or nothing from China. Continued exploration, particularly of China and adjacent Indochina and Burma, has brought in new collections that have made possible better tax- onomic decisions, and many new examples of close relationships between the floras of China and Japan and of China and eastern North America have been found. Even relatively recently, the discovery in eastern China of a white- flow^ered species of Calycanthus (described as C. sinensis in 1963, and raised to generic rank as Sinocalycanthus by Cheng and Chang, 1964) added a new mor- phological dimension to the genus and confirmed the relict relationship suggested by the occurrence of Chimonanthus (the only other genus of the family) in China and of species of Calycanthus in the eastern United States and in California (Map 5). As a result of his studies of Asiatic plants Gray suggested that relatives of "monotypes" of eastern North America should be sought in Asia, and, with the con- tinuing botanical exploration of Asia, numerous examples of this relationship have been found {e.g. in Saururus, Decumaria^ Itea^ Berchemia^ and Gordonia^ to name only a few). One of the more recently recognized instances of this relationship is that of the monotypic endemic genus Anemonella (Ranunculaceae). Anemo- nella thalictroides, of wide distribution in open deciduous woods over much of the eastern United States, is a low, herbaceous, tuberous-rooted plant with Thalictrum-like foliage and a reduced inflorescence of Anemone-like flowers with a perianth of five to ten conspicuous white or pinkish tepals. Although usually treated as a distinct genus, it has occasionally been referred to Anemone. Boivin (1957), however, prompted by comments of A. J. Eames that Anemonella and Thalictrum are the only genera of the Ranunculaceae with multiple carpellary traces, found that Anemonella should be referred to Thalictrum section Physo- carpum. Aside from T. clavatum, of the southern Appalachians, and the very similar T, mirahile, of Kentucky and Alabama, both of which seem to be more closely related to Asian taxa than to Anemonella^ this section is composed of twelve species of eastern Asia. The tuberous roots, the compound leaves, and the tepals, stamens, carpels, and fruits of Anemonella fit well with those of species of this section. A number of the species have subopposite or opposite leaves; T. tubiferum and T. filumentostim have two opposite leaves, as in Anemonella; and T. coraneum and T. chiaonis have reduced inflorescences. The tendencies toward 1972] WOOD— CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS 121 reduction in leaves and inflorescences seem only to have been carried somewhat further in Anemonella, and the plant fits well in Thalictrum (but not at all in Anemone^ from which it differs fundamentally in leaves, inflorescences, and carpels) . Accordingly, Anemonella became, rather redundantly, Thalictrum thalictroides^ and a further disjunct relationship between eastern North America and eastern Asia became evident. Discoveries that alter the distributional patterns that led Fernald to think in terms of nunataks have already been mentioned, and a few of these are seen in the maps of Agoseris aurantiaca, Lesquerella arctica^ and Astragalus ahoriginum (Maps 12-14). In another geographic direction, work on the flora of the State of Santa Catarina, Brazil, has added species such as Arenarui groenlandica, Xyris brevifolia, Hypericum gentianoides^ and Proserpinaca palustris (Map 24) to the list of plants that are disjunct in Brazil from the main body of their distribution in the eastern United States {of. Raven, 1963); and other examples of this dis- junction have been found. Beyond these considerations, one of the principal problems in understanding disjunctions is our meager knowledge of the past distributions, the expansions and contractions in the ranges, of most plants. The fragmentary palaeobotanical record will always leave much to be desired, and the usefulness of palaeobotanical data is further impaired by misidentifications made by uncritical workers. Many of these misidentifications are based upon fossil leaves that it may or may not be possible to identify on the basis of detailed studies, and many of the reports based on fossils of this type must be discounted, at least for the present. The comments of Tralau (1963: 40) in his review of Eurasian fossils attributable to Magnolia kohus are pertinent: "It is, nevertheless, uncertain and can not be decided here whether all these finds [of leaves] really belong to the genus Magnolia or not. On the other hand fossil foliar specimens referred to the genera Juglans^ Ficus, Anona^ Persea, Lindera, Laurus, Eriobotrya, Evodia, Rhododendron, Ardisia, Solandra, and possibly even others actually may be remains belonging to Magnolia. Because of this obvious uncertaint}^ the prospect of unravelling the problems asso- ciated with these remains and reconstructing the Tertiary history of M. Kohus with them is most unattractive." Some groups, however, have been reviewed carefully and now have relatively well-documented fossil records based on wood, fruits, and pollen {e.g. Nyssa, Map 8, after Eyde & Barghoom, 1963; Fagus, Map 9, after Tralau, 1962) that show that the present disjunctions are the remnants of formerly broad ranges. Fossil pollen, in particular, is proving to be helpful in showing the presence of various genera of flowering plants in areas where they no longer occur {e.g. the well-documented pollen record of Liquidamhar in Eurasia, and the growing pollen record of Pachysandra in western North America). Contrary to the assertions of many plant geographers, present distributions do not necessarily indicate where a group of plants has occurred in the past, although there are exceptions. On the basis of similarities in the modem distribu- tions of plants that are restricted in their occurrence to the four areas where Tertiary relicts occur it can be inferred that genera such as Narthecium, Erythro- ]^22 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 niiim, Asarum^ Aesculus^ and Waldsteinia (see Maps 10-11) were formerly dis- tributed around the Northern Hemisphere as Nyssa and Fagus were. However, the modern distributions of most plants may not show where they have occurred in the past, and many speculations about origins, migrations, and disjunctions can be nothing more than speculations unless other evidence can be found. For example, the sedge Dulichium arwidinaceum is widely distributed in the eastern United States and occurs quite disjunctly in the Pacific Northwest and in a few localities in California. One cannot suspect from its extant range that it formerly occurred in western Europe, where fossil remains of it are known from a number of localities {cf. Map 23, after Tralau, 1959). Further, in the aquatic genus Proserpinaca (Haloragaceae), P. pectinata is mostly confined to the Coastal Plain of the southeastern United States, while the other species of the genus, P. palustris^ is more widely distributed in eastern North America and has disjunct occurrences in Central America, Colombia, and southern Brazil (Map 24). Cer- tainly there is nothing in this distribution to suggest that Proserpinaca formerly occurred in Europe and Asia, but Dorofeev (1958) has illustrated fossil endocai'ps that he assigns to two species analogous to the living ones. Although on the basis of morphology of the all too similar endocarps, the recognition of two ex- tinct species hardly seems justified, these endocarps certainly represent Proser- pinaca and document a formerly wide Eurasian range for the genus. These two examples, to which many others can be added, show that distributions of individ- ual species have changed in the past — and undoubtedly are still changing — and strongly suggest that in dealing with disjunctions the phytogeographer must be wary of interpretations based only on extant distributions. Morphology, then, is basic to the taxonomy of both extant and extinct plants, and, consequently, to the study of disjunctions. As in taxonomy, however, there are questions that cannot be answered by morphology alone, and evidence from other disciplines must be sought. For most species of living plants it is unlikely that we shall ever have any sort of fossil record that will show where their an- cestors grew, but detailed studies involving morphology coupled with ecology, cytology, genetics, physiology, and other disciplines will undoubtedly provide many clues that will be of great importance in explaining the many patterns of disjunction. Some data will provide answers, some will only be suggestive, and some will raise still further questions. The other papers in this symposium show some of the varied ways of approaching these problems, but in all of them mor- phology, the foundation of taxonomy and plant geography, is always the comer stone, no matter how elaborate the structure it supports. Literature Cited Barneby, R. C. 1964. Atlas of North American Astragalus. Mem. New York Bot. Card. 13: 1-1188. Boivix, B. 1957. fitudes thalictrologiqiies III. Reduction du genre Anemonella Spach ( Ranunculaceae ) . Bull. Soc. Roy. Bot. Belgique 89: 319-321. Castiglioni, L. 1790. Viaggio negli Stati Uniti deirAmerica Settentrionale fatto negli anni 1785, 1786 e 1787 con alcune osser\azioni sui vegetabili piu utili di quel paese. 2 vols. Milano. Chambers, K. L. 1955. A biosystematic study of the annual species of Microseris, Contr. Dudley Herb. 4: 207-312. 1972] WOOD— CLASSICAL APPROACH TO STUDY OF DISJUNCTIONS J23 , 1963. Amphitropical species pairs in Microseris and Agoseris (Compositae: Cicho- rieae). Quart, Rev. Biol. 38: 124-140. Cheng, W. C. & S. Y. Chang. 1964. Genus novuni Calycanthacearuni Chinae orientalis. Acta Phytotax. Sin. 9: 137-138. PI, 9. Chuang, T. I. & L. Constance. 1969. A systematic study of Ferideridia (Umbelliferae- Apioideae). Univ. Calif. Publ. Bot. 55: 1-74. Constance, L. & R. W. Shan. 1948. The genus Osmorhiza (Umbelliferae), a study in geographic affinities. Univ. Calif. Publ. Bot. 23: 111-156. Pis. 26-29. DoROFEEv, P. I. 1958. A new species of Proserpinaca L. from the Tertiary flora of the USSR. Bot. Zum. 43: 1337-1340. (In Russian.) Eyde, R. H. 1963. Morphological and palaeobotanical studies of the Nyssaceae, I. A survey of the modem species and their fruits. Jour, Arnold Arbor. 44: 1-59. & E. S. Barghoorn. 1963. Morphological and palaeobotanical studies of the Nyssaceae, 11. The fossil record. Jour. Arnold Arbor. 44: 328-376. Fassett, N. C. 1953. Proserpinaca. Comm. Inst. Invest. Ci. Univ. El Salvador 2(5-6): 139-162. Fernald, M. L. 1925. Persistence of plants in imglaciated areas of boreal America. Mem. Amer. Acad. Arts Sci. 15: 239-342. (Mem. Gray Herb. 2.) . 1926. Two summers botanizing in Newfoundland. Rhodora 28: 49-63, 74-87, 89- 111, 115-129, 145-155, 161-185, 203-223, 230-247, 265-283, 298-315, 327-346, 364-386, 395-403. — . 1933. Recent discoveries in the Newfoundland flora. Rhodora 35: 1-16, 47-63, 80-107, 120-140, 161-185, 203-223, 230-247, 265-283, 298-315, 327-346, 364-386, 395^03. — . 1935. Critical plants of the Upper Great Lakes region of Ontario and Michigan. Rhodora 37: 197-222, 238-262, 272-301, 324-341. Graham, A, 1966. Plantae rariores camschatcenses; A translation of the dissertation of Jonas P. Halenius, 1750. Brittonia 18: 131-139. Gray, A. 1840. Dr. Siebold, Flora Japonica; sectio prima. Plantas ornatui vel usui inservientes; digessit Dr. J. G. Zuccarini; fasc. 1-10, fol. [A review]. Amer. Jour. Sci. Arts 39: 175-176 . . 1846. Analogy between the flora of Japan and that of the United States. Amer. Jour. Sci. Arts II. 2: 135-136. — . 1856. Account of the botanical specimens. In Narrative of the Expeditions of an American Squadron to the China seas and Japan, performed in the Years, 1852, 1853, and 1854, under the command of Commodore M. C. Perry, United States Navy, by order of the government of the United States. 2: 305-332. ^. 1859. Diagnostic characters of new species of phaenogamous plants, collected in Japan by Charles Wright, Botanist of the U. S. North Pacific Exploring Expedition. ( Pub- lished by request of Captain John Rodgers, Commander of the Expedition.) With obser- vations upon the relations of the Japanese flora to that of North America, and of other parts of the Northern Temperate Zone. Mem. Amer. Acad. Arts Sci. II. 6: 377-452. (Re- printed, in part, in C. S. Sargent (editor), "Scientific papers of Asa Gray." 2; 122-141. 1889. ) — . 1860. Illustrations of botany of Japan and its relation to that of central and northern Asia, Europe, and North America, Proc. Amer. Acad. Arts Sci. 4: 131-135. — . 1873. Address of ex-president of the Association. Proc. Amer. Assoc. Advanc. Sci. 21: 1-31. (Reprinted as ''Sequoia and its history" in C. S. Sargent (editor), "Scientific Papers of Asa Gray " 2: 142-173. 1889.) — . 1878. Forest geography and archeology. Amer. Jour. Sci. III. 16: 85-94, 183-196. (Reprinted in C. S. Sargent (editor), "Scientific Papers of Asa Gray." 2: 204-233, 1889. ) Hardin, J. W. 1957. A revision of the American Hippocastanaceae II. Brittonia 9: 173-195. — . 1960. Studies in the Hippocastanaceae, V. Species of the Old World. Brittonia 12: 26-38. IsLEY, D. 1951. Desmodium: Section Podocarpium Benth, Brittonia 7: 185-224, Jones, Q. 1954. Monograph of Ago.vem, tribe Cichorieae. Unpublished Ph. D. Thesis, Harvard University. Kruckeberg, a. R. 195 L Intraspecific variability in the response of certain native plant species to serpentine soil. Amer. Jour. Bot. 38: 408-419. 124 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 . 1967. Ecotypic response to ultramafic soils by some plant' species of northwestern United States. Brittonia 19: 133-151. . 1969. Soil div^ersity and the distribution of plants, with examples from western North America. Madrono 20: 129-154, Li, H. L. 1952. Floristic relationships between eastern Asia and eastern North America. Trans. Amer. Philos. Soc. Philadelphia 42: 371-429. (Reprinted 1971, with a new fore- word, as a Morris Arboretum Monograph.) 1955. Luigi Castiglioni as a pioneer in plant introduction. Proc. Amer. Philos. Soc. 99: 51-56. Linnaeus, C. 1750. Plantae rariores camschatcenses. Uppsala. (Reprinted in Amoen. Acad. 2: 332-364. PL 1751.) Little, E. L., Jr. 1971. Atlas of United States Trees. Volume 1. Conifers and Important Hardwoods. U. S. D. A. Forest Serv. Misc. Publ. 1146. Nicely, K. A. 1965. A monographic study of Calycanthaceae. Castanea 30: 38-81. Ornduff, R. 1970, The systematics and breeding system of Gehemium Loganiaceae. Jour. Arnold Arbor. 51: 1-17. Platt, R. B. 1951. An ecological study of the Mid-Appalachian shale barrens and of the plants endemic to them. Ecol. Monogr. 21: 269-300. Raven, P. H. 1963. Amphitropical relationships in the floras of North and South America. Quart. Rev. Biol. 38: 151-177. Rollins, R. C. & E. A. Shaw. The genus Lesquerella (Cruciferae) in North America. Cam- bridge, Mass. (In press.) Rune, O. 1954. Notes on the flora of the Gaspe Peninsula. Svenslc Bot. Tidslcr. 48: 117- 136. Russell, N. H. 1956. Regional variation patterns in the stemless white violets. Amer, Midi. Naturalist 56: 491-503, ScHOFiELD, W. B. 1969. Phytogeography of northwestern North America: Bryophytes and vascular plants. Madrono 20: 155-207. Shaw, H. K. Arav. 1940. Studies in the Ericales. IV. Classification of the Asiatic species of Gaultheria. Bull. Misc. Inform. Roy. Bot. Card. Kew 1940: 306-330. Smith, A. C. 1947. The families Illiciaceae and Schisandraceae. Sargentia 7: 1-224. Stern, K. R. 1961. Revision oi Dicentra ( Fumariaceae ) . Brittonia 13: 1-57. Teppner, H. 1968. Zur Kenntnis der Gattung Waldsteinia. Graz, Austria. Thunberg, C. p. 1784. Flora laponica sistens plantas Insularum laponicarum. Lipsiae. Tralatt, H. 1959. Extinct aquatic plants of Europe. Bot. Not. 112: 385-406. 1962. Die Spattertiaren Fagus-Arten Europas (The Late-Tertiary Fagus species of Europe). Bot, Not. 115: 147-176. 1963. Asiatic dicotyledonous affinities in the Cainozoic Flora of Europe. Kungl. Svenska Vetenskapsakad. Handl. IV. 9(3): 1-87. 5 pis. Wolfe, J. A. 1969. Neogene floristic and vegetational history of the Pacific Northwest. Madroiio20: 83-110. Wood, C. E., Jr. 1971. Some floristic relationships between the Southern Appalachians and Western North America. Pp. 331-404, in P. C. Holt (editor), "The distributional history of the biota of the southern Appalachians. Part II. Flora. Virginia Polytech. Inst. State Univ. Res. Div. Monogr. 2. Wynne, F. E. 1944. Drosera in eastern North America. Bull. Torrey Bot. Club 71: 166- 174. APPROACHES TO DISJUNCT POPULATIONS: THE CONTRIBUTION OF PALYNOLOGY Donald R. Whitehead^ Abstract An examination of the palynological (and other paleoecological ) literature provides interesting perspectives on a number of North American disjunction problems, A great body of paleoecological data demonstrates that significant southward displacement of biota took place during glacial maxima. Thus necessary conditions for a southward movement of "northern" disjuncts can be verified. In a few cases we have direct evidence of the disjuncts being dis- placed into or beyond regions where they now occur as isolated populations. In other cases we have indirect evidence in the form of modern associates of the disjuncts. The data suggest the probability of arrival of the disjuncts during the full-glacial and late-glacial with survival until the present. The same body of information suggests tliat components of the mixed-mesophytic forest (often considered as a disjunct community) were displaced some distance south and that the present community is a function of reimmigration and reassociation during the late-glacial and postglacial. The relict populations in the Gulf of St. Lawrence are probably not glacial survivors; some probably dispersed eastward in a discontinuous zone of instability south of the ice margin during full-glacial and late-glacial time. Such an eastward dispersal is known for many vertebrates. The evidence suggests that the prairie disjuncts could have arrived either during the late-glacial when boreal woodland and/ or tundra bordered the retreating ice or during the hypsithermal when there was a pronounced eastward movement of the prairie-forest border (at least in the upper Midwest (Minnesota)). Many explanations are possible for the Adantic Coastal Plain disjuncts, some of which would have found favorable sites for inland dispersal during full-glacial times, others of which, notably the strand and dune plants, may have dispersed inland along the shores of the Champlain sea during the early late-glacial. Paleoecological approaches are clearly essential to illuminate the historical environmental context of the disjunction, but must be used in conjunction with other disciplines to provide definitive insights. Discontinuous distributions of species and communities have long fascinated both ecologists and evolutionary biologists. This fascination has led to an enor- mous literature and much speculation. A multitude of intriguing questions can be posed concerning such disjunct occurrences; for example: Do the disjuncts repre- sent relict populations (or communities)? If so, at what time and under what condi- tions did the disjunction develop? Have the disjuncts always been where we find them? How similar are the disjuncts (genetically and otherwise) to the presumed parent populations? Alternatively, are the disjuncts instead indicative of relatively recent dispersal from some distant source? If so, when and under what conditions did they arrive? What changes have taken place in the populations (or commu- nities ) since arrival? It is readily apparent that a great many factors may have contributed to the development and maintenance of any disjunction. It is equally apparent that an adequate understanding of such biogeographic phenomena will require a broadly interdisciplinary approach. For example, it is imperative that we understand the structure and historical evolution of the environmental "stage" on which this bio- geographic "play" has been enacted. Such an understanding can be derived, at ^ Division of Biological Sciences, Indiana University, Bloomington, Indiana 17401 Ann. Missouri Hot. Card. 59: 125-137. 1972, 126 THE [Vol. 59 least in part, from application of a variety of paleoccological techniques, notably palynology. In this manner a number of the questions posed above can be dealt with a little less speculatively. Similarly, many of the insights of mathematical ecology will be essential to develop an understanding of the dynamics of disjunction. For example, can disjuncts be thought of as occupying "islands" within a complex environmental mosaic? ( See Levins, 1968. ) Can we think of disjunct populations in equilibrium terms — as involving an interaction between definable immigration and extinction rates (e.g. MacArthur & Wilson, 1967; Wilson, 1969; Whitehead & Jones, 1969). Much potentially valuable information could derive from such approaches to dis- continuous distributions. In this paper I will limit myself to the potentialities of paleoccological investi- gations and consider the extent to which the information available from such studies can enhance our understanding of several intriguing eastern North Amer- ican disjunction questions. The issues we will approach include: (1) isolated populations of northern or boreal species south of their region of continuous distribution; (2) "relict" populations in the region of the Gulf of St. Lawrence; (3) occurrence of prairie species east of the forest-prairie boundary; (4) presence of "coastal plain'* species in the interior; and (5) the mixed-mesophytic forest as a disjunct community. ^ of southeastern North Carolina. Ecolog>'36: 762-763. Gleason, H. a. 1952. Illustrated Flora of the Northeastern United States and Adjacent Canada. Bronx, New York. Gruger, E. The development of the vegetation of southern Illinois since late Illinoian time (preliminary report). Rev. Geogr. Phys. Geol. Dynam. 12: 143-148. GuiLDAY, J. E., P. S. Martin & A. D. McCrady. 1964, New Paris No. 4; a late Pleistocene cave deposit in Bedford County, Pennsylvania. Natl. Speleol. Soc. Bull. 26: 121-194. Harbin, J. W. & A. W. Cooper. 1967. Mountain misjuncts in the eastern Piedmont of North Carolina. Jour. Elisha Mitchell Sci. Soc. 83: 139-150. Harrison, W., R, J. Malloy, G. A. Rusnak & J. Terasmae. 1965. Possible late-Pleistocene uplift, Chesapeake Bay entrance. Jour. Geol. 73: 201-229. Levins, R. 1968. Evolution in changing environments. Princeton. Lichti-Federovich, S. & J. C. Ritchie, 1968. Recent pollen assemblages from the western interior of Canada. Rev. Paleobot. Palynol. 7: 297-344. Livingstone, D. A. & B. G. R, Ltvingstone. 1958. Late glacial and post-glacial vegetation from Gillis Lake in Richmond County, Cape Breton Island, Nova Scotia. Amer. Jour. Sci. 256: 341-359. McAndrews, J. H. 1966. Postglacial history of prairie, savanna, and forest in northwestern Minnesota. Mem. Torrey Bot. Club 22(2). MacArthur, R. H. & E. O. Wilson. 1967. The Theory of Island Biogeography. Princeton. McLaughlin, W. T. 1932. Atlantic coastal plain plants in the sand barrens of northwestern Wisconsin. Ecol. Monogr. 2:335-383. Martin, P. S. 1958. Taiga-timdra and the full-glacial period in Chester County, Pennsyl- vania. Amer. Jour. Sci. 256:470-502. & P. J. Mehringer, Jr. 1965. Pleistocene pollen analysis and biogeography of the Southwest. Pp. 433-451, in H. E. Wright, Jr. & D. G. Frey (editors), "The Quaternary of the United States." Princeton. Moldenke, H. N. 1960. Curlygrass in New York State. Rhodora 62: 294. MuNNS, E. N. 1938. The distribution of important forest trees of the United States. U. S. D. A. Misc. Publ. 287. Ogden, J. G., III. 1966. Forest history of Ohio. I. Radiocarbon dates and pollen stratigraphy of Silver Lake, Logan County, Ohio. Ohio Jour. Sci. 66: 387-400. OosTiNG, H. J. & D. W. Hess. 1956. Microclimate and a rehc stand of Tstiga canadensis in the lower Piedmont of North Carolina. Ecology 37: 28-39. Parizek, E. J. & J. F. Woodruff. 1956. The apparent absence of soil creep in the Georgia piedmont. Geol. Soc. Amer. Bull. 67: 1111-1116. & . 1957. Description and origin of stone layers in soils of the southeastern states. Jour. Geol. 65: 24-34. Peattie, D. C. 1922. The Atlantic coastal plain element in the flora of the Great Lakes. Rhodora 24: 57-70, 80-88. Ritchie, J. C. & F. K. Hare. 1971. Late-quaternary vegetation and climate near the arctic tree-line of northwestern North America (abstract). Amer. Jour. Bot. 58: 471. RosENDAHL, C. O. & J. W. MooRE. 1947. A new variety of Sedum rosea from southeastern Miimesota and additional notes on the flora of the region. Rhodora 49: 197-202. Segars, C. B., L. C. Crawford &. A. M. Harvill. 1951. The occurrence and distribution of hemlock in Alabama. Ecology 32: 149-15L SopER, J. H. & P. F. Maycock. 1963. A community of arctic-alpine plants on the east shore of Lake Superior. Canad. Jour. Bot. 41: 183-198. Watts, W. A. 1967. Late-glacial plant macrofossils from Minnesota. Pp. 89-97, in E, J. Gushing & H. E. Wright, Jr. (editors), "Quaternary Paleoecology." New Haven. 1970. The full-glacial vegetation of northwestern Georgia. Ecology 51: 17-33. Whii-ehead, D. R. 1963. "Northern" elements in the Pleistocene flora of the Southeast. Ecology 44: 403-406. 1964. Fossil pine pollen and full-glacial vegetation in southeastern North Carolina. Ecology 45: 767-776. — . 1965. Palynology and Pleistocene phytogeography of unglaciated eastern North America. Pp. 417-432, in H. E. Wright, Jr. & D. G. Frey (editors), "Tlie Quaternary of the United States." Princeton. — . 1967. Full-glacial vegetation and climate in southeastern United States. Pp. 237- 1972] WHITEHEAD—PALYNOLOGY AND DISJUNCT POPULATIONS ^37 248, in E. J, Gushing & H. E. Wright, Jr. (editors), "Quaternary Paleoecology." New Haven. — . 1968. Developmental and environmental history of tho Dismal Swamp. S. E. Sect. Geol. Soc. Amer. (abstract). — & E. S. Barghoorn. 1962, Pollen analytical investigations of Pleistocene deposits from western North Carolina and South Carolina. Ecol. Monogr. 32: 347-369. — ^ & D. R. Bentley. 1963. A post-glacial pollen diagram from southwestern Vermont, Pollen et Spores 5: 115-127. — & C, E. Jones. 1969. Small islands and tlie equilibrium theory of insular biogeog- raphy. Evohition23: 171-179. — , K. J. KtTRTz & J. G. Secrist. 1967. Late-glacial and post-glacial environmental changes in western Massachusetts, N. E. Sect. Geol. Soc. Amer. (abstract). Wilson, E. O. 1969. The species equilibrium. Pp. 38-47, in "Diversity and Stability in Ecological Systems." Brookhaven Symp. Biol. No. 22. Wright, H. E., Jr, 1968a. History of the prairie peninsula. Pp. 78-88^ in "The Quaternary of Illinois." Univ. Illinois Coll. Agric. Spec. Publ. 14. . 19G8b. The roles of pine and spruce in the forest history of Minnesota and adjacent areas. Ecology 49: 937-955. Wynne-Edwards, V. C. 1937, Isolated arctic-alpine floras in eastern North America: a discussion of their glacial and recent history. Trans. Royal Soc. Canada III, 3: 1-26. . 1939. Some factors in the isolation of rare alpine plants. Trans. Roy. Soc. Canada 111,33: 35-42. GENETICAL FEATURES OF FERNS AS CONTRASTED TO SEED PLANTS' Edward J, Klekowski, Jr. Many disjunct populations of plants are the result of the dispersal and germi- nation of a single propagule; the seed in seed plants; the spore in homosporous ferns. Although both of these propagules are generally an aspect of sexual repro- duction in these organisms, populations established from either a single spore or seed are genetically very distinct. Populations which differ genetically may be expected to exhibit different evolutionary prospects. This paper will attempt to elucidate these genetic differences and speculate upon their evolutionary pros- pects. As Figure 1 shows, an angiosperm seed is a propagule which develops after meiosis and syngamy have occurred. This propagule contains the results of double fertilization; endosperm tissue and an embryo. Upon seed germination this embryo will develop into a mature sporophyte. The genotype of this sporophyte does not reflect its current isolation but rather is related to the breeding system of the parental sporophyte population. Thus it may be homozygous or very hetero- zygous. In contrast, an homosporous fern spore is a propagule that develops after meiosis but prior to syngamy. In this case only a single haploid cell is dispersed and upon germination undergoes a series of mitotic cell divisions leading to the development of a haploid, generally autotrophic plant, the gametophyte. This organism is functionally hermaphroditic, simultaneously developing male and female gametangia at some point during its life. Self-fertilization ( intragameto- phytic selfing) results in a single diploid cell, the zygote, which is completely homozygous. This homozygous cell undergoes mitotic cell divisions leading to the > and eventuallv a mature sooronhvte. Thus the estab- mbry gi in a completely homozygous individual, whereas in a seed plant the sporophyte typ the ferns might predictably be very different. The angiosperm sporophyte, if it is perennial and dioecious, may live long enough to encounter a sporophyte of the opposite sex, outcross, and form progeny. These resulting progeny in all likelihood would be heterozygous to some degree. ^ This work was supported by National Science Foundation GB-8721. I should like to extend my thanks to Mr. Steve Garanln for technical assistance; to Dr. Roger Perry of the Charles Darwin Research Station for the collections of Galapagos Pteriditim; to Dr. R. M. Lloyd for the collections of Hawaiian Pteridium, I should especially like to extend my thanks to Dr. R, M. Try on and Mr. Leslie Hickok for many useful and stimulating discussions, ^ Department of Botany, University of Massachusetts, Amherst, Massachusetts 01002. Ann, Missouri Bot. Card. 59: 138-151. 1972. 1972] KLEKOWSKI— GENETICAL FEATURES OF FERNS AND SEED PLANTS 139 SEED PLANT HOMOSPOROUS FERN C2n) 2 tically disposed even before the development of forces leading to continental drift, or else these betalain-containing families would not so often dominate such habitats on these widely separated continents today. This is to say nothing of the obviously xerophytic families Fouquieriaceae, Koeberliniaceae, Krameriaceae, Welwitsch- iaceae, etc., which are desert groups without close peripheral mesophytic relatives. That is, these families have probably been a feature of tlie deserts in which they occur for many millions of years, or else one must assume that nature is selectively capricious in her destruction of the presumably more stable, mesic relatives from which they have had to arise; lacking this, one must assume a degree of adaptational divergence from some once closely related ancestor, the like of which baffles the imagination. 258 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 generic or familial level recognition will depend heavily upon amino acid sequence studies of proteins or else must await the development of more sophisticated methods for the study of nucleotide sequences in DNA. Until such a time, morpho- logical features will probably continue to serve as the primary criteria for their recognition. At the species level or below, the establishment of disjunct populations as auto junctional will probably rely quite heavily upon microchemical data, for it is difficult to believe that purely morphological data might do more than provide a model against which to test the chemical data. Of the various microchemical components which have been used for systematic purposes, the most common have been flavonoids and volatile oils. The former are relatively easy to isolate and identify, but since they are usually detected on a presence or absence basis only, they are not particularly useful in the study of disjuncts, especially in those cases which appear to be auto junctional. Ideally, chemical characters used in the study of putative autojuncts should be 1 ) rapidly and easily surveyed for, preferably by tests on a single individual or plant part; 2) readily quantified, preferably by automatic methods; 3) under genetic control so that they might respond to adaptational changes under selective influences. At present, the group of compounds which best fills all three criteria are those which are easily identified by gas liquid chromatography (GLC), for example, terpenes. In fact, using combined GLC-mass spectroscopy and appropriate com- puter methods, it is possible to identify and quantitate up to 100 volatile com- ponents within a 60 minute period. All of this is done by the hardware, of course, leaving little, if anything, to procedural bias, except for the collection of the specimens from which the compounds are obtained, that itself being a kind of intuitive bias since one cannot observe chemical constituents in the field. Further, the information, once assembled, lends itself to appropriate algorithms which purport to do for the chemical data what the taxonomist does for the mor- phological, with the exception that use of the former permits the objective (numerical) presentation of data along with statistical tests as to significance (Flake, von Rudloff & Turner, 1969). This is not generally true for the morpho- logical data, and as indicated above, this cannot be done with chemical compounds such as flavonoids where quantitation is not so readily obtained. In my consideration of the examples that follow, I would like to emphasize that their selection has been due primarily to their hueristic value in presenting the potential of such techniques, or else they have been selected for their enigmatic nature, inquiry into which must require ultimately some knowledge of the volatile constituents of the plants themselves. Autojunct populations in Picea glauca: Wilkinson et at (1971) have studied in considerable detail the monoterpene composition from 16 localities of Ficea glauca (White Spruce) in North America (Fig. 1). While the study was confined to tlie analysis of only nine monoterpenes from ten trees from any one locality, it is remarkable in that all of the trees sampled were taken from nature and grown 1972] TURNER — CHEMOSYSTEMATIC DATA IN THE STUDY OF DISJUNCTIONS 159 Figure 1. Aggregation contours of populations of white spruce (Picea glotica) using weighted chemical characters (Wilkinson et ah, 1971). The disjunct populations 30 and 28 clustered at a level below those for which contours are drawn and are discussed in more detail in the text. in an experimental garden so that environmental factors affecting tlie variation were presumably minimal. The study also pertains to this symposium, for it illustrates nicely the potential of chemical approaches to problems involving disjunction. Picea glauca exists in relatively extensive populations from northwestern North America to Alaska, south to the Great Lakes region of northern Minnesota and Illinois. There are^ however, at least two well-defined disjunct populations, one in central Montana and the other in southwestern South Dakota. The latter, in particular, is isolated from the main populational mass by approximately 400 miles. Using replicated data and a character-weighting algorithm developed around the assumption that those char- acters showing the least variance from population to population ought to be given more weight in cluster analysis, Wilkinson et at ( 1971 ) were able to show that the species consisted of two divergent chemical clines: one extending from the Lake States north-westward to Alaska, and the other north-eastward into Labra- dor. Furthermore, they suggested that these clines perhaps resulted from the migration of P. glauca both north-eastward and north-westward from populations to the south of the ice following glacial recession during post-Wisconsin time. With knowledge of the chemical structure of the main populational mass, it was possible to consider the status of the disjvmct populations 28 and 30 (Fig, 1). jgQ ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 It is interesting to note that these two populations entered the aggregations at a level below that for most of the more northern populations, but when the South Dakota population (28) entered the aggregation, it did so with a set of the typical "western" populations 54, 64, and 65 (B. H. Flake, personal communication). This knowledge permits several interpretations, the most reasonable being that of the authors themselves : 'The divergence of the South Dakota source could possibly be related to its isolation and perhaps unique evolutionary history in that the source is often considered to be a remnant from Pleistocene glaciation," The autojunctional population 28 is presumably a relict stand of Picea glauca whose origin goes back to a time when the main populational mass had a much more extensive distribution to the south; that is, its isolation does not appear to relate to the more northern populations, the latter presumably having dispersed from populations arranged along the front of the most recently glaciated regions. At least, there is no evidence that population 28 has been established through recent long-range dispersal from its more northern neighbors. Surprisingly, the Montana disjunct population 30 entered the aggregation at a slightly lower level than did population 28, clustering instead with a large set of "eastern" populations (44, 57, 60, 69, 76, 86, and 87)! This is anomalous in that population 30 is northwest of 28 and would be expected to cluster with the western set. However, this is believed to be readily explicable in that the Mon- tana population occurs in a region of sympatiy with Picea engelmanniiy and chemical examination of populations in this area suggests that they are involved in hybridization (Habeck & Weaver, 1969). In short, the relationship of population 30 to the more eastern population seems to be spurious, resulting from peripheral perturbations in the gene pool of P. glauca as a result of hybridization with P. engelmannii. In fact, it was shown by Wilkinson et dl. ( 1971 ) that hybridization between these two taxa should result in the kind of chemical variation actually found in the Montana population. In my opinion, it would have been difficult, if not impossible, to assemble morphological evidence bearing on the origin of these two autojunctional popula- tions vidthout the use of chemical data. Differences among the various populations of Picea glauca are simply not sufficiently distinctive to permit any sort of objective structuring of the populations using morphological characters, Hymenoxys odorata-Hymenoxys anthemoides: According to Parker (1962), these two species are thought to be disjuncts, H, odorata occurring in the desert regions of North America; H. anthemoides occurring in the drier regions of Argen- tina. The genus Hymenoxys is centered in North America where two well-defined subgenera are recognized: Hymenoxys and Tetraneuris, some workers preferring to treat them as distinct genera. Both H. odorata and H. anthemoides belong to the subgenus Hymenoxys; the somewhat larger subgenus, T etranetiris , is not repre- sented in South America, while Hymenoxys is represented by four species. The suggestion that Hymenoxys anthemoides is most closely related, on mor- phological grounds, to H. odorata is interesting because it bears upon the origin 1972] TURNER— CHEMOSYSTEMATIC DATA IN THE STUDY OF DISJUNCTIONS J61 of a whole group of interior arid land disjuncts with amphi-tropical distributions in the New World. In fact, this group of plants constitutes one of the most enig- matic assemblages of floristic disjunctions known to me, and there have been a number of hypotheses put forward to explain their occurrence (Bray, 1900; John- ston, 1940; Raven, this symposium). The case for the disjunctive nature of Hymenoxys odorata and H. antJiemoides received impetus with the preliminarily cytological findings that H. odorata was dibasic with chromosome numbers of n = 11 and 15, while H. anthemoides was monobasic with x = 15. This suggested that the South America populations were derived from those populations in North America with x = 15 or vice versa. This is about where the problem stood until about a year ago at which time Mr. Stewart Sanderson, graduate student at the University of Texas, began an intensive chemosystematic study of the subgenus Hymenoxys using terpenoid characters. His preliminary data bearing on the possible disjunctive nature of Hymenoxys anthemoides and H. odorata are quite interesting. The two species together possess over 40 volatile components, and examination of populations of both species grown in the same experimental garden has shown that only a few of these compounds are shared. In fact, by terpenoid chemistry, H, anthemoides is more closely related to some of the more restricted, mountainous species of North America than it is to H, odorata. Finally, the North American cytological races (n = 11 and 15) were clearly quite closely related by their terpenoid chem- istry (and morphology), thereby denying any likely relationship between the races of H. odorata with n = 15 and //. anthemoides with n = 15. Of course, the subgenus Hymenoxys must still be counted as a disjunctional element, but it seems clear from the chemical evidence available that the South American H, anthemoides is not particularly close to H. odorata^ and the two species could scarcely be termed an auto junctional pair. A more likely explanation for their disjunction, if the taxonomic suggestions put forth by Parker are valid, is that they are allojuncts, the South American populations having reached the arid regions of tlie Southern Hemisphere via long-range dispersal at some distant time and that subsequent divergence of the North American parental population has occurred. The reason for asserting that North America is the ultimate place of origin for Hymenoxys is that this genus and related genera are far and away best developed on that continent ( Bierner, 1971 ) . Other interpretations of these preliminary data are possible, of course, but it appears likely that a more comprehensive chemical study of the numerous species of Hymenoxys on both continents will do much to clarify the nature of these interesting disjuncts. In fact, taken together, a series of such studies on a wide spectrum of plant groups might do much to resolve the particular problem of whence came these amphi-tropical patterns and how. Indeed, such a collaborative study is already underway, and it will be most interesting to consider the impact of data assembled from these studies. Larrea divaricata-Larrea tridentata: Of the numerous disjunct taxa worthy of chemosystematic study, the genus Larrea ( Zygophyllaceae ) seems unusually qualified for detailed investigation, since it is a dominant element of the desert vegetation in both North and South America; and at least one species, L. divari- 262 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 caia^ is so similar to its southern hemisphere counterpart, L. tridentatay that some authors have considered them to be the same species. In fact, this is the best known example of an amphi-ti'opical autojunct. Since its distribution and eco- logical position in the deserts of both North America and South America are treated in detail by Solbrig (this symposium) and Hunziker et at (this sym- posium), I will confine my remarks here to the potential contribution that com- parative biochemistry might hold for the study of disjuncts. On the surface, Larrea would seem to be an "ideal" genus to study chemo- systematically. It possesses a large number of volatile compounds which are readily detected by gas chromatography. The plants occur in very large popula- tions, and a single leafy branch will suffice for distillation purposes. Thus, one can sample individual shrubs over a large region, marking the plants with perma- nent tags so that re-samples might be made, if needed. However, there are many problems, foremost of which is the quantitative variability likely to be induced in the volatile constituents by environmental factors. It has been amply demonstrated that the volatile fraction of a plant will vary significantly from month to month depending on the metabolic stage of the organism, which in turn presumably depends upon moisture and temperature conditions affecting the populational site itself. Much of this plant-to-plant vari- ation can be "smoothed out" by statistical treatment of the populational data but, except for trees growing under conditions of reduced metabolism in the field (e.g. Juniperus virginiana and presumably other conifers which show little or no growth activity during the winter months; Flake, von Rudloff & Turner, 1969), chemical analyses should come from plants grown in a uniform garden. This is particularly important with a desert shrub such as Larrea, for preliminary work on field populations of L. divaricata has shown quite extraordinary quantitative variation from individual to individual at a given site, and from population to population, presumably depending upon the amount of ground water available to the plants. Such variation might be dvie to the extraordinary sensitivity of the volatile compounds themselves, most of which are guanilides, a group of peculiar constituents unrelated to the terpenes. Since Larrea is an evergreen shrub, more or less restricted to temperate regions, it might prove feasible to sample popula- tions in the winter, much as in Juniperus, but this is an aspect of the problem which has not been investigated. Finally, I cannot help but note that three views prevail as to the origin of the more xeric amphi -tropical elements found on the North and South American conti- nents. These are: 1) that the species (and deserts) are ancient (60 million years or more) having a common origin; 2) that the floras evolved separately through parallel selective influences from the surrounding "vegetation, i.e. the floras have no historical connections as such; or that the two If the latter view (3) is held, then one would expect a very close similarity in the secondary compounds of the vaiious disjunct taxa. On the other hand, if 1972] TURNER^HEMOSYSTEMATIC DATA IN THE STUDY OF DISJUNCTIONS JgS very similar species have very different secondary compounds this might indicate a long-time isolation or different phyletic history of the taxa concerned, the morphological traits {e.g. habit and vegetative features), through parallel selective influences, remaining at least superficially similar; the secondary compounds, inasmuch as they are presumably under weak selective pressures probably reacted differently to the different parasitic and grazing fauna of the South American region. In other words the secondary compounds, being under quite different selective forces, should have diverged considerably from their disjunct counterpai^ts in spite of the quite similar climatic regimes in which they presently occur (Turner, 1969). As yet we do not know the origin of the apparently autojunct Larrea dwaricata. Hunziker (personal communication, this symposium) interprets his preliminary protein-band data as suggestive of a South American origin for the group as a whole. I am more cautious, for while I recognize the better development of Larrea (as to number of species) in South America, I am also aware that the North American element is made up of diploid, tetraploid, and hexaploid popu- lations, all of which show considerable variation, indicative of a species with some considerable tenure in the North American deserts. I suggest, therefore, that it is equally likely that L. divaricata developed as a diploid population in North Amer- ica millions of years ago, and that subsequently it was established in South Amer- ica through long-range dispersal. The fact that there are two other species of Larrea in South America implies little, for L. divaricata may be only remotely related to these. Of course, it is also possible that the genus itself, and L, divaricata specifically, has existed since early Tertiary time, such as suggested by Bray (1900) and Johnston (1940). But this would seem to raise serious problems as to the origin of the desert areas themselves, for it seems unlikely that the desert areas were ever connected such that Larrea divaricata achieved its disjunct status from a once intact gene pool with only subsequent physical events responsible for its isolation. Or does it? I tend to believe that the disjuncts making up the various amphi-tropical elements have a multiplicity of origins. They may be allojunctional, as appears to be the case with Hymenoxys odorata-H. anthemoides, or they may be auto- junctional, as appears to be the case with Larrea divaricata. Further, I tend to believe that the desert areas on the two continents developed early enough (at least beginning with Miocene) to have permitted the establishment of many kinds of disjuncts. Some of these are relatively old, and some relatively new, but pre- sumably all have passed over the tropics as long-distance dispcrsants in one direc- tion or the other. Clearly, more data of every sort will be needed before these abominable amphi-tropical disjunctions can be explained or understood. No doubt compara- tive chemistry will make up an important part of this information. Literature Cited Alston, R. E, 1967. Biochemical Systematics. Evol. Biol. 1: 197-305. — & B. L. Turner. 1963. Biochemical Systematics. Trenton, New Jersey. AxELROD, D. I. 1950. Evolution of desert vegetation. Publ. Carnegie Inst, Washington 590 215-306. 254 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Bray, W. L. 1900. The relation of the North American flora to that of South America. Science 12: 709-716. Behnke, H.-D. & B. L. Turner. 1971. On specific sieve-tube plastids in Caryophyllales, Taxon20: 731-737. BiERXER, M. 1971. A chemosystematic and cytotaxonomic study of Helenium, sect. Tetrodus (Compositae). Doctoral Dissertation, The University of Texas, Austin. Cain, S. A. 1944. Foundations of Plant Geography. New York. Croizat, L. 1952. Manual of Phyto-Geography. The Hague. Flake, R. H., E. von Rudloff & B. L. Turner. 1969. Quantitative study of clinal variation in Juniperus virginia7m using terpenoid data. Proc, Natl. Acad. Sci. (U. S.) 64: 487-494. Good, R. 1953. The Geography of the Flowering Plants. New York. Habeck, J. R. & T. W. Weaver. 1969. A chemosystematic analysis of some hybrid spruce (Picea) populations in Montana. Canad. Jour. Bot. 47: 1565-1570. Johnston, I, M. 1940. The floristic significance of shrubs common to North and South American deserts. Jour. Arnold Arbor. 21: 356-363. Love, A. 1954. Cytotaxonomical evaluation of corresponding taxa. Vegetatio 5: 212-224. Marshall, D. R. & R. W. Allabd. 1969. Genetic polymorphism in natural populations of Avena fatua and A. harhata. Nature 221: 276-278. McElhixny, M. W. & G. R. Luck. 1970. Paleomagnetism and Gondwanaland. Science 168: 830^832. Parker, K. F. 1962. The South American species of Hymenoxys (Compositae). Leafl. W. Bot. 9: 197-209. ScoGiN, R. L. 1968. Isoenzyme polymorphism in selected enzymes in natural populations of the genus Baptisia ( Leguminosae ) . Doctoral Dissertation, The University of Texas, Austin. Setchell, W. a, 1935. Pacific insular floras and Pacific paleogeography. Amer. Naturalist . 69: 289-310. Turner, B. L. 1967. Plant chemosystematics and phylogeny. Jour. Pure Appl. Chem. 14: 189-213. . 1969. Chemosystematics: recent developments. Taxon 18: 134-151. . 1971, Molecular approaches to populational problems at the infraspecific level. In "Phytochemical Phylogeny," London. Vierhapper, F. 1919. Uber echten und falschen Vikarismus. Oesterr. Bot. Zeitschr. 68: 1-22. Wilkinson, R. C., J. W. Hanover, J. W. Wright & R. H. Flake. 1971. Genetic variation in the monoterpene composition of white spruce. Forest Sci. 17: 83-90. WoHLPART, A. & T. J. Mabry. 1968. The distribution and phylogenetic significance of the betalains with respect to the Centrospermae. Taxon 17: 148-152. DISJUNCTIVE DISTRIBUTIONS IN THE LICHEN-FORMING FUNCr William Louis Culbkrson^ There is a widespread notion among botanists that the nonvascular cryptogams are easily disseminated, ubiquitous plants without well defined geographic ranges. ■ Certainly for many groups, such as the fresh-water algae and the nonpathogenic or nonsubstrate-specific microfungi, this view seems to be justified, for few meaning- ful geographic-to-taxonomic correlations among these organisms have been perceived. In other groups, for example the higher Basidiomycetes, our ignorance of phytogeographically significant ranges may reflect only the magnitude of the problems inherent in the study of organisms known only from ephemeral sporo- carps. There are, however, two large groups of cryptogams in which species and taxa of higher than specific rank show well marked geographic distributions, in- cluding disjunctive ones, like those known in the vascular plants. These are the bryophytes and the lichens, the only major groups of terrestrial nonvascular cryp- togams in which the vegetative plant body is exposed and perennating. The aim of this paper is to examine disjunction as it is known with regard to the second group, the lichen-forming fungi. Reliable distribution maps have been published for few ( only hundreds ) of the approximately 20,000 recognized species of lichen fungi. Most of the best sub- stantiated distributions are for conspicuous foliose or fruticose species from temperate regions of the northern hemisphere. To underline the point that lichen fungi may have geographic ranges comparable to those of vascular plants, Table 1 gives eight pairs of common, locally abundant, native North American plants, in each case a lichen fungus and a tree, for which the total geographic ranges are almost congruent or very highly comparable. It should be pointed out that al- though all but one of the lichen examples are epiphytes, none is ecologically restricted to the bark of the ti*ee species to which its range is compared. Many more such examples could be given from the flora of North America and the floras of Europe and the Far East. So similar in fact are the well documented ranges of lichens to the ranges of vascular plants that the conclusion that both result from the same physioecological and historical factors is inescapable. Types of Disjunctive Distributions in Lichen Fungi Most plant sj at continuity i junctions that typ modern ranges are on different continents. Disjunction could as well be discussed in a more restricted framework for the examples of North American and European ^This paper was prepared under the tenure of National Science Foundation grant GB 8359. ^ Department of Botany, Duke University, Durham, North Carolina 27706. Ann. Missouri Bot. Card. 59: 165-173. 1972. 166 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Table 1. Eight pairs of common, native North American species, a lichen and a tree, in which the total geographic ranges are nearly congruent or highly similar. [Ranges of trees from Little ( 1971 ); ranges of lichens from various authors.] Lichen Species Actinogyra muehlenbergii (Ach.) Schol. Anzia colpodes (Ach.) Stizenb. Dermatocarpon iuckermanii (Rav.) Zahlbr. Parmelia caroliniana Nyl. Parmelia livida Tayl. Physcia millegrana Degel. Platismatia herrei (Imsh.) Culb. & Culb. Pseudevernia cladonia ( Tuck. ) Hale & Culb Tree Species Tsuga canadensis (L.) Carr. Qiiercus paltistris Muenchh. Carya ovata ( Mill. ) K. Koch Diospyros virginianu L. Cormis florida L. Juniperus virginiana L. Picea sitchensis ( Bong. ) Carr, Picea rubens Sarg. lichen species disjunctive in their own native continent are numerous. I have selected my examples of disjunctive ranges from those species with distributions including the New World in the hope that comparisons to species of other plant groups discussed in this symposium may become apparent. I have attempted to minimize the inaccuracies that accompany insensitive taxonomies by restricting my examples to taxa 1) that are well-known systematically (even if their taxo- nomic rank is in dispute), 2) that have been recently studied, 3) that are common in at least part of their range, and 4) that are large enough to be conspicuous (and consequently often collected). Some of my examples may not satisfy all these criteria, but most of them do. 1. Widesp M extremely broad ranges and are represented on most continents. This is often pointed to as evidence that evolution in the lichens is ( or has become ) very slow. The following species occur on at least fovir continents: Cladonia merochlorophaea Asah. Peltigera polydactijla (Neck.) Hoffm. Heterodermialeucomelaena (L.) Poelt Sphaerophorus melanocarpus (Sw.) Menegazziaterebrafa (Hoffm.) Mass. N ormandina pulchella (Borr.) Nyl. Parmelia crinita Ach. DC. Thamnolia vermicularis ( Sw. ) Ach. ex Schaer. There are also large numbers of pantropical species, for example: Baeomyces absolutus Tuck. Cladonia aggregata (Sw.) Ach. Cladonia balfourii Del. Dirinaria aegialita (Ach.) B. Moore Heterodermia tremulans (Miill. Arg.) Culb Graphis afzelii Ach. Parmelia cristifera Tayl. Parmelia tinctorum Del. ex Nyl, two must find an appropriate algal partner reproduce primarily by asexual propagules — soredia or isidia, structures that hyphal source of "instant" lichens. the duction of asexual propagules is doubtless more common than among the lichens 1972] CULBERSON — DISJUNCTIONS IN LICHEN-FORMING FUNGI 167 as a whole. The recently monographed Parmelia subgen. Amphigymnia (Hale, 1965) provides data that may well prove to be typical for many widely distributed genera of lichen fungi: There are 26 pantropical species of Parmelia subgen. Amphigijmnia of which only 5 (19%) lack sorcdia or isidia. Although circum- stantial evidence confirms the great importance of soredia and isidia as effective propagules, direct observation of the dissemination of these bodies is infrequent. It has been shown experimentally that soredia arc indeed detached and carried by wind (Bailey, 1966). Soredia have also been recovered from samples of air-borne plant fragments (Rudolph, 1970), but most soredia in samplings from the air are probably recorded as algae and not recognized for what they are. In the category of widespread disjunctions one might include very large numbers of lichen species with enigmatic disjunctive ranges — for example Par- melia Are such distri- butions the result of long-distance dispersal or are they the relicts of earlier and broader ranges? Definitive explanations of such ranges are simply not possible in groups such as the lichen fungi that lack a fossil record. 2, Circumboreal disjunctions. — The following arctic and north-temperate species have circumboreal, disjunctive ranges: I nidulifi halei Ci Nephroma arcticum (L.) Torss. Parmeliopsis amhigua (Wulf.) I Cetrelia cetrarioides (Del. ex Duby) Peltigera horizontalis (Huds.) Baumg. Culb. & Culb. Solorina saccata (L.) Ach. Evernia mesomorpha Nyl. Umbilicaria arctica (Ach.) Nyl. Lobaria linita (Ach.) Rabenh. Xanthoria fallax (Hepp) Arn. The disjunctions of some circumboreal species are maintained by narrow amplitudes of ecologic tolerance. For example, the extremely oceanic species Platismutia norvegica (Lynge) Culb. & Culb. occurs in the oceanic parts of Scan- dinavia (main European range), in Scotland, in Newfoundland, and along the North American West Coast from Oregon to southern Alaska (main American range). However, no obvious ecological factors limit the distribution of the tundra lichen Asahinea chrysantha (Tuck.) Culb. & Culb., which occurs continu- ously over large parts of Siberia and Alaska (main range) and locally on Baffin Island and in northern Scandinavia. The main range of this species would seem to have developed from gradual spread from a center of origin with the outliers the result of long-distance dispersal. Unlike Platismatia norvegica^ Asahinea chrysanthu does not seem to be limited to a specialized environment by a restric- tive physiology. 3. Bipolar disjunctions. — ^Although some of the examples of lichens with bipolar disjunctive distributions pointed out by Du Rietz 30 years ago are perhaps to be challenged upon faulty taxonomy, the fact that this distribution type exists among the lichens is irrefutable. The best documented example is in the recently monographed genus Platismatia (Culberson & Culberson, 1968). The temperate species P. glauca (L.) Culb. & Culb. is locally very abundant in northern and western Europe, in northern North America, and in extreme southern Argentina 168 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 and Chile. The species is apparently absent from the high mountains of western South have been the origin of the antipodal population. This explanation is supported by the occurrence of P. glauca on Mt, Aberdare, Kenya, and Mt. Kilimanjaro, Tanzania, localities in east-ccnti'al Africa about 5,000 km south of the southern limit of the European range of the species. Curiously, this widespread species J very 4. Europe-North America disjunctions. — Some species occur in Europe and North America but are unknown elsewhere. Good examples are Parmelia quercina T (Willd. ) Vain, (widespread in southwestern Europe, locally common in Califor- nia), P. htjpoleucina J. Stein, (occasional in the Mediterranean Basin, common in the North American Atlantic Coastal Plain, rare in southern California), and Letharia vulpina (L.) Hue (locally common in the mountains of Central Europe, Morocco, and western North America). It is never possible to exclude long- distance dispersal as the explanation, but it is tempting to see these present ranges as the relicts of more extensive distributions before the separation of the conti- nents. 5. Eastern North America-Japan disjunctions. — The early appreciated and now classic affinities between the floras of eastern Asia and eastern North America are well represented among the lichen fungi. In the vascular plants the floristic affinities are usually revealed through vicarism, but in the lichens some species are common to both regions. The following are all lichens known to me to be restricted to the regions in question and to be so uniform that both the eastern Asiatic and the eastern North American representatives are considered to be conspecific: Anaptychia palmulata (Michx.) Vain. Lobar ia quercizans (Michx.) Vain. Parmelia aurulenta Tuck. Parmelia formosana Zahlbr, Parmelia galbina Ach. Pyxine sorediata ( Ach. ) Mont. Stereocaulon tennesseense Magn. Anziaornata (Zahlbr.) Asah. Candelaria fibrosa (Fr.) Miill. Arg. Cladonia submitis Evans Cladonia clavulifera Vain, Lobariaerosa (Eschw.) Nyl. It will be noted that all twelve of the above species are either foliose or fruticose. Surely as the crustose lichens become better known, the number of examples of this distribution type will be multiplied. A relatively large number of other species have their main distributions in Lpan and eastern North America but are also known elsewhere. Actinogyra muehlenbergii (Ach.) Schol. and Heterodermia hypoleuca (Ach.) Trev. are exam- ples, the former occurring additionally in Siberia and the latter occurring addi- tionally in Himalaya. Umbilicaria caroliniana Tuck, occurs on the highest peaks the locality Cetrelia chicitae (Culb.) Culb. & Culb. is one of the commonest lichens in the mountains of tlie eastern United States and of central Japan. It is also known 1972] CULBERSON— DISJUNCTIONS IN LICHEN-FORMING FUNGI 169 Figure 1. Comparison of tlie Cladonia flora of Japan (90 species) wilii diat in various parts of North America. From left to right the colmnns refer to the number of species present in Japan but absent from the indicated region (cross hatched), the number of species common to Japan and the indicated region (stipple), and the number of species absent from Japan but present in the indicated region (vertical lines). Redrawn from Yoshimura (1968), from a single locality in Europe ( in tlie Vosges ) and from a few mountain stations in Southeast Asia. us Cladonia is well represented in Japan there than elsewhere for having attiacted tion of perceptive chemotaxonomists in both countries. Summarizing our present knowledge of the genus, Yoshimura (1968) showed tliat the number of species common to floras in various parts of North America and to the flora of Japan varies greatly. The number of species shared is greater in Alaska, Canada, and the East and smaller in California and Mexico, underlining the classic relationsliip among the species of an entire genus ( Fig. 1 ) . The explanation of the origin of the eastern Asiatic-eastern North American vicarism is surely the same as that for the viearism seen in the vascular plants of these regions chly diverse and once continuous Tertiary tation that extended from North America through the then adjoining continent of Europe and across Asia to Japan was subsequently reduced to widely separated yrQ ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 relictual populations by wholesale extinctions. The role of glaciation in eliminat- ing from Europe the Tertiary hchen species now shared by Japan and eastern North America has been well discussed by Poelt ( 1963) , Disjunction with Differenttation : The Origin of Vicariads Divergent evolution in populations that have become disjunct for whatever reason leads to speciation if the process goes far enough. It must always be borne in mind however, that apparent vicariads might owe their origin to the extinction in the disjunct populations of reciprocal members of a once sympatric species pair. i. Vicariads in Japan (and /or eastern Asia) and eastern North America. — The most striking example of vicarism in lichen fungi comes again from the relation- ships of the eastern Asiatic and eastern North American floras. Such vicarious species may be differentiated by morphology or by chemistry of secondary natural products or by both. Morphologically differentiated: L I Anzia colpodes (Ach.) Stizenb. in eastern North America; A. colpota Vain, in Japan. Cladonia caroliniana (Schwein.) Tuck, in eastern North America; C. nip- ponica Asah. in Japan, Parmelia rudecta Ach. in eastern North America; P. ruderata Vain, in Japan. Umbilicaria mammulata (Ach.) Llano in eastern North America; U. escu- lenta (Miyoshi) Minks in Japan. Chemically (and in some also morphologically) differentiated: Cladonia evansii Abb. (atranorin and accessory usnic acid) in eastern North America; C. pseudevansii Asah. (usnic acid) in Japan. Cladonia atlantica Evans (baeomycesic and squamatic acids) in eastern North America; C. hondoensis Asah. (barbatic acid) in Japan. Cladonia cristatella Tuck, (barbatic and didymic acids) in eastern North Wi J America; C. pseudostellata Asah. ( hypothamnolic acid) in Japan and Alaska. Cladonia cylindrica (Evans) Evans (grayanic acid) in eastern North Amer- ica; C. norikurensis Asah. (merochlorophaeic acid) in Japan. Cladonia didijma (Fee) Vain, (barbatic acid) in eastern North America; C. pseudodidyma Asah. (accessory bellidiflorin ) in Japan. The title of this section implies that the divergence that produced the ten species pairs listed above took place after rather than before isolation by disjunc- tion. It could be argued, however, that all these species were present throughout the extent of the Tertiary forest and that in each pair, A and B, A became extinct in Asia and B became extinct in North America. This explanation cannot be categorically ruled out for some of the vicariad pairs. But it makes an unconvinc- ing explanation for the whole list, because one would not expect every pair of 1972] CULBERSON— DISJUNCTIONS IN LICHEN-FORMING FUNGI 171 micro- or sibling species to have followed such a pattern of reciprocal extinctions. There should be at least a few pairs of microspecies still present in both places but not found elsewhere — yet I know not a single example. 2. Ubiquitous collective species. — Disjunctions in wide-spread collective spe- cies may lead to sufficient differentiation of the isolated populations that the regional elements are recognized taxonomically, Moiphological (as opposed to chemical) differentiation can be seen in the common reindeer lichen, Cladonia rangiferina (L.) Web. subsp. rangiferina ( circumborcal ) , and its regional deriv- atives subsp. grisea Ahti (eastern Asia), var. ahbayesii Ahti (Mexico, Central America), and var. vicaria (Sant.) Ahti (Patagonia, Antarctica). Three common, closely related umbiHcate lichens appear to owe their differentiation to past disjunction: Lasallia pustulata (L.) Mer. is common in Europe and rare in north- ern North America, L. papulosa (Ach.) Llano is common in subboreal and higher- elevation temperate eastern North America, and L. asiae-orientalis Asah. represents the complex in Japan. The most enigmatic differentiations to be seen in widely distributed collective species are chemical ones. The already mentioned Parmelia hypoleucina J. Stein, occurs in the Mediterranean Basin and southern California and very abundantly on the Atlantic Coastal Plain from Cape Cod to Mississippi. But throughout the east-central section, from central Texas to southern New York and northern Georgia, the morphologically indistinguishable P. htjpotropa Nyl. abounds and is one of the commonest of all eastern North American lichens outside the Coastal Plain. Phenotypically P. hypotropa differs from P. hypoleucina in being unable to convert its most abundant secondary natural product, norstictic acid, to stictic acid as P. hypoleucina does. Although this chemical change is a minor one — a one-step methylation that could conceivably be mediated by a single enzyme controlled by a single gene — the demonstrable but small chemical difference in the two populations must be linked to factors governing prodigious differences in physiological potential. Parmelia hypotropa, the dominant North American representation of the complex, seems to have been derived from the much older and (on a world-wide scale) still more widely distributed one. An even more enigmatic case of chemical differentiation perhaps related to disjunction can be seen in the genus Thamnolia^ which is monotypic or dit}^ic depending upon the weight assigned to chemistry. In spite of being totally asexual, Thamnolia is one of the most successful of all lichen genera, occur- ring abundantly on bare soil and among herbs throughout the arctic and alpine regions of the world. Individuals belong to one or the other of two chemical types: they produce either a pair of /?-orcinoI-type para-depsides (baeomycesic and squamatic acids) or a ^-orcinol-type 7n^to-depside (thamnolic acid). The constituent(s) of a given specimen can be determined only by appropriate chem- ical analysis, and consequently the abundant representation of the genus in herbaria constitutes a world-wide random sample of the chemical variation. By testing herbarium specimens Sato (1968) demonstrated that the pam-depsides are produced exclusively or more commonly in populations in the northern hemi- sphere while the meta-depside is produced exclusively or more commonly in populations in the southern hemisphere, even though most populations in both 172 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Figure 2. The proportion of the Thamnolia vegetation in various parts of the world con- sisting of para-depside producers (stipple) [=: T. subuliformis (Ehrh. ) Culb.] and meta-depside producers (black) [= T. vermicularis (Sw.) Ach. ex Schaer.]. Redrawn from Sato (1968). hemispheres contain both chemical types (Fig. 2). Are these differences to be interpreted as coming from an allopatric differentiation in populations that have subsequently become extensively sympatric? The chemical differences involved are sufficiently complex that they would seem to reflect more than a superficial genetic modification and to require more than a simplistic explanation. Disjunction and Continental Drift Although certain Old World-New World lichen distributions may owe their origin to the ancient connection of these land masses, long-distance dispersal usually seems a more reasonable explanation. One body of information, however, can be interpreted only in relation to the now known geological events that pro- duced the Atlantic Ocean. Parmelia subgcn. Amphigtjmnia is a natural, well defined taxon of 106 species. Its botany and chemistry are well known tlianks to Hale's (1965) recent mono- graph, and the plants are so large and conspicuous that they are well represented in herbaria. Figure 3 shows the number of species that occur in various parts of the world and the number of species endemic there. It will be noted that most of the species occur in Africa (64 species: 60% of the total). South America (51; 48%), Central America and Mexico (47; 44%), and North America (25; 24%). Thirty (28%^)* of the species are endemic to the New World, 22 (21%) are endemic to Africa, and 7 species are found only in tropical America and Africa. Fewer spe- cies occur in Europe (8; 8%) and Australia-New Zealand (17; 16%), and none of these are endemic. Most of the southern Asiatic and Pacific representation is drawn from the grovip of 26 pantropical species. Significantly 21 ( 81%> ) of the pantropical 1972] CULBERSON— DISJUNCTIONS IN LlCHEN-FOR\[ING FUNGI 173 Figure 3. The numbers of species of Parmelia subgen. Amphigymnia in various parts of the world (larger-size numbers) and the number of endemic (£) species in each region (smaller-size numbers). Seven species are found only in Africa and South America. Data taken from Hale (1965). species produce asexual propagules. Parmelia subgen. Amphigymnia must have been well differentiated before the New World separated from the Old some 100 million (?) years ago and its present disjunctive distribution is due primarily to historical events and secondarily to long-distance dispersal. Literature Cited Many publications, too numerous to he cited here, were used in documenting die various ranges used as examples in the text. Bailey, R. H. 1966, Studies on the dispersal of lichen soredia. Jour. Linn. Soc, Bot. 59: 479-^90. Culberson, W. L. & C. F. Culberson. 1968. The lichen genera Cetrclia and Flatismatia (Parmeliaceae). Contr. U. S. Natl. Herb. 34: 449-558. Hale, M. E., Jr. 1965. A monograph of Parmelia subgenus Amphigymnia, Contr. U. S. Natl Herb. 36: 193-358. Little, E. L., Jr. 1971. Atlas of United States Trees. Volume 1. Conifers and Important Hardwoods. U. S. Department of Agriculture, Miscellaneous Publication No. 1146. Washington, D. C. PoELT, J. 1963. Flechtenflora und Eiszeit in Europa. Phyton, Ann. Rei Bot. 10: 206-215. Rudolph, E. 1970. Local dissemination of plant propagules in Antarctica. Pp. 812-817 in M, W. Holdgate (editor), "Antarctic Ecology^, Volume 2," London and New York, Sato, M. 1968. The mixture ratio of the lichen genus Thamndlia in Tasmania and New Guinea. Jour. Jap. Bot. 43: 328-334, YosHiMURA, L 1968, The phytogeographical relationships between the Japanese and North American species of Cladonia. Jour. Hattori Bot. Lab. 31: 227-246. DISJUNCTIONS IN BRYOPHYTES W. B. SCHOFIELD^ AND H. A. CrUM* Abstract In spite of the more general distribution of many bryophytes, dramatic disjunctions exist, many of them similar to those shown by vascular plants. Various explanations have been offered to explain these disjunctions including continental drift, long-distance dispersal, and the fragmentation of a once more continuous distribution. No single hypothesis is sufficient to accommodate all species within any disjunctive pattern. The most serious difficulty is tlie inadequacy of exploration of considerable areas of the globe. Various bryophyte disjunctions are mapped in 51 distribution maps, and details of sexual patterns and dispersal mechanisms are assessed. Most bryophytes are widely distributed. In the Northern Hemisphere more than 60% of the flora of arctic and boreal regions is made up of the same species. Within this wide range, however, each species has highly specific requirements and some are exceedingly local. Becavise bryophytes have air-borne diaspores their means of dissemination would appear to guarantee a wide distribution of all species. That disjunctions exist at all would seem somewhat anomalous, yet such disjunctions do exist, some of them very dramatic. The explanation of these disjunctions has led to numerous intriguing hypotheses, many of which have been derived from similar studies of flowering plant disjunctions. In North America tlie disjunctions that have received the greatest attention are eastern American-East Asian disjuncts (Iwatsuki, 1958a, 1958fe, 1958c; Iwat- suki & Sharp, 1967, 1968; Sharp & Iwatsuki, 1965. See Maps 1-7); tropical and subtropical taxa in the Southern Appalachians (Anderson, 1951; Andrews, 1920; Billings & Anderson, 1966; Crum, 1966; Sharp, 1936, 1938, 1939, 1941. See Maps 8-9); amphi-Pacific taxa (Ando, 1966; Ando, Persson & Sherrard, 1957; Crum, 1965; Hattori, 1952, 1963, 1966b; Hattori & Sharp, 1968; Hattori et al, 1968; Hori- kawa & Ando, 1957; Lazarenko, 1957; Noguchi & Saito, 1970; Persson, 1946a, 1946b, 1947, 1949, 1952, 1958, 1962, 1968; Persson & GjaervoU, 1957; Schofield, 1962, 1965, 1966a, 1966b, 1966c, 1968a, 1968b, 1969; Sharp & Hattori, 1967; Steere, 1969; Steere & Schofield, 1956; Steere & Schuster, 1960. See Maps 10-20); bipolar disjuncts (Martin, 1946, 1949, 1952a, 1952b; Du Rietz, 1940; Sainsbury, 1942; Schuster, 1969. See Maps 21-22); taxa disjunctive between Europe and western North America (Abramova & Dildarin, 1969; Evans, 1914; Harvill, 1950; Haynes, 1915; Koch, 1956; Paton, 1966; Schofield, 1968a, 1968b, 1969; Whitehouse, 1961, 1963. See Maps 23-^3); amphi-Adantic taxa (Andrews, 1938, 1961; Maass, 1965, 1966a, 1966b; Schuster, 1962; Sharp, 1941. See Maps 34-37); and arctic-alpine taxa, that have received surprisingly littie attention (Gams, 1955; Schuster, 1958a, 1958b; Steere, 1937, 1938, 1953, 1965. See Maps 38-40), ^ Department of Botany, University of British Columbia, Vancouver 8, British Columbia, Canada. ^The Herbarium, University of Michigan, Ann Arbor, Michigan 48104. Ann. Missouri Box. Card. 59: 174-202. 1972. 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 275 Irmsclier (1929) studied moss disjunctions particularly in their contribution to an insight reflecting plant distributions as influenced by continental drift. Herzog (1926) in his monumental Geographie der Moose treats bryophytc disjunctions briefly. Details from other areas are treated by Abramova & Abramov (1969), Bowers (1970), Croizat (1962, 1966), Crum (1966), Crum & Anderson (1964), Fulford (1951), Greig-Smith (1950), Horikawa (1955), Koch (1954), Lazarenko (1957), MiiUer (1916, 1954), Schelpe (1969), Schofield (1969), and Schuster ( 1967, 1968, 1969 ) . Hypotheses explaining these disjinictions have been based mainly on historical factors. In most cases the opinion is that these various disjunct populations repre- sent remnants of a more continuous distribution of the past, always placed preceding the Pleistocene glaciations and usually suggested to be as early as the Tertiary. Information from fossil bryophytc material is generally sparse; thus the assumption is made that bryophytes were associated with a vascular vegetation which exhibited a more continuous distribution during the Tertiary. One bryophytc showing disjunctive European-western North American distri- bution is Claopodium tvhippleanum (Noguchi, 1952. See Map 33). The species is also present in the Hawaiian Islands. Abramova and Abramov (1969) indicate that they discovered it as fossil material from Middle Pliocene deposits in the Caucasus region. In Europe the species is pi'csently restricted to a number of localities in Portugal and Spain. Another species of similar distribution, but with the Caucasus showing the only extant European population, is Crumia latifolia (see Map 26). This is discussed by Abramova and Dildarin (1969). These distributions strongly suggest a more continuous distribution of the past. Remarkable features of a high proportion of disjunctive species is that they are dioicous and in many cases produce no specialized propagula that would make them readily disseminated. This complete reliance on asexual reproduction by simple fragmentation leads to a paucity of biotypcs. These combined features mean that the disjuncts tend to be, within their disjunctive areas, highly localized as relatively small populations with very specific environmental requirements. In most cases they are very poor competitors with the more generally distributed flora and are often confined to sites that are continually undergoing ecesis (Lye, 1967). Areas particularly rich in disjunctive taxa often have hyperoceanic climates. This subject has been treated thoroughly by Ratcliffe (1968) and St0rmer (1969) in particular, but also discussed by Amann (1929), Billings and Anderson (1966), Courtejaire (1962), Gaume (1952-1954), Ivvatsuki (195Sa), Lye (1967), Nichol- son (1930), Persson (1949), Schuster (1962), and Touffet (1964). The areas richest in disjunctive species confined largely to hyperoceanic cli- mates are: Britain and Norway, the Faeroes, the Alps, the Himalayas, high mountains of Japan and Taiwan, North Pacific North America, and to a certain degree, high mountain slopes in the Hawaiian Islands. In some cases identical disjunctive species are found in many of these widely separated areas, for example: Geheebia gigantea (Map 41), MastigopJiora woodsii (Map 42), Scapania omitho- podioides (Map 43), Anastrepta orcadensis (Map 44), Aruisirophtjllum donianum (Map 45), Pleurozia purpurea (Map 46), Bazzania pearsonii, Cephcdoziella pear- 176 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 sonii, Campylopus atrovireiis, and others. Species of the East Asian-eastern American disjunction, the European-western American disjunction and amphi- Pacific disjunction are also largely of oceanic climates. In the Southern Hemisphere disjunctions are also richly represented in hyper- oceanic climates: the southern Australasian-southern South American disjunc- tion is the most pertinent example, shown in Maps 47^8 (see especially Schuster, 1969). Disjuncts of more arid climates are equally dramatic: thus the isolation in Australasia and South Africa represented by Carrpos sphaerocarpos (Map 49) of arid salt pans and Pottia maritima (Map 50) of sandy sites. These species, how- ever, are either very rare or are overlooked because of their inconspicuousness. As Schelpe (1969) has shown, the number of bryophytes of this disjunctive pattern may increase as both areas become better collected. Species of mediterranean climates exemplify yet another type of disjunction, Neckera menziesii, Antitrichia calif ornica (Map 32), Funaria muehlenbergii, and Bartramia stricta are particularly good examples. The Neckera is largely mediter- ranean in Eurasia and North Africa but in western North America is both in med- iterranean and more humid climates, but in the more humid climates is confined largely to edaphically dry calcareous rock; it is rare in Japan. Antitrichia shows a similar pattern but is absent from Japan. As has been noted (Sharp, 1938; Schornherst, 1943), many bryophytes of tropical affinity disjunctive in more northern areas tend to be calcicoles. Consider- ing the Florida moss flora, Schornherst (1943) suggests that this may be the result of the frequency of this habitat in tlic tropics, thus the selection favoring bryophytes of this specificity. In spore-producing plants there is an obvious temptation to state that long- distance dispersal is especially important. Spore size enhances wind-dispersal and air transport of spores is necessary even in short-distance dissemination of most spore-bearing plants. In bryophytes it is apparent that long-distance dispersal is not only possible, but in some cases probable. Certainly the taxa of volcanic oceanic islands reached their present sites via long-distance dispersal. Although published analyses of the Hawaiian bryoflora are lacking, this archipelago could serve as a particvilarly important source of information concerning long-distance dispersal in bryophytes. The archipelago is relatively youthful, emerging no earlier than the Late Tertiary, thus has been available for colonization for approximately ten million years. A number of disjunctive bryophytes are of particular interest: Cyrtopus setosus (Map 48) is otherwise a species of the Southern Hemisphere. Dixon (1922) has established the authenticity of the Hawaiian collection of the species, but it has not been recollected recently. The hyperoceanic taxa Scapania ornithopodioides (Map 43) and Anastrept a orcadensis (Map 44), in particular, are of considerable significance. These are both dioicous and sporophytes are rare or unknown. Gemmae are present in only the Anastrepta, and the Scapania has no specialized vegetative disseminules. It must be assumed that the species did produce sporo- phytes in the past, and it would be reasonable to assume that the species were then more widespread. If rare, even in the past, their chances of establishing their 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 177 many widely disjunct populations would be so greatly reduced as to be unreason- able. The added problems of disseminules taken from the parental population of a humid environment, drawn into updraughts of air and then carried to a favour- able site (which for these species is highly specialized) make the chances of long- distance dispersal even more unlikely. Yet the species are present in areas that they could not have reached except by wind-dispersal from sources a considerable distance away. St0rmer (1969) has outlined a clear instance of long-distance dispersal in the moss Orthodontium lineare. Although not so spectacular as the Hawaiian disjunc- tion, this case is well documented. Orthodontium lineare was inadvertently Intro- r duccd to the Liverpool district of Great Britain around 1911, The species has spread rapidly through Britain, by 1963 being present in most vice counties in England and reported also in Scotland and Ireland. In 1952 the species was re- ported from Holland, where "tlie spores must have been carried 300 km or more before they found suitable habitation" (St0rmer, 1969). In 1939 the species was discovered in Germany, presumably derived from the British populations. In many cases, at least, the expansion of the range of this species must be ascribed to wind dispersal of the spores. Considerable" discussion was elicited by the paper of Pctterson (1940) con- cerning the long-distance dispersal of the mosses Aloina brevirostris and A. rigida from Siberia to southwestern Finland, where he had recovered numerous spores of the species in rainwater. Persson (1944) and Bergeron (1944) have demonstrated tlie greater likelihood that the spores originated from nearby Scandinavian pop- ulations. The fact that moss spores serve as nuclei for rain drops is of particular importance, however, and should not be overlooked. The discussions of both Bergeron (1944) and Gregory (1945) are especially rich in information concerning dispersion of air borne spores. In a few cases bryophytes are disjunctive because they have been introduced to their widely separated localities through mans activities. The hepatics Lunu- laria cruciata and Marchantia polymorpha are common greenhouse weeds and are widely dispersed throughout the world because of this. More precisely docu- mented cases are those for Tortula stanfordensis (Map 31) and Pseudoscleropo- dium purum (Map 51). The Tortula was described by Steere (1951) from the San Francisco Bay Region of California where it is widely distributed on hard clayey soil both near habitations and in the native vegetation. Paton (1966) reported this species from southern Britain. More recently further details have ap- peared concerning its British distribution ( Whitehouse, 1961; Whitehouse & Paton, 1963). In the latter publication the authors state "since both Mousehole and Gulval are centres for the horticultural and market gardening industries, it seems possible that Tortula stanfordensis may have been accidentally introduced to one or both of these areas from California." It is suggested that it might have been introduced in soil of planted trees originating from a Californian nursery. For Pseudoscleropodium the disjunctions are equally interesting and are dis- cussed by Dickson (1967); details for the western American localities are given by Lawton (1960) and Schoficld (1965), while the New Zealand populations are discussed by Sainsbury ( 1935, 1955). Dickson states tliat the species was probably 178 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Table 1. Sexuality of disjunct bryophyte disjunctive species are sterile. An asterisk ( * ) usually sterile. taxa. A plus sign ( + ) indicates that most indicates that the species is consistently or Monoicous Desmntodon randii Pottia maritiriia Sphagnum junghuhnianum (also dioicous) Grimmia olympica Atdacomnium heterostichum Buxhaumia mhiakatae Hookeria lucens Cephaloziella turneri (also dioicous) Pleurozia purpurea \ Dioicous Pterogonium gracile Echinodium (the genus) 4- M(/wnw77i (the genus) Phyllogonimn (the genus) Drepanocladus uncinatus Hylocomium splendens Ctjrtopus setosus * Tortula caroliniana Leptodontium orcuUii Gollania turgens Sphagnum junghuhnianum (also monoicous) + Acanthocladium (sect. Tanythrix) Oligotrichum hercynicum Geheebia gigantea Drummondia prorepens Homaliadelphus sharpii Schwetschkeopsis fabronia Atrichum crispum Sphagnum S. pylaesii Ditrichum A- onatum Plagiothecium undulatum Crumia latifolia Leptodontium recurvifolium Antitrichia calif arnica Polytrichum sphaerothecium Acrobolbus ciliatus Mastigophora woodsii Anastrepta orcadensis Scapania ornithopodioides Bucegia romanica Porella cordaeana Cephaloziella turneri ( also n Plagiochila japonica Radula auriculata Carrpos sphaerocarpos Ascidota hlepharophylla Ana^trophyUum donianum Haplomitrium hookeri Takakia ceratophylla Macrodiplophyllum pUeatum Plagiochila carringtonii awn introduced to the south Atlantic islands of St. Helena and Tristan da Cunl packing material of young trees. In western North America the species is a weed and is always associated with human habitation, thus a probable introduc- tion with nursery stock from Europe. The case for New Zealand is less clear, most confi "the Tasman finding was in Leptospermum scrub and would be more likely to indicate an indigenous moss" (Sainsbury, 1955). 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 179 As mentioned earlier, a high proportion of disjunct bryophytes are dioicous. In some cases, the male plant is in one of the areas of disjunction and the female plant in another: Acrobolbus ciliatus (Map 7) is male in Japan, female in the Southern Appalachians. In such cases the only reasonable explanation for the disjunction is to assume that both arose from an originally continuous population. Why one sex should survive in one region and the other in the second is not readily explained. H omaliadelphus sharpii (Map 2) would appear to be sterile in North America bvit fertile in eastern Asia. However, dwarf male plants have been found in two North American populations, and therefore it is obvious that spores have been produced there, though indeed rarely. In dioicous species of mosses specialized means of asexual production are no more frequent than in monoicous species whose ranges tend to be wider ( Gcm- mell, 1952). The wider range of the monoicous species is presumably more a reflection of their spore dispersal than their greatly increased variability. As Gemmell (1950) has noted, it appears that many of the monoicous species are self -fertilized, greatly decreasing the possible variability that would be available through cross fertilization. For species occupying open areas and in which either spores or gemmae are produced, the chances of wind dispersal are greatly improved, and in areas of relatively dry climates, the possibility of getting propagula air-borne is even greater. As Persson (1944) and Petterson (1940) have shown, such spores are air transported and can come to earth in rainfall some distance from their place of origin. The problem if their germination and survival in the place where they are deposited is much more uncertain. Since most bryophytes are ecologically re- stiieted and conditions favoring germination and survival of propagula are higlily critical, the chances that many air-transported propagula survive to establish a colony are very remote. The factors severely limiting the effectiveness of distance dispersal are discussed by Crum ( 1966, 1972) . A further complication is the fact that many disjunctive taxa are in hyper- oceanic climates. This further decreases the opportunities for propagula to be- come air-borne in the rare cases when they are produced. Added to this is their restrictive ecology. Lye (1966) has emphasized that oceanic bryophytes are commonest "in localities where topography prevents both wind and sunshine from reaching high intensities. It might be suggested that some bryophytes appear to be disjunctively distrib- uted merely because collections have been inadequate in the intervening areas. Although this may prove to be true for a number of species now considered dis- junctive, it is not likely to greatly decrease the number. In areas relatively thoroughly explored as for example, Europe, the disjunctive patterns have been maintained even after increased knowledge of the area. The absence of suitable habitats in the intervening areas also makes their possible presence in them very doubtful. A knowledge of the details concerning continental drift is unlikely to be helpful in explaining most disjunctions in the Northern Hemisphere. In the Southern Hemisphere, however, Schuster (1969) suggests that the disjunctions are readily accommodated by the notion of a large Gondwanaland continent. It is y> 180 ANNALS [Vol. 59 also believed that many of the disjunctive taxa are extremely ancient, perhaps dating back to these earliest times. As freqviently emphasized, bryophytes appear to be extremely conservative, controlled both by their inbreeding and by infre- quency of sexual reproduction in many. No single hypothesis can explain all disjunctions. Each hypothesis when ap- plied to all disjuncts raises a number of serious questions. First, assuming that long-distance dispersal is tlie answer: 1. Why has dispersal been so selective for taxa that now possess such ineffi- cient means of dissemination and are presumably poor in biotypes? 2. Why should so many disjuncts be confined to oceanic environments? 3. Two areas possessing disjunctive species also have many species endemic to them that are widespread there, ecologically diverse and have ready means of air- dissemination, yet in both cases these species have not reached the other disjunc- tive area. The problem is : why not? Second, assuming that the disjunctive bryophytes represent remnants of an ancient flora that has persisted in or near the present location of the disjunctive population: 1. Why should some of the disjunctive species now exist in areas that have been available for only the time since the Pleistocene glaciations? 2. Since a number of disjuncts that presently lack any diaspores that are readily air-transported are now present on oceanic islands where they must have arrived by air, why could not other disjuncts have reached their sites in the same manner? Third, assuming that continental drift has led to the establishment of the dis- junctions: Although this would be satisfactory for the amphi-Atlantic species, it creates further difficulties in attempting to explain the western North American-western European disjunction and the eastern North American-eastern Asian disjunction. It also poses serious problems in explaining the amphi-Pacific taxa, since preced- ing continental drift, these continental masses would have been even more distant. At present, then, there is no conclusive hypothesis that will explain all disjunc- tive distributions. 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 181 Maps 1-3. — Distribution of Plagiochila japonica Sande Lac. ex Miquel. — 2. Distribution of H omaliadelphus sharpii ( Williams) Sharp. — 3. Distribution of Drummondia prorepens (Hedw.) Brid. ( after Iwatsuki, 1958, in part). 182 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 4-6. — 4. Distribution of Aulacomnium heterostichum (Hedw.) B.S.G. (after Iwatsuki, 1958, expanded).— 5. Distribution of Buxbaumia minakatae O. Okam. (after Iwat- suki & Sharp, 1967). — 6. Distribution of Schwetschkeopsis fabronia (Schwaegr.) Broth, (after Iwatsuki & Sharp, 1967). 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 183 Maps 7-9. — 7. Distribution of Acrobolbus ciliatus (Mitt.) Schiffn, (from information in Sharp & Hattori, 1968).— 8. Distribution of Leptodontitnn orcnttii Bartr.— 9. Distribution of Tortula caroliniaiia Andn 184 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 10^12.-10. Distribution of Radiila auriculata Steph.— 11. Distribution of Gtimmia olympica Britt. ex Frye.— 12. Distribution of Gollania turgens (C. Miill.) Ando (after Ando, Persson & Sherrard, 1957; Ando & GjaervoU, 1961 ). 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 185 Maps 13-15. — 13. Distribution of Ascidota blepharophylla Mass. — 14. Distribution of Acanthocladium (Sect. Tanythrix), — 15. Distribution of Sphagnum junghuhnianum Dozy. & Molk. 186 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 16-18. — 16. Distribution of Takakia ceratophtjlh (Mitt.) Grolle (after Ilattori ef al, 1968). — 17. Distribution of Rhizomnium nudum (Williams ex Britt. & Williams) Koponen. — 18. Distribution of MacrodiplophijUum plicatum (after Horikawa, 1955, ex- panded ) . 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 187 19-21. — 19, Distribution of Polyirichum sphaerothecium (Besch.) Broth. — 20. 1 of Treubia nana Hatt & Inoue (after Hattori et al., 1966, expanded). — 21. Dis- ■ Drevanocladus tincinafus (Hedw. ) Warnst. (after Irmscher. 1929. exnandpflV 188 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Voi,, 59 Maps 22-24. — 22. Distribution of Hijlocomiwn splendens ( Hedw. ) B.S.G.— 23. Distribution of Ditrichum zonatum (Brid.) Kindb. — 24. Distribution of Leptodontiiun re- ctirvifolium (Tayl.) Lindb. 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 189 Maps 25-27.-25. Distribution of C ephaloziella turneri (Hook.) MiilL— 26. Distribu- tion of Crumia latifolia (Kindb. ex Mac.) Schof. (after Abramova & Dildarin, 1969).— 27. Dis- tribution of Plaaiothecium undulatum (Hedw.) B.S.G. {after Ireland, 1969, Stdrmer. 1969). 190 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 28-30. — 28. Distribution oi Porclla cordaeana (Hueb.) Evans (after Szwckowsl^i, 1962). — 29, Distribution of Hookeriu lucens (Hedw.) Sm. (after Irmscher, 1929, modified: St0nner, 1969).— 30. Distribution of Bucegta romanica Radian (after Szwekowski, 1964). 1972] SCHOFIELD & CRUM — DISJUNCTIONS IN BRYOPHYTES 191 Maps 31-33. — 31. Distribution of Tortula standfordenm Steere. — 32, Distribution of Antiirichia californica Sull. ex Lesq. — 33. Distribution of Claopoditim whipplearmm (Sull.) Ren. & Card, (from Noguchi, 1952, expanded). 192 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 35 Maps 34-36.-34. Distribution of Sphagnum pylaesii Brid. (from Maass, 1966a, 1966fo). angermanicum Melin (supplied bv W.S.G. Maass). — 36 ames 1966) 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 193 Maps 37-39 — 37. Distribution of Cladopodiella francisci (Hook.) Dum. (after Szwey- kowski, 1964, expanded). — 38. Distribution of Oligotrichum hercynicum (Hedw.) Lam. & DC. — 39. Distriliution of Haplomitrium hookcri (Sm.) Nees (after Szweykowsld, 1966, ex- panded; Corely, 1970). 194 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 40-42. — 40. Distribution of Fleuroclada albescens (Hook.) kowski, 1966). — 41. Distribution of Geheebia f^iganiea (Funck.) Boul.- Mastigophora woodsii (after Ratcliffe, 1963, in part). Spr. (after Szwey- 42. Distribution of 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPIIYTES 195 Maps 43-45. — 43. Distribution of Scapania ornithopodioides (With.) Pears, (after Ratcliffe, 1965, in part). — 44. Distribution of Ana.strepta orcadcnsis (I look,) Schiffn. (after Szweykowski, 1964, expanded). — 45. Distribution oi Anastrophijllum donianum (Hook ) Spr (after Szweykowski, 1966, emended; Ratcliffe, 1965, in part). 196 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Maps 46-48. — 16. Distribution of Pleurozia purpurea (Lightf.) Lindb. (after Ratcliffe, 1969, in part). — 47. Distribution of the genus PhyllothalUa Hodgs. (after Schuster, 1969). 48. Distribution of Ctjrtopus setosiis (Hedw. ) Hook. 1972] SCHOFIELD & CRUM^DISJUNCTIONS IN BRYOPHYTES 197 Maps 49-51. — 49. Distribution of Carrpos sphaerocarpos (Carr) Prosk. — 50. Distribu- tion oiPottia maritima (R. Br.) Broth, (after information in Scheipe, 1969).— 51. Distribution of PseudoscleropoJium purum (Hedw.) Fleisch. (after St0rmer, 1969, in part; Dickson, 1967). 298 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Literature Cited Abramova, a, I. & B. I. DiLDARiN. 1969. Cnimia lafifolia (Kindb.) Schof. in the moss flora of the U.S.S.R. Biol. Ziim. Armen. 22: 43-48. (In Russian.) & I. I. iVuBAMOv. 1969. Eastern-Asiatic affinities of the Caucasian bryoflora. Jour. Hattori Bot. Lab. 32: 161-154. Amaxn, J, 1929. L'hygrothermie du cliniat, facteur deteniiinant la repartition des especes atlantiques. Rev. Bryol. 2: 126-133. Anderson, L. E. 1951. The mosses of North Carolina VL Encalyptaceae to Pottiaceae. Bryologist54: 145-161. Ando, H. 1966. A revision of the Chinese Cupressinae described by C. Miiller. Bot. Mag. Tokyo 79: 759-769. , H. Persson & E. M. Sherrard. 1957. The first record of CoUania in North America. Bryologist 60: 326-335. Andrews, A. L. 1920. Tortula caroliniana, new species. Bryologist 23: 72-76. . 1938. The North American Atlantic species of Sphagnum, Ann. Br>ol. 11; 15-20. . 1961. Notes on North American Sphagmim XIIL Sphagnum pylaesii. Bryologist 64: 208^214. Bergeron, T. 1944. On some meteorological conditions for the dissemination of spores, pollen etc., and a supposed wind transport of Aloina spores from the region of Lower Yenisey to Southwestern Finland in July 1936. Svensk. Bot. Tidskr, 38: 269-292. Billings, W. D. & L. E. Anderson. 1966. Some microclimatic characteristics of habitats of endemic and disjunct bryophytes in the southern Blue Ridge. Bryologist 69: 76-95. Bowers, F. D. 1970. High elevation mosses of Costa Rica. Jour. Hattori Bot. Lab. 33: 7-35. Carr, D. J. 1956. Contributions to Austrahan Bryology. I. The structure, development, and and systematic affinities of Monocarpns sphaerocarpns gen. & sp. nov. (Marchantiales). Austral. Jour. Bot. 4: 175-191. CoRLEY, M. F. V. 1970. Distribution maps of bryophytes in Britain: Haplomitrium hookeri (Sm.) Nees. Trans. Brit. Bryol. Soc. 6: 142. Courtejaire, J. 1962. La microaire frangaise de Sphagnum pylaiei, Bryologist 65: 38^7. Croizat, L. 1962. Les hepatiques par devers la biogeographie mondiale. Rev. Biyol. Lichenol. 31: 5-22. . 1966. La repartition des Trachypodaceae (Musci). Rev. Bryol. Lichenol. 34: 771-798. Crum, H. a. 1956. Notes on Hypnodon, a genus of the Orthotrichaceae new to North Amer- ica. Bryologist 59: 26—34. . 1965. Mmum nuJum in Japan. Bryologist 68: 118-119. . 1966. Evolutionary and phytogeographic patterns in the Canadian moss flora. Pp. 28-42, in Taylor & Ludwig, "The Evolution of Canada's Flora." Toronto. ~ . 1972. The geographic origins of the mosses of North America's eastern deciduous forest. Jour. Hattori Bot. Lab. 35: 269-298. Crundwell, a. C. 1957. Some neglected British moss records. Trans. Brit. Bryol. Soc. 3: 174-179. Dickson, J. H. 1967. Pseuchscleroiwdium punim (Limpr.) Fleisch. on St. Helena and its arrival on Tristan da Cunha. Bryologist 70: 267-268, Dixon, H. N. 1922. Miscellanea bryologica 8. Jour, Bot. 60: 281-291. Evans, A. W. 1914. Report on the Hepaticae of Alaska. Bull. Torrey Bot. Club 41: 577- 616. Fulford, M. 1951. Distribution patterns of tlie genera of leafy Hepaticae of South America. Evolutions: 243-2(>4. Gams, H. 1955. Zur Arealgeschichte der arktischen und arktisch-oreophytischen Moose* Feddes Repert. 58: 80-92. Caume, R, 1952-1954. Les elements de la flore bryologique de Bretagne, Rev. Bryol. Lichenol. 21: 229-234; 22: 20-21, 141-147; 23: 291-294. Gemmell, a. R. 1950. Studies in the Bryophyta I. The influence of sexual mechanism on varietal production and distribution of British Musci. New Phytol. 49: 64-71. . 1952. Studies in the Bryophyta H. The distribution of the sexual groups in British mosses. New Phytol. 51: 77-89. Gregory, P. H. 1945. The dispersion of air-borne spores. Trans. Brit. Mycol. Soc. 28: 26- 72. 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES IQQ Greig-Smith, p. 1950. Evidence from hepatics on the history of the British flora. Jour. Ecol. 38: 320-344. Grolle, R. 1963. Takakia im Himalaya. Oesterr. Bot. Zeitscher. 110: 4U-AA1. ■ -. 1964. Jamesoniella carrmgtonii eine Plagiochila in Nepal mit Perianth. Trans. Brit. Bryol. See. 4: 65.3-663, 1966. Gymnomitrion crenulatum und Verwandte. Trans. Brit. Bryol. Soc. 5: 86-94. Harvill, a. M. 1950. The western American-European element in the Alaskan moss flora. Rev. Bryol. Lichenol. 19: 32-34. Hattori, S. 1951. On the distribution of the Hepaticae of Shikoku and Kiushiu (southern Japan). Bryologist 54: 103-118. . 1952. Ftilidhim caJiforniciim and other nearctic hverworts in Japan. Bryologist 55: 147-149. 1963. Takakia of North Borneo. Jour. Jap. Bot. 215-217, 241-243. 1966a. Anthocerotae and Hepaticae. Pp. 501-536, in Hara (editor), "The Flora of Eastern Himalaya. j> — . 19666. Hepaticae and Anthocerotae of Mt. Chokai, Northern Japan. Jour. Hattori Bot. Lab. 29: 266-278. — & A. J. Sharp. 1968. Takakia ceratophtjlla and T. lepidozioides of Pacific North America and a short history of the genus. Misc. Bryol. Lichenol. 4: 137-149, — , M. MizuTANi & Z. IwATSUKi. 1966. The systematic position and distribution of Treubia nana. Bryologist 69: 488-492. , , & . 1968. Takakia ceratophylla and T. lepidozioides of Pacific North America and a short history of the genus. Misc. Bryol. Lichenol. 4: 137-149. Haynes, C. C. 1915. Bucegia, a new genus for North America. Bryologist 18: 93-94. Herzog, T. 1926. Geographic der Moose. Jena. Horikawa, Y. 1955. Distributional studies of bryophytes in Japan and the adjacent regions. Hiroshima. & H. Ando. 1957. Phytogeographical notes on Hypnum sidnmponens Lesq. and H. dieckii Ren, et Card, Jour. Jap. Bot. 32: 225-231. Inoue, H. 1958. Regeneration of the leaf of Acrobolbus ciliatuH. Jour. Jap. Bot. 33; 14-18. . 1958. The family Plagiochilaceae of Japan and Formosa II. Jour. Hattori Bot. Lab. 20: 54-106. Ireland, R. R. 1969. A taxonomic revision of the genus Tlagiotliecium for North America^ north of Mexico. Natl. Mus. Nat. Sci. Publ. Bot. 1, Irmscher, E. 1929. Pflanzenverbreitung und Entwicklung der Kontinente II Teil, Weitere Beitrage zur genetischen Pflanzengeographie unter besondere Beruchsichtung der Laub- moose. Mitt. Inst. Allg. Bot. Hamburg 8: 171-364 + 16 pi. Iwatsuki, Z. 1958a. Correlations between the moss floras of Japan and the Southern Appa- lachians. Jour. Hattori Bot. Lab. 20: 304-352. , 1958Z?. Two interesting Fissidens species common to Japan and eastern North Amer- ica. Jour. Jap. Bot. 33: 245-250. — . 1958c. Review of the genus Homaliadclphus. Bryologist 61; 68-78. — & A. J. Shai^p. 1967. The bryogeographical relationships between eastern Asia and North America I. Jour. Hattori Bot. Lab. 30: 152-170. — ^ 1968. The bryogeographical relationships between eastern Asia and North America II. Jour. Hattori Bot. Lab. 31: 55-58. Johnson, A. 1960. Variations in Sphagnum junghuhnianum subsp. junglndinianum Dz, & Molk. Trans. Brit. Bryol. Soc. 3: 72,5-728. Kitagawa, N. 1966. A revision of the family Lophoziaceae of Japan and its adjacent regions II. Jour. Hattori Bot. Lab. 29: 101-149. Koch, L. F, 1954. Distribution of California mosses. Amer. Midi. Naturalist 51: 515-538. . 1956. Mosses common to California, U.S.A. and the Basque Country but not to Baja Cahfornia, Mexico. Rev. Bryol. Lichenol. 25: 285-287. Lawton, E. 1960. Pseudoscleropodium purum in the Pacific Northwest. Bryologist 63: 235-237. Lazarexko, a. S. 1957. Versuch einer Analyse der Laubmoosflora vom nordostlichen Asien. Rev. Bryol. Lichenol. 26: 146-157. Love, A. & D. Love. 1953. Studies on Bnjoxiphium, Bryologist 56: 73-94; 183-203. LowRY, R. J. & W. C. Steere. 1946. A propaguliferoiis form of Aulacotnninm heterostichwn. Bryologist 49: 30-32. 200 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Lye, K. a. 1967, Studies in the growth and development of oceanic bryophyte commimities. Svensk Bot. Tidskr, 61: 297-310. Maass, W. S. G. 1965. Zur Kenntnis des Sphagnum angermanicum in Europa. Svensk Bot. Tidskr. 59: 332-344. . 1966flr. Studies on the taxonomy and distribution of Sphagnum I. Sphagnum pylaesii and Sphagnum angermanicum in Quebec and some phytogeographic considerations. Bry- ologist69: 95-100. — . 1966fo. Untersuchungen iiber die Taxonomie und Verbreitung von Sphagnum VL Sphagnum pylaesii Brid. und das boreo-atlantische Florenelement unter den Torfmoosen in Siidamerika, Nova Hed\\igia 12: 81-105 + Tab. 13-19. — . 1967. Studies on the taxonomy and distribution of Sphagnum III. Observations on Sphagnum macrophyllum in the northern part of its range, Bryologist 70: 177-192. Martix, W. 1946. Geographic range and internal distribution of the mosses indigenous to New Zealand. Trans. Roy. Soc. New Zealand 76: 162-184. . 1949. Distribution of the mosses indigenous to New Zealand. Supplement No. I. Trans. Roy. Soc. New Zealand 77: 355-360. 1952a. Distribution of the mosses indigenous to New Zealand. Supplement No. 11. Trans. Roy. Soc. New Zealand 80: 197-205. . 1952&. New records of Northern Hemisphere mosses in New Zealand. Trans. Roy. Soc. New Zealand 80: 233-235. MiJLLER, K. 1916. Zur geographischen Verbreitung der europaischen Lebermoose und ilirer Verwertung flir die allgemeine Pflanzengeographie. Ber. Deutsch. Bot. Gesell. 34: 587. . 1954. Die pflanzengeographischen Elemente in der Lebermoosflora Deutschlands. Rev. Bryol. Lichenol. 23: 109-122. NiCHOLsox, W. E. 1930. ''Atlantic" hepatics in Yunnan. Ann. Bryol. 3; 151-153. NoGUCffl, A. 1952. A revision of the genus Claopodium. Jour. Hattori Bot. Lab. 27: 20^6. . 1966. Musci. Pp. 357-591, in Hara (editor), "The Flora of Eastern Himalaya." & K. Saito. 1970, Grimmia olympica E. G. Britton occurs in Japan. Misc. Bryol. Lichenol. 5: 104-105. Patox, J. A. 1966. Tortula stanfordensis in Yorkshire. Trans. Brit. Bryol. Soc. 5: 147-148. . 1966. Distribution maps of bryophytes in Britain: C ephaloziella iurneri (Hook.) K. Miill, Trans. Brit Bryol. Soc. 5: 159. Persson, H. 1944. On some species of Aloina, with special reference to their dispersal by the wind. Svensk Bot. Tidskr. 38: 260-268. . 1944a. The genus Hahrodon discovered in North America. Svensk Bot. Tidskr. 40: 317-324. — . 1946Jb. Some Alaskan and Yukon bryophytes. Bryologist 49: 41-58. — . 1947, Further notes on Alaskan-Yukon bryophytes, Bryologist 50: 279-310. — . 1949. Studies on the bryophyte flora of Alaska-Yukon. Svensk Bot. Tidskr. 43: 491-533. — . 1952. Critical or otherwise interesting bryophytes from Alaska- Yukon. Bryologist 55: 1-25, 88-116. — . 1958. The genus Takakia found in North America. Bryologist 61: 359-361. . 1962. Bryophytes from Alaska collected by E. Hulten and others. Svensk Bot. Tidskr. 56: 1-36. — . 1968. Bryophytes from die Aleutian Islands, Alaska, collected mainly by Hansford T. Shacklette. Svensk Bot, Tidskr. 62: 369^387. & O. GjAERvoLL. 1957. Bryophytes from the interior of Alaska. Det. Kgl. Norske Vidensk. Selsk. Skr. 1957(5): 1-74. Petterson, B. 1940. Experimentelle Untersuchungen iiber die Verbreitung der Sporen- pflanzen. Acta Bot. Fenn. 25; 1-103. Proctor, M. C. F. 1964. Distribution maps of bryophytes in Britain. Trans. Brit. Bryol. Soc. 4: 744. Ratcliffe, D. a. 1963. Distribution maps of bryophytes in Britain; Mastigophora woodsii (Hook.) Nees, Trans. Brit. Bryol. Soc. 4: 511. . 1963, Distribution maps of bryophytes in Britain: Leptodontium recurvifolium ( Tayl. ex Wils. ) Lindb. Trans. Brit. Bryol. Soc. 4: 521. 1965. Distribution maps of Bryophytes in Britain : Anastrophyllum donianum (Hook.) Steph. Trans. Brit. Bryol. Soc. 4: 874. 1972] SCHOFIELD & CRUM— DISJUNCTIONS IN BRYOPHYTES 201 . 1965. Distribution maps of bryophytes in Britain: Scapania ornithopodioides (With.) Pears. Trans. Brit. Bryol. See, 4: 881. -. 1968. An ecological account of Atlantic bryophytes in tlie British Isles. New Phytob 67: 365-439, DuRiETZ, G. E, 1940. Problems of bipolar plant distribution. Acta Phytogeogr. Suec. 13: 215-282. Sainsbury, G. O. K. 1942. Northern mosses in New Zealand. Bryologist 45: 40-43. . 1955. A handbook of the New Zealand mosses. Roy. Soc. New Zealand Bull. 5: 1-490. . 1965. Introduced mosses in New Zealand. Bryologist 68: 91-92. ScHELPE, E, A. C. L, E. 1969. Three new records of southern hemisphere Bryophyta for South Africa. Jour. S. Afr. Bot. 35: 109-112. ScHOFiELD, W. B. 1962. Treubia nana in North America. Bryologist 65: 277-279. . 1965. Correlations between the moss floras of Japan and British Columbia, Canada. Jour. Hattori Bot. Lab. 28: 17-42. 1966a. Acanthocladium (Sect. Tanythrix) in North America. Bryologist 69: 334- 338. 1966^^. Crumm, a new genus of the Pottiaceae endemic to western North America. Canad. Jour. Bot. 44: 609-614. — . 1966c. The identity of Polytrichum sphaerothecium (Besch.) Broth. Misc. Bryol. Lichenol. 4: 33-35. — . 1968(3. Bryophytes of British Columbia I. Mosses of particular interest. Jour. Hattori Bot. Lab. 31: 205-226. — . 1968&. Bryophytes of British Columbia II. Hepatics of particular interest. Jour. Hattori Bot. Lab. 31: 265-282. — . 1969. Phytogeography of northwestern North America: bryophytes and vascular plants. Madrono 20: 155-207. ScHORNHERST, R. O. 1943. Phytog CO graphic studies of the mosses of northern Florida. Amer. Midi. Naturalist 29: 509-532. Schuster, R, M. 1958a. Notes on nearctic Hepaticae VI. Phytogeographical relationships of critical species in Minnesota and adjacent areas of the Great Lakes. Rhodora 60: 209- 234, 243-256. . 1958&. Boreal Hepaticae, a manual of the liverworts of Minnesota and adjacent regions. III. Phytogeography. Amer. Midi. Naturalist 59: 257-332. — , 1959. A monograph of the nearctic Plagiochilaceae II. Sectio Zonatae through Sectio Parallelae, Amer. Midi. Naturalist 62: 257-395. — . 1962. A study of Cephaloziopsis with special reference to C. pearsoni and its distri- bution. Trans. Brit. Bryol. Soc. 4: 230-246. — . 1967. Studies on antipodal Hepaticae IX. Phyllothalliaceae. Trans. Brit. Bryol. Soc. 5: 283^-288. — . 1968. Studies on Hepaticae XLIV. On Fhtjllothallia Hodgs. Nova Hedwigia 15: 517-522 + pi. 66-67. — : 1969. Problems of antipodal distribution in lower land plants, Taxon 18: 46-91. & G. A, M. Scott. 1969. A study of the family Treubiaceae (Hepaticae: Metz- geriales). Jour. Hattori Bot. Lab. 32: 219-268. Sharp, A. J, 1936. Interesting bryophytes, mainly of the Southern Appalachians. Jour. S. Appal. Bot. Club 1:49-59. — . 1938. Tropical bryophytes in tlie Southern Appalachians. Ann. Bryol. 11: 141-144. — , 1939, Taxonomic and ecological studies of eastern Tennessee bryophytes. Amer. Midi. Naturahst21: 267-354. — , 1941a. Some historical factors and the distribution of southern Appalachian bryo- phytes, Bryologist 44; 16-18. — . 1941Z7. Southern Appalachian bryophytes in Europe. Bryologist 44; 65-68. — & S. Hattori. 1967. Takakia ceratophylla found in the Aleutians, Misc. Bryol, & Lichenol. 4: 120. — gj . 1968. Acroholhus ciliaius from Attu Island of the Aleutian Islands. Jour. Jap. Bot. 43; 311-315. & Z. IwATsuKi. 1965. A preliminary statement concerning mosses common to Japan and Mexico. Ann. Missouri Bot Card. 52: 452-456, 202 ANNALS [Vol. 59 & , 1969. Crumia deciduidentata, a new species in Alaska. Jour, Hattori Bot. Lab. 32: 95-98. Smith, A. J. E. 1966. Distribution maps of bryophytes in Britain: Atrwhum crisptim (James) SulL Trans. Brit. Bryol Sec. 5: 362. Steere, W. C. 1937. Critical bryophytes from the Keweenaw Peninsula, Michigan. Rho- dora39: 1-14,33-46. . 1937. Bryoxiphhim norvegiciim, the sword moss, as a preglacial and interglacial relic. Ecology 18: 346-358. — -. 1938. Critical bryophytes from the Keweenaw Peninsula II. Ann. Bryol. 11: 145- 152. — . 1951. Tortula stanfordensis, a new species from California. Bryologist 54: 119-123. — . 1953. On the geographic distribution of arctic bryophytes, Stanford Univ. Publ. Biol. Sci. 11: 30-47. — . 1965. The boreal bryophyte flora as affected by Quaternary glaciation. Pp. 485- 495, in Wright & Frey, ''The Quaternary of the United States." Princeton. — . 1969. Asiatic elements on the bryophyte flora of western North America. Bryol- ogist 72: 502-512. & W. B. ScHOFiELD. 1956. Mtjitroclada, a genus new to North America. Bryologist 59: 1-5. & R. M. Schuster. 1960. The hepatic genus Ascidota Massalongo new to North America. Bull. Torrey Bot. Club 87: 109-215. ST0RMER, P. 1960. Antitrichia calif omica in the Canary Islands. Rev. Bryol. Lichenol. 29; 254-255. . 1969. Mosses with a Western and Southern Distribution in Norway. Oslo. Suzuki, H. 1956. Variations in Sphagnum jun ghuhnianum var. pseudomolle Wamst. and the status of Sphagnum kiiense Warnst. Jap. Jour. Bot. 15: 186-198. Szweykowski, J. 1962. Atlas of Geographical Distribution of Spore-Plants in Poland. Series 4. Liverworts (Hepaticae), Parti. . 1964. Atlas of Geographical Distribution of Spore-Plants in Poland. Series 4. Liverworts ( Hepaticae ) . Part 2. TouFFET, J, 1964. Les localites du Sphagnum pylaiei Brid. dans les Montagues Noires de Bretagne. Rev. Bryol. Lichenol. 33: 501-504. Whitehouse, H. L. K, 1961. The occurrence of Tortula stanfordensis Steere in Cornwall, new to Europe. Trans. Brit. Bryol. Soc. 4; 84-94. & J. A. Paton. 1963. The distribution of Tortula stanfordensis Steere in Cornwall. Trans. Brit. Bryol. Soc. 4: 462-463. DISJUNCTIONS IN HOMOSPOROUS VASCULAR PLANTS' W. H. Wagner, Jr. Pteridopliyte geography has been examined from a number of standpoints ( cf. Christ, 1910; Winkler, 1938; Tryon, 1969, 1970). A complete assessment of tlic significance of their geographical disjunctions would be difficult to make because of the numerous factors involved. The following discussion will concentrate upon special problems and the examples will have a strongly North American bias. Homosporous vascular plants are pteridophytes which lack micro-megaspore differentiation; all of the spores are presumably bipotcntial and capable of pro- ducing gametophytes with both sex organs. These pteridophytes include some of the widest ranging of all vascular plants, such as Ltjcopodium clavatum^ Osmunda regalisy Cystopteris fragilis, and Asplenium trichomanes. The distribution patterns of pteridophytes in general are basically like those of seed plants: Narrow ende- mism is common, and long distance disjunctions frequently occur. One of our major concerns is whether a given disjunction may not be the result of a casual spore introduction. As Klekowski (this symposium) has pointed out, there are significant contrasts between ferns and seed plants, not only in means of dispersal but in their genetic apparatus as well. Some patterns of distribution, e.g. eastern Asia- North America and the amphitropical ones, may be related to major events in earth history. Others may be merely recent occurrences resulting from chance spores that traveled long distances. Their popularity with researchers and field botanists has caused the pterido- phytes to be well collected and represented in herbaria. In the eastern United States, for example, we know the ranges of these plants so well that finding a disjunct population only 100 miles from previously known stations is considered an event. As to what we may call a disjunction, there is, of course, no set definition. As Erickson (1945) showed in a species of Clematis, the range is made up of thousands of spatially separated populations. Ehrlich and Raven (1969) assert that distances of only a few miles or less may suffice to isolate plant populations from gene flow. Here I give as examples of disjunctions separations in range of as little as 100 miles. Most of my examples (Table 1) deal with situations in which there is a large center of population and a small strongly disconnected population or group of populations. The principal seat or center will here be designated the ''metropolis'' and the small disjunct stand the ''outlier'' Although the metropolis was not neces- sarily the ancestral area, in some cases it probably was. Some taxa, as we go away from their centers of abundance, simply "fade out/' their occurrences becoming ^Written under National Science Foundation grant GB-8113, "The Evolutionary Charac- ters of the Ferns." Among those who helped I wish to thank especially C. E. Delchamps, Thomas Duncan, Hugh litis, David B. Lellinger, T. Lawrence Mellichamp, Dale J. Hagenah, Elbert L. Little, Ronald L. Stuckey, and Florence S. Wagner. 2 Department of Botany, University of Michigan, Ann Arbor, Michigan 4S104. Ann. Missouri Box. Card. 59: 203-217. 1972. 2Q4 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 more and more scattered, but lacking extreme peripheral disjunctions. The best example of transition from common to rare or sporadic in eastern United States is probably among the species of woodferns, Dryopteris, e.g. Crested woodfern, D. cristata; Spinulose woodfern, D. spinulosa, as we proceed down the Appa- lachian mountain chain. Common in the north, frequent in the middle range, and rare in the south, these species probably have no large gaps of 100 or more miles. Really large gaps are expectedly much more common in the western mountains because of their more xeric conditions and their more precipitous and isolated peaks. The eastern North American-eastern Asiatic pattern of disjunction, first promulgated by Asa Gray, is illustrated by such examples as Shining clubmoss, Lycopodium lucidulum; the "J^P^^^^^^ grapefern," Botrychium ternaturn (but sec below); Interrupted fern, Osmunda claytoniana; and Sensitive fern, Onoclea sensibilis — all of which appear in more or less identical form in the eastern parts of the two continents (c/. Broun, 1938; Fernald, 1950; Wherry, 1961). Amphitropical T distributions have been discussed by Raven (1963), who lists several pterido- phytes. Probably the best example involving western North America and southern South America is Polystichum mohrioides (syn. P. lemnionii; also confused with P, X scopulinum; see Taylor, 1970: 172-187), The plants from Washington, Oregon, and northern California are closely similar to those from the Andes of South America and certain of tlie antarctic islands, and they are evidently conspe- cific. An apparent near relative of the circumboreal and familiar Moonwort fern, is also amphitropical. The plant in question, which ranges )., California, far to the north in British Columbia and Alberta, lunar ia »_-^ is evidently the same as the species distinguished by A. H. G. Alston as B. dusenii, heretofore known only from Argentina {cf. Alston, 1960; Wagner & Lord, 1956). In at least some North American localities, B. dusenii grows sympatrically with B. lunaria {e.g. Snohomish Co., Washington). Probably all of the plants from western North America which were formerly identified as B. mh actually B. dusenii. Although there is much interest in disjunctions of pteridophytes over water (e.g. Hawaii, Fiji, Samoa, Canary Islands, Galapagos — c/. Tryon, 1970), overland disjunctions across continents are equally intriguing, and they present different problems. Many intracontinental disjunctions are vexing because of seemingly suitable habitats in intervening areas where one would expect to find the plants in question. Oceanic disjunctions cause no problems from this standpoint because atolls. the nature of plant distribution conjured up by the famihar phrase "rare and local." Some Major Disjunctive Trends In Table 1, I have listed a number of disjunctive "trends" in North American pteridophytes, giving the metropolis, estimated distance of the disjunction, ploidal level, and references. The following discussion refers to that table. 1972] WAGNER — DISJUNCTIONS IN HOMOSPOROUS VASCULAR PLANTS £05 A. North to South Trend. In this the southernmost occurrences become widely separated, either "fading out" as they go southward, finally appearing only on the highest peaks, or forming more or less large disjvmctions. Climate seems to be the major controlling factor, but often a given species is found upon one peak and then skips a number of others that would seem to be appropriate. When this happens we get disjunctions like those in Table 1. The Slender rock brake, Crypto gramma stelleri^ was known only as far south as the ravines of northern Pennsylvania until Wherry (1961) found it in West Virginia: "Unexpectedly, there in a crevice so sheltered that the sun s rays never entered, and kept cool by evaporating moisture, was a colony of this tiny rock fern; its known range was thus extended some 200 miles southward." The most peculiar north-south pattern is that of American Hart's -tongue, Phijllitis scolopendrium var. americanum. This fern has a strongly disrupted range except on the Niagara Escarpment of Bruce and Grey counties in Ontario ( Soper, 1954), the center of its metropolis. All other stations are much further separated, and especially those far to the south in Tennessee. As early as 1878, Hart's-tongue plants were encountered in cool, damp limesinks there (Shaver, 1954). (The Mexican and Caribbean plants assigned to this species apparently represent a distinct subspecies.) In connection with investigations of the spontaneous floras of pine plantations, a curious north-south disjunction was recently discovered involving the Braun s holly fern, Polystichum braunlL The "Marshall Tract" near the town of Ann Arbor, Michigan, was planted with species of spruce and pine on an old pasture during the 1920's and since, as in practically all similar plantations of the region, a distinctive, though small, spontaneous flora has arisen, including Running pine, Lycopodium flabelliforme; Spinulose woodfern, Dryopterus spinulosa; Sensitive fern, Onoclea sensihilis; and Ebony spleenwort, Asplenium platyneuron, A soli- tary plant of Braun's holly fern was discovered here in June, 1971, by Florence S. Wagner, growing in a deeply shaded valley witli the foregoing pteridophytes. This constitutes a range extension of over 200 miles from the nearest locality to the north on South Manitou Island. Hagenah (1955: 75-76) reported a disjunction of the sterile woodfern hybrid, Dryopteris filix-mas X margirmlis in southern Lower Michigan — a single collection, some 200 miles south of the nearest localities to the north. Perhaps, in this case, certain "sterile" hybrids are able to produce occasional viable spores (Morzenti, 1967). Such spores are unreduced and capable of forming apogamous sporophytes. Michitran which fits this nattern is the Min anense which I discovered south of Detroit in 1962. Comment- ing on this occurrence, Hagenah (1966: 159) wrote as follows: ''The most unusual locality for this species in Michigan is that in Wayne County ... far to the south of all other stations. Although only one plant was fovmd here originally, it was observed for several years and now there are two." Since 1966 both plants, prob- ably offsets of the same one, have disappeared. It probably represents an unsuc- cessful single spore introduction, the site being too far south for continued survival. B. South to North Trend. Examples in which the metropolis is southern rather than northern are very few in homosporous pteridophytes. A prominent south- Table 1, Some examples of outlying North American over-land pteridophytic disjunctions with rough estimates of mileages away from their K> meti"opolises. polyploid (level given if known), f = outliers suspected of being casual spore establishments rather than relicts of past floras. Taxon Metropolis Estimated Disju: Ploidy Boinjchium muUifidum Woodsia ihensis Boinjchium lanceohUim var. angustisegmentiim Gymnocarpium dryopteris Thchjptcris phegoptcris Botrychium simplex Crijpto gramma stelleri n. N. Amer. n. X. Amer. n, e. N. Amer. A. North to South Trend 100 mi. 100 150 2x 2x 2x n. N. Amer. n. N. Amer. n. N. Amer. n. N. Amer. t Dryopteris filix-mas X marginalis n. e. N. Amer. t Polysiichum bravnii vat, purshii n. N. Amer. n. e. N. Amer, n. w. N. Amer. Botrychium mingancnse Polysiichum X scopuUivim PhyllitLs scolopendriiim var. amcricanum Bruce Peninsula, Ontario 150 150 200 200 200 200 200 450 650 4x* 3x* 2x 2x 3x* 4x* 4x* 4x* 4x* B. South to North Trend Ophioglossiim vulgatum s. e. U.S. 100 « *Var. pycnostichum 99 t Dennstacdtia punctilohida f Adianium capiUus-vencris s. e. U.S. s. U.S. 150 700 2x 2x, 4x C. Trend to Southwestern Michigan t Lygodium palmatum t Woodivardia arcolata Dryopicris X celsa Lycopodium appresstim s. e. U.S. s, e. U.S. s. e. U.S. e. Coastal Plain 300 300 400 600 2x 2x 4x* 2x? D, Trend to "Driftless Area" A^lenium X pinnatifidum Lycopodium porophilum Thclyptcris simtdata s. highlands s. highlands e. Coastal Plain 300 350 700 4x* 4x* Outlier & Reference Madison Co., Va. (Wagner, 1946) Ashe Co., N.C. (Bozeman, 1968) Macon Co., N.C. (Wagner et al, 1970) Ashe Co., N.C. (Bozeman, 1968) Macon Co., N.C. (Wagner et al, 1970) Giles Co., Va. (Wagner, 1963) Pendleton Co., W. Va. (Wherry, 1939) Barry Co., Mich. (Hagenah, 1955) Washtenaw Co., Mich. (Wagner, unpubl.) Wayne Co., Mich. (Hagenah, 1966) Cochise Co., Ariz. (Wagner, unpubl.) Marion Co., Tenn. (Shaver, 1954; Fernald, 1935) Lenawee Co., Mich. (Wagner, 1971) Jackson, Shiawassee Cos., Mich. (Wagner, unpubl.) Fairmont, B.C. (Taylor, 1970) Kalamazoo Co., Mich. (Pippen, 1966) Van Buren Co., Mich. (Billington, 1952) Kalamazoo Co., Mich. (Wagner et ah, 1969) Van Buren Co., Mich. (Wagner & Hagenah, un- publ. ) Iowa Co., Wise. (H. litis, pers. comm. ) Grant Co., Wise. (Hartley, 1966) Jackson Co., Wise. (Hartley, 1965) 05 > o C/3 O o H > "Z n > a 2 < CO t3 t t Table 1. (Continued) Taxox Metropolis Estimated Disjunction Ploidy E, East to West Trend Cystopfcris bttlhifcra Ophioglossum vulgaium "var, pseudopodunC* n, e. N, Amer. n. e. N. Amer. 650 1300 2x F. West to East Trend t Ctjstopteris X tennesseensis t Polystichuin tutmiium Cumberland Plateau w. N, Amer. t Gymnocarpiuni X hctcrosporuni Lake Superior Crypfogramma cri$pa van acrostichoides t ClwiJaulJics caaianca t PcUaca X ivrightiana t A^lcnittm scptmtrionale t Aspidotis densa t PolyilicJiiun X scopulinutn Vittaria sp. Gravuiiitis nhnhaia t HymcnophyUum wrighiii w. N. Amer. 400 550 650 900 4x 2x 3x* 2x w. N. Amer. 1000 3x s, w. N. Amer. w. N. Amer, w. N. Amer. n. w. N. Amer. 1000 1200 2200 2400 4x* 4x* 2x 4x* G. Sporophyte to Gametophyte Trend (Outliers all or mObtly gametophytes ) Tropics Cuba, Jamaica S, Japan 250 800 4500 ? ? ? Outlier & Reference Guadalupe Mts., N. Mex. (Blasdell, 1963) Kittatas Co., Wash. (Taylor, 1970) Jones, Craig Cos., N.C. (Wagner, 1965) Pennington Co., S.D. (Brooks, 1968) Blair Co., Pa. (Wagner, 1965) Isle Royale, Mich. (Soper, 1963; Fernald, 1950) Montgomery Co., Va.; additional localities in Va., W. Va. (Knobloch & Lellinger, 1969) Alexander Co., N.C. (Wagner, 1965) Monroe Co., W. Va. (Emory, 1971 ) Megantic Co., Quebec (Fernald, 1950) Gaspe Co., Quebec (Fernald, 1950) Rabun Co., Ga. (Farrar, 1971) Macon Co., N.C. (Farrar, 1971) Queen Charlotte I., B.C., Biorlca I., Alaska (Taylor, 1970) I M H o CO O o O O < > n > r > H CO to M 208 ANNALS [Vol. 59 veneris Towin th of its normal latitudinal limit. This soutli-temperate to subtropical fern is flourishing in the runnels of a system of hot springs, an obviously unusual site. Limitations of north extensions of many species into the Great Lakes area are probably due to cold winters and short growing seasons. Hay-scented fern, Dennstaedtia punctilohula^ of the southern and eastern United States has been taken in Michigan on only two occasions, these widely separated in time. (A third collection, allegedly from the Keweenaw Peninsula of the Upper Peninsula, by O. A. Farwell, is under question.) In 1889, G. H. Hicks recorded a collection from Shiawassee Co. (specimen in University of Michigan Herbarium), and in 1954 I discovered it in Jackson Co. Of special interest for our subject is the fact that the latter collection comprised two juvenile plants, probably offsets from a single plant, found growing in the depression made by the overturning of a large tree a transient habitat ideal for the growth of fern prothallia. In the loose, exposed sand there we found hundreds of gametophytes and young sporophytes of other fern species. One of the offsets of the Hay-scented fern was left in situ, but was gone the following year, having failed to survive the winter. Our best example of "fading out" of species going northward is Ebony spleen- wort, Asplenium plafyneuron. Among the most abundant of ferns in southern and eastern United States, it decreases rapidly in numbers north of the Wisconsin glaciation in Ohio, Indiana, and Illinois (well illustrated in the last state by Mohlenbrock, 1967: 157). By the time we reach Michigan and Wisconsin, the plant becomes very sporadic. Some "colonies" are but a single plant. Of 21 north of Saginaw Co. Michigan, only two University of Michi Station, Cheboygan Co., and has been found twice "pavements" in Chippewa Co. The suggestion seems plausible that this fern may be constantly re-established from wind-blown spores coming from the south. Single plants or small colonies may survive for a while and then perish. C. Eastern U. S. to Southwestern Michigan Trend, In the Upper Great Lakes area, two regions have special interest for their pteridophytic disjunctions, the area along the southeastern shores of Lake Michigan and the "Driftless Area" of Wisconsin. The former is the region involving the counties of three states running from approximately Chicago, Illinois, to Muskegon, Michigan. From the stand- point of plant geography, this locality is famous for the most spectacular disjunc- tion in the United States — Thismia americana ( Burmanniaceae ) which once grew near Lake Calumet, Illinois, separated from its nearest relatives in New Zealand and Tasmania (Swink, 1969; 411). Of disjunct pteridophytes, several have been detected only during the past few years, viz, American climbing fern, Lijgodium palmatum; the Log fern, Drtjopteris X celsa; and the Southern clubmoss, Lycopodium appressum. All are at least several hundred miles from their meti'opolises. The clubmoss has only recently been distinguished, because it grows together with the closely similar Bog club- moss, Lycopodium inundatum, with which it is readily confused. 1972] WAGNER — DISJUNCTIONS IN HOMOSPOROUS VASCULAR TLANTS £09 D. Eastern U. S. to ''Driftless Area' Trend, This well known focus of phytogeo- graphical interest boasts three outstanding examples of pteridophyte disjuncts, namely the Massachusetts fern, Thelypteris simuluta, a plant of low, sandy woods (Hartley, 1965); the Rock clubmoss, Lycopodium porophilum (also known as L. selago var. patens, but probably not the same; cf. Hartley, 1966; Wherry, 1961); and Lobed spleenwort, Asplenium X pinnatifidum (litis, personal communica- tion) — both of the latter plants of sandstone cliffs. E. East to West Trend. Very few pteridophytes show this pattern, in which the outliers are western. Most species which occur in both tlie East and West tend to skip the Great Plains and are actually "bimetropolitan" in that the eastern and western populations are more or less equivalent, neither being obviously the outlier of the other. Of bimetropolitan taxa I include such plants as Leather grapefern, Botrychium multifidum; Moonwort, B. lunaria; Least moonwort, B. simplex; Five- fingers or "American" maidenhair, Adiantum pedatum; Fragile fern, Cystopteris fragilis; Crested fern, Dryopteris cristata; Spinulosa woodfem, D. spinulosa; Spreading woodfern, D. assimilis; and Maidenhair spleenwort, Asplenium trichom- anes. Some of these are undoubtedly connected in the forested areas at the north of the Great Plains. All of them are wide-ranging and are known also in Europe and (or) Asia. They are probably disjunctive due to past historical events. Two examples which seem clearly to involve a metropolis in the East with ovitliers in the West may be cited. One is the Northern Adder's-tongue, Ophio- glossum vulgatum var. pseudopodum, which is a frequent and widespread plant north of the line of maximum Wisconsin glaciation from the eastern coast ( New Jersey to Quebec and Nova Scotia) west to Minnesota and western Ontario. This Adder's-tongue reappears in the Northwest, where it is exceedingly rare and local. Although it is possible that the western plants may represent a distinct variety, I doubt it. A second example, the familiar eastern Bulblet fern, Cystopteris bulb- ifera, occurs in a few very widely scattered localities in southwestern United States, the outlying stations hundreds of miles from each other. Bulblet fern is enabled to produce enormous colonics by forming gemmae, subspherical struc- tures, along the costae, which abscise and roll, and will germinate into new sporophytes if they reach suitable sites. The Adders-tongue, on the other hand, forms extensive colonies by root proliferation. Thus both of these plants are capable of developing large stands without sexual processes. F. West to East Trend. In terms of number of examples and distances, this * pattern of metropolis to outlier is the most striking of the directional trends. Some reach 2,000 or more miles in length of disjunction. Over half of those listed in Table 1 have been recognized only during the past 20 years, the outliers being so rare and local that they were overlooked. Wright^s cliffbrake, Pellaea X wrighti- ana, is known from a single granite outcrop in North Carolina. The solitary locality in Pennsylvania for the Hybrid oakfcrn, Gijmnocarpium X heterospornm, no longer supports a population, the limestone upon which it grew being now destroyed in the excavation of a quarry. The Forked spleenwort, Asplenium septentrionale y is so far represented in the eastern United States on only two rock cliffs in Virginia one-fifth of a mile apart. In view of the present incidence of pop- 210 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 ulations in the outlying eastern region of these and others of those Hsted, the conclusion seems inescapable that at least some of them may not represent relicts of former floras in which the taxa were once more abundant and widespread. Rather they have arisen by long-distance spore dispersal following the prevalent winds from the west. Alternatively, some may have originated from spores on nursery or garden stock transferred from one part of the country to another, or escapes of western plants cultivated in the east. G. SporopJiyte to Gamefophyte Trend. Surely the strangest disjunctions of homosporous vascular plants are those ferns that show dominant sporophytes in the metropolis and dominant gametophytes in the outliers. All of the examples Wa turn the world, especially in temperate rock outcropping regions. The examples involve tropical rainforest fern genera with gemmiferous gametophytes capable of dis- persal by few-celled bodies which abscise and germinate, if deposited in appro- priate sites. latitudes their sporophytes form sporadically if at all, according to the researches of Farrar at Iowa State University (personal communication) involving examples of several groups of Filmy-ferns ( Hymenophyllaceae ) , Shoestring ferns (Vittarioideae), .lyp In the wild these gametophytes are y> almost totally overlooked by vascular plant field botanists. Even bryologists, assuming presumably that the plants were algae, tended to disregard them in the past. Phycologists, upon encountering them, probably treated them as bryophytes. In view of our present knowledge, field botanists of all persuasions are urged to look out for these "ferns without sporophytes.*' They differ from "standard cordate fern prothallia in being narrower and more profusely branched. Some are mere branching filaments. All produce more or less specialized gemmae, and even if they form tiny sporophytes, they usually do not complete the alternation of generations. The classical generalization that homosporous pteridophytes have dominance of the sporophyte generation does not apply in the outlying popula- tions, and the plants behave more or less as do algae or bryophytes, occupying similar microhabitats on shaded cliffs and on damp tree bark. In the outlying populations the gametophytes no longer seem to "need" their sporophytes and spores, being able to propagate and spread entirely by their gametophytic gemmae. H. Indigenous to Naturalized Trend. More and more pteridophytic disjunctions are being reported in which the populations of the metropolis are indigenous members of the flora and the outlying populations have resulted from mans activities. Two illustrative areas of such disjunctions in the United States are given in Table 2 listing the suspected artificial introductions in the states of Florida and Hawaii. Some of the taxa included, e,g. Ladder brake, Pteris vittata; and so- called "Garden maidenhair fern, ?? atural in many places from cultivated garden and greenhouse plants. Some are con- servatory and hothouse weeds. In some cases the same species may be both native and naturahzed. 1972] WAGNER— DISJUNCTIONS IN HOMOSPOROUS VASCULAR PLANTS 211 Table 2. Artificial pteridophyte outliers in Florida and Hawaii. Species listed in theses probably native in that state. Metropolises: O — Old World tropics, N = New tropics, X = Pantropics. paren- World Selaginellaceae Ophioglossaceae Marattiaceae Polypodiaceae SCHIZAEACEAE Adiaxtaceae ASPLENIACEAE Azollaceae Florida*^ Hawah Sclaginella krausiana Ophioglossum petiolahim (May O, petiolatum (May be intro- be introduced in part.) Phijmatodes scolopendria {Phlebodium aurenm.) Lijgodium japonicum L. scandens (syn. L. micro- pJiyUutn, ) Adiantum hispidulnm Pellaea viridis ( Volunteer in orchid houses.) Pityrogramma calomelanos ( Probably introduced. ) Pteris cretica P. ensiformis P, multifida P. vittata P. tripartita duced in part. ) Angiopteris evecta P. scolopendria P. aurenm L, japonicum A. hispidulnm P. viridis F. calomelanos (P. cretica) P. vittata Ceraiopteris thalictroides Athyrium esculenfiim (In and A. esculentum (In and near near gardens.) A. japonicum Cyiiomium falcatum Thehjpteris torresiana T, deniata gardens? ) A. japonicum C. falcatum T. torresiana T. dent at a T, parasitica B. occidentale (Blechnum occidentale) Nephrolepis hirsutida (s.l.) N. hirsutiila (s.l.) N. cordifolia (Introduced in N. cordifolia (Introduced in part? ) part. ) N. hiserrata cv. "Furcans" Azolla filiculoides " Records from C. E. Delchamps in part. Metrop- olis O X (?)o o N o o o o N X o o o o o o o o o o o N o X ? N It is remarkable that many naturalized ferns seem now to be fully "at home" in their newly adopted countries. In Hawaii particularly certain introduced species are spreading well into the native forest, especially along the foot trails, and some appear as if they were native (Wagner, 1950). To analyze the phyto- geography of Hawaiian pteridophytes it is essential that we recognize which species are adventive through man's commerce and which indigenous. Failure to do so has given strange results: For example, MacCaughey (1918) listed six species as native (including one he called "endemic"!) which were actually brought in by man. Although ferns and other pteridophytes have not been especially popular garden plants in temperate zones, they flourish in tropical plantings. Green- house gardeners are familiar with the experience of receiving specimens of some 212 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Tabl£ 3. Considerations to determine whether disjunctive populations arose by (A) natural means or by (B) actions of man. Standpoint A. Natural B. Artificial 1, Documentation Data confirmed and dependable 2. Transport 3. Flora 4. Prominence 5. Abundance 6. Taxonomy 7. Remoteness 8. Dispersal 9. Pattern 10, Habitat 11. Associates 12. Genetics Unlikely to be carried by man Area in question poorly known Plant easily overlooked or confused Data unconfinnable and probably erroneous Likely to be carried by man Area in question well known Plant difficult to overlook or con- fuse Plants numerous and in more than Plants solitary or in only one colony one colony Population showing divergence Disjunction only a short distance Propagules capable of wide dis- persal Distribution pattern fits in with other, similar ones Physical environment typical for taxon Adjacent species native ones Reproductively self -compatible or apomictic Population showing no divergence Disjunction over a long distance Propagules incapable of wide dis- persal Distribution pattern entirely pe- culiar Physical environment atypical Adjacent species adventive exotic ones Reproductively self -incompatible and sexual the rare orchid or exotic gesneriad, for example, and losing the seedplant only to have it replaced by a foreign fern arising from spores in the soil. Tropical situations are probably more conducive to artificial introductions than temperate. However, in northern United States and southern Canada the Ostrich fern, Matteuccia, is the most popular garden fern. I wonder how many of the presumed "native" popula- tions of Ostrich fern are actually escapes from cultivation, from either broken rhizome fragments or wind-blown spores of nearby plantings. Even some of disjunctions listed in Table 1 may actually have resulted from man's activities. In this connection, Fosberg (1967) repeats a wonderful story of a disjunct population of an Asiatic water lily found in a lake in Idaho : "It was assumed that here was a remarkable instance of a wide disjunction in range until someone pointed out that some years earlier there had been a Chinese laundry on the shore of this lake.** What tests can we apply to evaluate whether a given disjunction was the result of artificial introduction or not? In Table 3 I have listed a number of considerations that must be applied in each case. Even mere carelessness or inad- vertent error in documentation (improper labeling especially) can lead to state- ments of range disjunctions that never existed and create serious problems in phytogeography. The data in herbarium or publication may be erroneous but 1 ature Mich example, several reports of Oliver A. Farwell are under question, these involving very large disjunctions in the conceme artificial introductions with increasing comr used as a sort of "Index of Autochthonism" Table 3 can be the exDandino: nroblems study of plant distribution. As Fosberg (1967) 1972] WAGNER— DISJUNCTIONS IN HOMOSPOROUS VASCULAR PLANTS £13 however, "it is frequently difficult to prove conclusively that a species was brought by man in the absence of documentary evidence directly involving the intro- duction." Intrinsic Factors Bearing Upon Pteridopiiyte Disjunctions Thus far we have considered mainly extrinsic factors bearing upon disjunc- tions. Geological history and floristic changes account for some patterns; climatic conditions for others; and human intervention for others. Prevailing winds may have produced the outliers of taxa with primarily western metropolises. Cold temperate conditions may have influenced primarily tropical rainforest species to change from the dominantly sporophytic expression of the metropolis to the dominantly gametophytic expression of the northern outposts. What of intrinsic factors such as chemical differentiation? Chromosomes? Examples of a few of these will be enumerated below. Chemistry, Comparison of chemical aspects of disjunct populations in pterido- phytcs is still in its infancy. The case of the rare "J^P^i^i^se grapcfern," Botrychium ternatum^ in the region of the St, Lawrence Seaway poses a number of questions. I had interpreted plants of the outlying American populations to be the same taxonomically as those of the metropolis in Japan, China, and India (Wagner, 1959; Wherry, 1961). I am still unable to separate them morphologically. My student, David M. Smith (1967), however, has shown that the American taxon differs in presence or absence of six phenolic compounds from the Japanese. In spite of these chemical conti'asts I am inclined to continue treating the American plant as only varietally distinct at best. Will we find similar chemical differentia- tion in such disjunctive plants as Lijcopodium lucidulum, Osmunda claytoniana, and Onoclea sensibilis? Should we treat separated populations which are morpho- logically alike but chemically unlike as distinct varieties? w Cytology, In Table 1, the current knowledge of level of polyploidy is indicated for the various patterns of disjunction. Most disjunctive taxa have the same chromosome number in their outliers as in their metropolis, but our knowledge is still limited and some of the examples in Table 1 may be shown to have different polyploid levels represented. One of the best known examples of polyploid differ- entiation is the Hart's-tongue, Phyllitis scolopendrium. Its metropolis in eastern Eurasia and north Africa comprises diploids {cf. Tryon, 1969: map 4), but its outliers in Japan and northern North America are tetraploids. The western vari- ety of the Oak fern, Gymnocarpium dryopteris var. disjunctum is practically indis- tinguishable morphologically from the eastern American plant, but the former possesses 2x = 80, the latter 2x ~ 160. The Hawaiian Palaa, Sphenomeris chinensis (S. chusana of authors) is a diploid with n — 47, but all those populations of this taxon so far studied in India, Ceylon, and Japan are polyploid with n = 94 or near that number. There are two important facets of polyploidy to consider in relation to dis- junction, namely the effect it would have on the fate of the resulting population if the arriving spore were polyploid, and what causes arriving populations to remain diploid or change to polyploid, Tryon (1970) noted that Hawaii constitutes the 214 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 C< cc only notable exception to the rule that ferns, in general, do not develop endemic species flocks" on islands. He concluded that in fern floras of oceanic islands adaptability of the genotype-phenotype of the single spore to a new environment and sensitivity to selection are characteristics that dominate success in migration and evolutionary potential under geographic isolation. Our data on Hawaii seem to fit this conclusion: the incidence of pol>^loidy is approximately twice as great among the non-endemic taxa as it is among the endemic (Wagner & Wagner, unpubl.). According to our present evidence of ca, 85 percent of the fern taxa (exclusive of Ophioglossaceae and Gleicheniaceae), over 60 per cent of the non- endemic indigenous species are polyploid (3x, Ax, 6.t, and 8x), but less than 30 per cent of the endemics are polyploid. Also, not one of the members of the endemic species flocks" in Hawaii displays polyploidy (namely Cibotium with 5 species, Sadleria with 4, Diellia with 5, and Adenophorus with 8). In contrast, however, among the solitary endemics, including some of Hawaii's most distinctive taxa, there are several {Pteris lidgatii, Thehjpteris {^'T oppingia' ) keraudreniana, and Microsorium spectrum) which are tetraploids. These may constitute last remnants of earlier "species flocks" which evolved high levels of differentiation from their ancestors and are now senescent, being represented by isolated vestigial species, Hybridity. Seven of the disjuncts discussed in this paper are interpreted as inter- specific hybrids rather than normal, divergent species. They show metropolis-to- outlier disjunctions ranging from ca. 200 to over 2,000 miles. The outliers may be separated from one or both presumed parents. Five of the intermediates are sexual allotetraploids, and two are "sterile" allotiiploids. There is no problem of accounting for long distance disjunction in the sexual hybrids, as they can spread by spores in the same manner as normal species. Our present inteipreta- tions of their origins are as follows: Polystichum X scopulinum — P. moJirioides x munitum Dryopteris X celsa — D. goldiana X ludoviciana Asplenium X pinnatifidum = A. montanum X rhizophylliim Pellaea X ivrightiana — P. longimucronata X ternifolia Cystopteris X tennesseensis = C bulbifera X protrusa The widest transcontinental disjunction of these is shown by Polystichum X scop- ulinum, the outlier skipping 2,400 miles of intervening country and confined to Gaspe Co., Quebec. This hybrid is a well known western American holly fern in its extensive metropolis where it occurs in both sterile diploid and fertile tetiaploid forms. It is intermediate between the two species with which it most commonly grows — the amphitropical montane P. mohrioides and the extremely abundant and widespread endemic North American P. munitum. The former, P. mohrioides, has apparently produced no outliers east of Washington, Oregon, and California. However, the latter, P. munitum, which is vastly more abundant in its metropolis, has recently had an outUer reported as far east as South Dakota. A much greater problem lies in the explanation of disjunctions of sterile triploids. Dryopteris filix-mas X marginalis, despite its abnormal spores, forms remarkably large local populations in certain bcahties in its metropolis, e.g. Bruce 1972] WAGNER— DISJUNCTIONS IN HOMOSrOROUS VASCULAR PLANTS 215 Peninsula of Ontario and a few offshore islands of Lake Superior. It normally occurs with the parents near by. Gijmnocarpium X heterosporum combines the characters of G. dnjopteris (n = 40, 80) and G. rohertiomim (n = 80), None of the populations of the intermediate Oak fern has yet been discovered with normal spores. How such plants can spread if at all we can only guess. Perhaps it is by dispersion of unreduced spores (c/. Morzenti, 1967). Morphology. More or less subtle anatomical differences are likely to be found between separated populations. These differences may be sufficient to designate them as different varieties or subspecies. For example, the Asiatic counterpart of Osmunda claijtoniana differs from its eastern American representative in its hairiness (it has been separated as "var. vestitumr) A particularly interesting taxonomic-geographic situation arises when we discover a disjunction of one variety in the area of another. No better example can be found than the situation in Bracken ferns, Pteridium aquilinum sensu lato. Most of the so-called "varieties" are allopatric and some are widely separated from the others, such as "var. decompositum' in the Hawaiian islands, an obviously highly disjunctive taxon. Brackens have been lumped by many authors into one species — "the only species of the monotypic genus Pteridium . . . ." (Tryon, 1941). However, a real question exists here, for two of the so-called "varieties" may also grow sympatrically. In March, 1970, I observed taxon arachnoideum intergrowing in abundance with taxon can datum over a very large area of the Blue Mountains of Jamaica. The clones of these ferns differ respectively from each other in many obvious characters and they produce mixed tangles which are almost impenetrable, but which afford an excellent opportunity to compare under like site conditions. The metropolis for arachnoideum is mainly South American, while that for caudatum is southern North America and Central America. However, both coin- cide over a very large area of the islands of the Caribbean and the land mass of Central America. Significance of Disjunctions in Homosporus PTERmoniYTES Do long-distance disjunctions in homosporous pteridophytes necessarily give evidence of major geological and biological events in floristic history? If the disjunctions are between large and broadly distributed populations, they may indeed be highly significant as indicators of past events and relict floras. On the contrary, where the disjunctions exist between a large population center or metrop- olis and one or a few widely separated small populations or outliers, the signifi- cance may be less. Perhaps some of these are actually mere casual and short-lived introductions, such as those occasionally noted, say, among the butterflies, which establish themselves in a given area for a few years only to disappear. Spores of the homosporus pteridophytes are capable of long-distance dispersal by wind (Ewan, 1945; Tryon, 1970). Even though the statistical chances of a given spore reaching a given point rapidly become smaller the farther away we go from the source point, at least some spores must be able to reach higlily isolated positions as in Hawaii where the nearest sources over perhaps millions of years have been well over 2,000 miles away. The question of whether a single spore, by itself, 216 ANNALS [Vol. 59 can initiate and establish a population now seems to be satisfactorily answered in the work of Klekowski and Baker (1966^ bibliography; present symposium), who find that ferns indeed are capable of intragametophytic selfing and that thus a single spore can start a colony. Many of the disjunctions cited in this paper (those marked with daggers in Table 1 ) may be casual introductions in fairly recent times rather than "relicts'" in the sense of persistent remains of ancient floras. Stanley A. Cain (1944) wrote, "Minor discontinuities of areas probably frequently result from recent migrations, but major disjunctions seem almost exclusively to have resulted from historical causes which have produced the disjunctions, in a once more nearly continuous area, through destruction or divergent migrations caused by climatic or some other changes." However, in homosporous pteridophytes, if it is ti'ue that spores can be carried long distances and produce self -fertilizing gametophytes, then we must be cautions not to read too much into long-distance disjunctions, especially those with well marked metropolis and outliers. Literature Cited Alston, A. H. G. 1960. Some new species of ferns from South America. Lilloa 30: 106- 112. Billington, C. 1952. Ferns of Michigan. Cranbrook Institute of Science Bull. No. 32. Blasdell, R. F. 1963. A monographic study of the fern genus Cystopteris, Mem. Torrey Bot. Club 21(4): 1-102. h BozEMAN, J. R. 1968. A note on Gyrnnocarpium dryopteris (L.) Newm. in North Carolina. Castanea 33: 258. Brooks, R. 1968. Polystichum munitum in South Dakota. Amer. Fern Jour. 58: 92. Broun, M, 1938. Index to North American Ferns. Lancaster, Pennsylvania. Cain, S. a. 1944. Foundations of Plant Geography. New York. Christ, H. 1910. Die Geographic der Fame. Jena. Clausen, R. T. 1938. A monograph of the Ophioglossaceae. Mem. Torrey Bot. Club 19(2) : 1-177. Ehrlich, p. R. & P. H. Raven. 1969. Differentiation of populations. Science 165: 1228- 1232. Emory, D. L. 1970. A major North American range extension for the Forked Spleenwort. Amer. Fern Jour. 60: 132-133. Erickson, R, O. 1945. The Clematis fremontii var. richlii population in the Ozarks. Ann. Missouri Bot. Card. 32: 413-460. EwAN, J. 1945. Sources of the fern flora of Colorado. Amer. Fern Jour. 35; 114-128. Farrab, D. R. 1967. Ganietophytes of four tropical fern genera reproducing independently of their sporophytes in the southern Appalachians. Science 155: 1266-1267. Fernald, M. L. 1935. Critical plants of tlie Upper Great Lakes region. Part I. Rlio- dora 37: 197-222. . 1950. Gray's Manual of Botany. 8th ed. New York. FoSBERG, F. R. 1967. Some ecological effects of wild and semi-wild exotic species of vascular plants. lUCN Publ. New Ser. 9: 98-109. Hagenah, D. J. 1955. Notes on Michigan pteridophytes. I. New county records in Os- mundaceae and Polypodiaceae. Amer. Fern Jour. 45: 65-80. . 1966. Notes on Michigan pteridophytes. Amer. Fern Jour. 56: 150-163. Hartley, T. G. 1965. Discovery of the Massachusetts fern in Wisconsin. Rhodora 67: 399-404. . 1966. The flora of the "Driftless Area." Univ. Iowa Stud. Nat. Hist. 21(1): 1-174. Klekowski, E. J., Jr. & H. G. Baker. 1966. Evolutionary significance of polyploidy in the Pteridophyta. Science 153: 305—307. Knobloch, I. W. & D. B. Lellinger. 1969. Cheilanthes castanea and its allies in Virginia and West Virginia. Castanea 34: 59-61. 1972] WAGNER— DISJUNCTIONS IN HOMOSPOROUS VASCULAR PLANTS 217 MacCaughey, V. 1918. An ecological survey of the Hawaiian pteridophytes. Jour. Ecol. 6: 199-219. MoHLENBROCK, R. H. 1967. Ferns. The Illustrated Flora of Illinois. Carbondale, Illinois. Morton, C. V. 1958. Diplazium japonicum naturalized in Florida. Amer. Fern Jour. 48 (no. 1): 28-30. MoRZENTi, Virginia M. 1967. Asplenhim plemitn: a fern which suggests an unusual method of species formation. Amer. Jour. Bot. 54: 1061-1068. PIPPEN, R. W. 1966. Lygodium palmatum, the climbing fern, in southwestern Michigan. Michigan Bot. 5: 64-66. Raven, P. H. 1963. Amphitropical relationships in the floras of North and South America. Quart. Rev. Biol. 38: 151-177. Shaver, J. M. 1954. Ferns of Tennessee. Nashville. Smith, D. M. 1967. Chemical characters as a guide to the taxonomy of Bolrychium, Ph.D. Dissertation, Univ. of Michigan. SoPER, J. H. 1954. The hart's-tongue fern in Ontario. Amer. Fern Jour. 44: 129-147. . 1963. Ferns of ManitouHn Island, Ontario. Amer. Fern Jour. 53: 28-40; 71-81; 109-123. SwiNK, F. 1969. Plants of the Chicago Region. Lisle, Illinois. Taylor, T. M. C. 1967, Mecodium torightil in British Columbia and Alaska. Amer. Fern Jour. 57: 1—6. , 1970. Pacific Northwest Ferns and tlieir Allies. Toronto. Tryon, R. 1941. Revision of the genus Pteridium, Rhodora 43; 1-70. — ' . 1969. Taxonomic problems in the geography of North American ferns. BioScience 9: 790-795. 1970. Development and evolution of fern floras of oceanic islands. Biotropica 2: 76-84. Wagner, W. H., Jr. 1950. Ferns naturalized in Hawaii. Occ. Pap. Bishop Mus. 20: 9-5- 121. . 1959. American grapefems resembling Botrychium ternatu?7i: A preliminary report. Amer. Fern Jour. 49: 97-103. — . 1965. Pellaea wrighilana in North Carolina and the question of its origin. Jour. Elisha Mitchell Sci. Soc. 81: 95-103. — . 1966. New data on North American oak ferns, Gymnocarpium. Rhodora 68: 121- 1970. Evolution of Dryopteris in relation to the Appalachians. Pp. 147-192 in Holt, 138. E. Perry (editor), "The Distributional History of the Biota of the Southern Appalachians." Virginia Polytech. Inst, and State Univ. Res. Div. Mon. 2. — . 1971. The southeastern Adder's Tongue, Ophioglossum vulgatum van pycno- sticum, found for the first time in Michigan. Michigan Bot. 10: 67-74. -^, D. R. Farrar & B. W. McAlpin. 1970. Pteridology of the Highland Biological Station area, southern Appalachians. Jour. Elisha Mitchell Sci. Soc. 86: 1-27, — & Lois P. Lord. 1956. The morphological and cytological distinctness of Botrychium minganense and B, lunaria in Michigan. Bull. Torrey Bot. Club 83: 261-281. — & A. J. Sharp. 1963. A remarkably reduced vascular plant in the United States. Science 142: 1483-1484, cover. — , F. S. Wagn-er & D. J. Hagenah. 1969. The Log Fern {Dryopteris celsa) and its hybrids in Michigan — A preliminary report. Michigan Bot. 8: 137-145. Walker, S. 1962. Further studies in the genus DrijojHeris: the origin of D. clinfoniana, D. celsa and related taxa. Amer. Jour. Bot. 49; 497-503. . 1969. Identification of a diploid ancestral genome in the Dryopteris spinulosa com- plex. Brit. Fern Gaz. 10: 97-99. Wherry, E. T. 1939. Recent fern finds in West Virginia. Castanea 4: 1-4. . 1961. The Fern Guide. Garden City, New York. Wlnkler, H. 1938. Geographic. Pp. 451-472, in F. Verdoom, "Manual of Pteridology, The Hague. » ££■» -• V--WT*. Tm-r-i?> THE FLORISTIC DISJUNCTIONS BETWEEN THE MONTE tfCi-^y-w^T-^^T-* 1 -mT T-*. T-1 r^ T-1 -r-» m" IN ARGENTINA AND THE "SONORAN DESERT IN MEXICO AND THE UNITED STATES' Otto T. Solbrig^ One of the best known but most puzzling disjunctions is that between the temperate regions of South America and the temperate regions of North America, It is puzzhng because it involves a large number of floristic elements, and because it is evident from the geological data available that the two regions have never been closer to each other than they are today ( Dietz, 1961 ) . Raven (1963, 1971) has addressed himself to this problem and has very accu- rately pointed out that we are dealing with more than one type of disjunction. There is the disjunction involving elements that grow in the cold regions of South and North America; the disjunctions involving plants that grow in areas with a Mediterranean type of climate in both continents; the disjunctions between areas with a mesic hydric and temperature regimes; and finally the disjunctions between the so-called "desert areas/' In his reviews^ Raven presents evidence that indicates that these disjunctions are relatively recent and that they probably have arisen almost exclusively by long-range dispersal. However, Raven did not discuss in detail the disjunction involving the desert areas. This is the best known one (Bray, 1898, 1900; Campbell, 1944) and prob- ably the most controversial. It also involves the largest percentage of the flora. In this paper I want to present data that lead me to believe that we are dealing with patterns of distribution that have been established over a long period of time and that no one single explanation can encompass them all. In order to clarify the discussion, only plants that grow in the phytogeographical province of the "Monte" in Argentina and the "Sonoran Desert" in Mexico and the southwestern United States will be discussed. Description of the Two Regions Monte characteristics (Hauman, 1947; Morello, 1958; Shreve, 1951; Shreve & Wiggins, 1964; Solbrig, 1972). Basically they are ecological semideserts with mild winters and hot summers, rainfall that rarely exceeds 300 mm, and negative evapotrans- piration throughout most of the year. The rainfall regime varies within both areas. Each region has areas of winter rainfall, winter and summer rainfall, and only summer rainfall. with predomi These show a series of adaptations to a defi- cient water economy such as small leaves, sunken stomata, thick cuticles with ^ Work done under a National Science Founda ^ Department of Biology and Gray Herbarium setts 02138. Ann. Missouri Bot. Card. 59: 218-223. 1972. 1972] SOLBRIG — DISJUNCTIONS BETWEEN THE "MONTE" AND THE "SONORAN DESERT" £^9 waxy and resinous exudates, etc. In addition to the shmbs, other conspicuous elements are succulents, particularly Cactaceae, Bromeliaceae, Agavaccae, and Fouquieraceae. A third element is the annual flora, which grows in conjunction with the rains and consequently, both winter and summer annuals can be identi- fied. In areas that span such latitudinal and altitudinal ranges as the Monte and the Sonoran Desert, by necessity variations in all kinds of physical and biological parameters will be encountered. There will be truly "desert" elements as well as riparian elements that because of their ability to tap underground water are able to escape to a degree the limitations that lack of water impose on plant growth. This should be borne in mind when the disjunctions are discussed in detail. It also applies to the density of the vegetation and total coverture that one is likely to find. Although basically both regions are characterized by less than 100% coverture, this is not always so, particularly in those areas where more humidity is locally available for one reason or another. The Floristic Similarities The flora of the Sonoran Desert has been recently studied by Shreve and Wig- gins (1964). Although they give no exact figures, a rough estimate is that it is formed by approximately 3,000 species. No comparable work exists for the Monte, the closest being Morello's (1958) study of the vegetation, which cites 307 species in a list that does not wont to be complete. The exact figure is prob- ably closer to 3,000. Of this total of roughly 5,000 species growing in both areas, only some 100 to 150 are disjunct or pairs of closely related species, roughly 2 per cent. The number of genera in common is much higher, the regions sharing approximately 50 of the roughly 500 genera growing in both areas, that is approx- imately 10 per cent. On a purely numerical basis the common elements are not very great. What is impressive is that among the elements that are shared by both areas are some of the most important species of both areas. First in importance is the creosote bush, gobernadora or jarilla, Larrea divari- cata, the most important perennial shrub in the Sonoran Desert and in the Monte. In both areas Larrea divaricata occupies hundreds of square miles witli an abun- dance of 90% or more. In the Monte two other species of Larrea, L. ctineifolia and L. nitida, often replace L. divaricata, but the latter is still the most important species. The second most important element with a disjunct distribution is the genus Prosopis, sect. Algarrohia, the "mesquitc" and "algarrobos." Although tliere are no common species between South and North America, P. chilensis from South America and P. jtdiflora from North America are very closely related and were considered conspecific at one time. These trees are the principal riparian element in both the Monte, where they have speciated extensively, and in the North Amer- ican deserts. A number of other floristically abundant shrubs and trees also show disjunct distributions or have pairs of closely related species at the northern and southern sites. Among them are tlie genus Celtis in the family Ulmaceae, represented in South America by C. spinosa and in North America by C. pallida and C. laevimta- 220 ANNALS OF TIIE MISSOURI BOTANICAL GARDEN [Vol. 59 ?? tlie genus Condalia in the Rhamnaceae, represented in the Sonoran Desert by Condalia lascioides and C. spathulata and in the Monte by C. microphylla; also in the Rhamnaceae the genus Ziziphus, represented by Z. mistol in the Monte and by Z. sonoriensis in the Sonoran Desert; in the Leguminosae we find among others Caesalpinia gillesii with a disjunct distribution and the genus Cercidium with closely related species at both sites; and finally in the Capparaceae, the species Atamisquea emarginata is a very conspicuous element of the flora of both the Monte and the Sonoran Desert. All these species are important elements of their respective vegetations, being widespread and abundant. Members of the family Cactaccae are very conspicuous elements of the flora of the Monte and the Sonoran Desert, Although several genera are shared by both regions, notably Opuntia and Cereus, no close relationship at the specific level is known that can not be traced to a recent introduction by man ( such as the "choya cactus Opuntia fulgida, introduced into Argentina in historic times). There is however a great ecological and morphological similarity, notably among the giant "saguaro" and "organ" type cacti. In addition to the species and genera mentioned, a great number of species and genera which occupy a much less conspicuous role in the vegetation have species in common, or pairs of closely related species. Although some of these genera are woody, most notably some of the compositac, the majority are herbaceous. The Distributional Patterns There are three main explanations that can be brought forward to explain the floristic similarities between North and South America. They are as follows: that the disjunction has arisen as a result: (1) of long-distance dispersal; (2) of a once continuous distribution; or (3) of parallel convergent evolution. All three expla- nations have been brought forward, and there are reasons to believe that all three are applicable in some instances. Long-range dispersal. It will never be possible to say conclusively that a certain pattern of distribution is due to long-range dispersal. However, with a judicious use of Okam's razor, circumstantial evidence may point to long range dispersal as the most likely explanation. For the Monte-Sonoran disjunction, long-range dispersal is obviously the simplest explanation if the dispersal agent is not considered. However, it becomes much more difficult to see how long-range dispersal can be effected if the possible dispersal agents are considered. Wind as a dispersal agent can be discarded from the onset. The flow of air, both at the surface and in the upper atmosphere, at every time of the year is such that it is impossible for seeds to be carried from one area to the other by wind across the equator. Neither is there any evidence that the flow of air masses was different in the past. We can therefore safely discard wind as an agent of long-range dispersal between the Monte and the Sonoran Desert. This leaves only animals as possible agents of dispersal. Obviously more knowledge is needed, and particularly experimental work to determine the effec- tiveness of birds as agents of dispersal. Species of birds are known that fly from the United States to Argentina in their migration flights. None of these inter- 1972] SOLBRIG— DISJUNCTIONS BETWEEN THE "MONTE" AND THE "SONORAN DESERT" £21 continental migrators are birds of the desert areas, however. To consider them as agents of dispersal it must be hypothesized that either regularly or occasionally they stop at the periphery or the center of the areas under consideration. A further complication is the well-known behavior trait of migrating birds of flying with an empty gut. Seeds still could be carried on feet or plumage. In any case, birds as agents of migration can not be ruled out, but neither are they a very likely source. What makes long-range dispersal attractive as an explanation in certain cases is the pattern of distribution of certain species. Axelrod ( 1950) has shown that the vegetation of the North American desert areas has two main sources: one is the so-called Arcto-Tertiary element, a northern, cold-temperate source; the other is the so-called Madro-Tertiary element from a warm-tropical source. Raven (1971) has pointed out that because of its geographical situation the southern hemisphere has not produced an equivalent of the Arcto-Tertiary flora. Although this interpre- tation is open to discussion and more work on the fossil floras of the southern hemi- sphere is needed to shed light on this point, it is undoubtedly true that most of the elements of the flora of the dry temperate areas of South America have a tropical origin as Raven (1971) has pointed out. This is also true of the Monte, where Leguminosae, Zygophyllaceae, Rhamnaceae, Cactaceae, etc. are the dominant elements. If this line of reasoning is accepted, then it is tempting to assume that disjunct species belonging to northern genera not represented in the tropics have been dispersed by long-range dispersal. Particularly some grasses such as Bephlaridagne henthammianae, Andropogon saccharoides^ Sfipa speciosa; Compositae such as Verbesina encelioides, Encelia farinosa, etc. that possess small seeds, are herba- ceous, either annual or short-lived perennials, might have dispersed this way. In some cases such as Leptochloa uniflora the pattern of dispersal involves a scries of steps across the equator. In summary, although it is hard to envision how exactly seeds are carried across the equator, certain patterns of distribution are easiest to explain assuming long- range dispersal in recent times. Continuous distribution across the tropics. The greatest intellectual block in accepting this explanation is the existence of a tropical belt between the Monte and the Sonoran Desert. It is well known that plants are good dispersers across physical gaps such as water barriers or mountain ranges, but poor dispersers across climatic barriers. Although this is an undeniable fact, it might not be as formidable a considera- tion as it appears. Plants are capable of tolerating a greater set of environmental conditions than those to which they are exposed, as witnessed by the variety of ambients where species can grow under cultivation. What usually happens is that although plants are able to grow, they are not able to compete under the warmer climatic regimes. At present, desert conditions exist scattered across the tropics of South Amer- ica, and what evidence exists points to the existence of such areas also in the past ( Harrington, 1962; Vuilleumier, 1971 ) . Such dry areas exist mostly in the lee of mountains in the western parts of the continent. Competition in such areas is 222 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 mostly for water, and it is possible that plants from more mesic, but dry areas, such as the Monte and the Sonoran Desert, could have used these areas as stepping stones in moving from North to South or vice versa. Indirect evidence for such an explanation is found in the present distribution of certain characteristic shrubs and trees. Cercidium is a genus of seven species. Two of these are found in the Monte and adjacent areas; four in the Sonoran and Chihuahuan Deserts; and the last, C praecox, has an almost continuous distribu- tion from Mexico to Peru, throvigh the Antilles. Cercidium praecox is so closely related to the Monte species that it is considered conspecific by some botanists, and although the Nortli American species are more distinct, it appears very likely that the genus spread from North America to South America or vice versa through the intervening dry tropical areas. The genus Bulnesia in the family Zygophyllaceae is formed by seven species. It is obviously of tropical origin and speciated in the Chaco and adjacent areas of South America. The species Bulnesia retama is a conspicuous element of the flora of the Monte. This is the most xeric species of the genus. The distribution of Bidnesia retama is of interest because it extends from Argentina to Venezuela, again hopping through the dry areas. Although Bulnesia is not found in the Sonoran Desert, it conceivably could get there in the future if populations can become established in the Antilles. Also in the Zygophyllaceae, the distribution oi Larrea divaricata is of interest. This dominant element of the Monte and So- noran Desert has a few intermediate stations in Bolivia and Peru. It is tempting to postulate that the species once had a broader range of distribution in the tropical areas. In brief, there are patterns of distribution particularly among the shrubs and trees that point to a more continuous distribution across the islands of xeric climate in the American tropics, and this explanation at least for cases such as Cercidium can not be ruled out. Convergent evolution. The final alternative explanation for the disjunct distribution says that we are not dealing with a disjunction at all, but with a case of convergent evolution. The proponents of this hypothesis (Barbour, 1969) state that those elements that are tropical in nature might have produced, under the selection of the similar climates of the desert region of the Monte and the Sonoran Desert, similar or identical types through convergence. The major problem with this hypothesis is to find an independent criterion to discriminate between it and the other two. Nevertheless, the distribution of the genus Prosopis is suggestive of the validity of the idea of convergence, at least in some cases. Prosopis is one of the most primitive genera of the Mimiosoid-Legumes. Spe- cies of Prosopis grow in the Old World (Asia Minor and East Africa) as well as in the Americas, where they are found from the southern United States to northern Argentina. The distribution suggests an early Tertiary or even Cretaceous origin for the genus, since at that time Africa and South America were joined together. It also suggests a tropical origin. This idea is reinforced by the distribution of the sections, with the primitive section Algarrobia being both tropical and subtropical in distribution, while all the other more specialized American sections are either found in the mountains or temperate regions. 1972] SOLBRIG — DISJUNCTIONS BETWEEN THE ^'MONTE** AND THE "SONORAN DESERT" £23 If we look at the various species of Frosopis^ sect. Algarrobia, we find that the less specialized species are found in the more mesic areas both in the tropics and in the temperate zones, while the more specialized species are found in the Monte and the Sonoran Desert. This is accompanied by the development of similar morphological specializations in leaf structure and habit. The convergence is^ however, not so great as to put any doubt on the separate origin of the North and South American desert species. However, presumably the process could proceed to the point where the species may become identical morphologically. If this were accompanied by the extinction of the parental species, the resulting distributional pattern would be suggestive of long-range dispersal. Summary and Conclusions The American tropics are flanked by areas with semidescrt climate. These xeric belts possess some floristic elements in common. Since they are separated by some 6,000 miles across tropical climate, this presents a formidable problem. Three hypotheses to explain those disjunctions have been proposed: (I) long- range dispersal; (2) continuous distribution across the tropics in the past; and (3) convergent evolution from tropical ancestors. There hypotheses cial cases. What is now needed is careful, detailed work with the species involved in order to obtain more insights into their genetics, cytology, chemistry, and eco- logical relationships. Literature Cited AxELROD, D. I. 1950. Evolution of desert vegetation, Publ. Carnegie Inst, Washington 590: 215-306. Barbour, M. 1969. Patterns of genetic similarity between Larrea divaricata of North and South America. Amer. Midi. Naturalist 81: 54-67. Bray, W. L. 1898. On the relation of the flora of the Lower Sonoran zone in North America. Bot. Gaz. 26. , 1900. The relation of the North American flora to that of South America. Science 12: 709-716. Campbell, D. H. 1944. Relations of the temperate floras of North and South America. Proc. California Acad. Sci. Ser, 4; 25. DiETZ, R, S, 1961. Continent and ocean basin evolution by spreading of the sea floor. Na- ture 190: 854-857. Harrington, H. J. 1962. Paleogeographic development of South America, Bull. Amer, Assoc. Petro. Geo!. 46: 1773-1814. Hauman, L. 1947, Provincia del Monte. In GAEA, "Geografia de la Republica Argentina." 8: 208-248. MoRELLO, J. 1958. La provincia fitogeografica del Monte. Opera Lilloana 2: 1-155. Ra\^n, p. H. 1963. Amphitropical relationships in the floras of North and South America. Quart. Rev. Biol. 38: 151-177. , 1971. The relationships between mediterranean' floras. Pp. 119-134, in P. H. Davis, P. C. Harper & I. C. Hedge (editors), "Plant Life of South-West Asia." Edinburgh. Shreve, F. 1951. Vegetation of the Sonoran Desert. Carnegie Inst. Washington Publ. 591. & I. L. Wiggins. 1964. Vegetation and Flora of the Sonoran Desert. 2 vols. Stan- ford. SoLBRiG, O. T. 1972. New approaches to the study of disjunctions with special emphasis on the American amphitropical desert disjunctions. Pp. 85-100, in D. Valentine (editor), "Taxonomy, Phytogeography and Evolution," London. VuiLLEUMiER, B. 1971. Plcistocenc changes in the fauna and flora of South America. Sci- ence 173: 771-780. SPECIES DISJUNCTIONS IN LARREA: EVIDENCE FROM MORPHOLOGY, CYTOGENETICS, PHENOLIC COMPOUNDS, AND SEED ALBUMINS' J. H. HuNziKER,- R, A. Palacios,2 Amalia G. de Valesi^ AND LiDIA POGGIO^ Larrea divaricata^ the "creosote busli " is often regarded as the most drought tolerant higher plant in North America (Barbour, 1969; Morello, 1955). It is probably a single species having an enormous disjunct distribution. In North America it extends from the State of Nevada to the State of Hidalgo in Mexico, where it is called "gobernadora'* on account of its dominance in the natural vege- tation (Fig. 1). In South America it occurs in several isolated places in Peru (lea, Nazca, Chuquibamba, Moquegua), in Bolivia (Tarija), and in Chile (Atacama, Concepcion). In Argentina it is called "jarilla," and it extends from Salta in the North to North Patagonia (Chubut) (Fig. 2). As pointed out by Barbour the disjunction reaches 36 degrees latitude, that is, nearly 4,000 km (Barbour, 1969). The question that immediately arises, when considering such an extense distri- bution is whether we are dealing with a single species or at least two vicariant species: Larrea tridentata (North America) and L. divaricata (South America). How much have these populations diverged since they became separated? Typ- ical Larrea divaricata was described from Mendoza, Argentina, South America, and the North American taxon has been named L. tridentata. Morphologically, they can easily be separated by the form of the stipviles, the North American populations having acute while the South American ones have obtuse stipules. They could be distinguished, therefore, as different subspecies as done recently by Felger and Lowe (1970) who recognize L. divaricata Cav. subsp. divaricata for the South American subspecies and its ecotypes alid subsp. tridentata (Sesse & Moc. ex DC.) Felger & Lowe for the North American subspecies and its eco- types. BlOSYSTEMATICS OF THE GeNUS LaRREA A brief consideration of the biosystematics of the genus will help us to have an idea of the degree of morphological diversity and the genomic differentiation of ^ The authors wish to express their gratitude to several persons that in one way or another helped their research. Mrs. Vikna B. de Fernandez perfonned the protein electrophoresis ex- periments; Miss Aida De Gregorio made part of the cytological preparations; Dr. J. Morello gave us some data on the present day distribution of L. divaricata in Santiago del Estero, Argen- tina, Dr. J. Rzedowski and Dr. Otto T. Solbrig collected seed samples in the field; Dr. Tien Wei Yang very generously supplied us with many seed samples identified insofar as chromosome number is concerned. Financial support through a National Science Foundation (U.S.A.) grant to Dr. O. T, Solbrig and two grants from the Consejo Nacional de Investigaciones Cientificas y Tecnicas to the senior author is acknowledged. ^ Departamento de Ciencias Biologicas, Facultad de Ciencias Exactas y Naturales, Ciudad Universitaria, Buenos Aires, Argentina. I Ann. Missouri Bot. Gard. 59: 224-233. 1972, 1972] HUNZIKER ET AL.— DISJUNCTIONS IN LARREA 225 Figure 1. Chi-omosome numbers of creosote bush populations in Nortli America (Larrea divaricata subsp. tridentata). Diploid populations 1-17 occur in the Chihuahuan Desert and adjacent desert grassland. Tetraploid populations 19-33 are situated in the Sonoran Desert. Hexaploid populations 35-41 are located in the Mojave Desert. Shaded area shows the con- temporary distribution of Larrea divaricata subsp. iridentata in North America. Reproduced from Yang ( 1970) by kind permission of the author. the species, the center of species diversity, the age of some species, and the pos- mi Table 1 summarizes some of the characteristics of the five taxa of the genus Larrea and their geographic distribution. The multifoliolatc species having small petals and many leaflets are supposed to be representatives of a more primitive group, since most species of die other genera in the family Zygophyllaceae are multifoliolatc. The bifoliolate species would represent a more specialized and re- cent group. All species have acute for nearlv so ) stipules except South American L. 226 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 FiGXJRE 2. Distribution of diploid Larrea divaricata subsp, dwaricata in South America. The triangles indicate localities where chromosome counts (n = 13 II) have been made. (Base map copyright University of Chicago.) divaricata subsp. divaricata which has obtuse stipules. North American L. divari- cata subsp. tridentata has acute stipules (Fig. 3) as has been pointed out by Porter (1963). Figure 4 shows the leaf morphology, the growth habit, and the phenolic patterns of the four South American species of Larrea and of their five natural interspecific hybrids (Hunziker, Palacios, Valesi & Poggio, in preparation). Each species has at least one characteristic marker phenolic compound which is ) of drawing. These genome marker compounds have typ the hybrids. They have been very helpful for the identification of the parental species of the hybrids ( Hunziker et ah, in preparation) . forming If oil 1972] HUNZIKER DISJUNCTIONS IN LARTiEA 227 A B L 1 5mm -I L J Figure 3. Petals, stamens, staminal scales, stipules, and seed cross sections of the five taxa of the genus Larrea. A = L. nitida, B = L. ameghinoi, C zr L. cuneifolia^ D = L. divari- cata subsp. divaricata, E =r L. divaricata subsp. tridentata. side of the centromere. The diploid South American species form 10.4-12.6 closed ame is highly fertile and forms nearly 13 bivalents. This points out the close relation- ship of L. ameghinoi to L. nitida. It is possible that introgression has occurred between some populations of these two taxa, especially of genes from L. ameghinoi nitida If cata, on the other hand, forms 13 bivalents + 13 univalents in nearly 507^ of the Table 1. Leaflet characteristics, chromosome numbers, and geographic distribution of the five taxa of the genus Larrea, Group Holate late Morphological characteristics Taxon Multifo- Flowers small, L. nitida Cav, mericarps pu- berulous L. ameghinoi Speg Bifolio- Flowers larger, L. divaricata Cav. mericarps hairy subsp, divaricata L, divaricata Cav. subsp. tridentata ( Sesse. & Moc. ex DC. ) Felg. & Lowe L. cuncifolia Cav Leaflet number Habit 8-16 3-7 2 2 2 Erect Pros- trate Erect Erect Erect Chromo- somes (n) 13 13 13 13 26 39 26 Distribution Argentina and Chile Argentina (Patagonia) Peru, Bolivia Argentina Chile Chihuahuan Desert Sonoran Desert Mohave Desert Argentina 228 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 @ ] (f o L. am. X div. ooo Figure 4. Leaves, growth habit, and phenohc compounds of North Patagonian species and hybrids of Larrea. cells ( Hunziker et al, in preparation ) . Larrea divaricata is, therefore, one of tlie parental species of tetraploid L. cuneifolia or very closely related to one of the ancestral progenitors. The other hybrids are highly sterile (L. ameghinoi X L. cuneifolia), completely sterile (L. cuneifolia X L. divaricata), or flower rarely and 1972] HTJNZIKER ET AL.— DISJUNCTIONS IN LARREA 229 Figure 5. Putative crossing relationships of the four species of Larrea and their five natural hybrids. Mean chromosome associations are indicated for species and hybrids. The connections between circles give an idea of fertility as measured by the percent of seed-bearing mericarps: black, > 70%; hatched, partially fertile (15-19%); solid line, sterile; broken line, flowering sporadical, almost sterile. are highly sterile (L, ameghinoi X L. dwaricata). Analysis of chromosome associ- ations in these hybrids could not be studied with success due to the occurrence of cytomixis, which produced cells with very different chromosome numbers and associations. There is evidence that L. divaricata subsp. tridentata from California and L. nitida or L. ameghinoi are self compatible species ( Twisselmann, 1956; Raven, 1963; Hunziker et aly in preparation). Larrea cuneifolia is possibly a relatively old species having a fairly wide distri- bution in Argentina. As we have seen, it shares a genome with L. divaricata, which would be one of its diploid progenitors. However, L. cuneifolia has lost many of the abundant phenolic markers produced by the South and North Amer- ican divaricata genomes (Figs. 4, 6). The other diploid has apparently become extinct and perhaps was responsible for the two most remarkable morphological characteristics of L. cuneifolia: a) the fusion of the leaflets, and b) the orientation of the epiphylls towards the east. It is probably this last characteristic, which enables the plant to avoid the full exposure of the leaves to noon sunrays, that makes Larrea cuneifolia more drought tolerant than L. divaricata. Since the divaricata diploid genome seems, therefore, 230 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 o o COo o New Mexico o O o ac?oo Catnrnarca oo oo O O o Arizona ^o"" Utah ^Oo San Luis O o Rio Negro Cb o °oO o O O cx>s o o Pd D cD o o O OO o Figure 6. Chromatograms and leaf and stipule forms of three populations each of Larrea divaricata subsp. tridentata (upper row) and L, divaricata subsp. divaricata (lower row). New Mexico, Eddy Co., diploid (NME); Arizona, Tucson, tetraploid (AP, WA); Utah, near Toquerville, hexaploid (UW). All three collected by Dr. T. W. Yang, who also has determined the chromosome numbers. The samples of L. divaricata subsp. divaricata are from the northwest (Catamarca Prov., Dep. Belen, near Hualfin), center (San Luis Prov., Lujan) and south (Rio Negro, Dep. San Antonio, Mina Gonzalito) of Argentina. to be quite an ancient one in South America it is possible that L. divaricata in South The Disjunction of Larrea divaricata and its Differentiation IN North America that varicata subsp. tridentata in North America is composed of 3 races: diploid in the Chihuahuan Desert (Mexico, Texas, New Mexico. Arizona), tetranloid in the M Mohave (California, western Arizona, Nevada, Utah), With tridentata few exceptions, these chromosome races are allopatric. Yang (1970) studie regarding chromosome number. His detailed geographic chromosome survey shows ( Fig, 1 ) : 1) There appears to be some sympatric contact in limiting areas of diploids and tetraploids (southeastern Arizona). In some places diploids and tetraploids are living together (Population 18). Also in west-central Arizona and south- eastern California hcxa- and tetraploid plants occur together. 1972] HUNZIKER ET AL.— DISJUNCTIONS IN LARREA 231 ^^ >■-■» f^. '.J ■ .r b d a b d b ^ ■■. ^^j^.v e -? ::. ^ d f g b Vt-^^Wl^ _^ ^ d J:^ "-? d e ■A. Diploid L. divaricata from South and North Amer- siibsp. tridentata, [a ^ Dep. San Antonio, Rio Figure 7. Polyacrylamide gel electrophoresis of seed albumins of South and North Amer- ican Larrea species. Anode to the right, ica; a and b z=: siibsp. divaricata; c and d Negro, Argentina; b — Dep. Valcheta, 73 km W. of Sa. Grande; c = Hidalgo Co., New Mexico; d = Vizarron, Queretaro, Mexico.] — B. Diploid L. divaricata subsp. tridentata from different localities in the U.S.A.; a =: Arizona, Pima Co.; b ^ idem; c = New Mexico, Eddy Co.; d — New Mexico, Luna Co.; e = N. Mexico, Chaves Co.; f — Texas, Maverick Co.; g = Texas, Crockett Co. — C. Tetraploid L. divaricata subsp. tridentata from different localities in tlie Arizona, Tucson or vicinity; d = Arizona, Pima Co. — D. Hexaploid L. rff- b = Nevada; c — Nevada, U.S.A. a, b, and c varicata subsp. tridentata from different localities in the U.S.A. a, Arizona, Mohave Co. — E. Different polyploid races of L. divaricata subsp. = tetraploid, Arizona, Pima Co.; c ~ tetraploid. Hoover Dam; d, e tridentata. a = diploid, Arizona, Pima Co.; b Arizona, Yuma Co.; d =: diploid. New Mexico, Eddy Co.; e = hexaploid, Nevada, Nye Co. th actual surprising the fact that in South America hybridization among four different species is frequent and five out of the six possible hybrid combinations have been found. It may the difficulty in distinguishing the the hybrids, if existing, might nol 3) Yang (1970) miles north-northea; mtral Arizona (30 It is isolated from tetra-, hexa- and other diploid populations, (The nearest diploids are approxi- mately 100 miles to the southeast. ) 4) Diploid and tetraploid populations can be recognized in the field on the 232 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 basis of habit, height and other morphological features. Tetra- and hexaploid indi- viduals are more difficult as already mentioned. On the other hand, several chromosome counts made along a stretch of 2^500 km from individuals collected in the provinces of Salta, Tucuman, Mendoza, and Rio Negro (from the northern, western and sovithem part of the area) indicate that subsp. divaricata is uniformly diploid in South America (Fig. 2), The chromatographic pattern of the phenolics of North and South American specimens of Larrea divaricata present some slight differences but on the whole are strikingly similar (Fig. 6). Electrophoresis of seed albumins has been found to support the arrangement of species within the genus made on morphological grovmds (Hunziker, 1971; Hunziker ^f at^ in preparation). Moreover, the elcctrophoregrams show that there are few differences among diploids of Larrea divaricata growing in distant places as northern Patagonia (subsp. divaricata) and New Mexico (subsp. tri- dentata); compare b and c from Fig. 7A, which are separated by 74° latitude or nearly 7,000 km. Within Larrea divaricata subsp. tridentata there are not marked differences in the protein patterns of the different chromosomal races. The diploid patterns present some slight differences among themselves (Fig. 7B). There are no obvious differences in the presence or absence of bands among the tetraploids except concentration of certain fractions (Fig. 7C). The hexaploids also show slight variation; they are almost identical except e, which shows an extra fast band on the extreme right ( Fig. 7D ) . As shown in Fig. 7E there are no major differences among di-, tetra- and hexa- ploid patterns. This strongly suggests that autoploidy is involved in the origin of tetra- and hexaploid populations of Larrea divaricata subsp. tridentata. If alloploidy were involved, the tetra- and hexaploid patterns would be more complex than the diploids because in amphiploids generally there is at least some addition of protein fractions from the original diploids ( Hall, 1959; Hall & Johnson, 1962; Johnson & Hall, 1965; Smith et at, 1971 ) . Conclusions On tlie basis of morphological, phenolic, and protein data it is likely that both diploid Larrea divaricata from South and North America should be regarded as conspecific^ but the final answer to this question should come from the analysis of the hybrid between both populations ( chromosome behavior, fertility ) . The evidence from morphology, phenolic patterns, and albumin electrophoresis suggests that intervarietal autoploidy is involved in the origin, differentiation, and northwest migration of the chromosome races in North America. The entire dis- tribution of the genus, with species diversity in northern Patagonia (primitive and more recent species) and the role played by diploid South American Larrea di- varicata subsp. divaricata in the origin of a relatively old species such as L. cunei- folia suggests a South-to-North America pattern of migration. Since L. divaricata appears to be self compatible, this characteristic might have helped in the expan- sion of its range as suggested by Baker ( 1955 ) . 1972] HUNZIKER ET AL.— DISJUNCTIONS IN LARREA 233 LiTEBATURE CiTED Baker, H. G. 1955. Self compatibility and establishment after 'long-distance" dispersal. Evolution 9: 347-348. Barbour, M. 1969. Patterns of genetic similarity between Larrea divaricata of North and South America. Amer. Midi. Naturalist 81: 54-67. Felger, R. S. & C. H. Lowe. 1970. New combinations for plant taxa in northwestern Mexico and southwestern United States. Jour. Arizona Acad. Sci. 6: 82-84. Hall, O. 1959. Immuno-electrophoretic analyses of allopolyploid rye-wheat and its parental species. Hereditas 45; 495-504. & B. L. Johnson. 1962. Electrophoretic analysis of the amphiploid of Stipa viridula X Onjzopsis hymenoides and its parental species. Hereditas 48; 530^535. HuNziKER, J. H. 1971. El uso simultaneo de datos citogeneticos y moleculares en taxonomia experimental. En R. H. Mejia & J. A. Moguilevsky (editors), "Recientes adelantos en biologia." Pp. 129-137. , R. A. Palacios, Amalia G. de Valesi & Lidia Poggio. Hybridization in Larrea ( Zygophyllaceae ) ; A morphological, cytogenetic and chemosystematic study. (In prep- aration. ) Johnson, B. & O. Hall. 1965. Analysis of phylogenetic affinities in the Triticinae by pro- tein electrophoresis. Amer. Join. Bot. 52: 506-513. MoRELLO, J. 1955. Estudios botanicos en las regiones aridas de la Argentina. Rev. Agron. Noroeste Argentine 1: 301-370. Porter, D. M. 1963. The taxonomy and distribution of the Zygophyllaceae of Baja Cali- fornia, Mexico. Contr. Gray Herb. 192: 99-135. Raven, P. H. 1963. Amphitropical relationships in the floras of North and South America. Quart. Rev. Biol. 38: 151-177. Smith, H. H., D. E. Harmill, E. A. Weaver & K. H. Thompson. 1971. Multiple molecular forms of peroxidases and esterases among Nicotiana species and amphiploids. Jour. Hered. 61:203-212. TwissELMANN, E. C. 1956. A flora of the Temblor Range and the neighboring part of the San Joaquin Valley. Wassmann Jour. Biol. 14: 161-300. Yang, T. W. 1970. Major chromosome races of Larrea divaricata in North America. Jour. Arizona Acad. Sci. 6: 41-45. / PLANT SPECIES DISJUNCTIONS: A SUMMARY^ Peter H. Raven^ Disjunctions in the ranges of plant species have fascinated biologists ever since they were first detected; their interpretation has long been regarded as one of the central problems of plant geography. It seems intuitively reasonable that follow- ing its origin, a species should migrate to occupy a more or less continuous range; but when we find it occupying two or more areas separated by hvmdreds or even thousands of miles, we wonder how this pattern could possibly have originated. If the disjvmction is expressed at a higher level — for instance, generic or familial the problem remains the same, but its solution may be even more difficult. Other papers in this symposium have touched on the major classes of disjunc- tions which have concerned biologists in the past. Before reviewing these, it is appropriate first to consider some general points that pertain to the study of all disjunctions and to consider some of the new and important kinds of evidence that have been brought into play in recent years. General Considerations In the study of disjunctions, as in all other questions of plant geography, an accurate taxonomic framework is a prerequisite, as stressed by Wood (this symposium). For example, the moss Macromifrium stillivantii C, Mull., thought to be an endemic of a small area of the southeastern Blue Ridge escarpment of the United States, has recently been shown to be identical with Macrocoma hymeno- sfomum (Mont.) Grout, a well known species that occurs throughout the American tropics and South America (Anderson, 1970). The range as now understood becomes a striking example of a disjunct distribution. On the other hand, the genus Boisduvalia ( Onagraceae ) , comprising six species of semiarid western North and South America, was until recently thought to include a species of the mountains of southeastern Australia and Tasmania, B. tasmanica (Hook, f.) Munz. With the demonstration that this species is actually an Epilobium, now known as E. curtisiae Raven, which forms natural hybrid populations with closely related Australian species (Raven, 1963^, and unpubL), the situation demands a very different interpretation. In efforts to clarify the relationships between disjunct populations, the methods of chemosystematics ( Hunziker et ah, Wagner, Turner, this symposium), biosystematics {e,g. Moore & Raven, 1970; Wagner, this sym- posium), and other modern techniques often have proven useful adjuncts to the more traditional morphological approaches (Wood, this symposium). If one considers the distributions of plants throughout the world, one will be able to find examples of almost any conceivable sort of disjunct range. Thus very ^ Support received from the U. S. National Science Foundation, most recently through Grant GB-29905, is gratefully acknowledged. D. I. Axelrod has kindly reviewed this manuscript. ^Missouri Botanical Garden, 2315 Tower Grove Avenue, St. Louis, Missouri 63110 and Washington University, St. Louis. Ann. Missouri Bot. Gard. 59: 234-246. 1972. 1972] RAVEN— PLANT SPECIES DISJUNCTIONS 235 few genera of semiarid habitats are common to Australia and South America, but Plagiobothrys (Boraginaceae), Calandrinia ( Portulacaccac ) , and Nicotiana (So- lanaceae) exhibit this sort of range. The fact that Nicotiana also occurs on some Pacific islands (van Steenis & van Balgooy, 1966: 78-9) may provide a clue to its pathway of dispersal, but hardly justifies the construction of a landbridge across the tropical Pacific, as proposed for example by Corner (1963). Similarly, the occwrrence oi Sibthorpia (B^osaceae) , Hijpochoeris ( Asteraccac ) , and Deschampsia setacea (Huds.) Hack. (Poaceae) in Europe and South America, but not in North America, must be taken as special cases and not used as the basis for postulating vast geological rearrangements. Nevertheless, when considerable numbers of ranges coincide in their disjunc- tions, we are justified in looking for some collective reason. The recurrent disjunction between eastern Asia and eastern North America has been treated elegantly by Wood in the present volume; that between semiarid North and South America by Solbrig and by Turner; and the famous disjunction in range between some of the plants and animals of temperate South America and temper- ate Australasia has been mentioned by Culberson and by Schofield and Crum. For these and other equally impressive patterns, some historical explanation must be sought, but it must always be kept in mind that most patterns have been af- that (Wh the habitats in the two disjunct areas. This fact in itself should serve as a warning when attempting to make summary judgments, for when two habitats are similar, there is always some probability of additional plants achieving the disjunction, regardless of its original historical basis (Simpson, 1952). For example, the dis- junctions in range between Nothofagus (Fagaceae) and Proteaceae in temperate South America and Australasia are certainly related to Eocene geography, as we shall discuss below. On the other hand, to assert, as Melville (1966) docs, that Juncus scheuchzerioides Gaud. (Juncaceac) achieved its range at such a remote time, is incredible. This species has dust-like seeds easily carried in the mud adhering to the feet of birds, and it occurs on many subantarctic islands known to have been completely glaciated in the Pleistocene. It is often possible to assign a maximum, but not a minimum, age to any particular example of disjunction, unless there is a fossil record (Whitehead, this symposium). Wlienever two habi- tats are similar, they may have similar plants and animals, regardless of how far apart they may be, and additional ones may be added at a rate consistent wit]> their powers of dispersal subsequently. Many plants have achieved striking dis- junctions in range as weeds since the initiation of human activities on a global Wa diat went on continuously before. explanation colored by the expectations and desires of the explainer. It is tempting to adduce th but it is unreasonable to expect these explanations to be correct for the whol(\ Each disjunction has its own explanation, and the reasons can be verified convinc- inslv for the narticular case only if there is a fossil record. f 236 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 LONG-DlSTANCE DiSPERSAL Islands such as Hawaii, Tahiti, and Samoa have never been connected with any continental area, and yet they have acquired significant numbers of plants and animals as a result of long-distance dispersal. In the case of Hawaii, the pattern of dispersal has been particularly well documented (Fosberg, 1948; Gemmell, 1954; Carlquist, 1967, 1970). South American bryophytes have spread to Tristan da Cunha, a million-year-old island in the middle of the South Atlantic (Schuster, 1969). On Marion and Prince Edward Islands, which lie in the Antarctic Ocean south of Africa and are no more than 276,000 years old, there occurs a flora of at least 80 species of mosses (van Zanten, 1971) and 36 liverworts (GroUe, 1971), including many "Antarctic rehcts** that are often taken as evidence of continental drift. On these same islands, a pollen grain of Nothofagus that must have come about 7,500 km from South America was collected from snow (Schalke & van Zinderen Bakker, 1971). Similar patterns are evident in both the Arctic and the Antarctic (e.g. Young & Klay, 1971), where recolonization of areas glaciated in the Pleistocene has taken place rapidly and recently. In the face of these observations, it is clear that long-distance dispersal must also take place with a certain frequency in continental areas, despite the attempts of authors such as Crum (1972) to minimize its importance. If islands that have never been connected to continents have acquired certain animals and plants, then animals and plants with similar characteristics must be dispersed between areas on the mainland also. The characteristics of seed plants that are dispersed to islands have been discussed in an elegant essay by Carlquist (1967). The question is whether they will, once they have reached new areas, become estab- lished. One of the most important aspects in the establishment and subsequent fate of a plant In a new locality is the breeding system of the population in question. Thus Baker (1955) pointed out the connection between self -compatibility and establishment following long-distance dispersal. In the present symposium, Scho- field and Crum have considered the effect of dioicism upon establishment by moss species, while Klekowski has stressed the genetic constitution of the sporo- pliyte as related to the characteristics of the gametophyte in ferns. In both bryo- phytes (Crum, 1972) and ferns (Tryon, 1970), the spores are light and easily airborne for great distances; despite this, and also in the lichens (Culberson, this symposium), patterns of distribution, while perhaps somewhat broader on the whole, are strikingly like those observed in the flowering plants. In bryophytes and ferns, the gametophytes and sporophytes may demand different conditions for survival; and in both, gametophytes may occur far beyond the range in which sporophytes are normally produced in nature (Schofield & Crum, Wagner, this symposium). This is especially likely in dioicous mosses, for obvious reasons. Sea-Floor Spreading During the past five years, the earth sciences have been the subject of a major revolution, as theories of plate tectonics have provided for the first time a basis sufficient to account for the kinds of movements of the continents first postulated 1972] RAVEN— PLANT SPECIES DISJUNCTIONS 237 We Since much of the information Mesozoic the continents is of direct appHcation to problems of disjunction in the ranges of plant taxa, it will be summarized briefly here. Among the salient points that are emerging from this new synthesis are the early Cretaceous (110 m.y. BP) separation of South America from Africa, the middle Cretaceous {^ 90 m.y. BP) separation of Africa from Antarctica, the early 55-47 45^9 Axelrod Some of the progressive climatic change that has taken place through the course of the Tertiary has been associated with sea floor spreading and changes in the position of the continents. For example, North America, Africa, and Australia have moved northward some 15"^ of latitude during the Cretaceous and Tertiary, and India some 50 ^^^ colliding with the mainland of Asia and throwing up the Himalayas in the process. All of these changes have had profound effects upon the disjunct distributions we observe at the present, as will be discussed below. Age of Taxa Involved in Disjunctions Increasing study of the fossil record has begun to provide valuable information on the probable age of angiosperm taxa, which can be brought to bear on problems of the age of particular disjunct ranges. For example, despite worldwide study and the common use of their pollen as indicators of particular strata, Aster- aceae have not been identified in the fossil record before the lower Miocene, some 25 m.y. BP (Muller, 1970), and the tribe Cichorieae of this family not until the upper Miocene, about 10 m.y. BP (Couper, 1960). Too much evidence is available now to continue ignoring this fact, and it is scientifically inadmissible to continue attributing "antarctic" distributions of members of this family to dispersal across Antarctica during the Cretaceous or Eocene, more than 40 m.y. BP. Indeed, it is doubtful that any living sympetalous genus, with advanced pollination systems, existed early enough to have taken advantage of such a route of dispersal. The species of Microseris (Asteraceae — Cichorieae) that occur in western North America, western South America, and temperate Australasia can only have achieved their present ranges by long-distance dispersal, regardless of how im- probable this may seem. It is worth noting in passing that we have very little hard evidence about the rates of evolution of particular taxa, and it is extremely dangerous to reason from a given degree of morphological divergence to a length of time thought necessary to produce that divergence. If the Hawaiian honeycreepers ( Drepanididae ) could have differentiated from a common ancestor within a million years ( Bock, 1970 ) , why should we assume that the Hawaiian silverswords and their relatives { Aster- aceae — Madiinae) have taken longer? On the other hand, certain plant species in Europe and western North America have scarcely changed since the late Eocene, some 40 m.y. (Axelrod, 1958, 1973) — much longer than tlic entire history of the family Asteraceae, in all likelihood. It is simply not justified to state that a certain degree of differentiation must have required a certain length of time, unless there 238 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 is a fossil record, yet arguments of this sort are commonplace in the literature on distributio Major Patterns of Disjunction Five major patterns of disjunction have been discussed repeatedly in the literature and mentioned frequently in the papers of this symposium. In addition, smaller disjunctions have been considered in other papers of this symposium and very often in the literature. The five major patterns distinguished here are treated in the sections that follow. 1. North Temperate Disjunctions It has long been noted that many of the forest plants of the north temperate region have disjunct ranges, commonly between eastern Asia and eastern North America, but also involving western North America, southeastern Europe-Asia Minor and the mountains of eastern and southern Mexico. In the light of recent geological information, the long-standing explanation first proposed by Asa Gray, that of migration through the Bering Straits in times of milder climate, must be regarded as highly questionable. About 47-55 m.y. BP, in the early to middle Eocene, North America was still broadly joined to Europe from abovit 50 "^N lati- tude northward (Dietz & Holden, 1970). At this time Arcto-Tertiary forest vege- tation was continuous across northern Europe and in much of western North (Axelrod, 1973), and it is clear that the ^ iie eastern United States (Mc- Kenna, 1972). By the start of the Tertiary (63 m.y. BP), the relationship between Asia and North America approached that of the present day (Pitman & Talwani, 1972; Churkin, 1972), so that migration into and out of North America was possible both from the east and from the west. The Bering Straits seem to have functioned as an important migration route between North America and Eurasia throughout the Tertiary (Simpson, 1947; Hopkins, 1967), although the geological history of Beringia needs further consideration in the light of plate tectonics. If these relationships are as assumed above, the close similarity of Miocene floras in Japan, Alaska, and Oregon (Wolfe & Leopold, 1967) is not surprising. In the Cretaceous, there seems to have been a greater similarity between the pollen floras of eastern Asia and eastern North America than to any other part of the north Temperate zone; Muller (1970) has recognized an East Siberian-North Pacific pollen province and a North Atlantic-European one, reviewing the perti- nent palynological literature. Smiley (1967) has considered Cretaceous leaf floras of Alaska to be closely similar to those of northeastern Siberia, although he later (Smiley, 1969) has stressed the homogeneity of tlie floras of the entire northern portion of Eurasia and North America at this time. These observations are the hypothes This suture is believed to have closed in early Cretaceous time. Nonetheless, direct migration via Europe apparently was feasible between Eurasia and North America into the early to middle Eocene ( 55-47 m.y, ) . Disjunct ranges in the north temperate forest, excepting the rapid expansion 1972] RAVEN— PLANT SPECIES DISJUNCTIONS £39 of Ai'ctic Species into formerly glaciated areas, are mainly Eocene in origin. For the boreal conifers, a more recent disjunction is implied. These relationships are interesting from an evolutionary point of view, because they imply that the related species of eastern Asia and the eastern United States have been evolving in isola- tion for approximately 47-55 m.y. As pointed out by Wood (this symposium), there are genera common to these two regions but very few species, a good indica- tion of evolutionary rate in the groups concerned. Plants that are now found in various favorable areas of the north temperate zone have, in general, been sepa- rated by the deterioration of the climate in intervening areas; in this process, Pleistocene glaciation in Europe, with its east-west trending mountains, and the uplift of the Sierra Nevada-Cascade system in western North America have been important. In evaluating such relationships, however, it is important to remember that for most of the Tertiary, the North Atlantic has been much narrower than at present, and probably was also dotted with islands. For many million years after the defin- itive separation of Europe and North America, birds probably flew back and forth regularly, and the probabilities of dispersal were very different from those obtaining at present. Some of the common species might have attained their disjunct ranges even more recently, as stressed by Wood (this symposium), as a result of long- distance dispersal; but it must be borne in mind that these are 'closed" communi- ties, where the chance of establishment may be relatively low, even if a seed does reach the disjunct area occasionally. For such entities as orchids, Phryma lepto- stachya L. ( Verbenaceae ) , and Circaea ( Onagraceae ) , all of which have readily dispersed seeds or fruits, it is difficult to imagine that they attained their disjunct ranges as early as the Eocene, and it would be dangerous to use their distributions to argue for or against such an antiquity for these plants. One special disjunct pattern in the north temperate zone involves persistence in areas of mild oceanic climate, documented in this symposium by Culberson (for lichens), Schofield and Crum (for mosses), and Wagner (for ferns). Such eco- logical requirements have resulted in disjunctions between western North America and Europe, both onshore from warm oceanic currents. Patterns of this sort are rare or absent among the flowering plants. 2. East-west Desert and Mediterranean Disjuncts Although there has always been a belt of reduced precipitation flanking the tropics, areas of desert and mediterranean climate in their present continental scale are a phenomenon of the past five million years (Axclrod, 1958, 1973; Raven & Axclrod, 1972), Some plant families that are restricted to semiarid or subhumid habitats have certainly been in existence since the Cretaceous, as pointed out by Rzedowski (1962) and others, but many species and genera confined to these areas have had a much more recent origin. In general, there is very little evidence for contact between the plants of the desert and mediterranean areas of North America and Eurasia (Raven, 1971, 1973a), despite the fact that these continents were broadly joined in the Eocene (Dietz & Holden, 1970; Tarling, 1971), Tliis implies strongly that the plant associations involved were not in exis- tence at the time, and also probably reflects the fact that the contincMits were joined 240 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 only at the north. It also suggests that the few genera and even fewer species common to the deserts and areas of mediterranean climate in the Old and New Worlds must have attained their present ranges by long-distance dispersal, a contention supported by the almost complete faunistic dissimilarity of the two areas. The several mosses (Schofield and Crum, this symposium) with this dis- junction in range were almost certainly spread from one area to the other by long- distance dispersal. 3. Tropical Disjuncts As reviewed by Axelrod (1970), there are many genera and infrafamilial taxa common to Africa and South America. Most of these presumably were present in the tropical vegetation of the two areas when they were contiguous in the early Cretaceous, or evolved in one or the other when the two continents were relatively near and partially connected by islands. Primary freshwater characoid fishes certainly spread overland between Africa and South America (Myers. 1967). On the other hand, present patterns of distribution suggest strongly that Cactaceae and Bromeliaceae evolved in South America subsequently, one genus of each having reached Africa, perhaps quite recently, by long-distance dispersal. If Bromeliaceae had been in existence in the Cretaceous, for example, it would be virtually unimaginable that they would not be represented in Africa at present. This strongly implies that the entire evolution of the family has taken place in the past 110 m.y. or less, a useful landmark in the fossil-poor monocots. Many other plants have been dispersed between Africa and South America during the Tertiary, and they are doubtless still being dispersed at present. This possibility has been stressed recently by litis (1967) and others, and it would be a serious error to relate every disjunction between the continents to Cretaceous geography. In the lichen ParmeUa subgen, Amphigymnia, discussed by Culberson (this symposium), it is especially difficult to date the time of dispersal between South America and Africa, in view of the presence of wind-dispersed soredia or isidia in many species; they may have been lost in others in the course of evolution. Prior to the middle Tertiary, migration between subtropical areas in southeast Asia, India, and Africa has been simple and direct. Only with the late Tertiary development of die deserts of Arabia and the Near East has this communication been interrupted, and it is therefore not surprising to find many links between the floras of the Indo-Malaysian region and Africa. Madagascar, now known to have been separated from the coast of Somalia in the Cretaceous or more recently (Heirtzlcr & Burroughs, 1971), is the home of many relict plants and animals that have become extinct on the African mainland subsequently. Such survival ac- counts for the otherwise inexplicable similarities between the flora and fauna of Madagascar and the West Indies (Stearn, 1971) and between Madagascar and New Caledonia ( Good, 1950 ) . Australasia, moving northward some 15° of latitude during the past 45-49 m.y., has approached tropical latitudes only within the past 10 million years, particularly with the emergence of New Guinea ( Oligocene and subsequently; Haven & Axel- rod, 1972). It has acquired its tropical flora during this period of time from the Indo-Malaysian region and as a result of the evolution of tropical representatives of archaic, austral groups of plants and animals as it moved northward. 1972] RAVExN— PLANT SPECIES DISJUNCTIONS 241 4. Southern Hemisphere Temperate Disjunctions Biologists have long been fascinated by the group of plants and animals that occupy disjunct ranges in the far-flung disjunct lands of the southern hemisphere. Many of these are related to the position of Australia and South America at the close of the Eocene, some 45-49 m.y. BP; both were directly connected with Ant- arctica (Dietz & Holden, 1970; Tarhng, 1971; Raven & Axelrod, 1972). All three continents were occupied by a continuous cool temperate forest of gymnospcrms and evergreen angiosperms that existed under equable conditions; in this forest were such plants as Podocarpaceae, Araucariaceae, Proteaceae, Winteraceae, Atherospermataceae, Epacridaceae, Loranthaceae, Myrtaccae, Nothofagus, and Gunnera, as well as many groups of lower plants and invertebrate animals that now have disjunct distributions in the south, including marsupials, hylid and leptodactylid frogs, and chelyid turtles, which crossed from South America to Australia by this route (review in Raven & Axclrod, 1972) . The gradual disruption of this once continuous forest by the movements of the continents was provided a predominant theme in southern hemisphere distribu- tions. Some distributions, however, may be related to the earlier connection of Africa and India with Antarctica, broken during the middle Cretaceous. The austral gymnospcrms (Florin, 1963), side-neck turtles, ratite birds, and galaxiid fishes (Darlington, 1948; Evans, 1958) seem to have reached Africa by this route, as did some leafy liverworts (Fulford, 1963). Among the angiosperms, there are a few patterns of distribution that suggest dispersal prior to the separation of Africa from Antarctica. Proteaceae, best developed in Australia, with closely related lines in South America and Asia and three very distinct lines in Africa, are suggestive of such a history (Johnson & Briggs, 1963); Xyridaceae and Restionaceae are also possible candidates (Cutler, 1972). It seems purely fanciful to relate the distri- butions of such living genera as Gossijpkim and Solarium (Hawkes & Smith, 1965), which provide no hint of such antiquity, to Cretaceous geography. Both genera have native species in Hawaii. The suggestion of Turner (this symposium) that the Centrospermae differentiated in a Gondwanaland that included Africa ( early Cretaceous) is almost ruled out by the Eocene origin of most families in this group (Muller, 1970). On the other hand, the families he discusses may well have differentiated in the south, after the separation of Africa, and svibsequently spread to the northern hemisphere. Many disjunct distributions involving southern lands, in contrast, are recent in origin (Moore, 1972); the case of Juncus scheuchzerioides and some other exam- ples have already been mentioned. A number of plants that are easily dispersed by wind have doubtless been spread by the prevailing westerlies, seasonally four times as powerful as the corresponding winds in the northern hemisphere ( Lamb, 1959). There is a great deal of evidence of such dispersal of small animals and seeds; recently data have been published which show that a balloon released at Christchurch, New Zealand, and held at approximately 40,000 feet elevation, made eight complete circuits of the southern hemisphere during 102 days (Mason, 1971). Other plants have been carried in the ocean currents ( Sykes & Godlcy, 1968) or by birds (review in Raven & Axelrod, 1972; Raven, 1973/;; see also Carlquist, 1970, 242 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Chapter 4). Like all other disjunct distributions, ones between the southern lands must be taken with caution as indications of the age of the groups involved. A single concrete example involves Fuchsia (Onagraccae), with more than 50 species in South America (northward to Mexico and the West Indies) and one section of four species in the Old World, three in New Zealand and one in Tahiti, It might at first be supposed that Fuchsia was a member of the An tarcto -Tertiary Geoflora and reached its disjvmct stations overland via Antarctica, but four lines of evidence suggest that this was not the case: 1) the family Onagraceae may be no older than the middle Eocene (Muller, 1970); 2) Fuchsia first appears in the fossil record in New Zealand in the middle Miocene, fully 25 million years after the separation of Australasia from Antarctica, and it is not known from the fossil record in either Australia or Antarctica; 3) Fuchsia is bird-pollinated, both in the Old World and the New, and it is most unlikely that the specialized lines of birds that regularly visit and pollinate flowers were in existence much, if at all, before the Miocene; and 4) Fuchsia occurs on Tahiti, which stands in the middle of the Pacific and has never been connected to any land, a station it must have achieved by long-distance dispersal. In summary, the weight of evidence suggests strongly that Fuchsia, despite its occurrence in cool-temperate forest in both South America and New Zealand, attained its present disjunct distribution long after Antarctica was no longer available as a migration route, by means of long- distance dispersal across the Pacific. 5. Trans-Tropical Disjunctions Disjunct distributions that span the tropics are not as likely to be remnants of formerly continuous distributions as the sorts of east-west disjunctions we have just discussed. Nevertheless, as Solbrig (this symposium) has indicated, there are opportunities for plants and animals of semiarid habitats to achieve only slightly interrupted ranges through the tropics even at the present day. Certain plants of temperate regions must also have passed through the tropics along elevated re- gions in the distant past, as suggested for example by the presence of Nothofagus as the only genus of Fagaceae in the southern hemisphere. There are relatively few trans-tropical disjunctions involving Europe and Africa, presumably because the temperate area of South Africa is so limited. More disjunctions are known involving Asia and Australasia, and most of these seem quite recent in origin. Australasia has come into contact with Asia only in the upper Miocene, some 10 million years ago, and the mountains of Malaysia, New Guinea, Australia, and New Zealand were all uplifted in the Pliocene and later. Consequently, the migration paths visualized by van Steenis (1934a, 1934fo, 1936) through Malaysia are no more than a few million years old, and the plants found on these movmtains reached their disjunct stations mainly during the late Pliocene and Pleistocene by long-distance dispersal. Many north- temperate groups of plants reached Australia and New Zealand only at this time, and some have evolved rapidly in the newly opened subalpine and alpine habitats, particularly in New Zealand (Raven, 1972). The best studied and most numerous transtropical disjuncts are those between North and South America, which have been in approximately their same relative 1972] RAVEN— PLANT SPECIES DISJUNCTIONS £43 positions since at least the Cretaceous (Raven 1963/;; Moore, 1972). Many of these involve identical or very closely related species of annual herbs in areas of medi- terranean climate that were set up only in the late Pleistocene and subsequently (Moore & Raven, 1970; Raven, 1973a; Axelrod, 1973); there seems no doubt that most of these have achieved their disjunct ranges by direct, long-distance dispersal within the past several hundred thousand years. Each year during this time mil- lions of individuals of the semipalmated plover, Charadrius vulgaris^ have migrated between the areas that have disjunct plant species on the two contincMits, and these birds have provided at least one obvious means for direct long-distance dispersal (Cruden, 1966). Carlquist (1967) has convincingly demonstrated the probabihty of dispersal by birds as a means for achieving many of the disjunct distributions between North and South America, comparing these patterns with those involv- ing dispersal to the Pacific Islands. Other plants have moved between North and South America by shorter jumps along the Andean chain, presumably in the main following its Pliocene uplift (Raven, 1973a). Disjunctions on range between the plants of desert and other semiarid areas of North and South America have received considerable attention in this symposium (papers by Hunziker et al.^ Solbrig, Turner). Even though areas of reduced pre- cipitation have existed on the margins of the tropics since the beginnings of angiosperm evolution, and provided the initial site of evolution of many of the plant groups that occur in these deserts at the present time, the deserts themselves, in their present continental scale, are a phenomenon of the latest Tertiary (Axel- rod, 1958). In other words, chances for dispersal between these areas, across the tropics or from east to west, are greater now than they have ever been at any time in the past ( Axelrod, 1952 ) . Solbrig has rightly re-emphasized the various times of dispersal that must have led to the many different patterns of disjunction at the present time. Some of the woody plants common to the deserts of North and South America have differentiated from tropical ancestors, whereas others extend more or less continuously through the tropics in ''islands" of subhumid vegetation. In general, the dissimilarity of the vegetation of semiarid areas in the Old and New World is a strong indication that many of the groups now involved in dis- junctions in the New World were rare or not in existence in the Eocene. At any time subsequently, they may have become dispersed between North and South America, almost certainly in a series of steps. Drier sites in the tropics may have provided stepping stones by which such plants may have become dispersed throughout the Tertiary, but the fact that less than 2% of the desert floras of Argentina and the southwestern United States and adjacent Mexico are common to both areas makes it quite impossible to imagine a direct and simple pathway through the tropics, also not possible on climatic grounds. In addition, the insects and other animals of the two areas are almost entirely different, something that would not be true were there direct communication. The level of similarity that is actually observed is consistent with a sporadic, stepwise migration, operating at different times throughout the Tertiary and by different path\\\ays, and resulting in a limited exchange of plant species and genera between North and South America, Detailed studies, such as that of Hunzilccr et at (this symposium) on Larrea (Zygophyllaccae), will bo necessary to clarify individual cases. 244 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Concerning trans-tropical disjunctions, however, Florins (1963) insistence on the long-sustained separation of north and south temperate floras still provides the most cogent generality. The uplift of mountains in the tropics of Africa, Malaysia, and America has facilitated movement between northern and southern hemisphere temperate areas, bvit such movement has apparently never been par- ticularly easy, despite indications (^.g., the origin of Nothofagtis; podocarps and araucariads in the Tertiary of Europe) that it has continued throughout the entire history of angiosperms. During Pleistocene and Recent time, the great expansion of open semiarid habitats on both sides of the tropics has apparently provided conditions especially favorable for the establishment of some plants following long-distance dispersal, but the only appreciable exchange of this sort seems to have been that between North and South America, perhaps because these areas lie on regular migration paths of birds. There has been almost no corresponding exchange between semiarid regions of the Old and New Worlds except that brought about as a result of human activities. There is a considerable amount of bird migration between Eurasia and South Africa, but almost none that crosses Wallace's line into Australasia (McClure, 1971). Could this possibly be attrib- utable to the relatively recent (15 m.y. BP) jvixtaposition of Asia and Australia? Literature Cited Anderson, L. E. 1970. Geographical relationships of the mosses of the southern Appalachian Mountains. Virginia Polytech. Inst. Res. Div. Monogr. 2: 101-115. AxELROD, D, I. 1952. Variables affecting the probabihties of dispersal in geologic time. Bull. Amer. Mus. Nat. Hist, 99: 177-188. . 1958. Evolution of the Madro-Tertiary Geoflora. Bot. Rev. 24: 433-509. 1970. Mesozoic paleogeography and early angiosperm history. Bot. Rev, 36: 277- 1973. History of the Mediterranean ecosystem in California. In H. Mooney & 319. F. DiCastri (editors), "Evolution of Mediterranean Ecosystems." In press. Baker, H. G, 1955. Self-compatibility and establishment after 'long-distance" dispersal. Evolution 9: 347-348. Bock, W. J. 1970. Microevolutionary sequences as a fundamental concept in macroevolu- tionary models. Evolution 24: 704-722. Cafj.quist, S. 1967. The biota of long-distance dispersal. V. Plant dispersal to Pacific Islands. Bull. Torrey Bot. Club 94: 129-162. . 1970. Hawaii. A Natural History. Garden City, Nev^ York. Churkin, M., Jr. 1972. Western boundary of the North American continental plate in Asia. Geol. Soc. Amer. Bull. 83: 1027-1036. Corner, E. J. H. 1963. Ficus in the Pacific region. In J. L. Gressitt (editor), "Pacific Ocean Biogeography." Pp. 233-245. Honolulu. Couper, R. a. 1960. New Zealand Mesozoic and Cainozoic plant microfossils. New Zealand Geol. Surv. Paleont. Bull. 32: 1-88, pi. 1-12. Cruden, R. W. 1966. Birds as agents of long-distance dispersal for disjunct plant groups of the temperate Western Hemisphere. Evolution 20: 517-532. Crum, H. 1972. The geographic origins of the mosses of North America's eastern deciduous forest. Jour. Hattori Bot. Lab. 35: 269-298. Cutler, D. F. 1972. Vicarious species of Restionaceae in Africa, Australia and South Amer- ica. In D. H. Valentine (editor), "Taxonomy, Phytogeography and Evolution," pp. 73-83. Daruxgtox, C. D. 1948. The geographical distribution of cold-blooded vertebrates. Quart. Rev. Biol. 23: 1-26, 105-123. DiETz, R. S. & J. C. HoLDEN-. 1970. Reconstruction of Pangaea: Breakup and dispersion of continents, Permian to present. Jour. Geophys. Res. 75: 4939-4956. Evans, J. W. 1958. Insect distribution and continental drift. Univ. Tasmania Geol. Dept. Symp., Mar. 1956, p. 134-161. 1972] RAVEN— PLANT SPECIES DISJUNCTIONS 245 Florin, R. 1963. The distribution of conifer and taxad genera in time and space. Acta Hort. Berg. 20: 121-312. FosBERG, F. R. 1948. Derivation of the flora of the Hawaiian Islands. In E. C. Zinnnerinan, "Insects of Hawaii." 1: 107-119. Honolulu. FuLFORD, M. 1963. Continental drift and distribution patterns in the leafy Hcpaticae. Soc. Econ. Paleont Minearol. Spec. Publ. 10: 140-145. Gemmell, a. R. 1954. Relationship and development of moss flora of Hawaii. VIH Congr. Internat. Bot., Vol. Prelim., pp. 90-91. Good, R. 1950. Madagascar and New Caledonia; a problem in plant geography, Blumca 6: 470-479. Grolle, R. 1971. Hepaticopsida. In E. M van Zinderen Bakker, Sr., J. M. Winterbottom & R. A. Dyer (editors), "Marion and Prince Edward Islands." Pp. 228-236, Cape Town. Hammond, A. L. 1971a. Plate tectonics: The geophysics of the earth's surface. Science 173: 40-41. . 1971&. Plate tectonics (II): Mountain building and continental geology. Science 173: 133-134. Hawkes, J, G. & P. Smith. 1965. Continental drift and the age of angiosperms. Nature 207: 48-50. Heirtzler, J. R. & R. H. Burroughs. 1971. Madagascar's paleo-position: New data from tlie Mozambique Channel. Science 174: 488-490. Hopkins, D. M. (editor). 1967. The Bering Land Bridge. Stanford. Iltis, H. H. 1967. Studies in the Capparidaceae. XI: Cleome afrospina, a tropical African endemic with Neotropical affinities. Amer. Jour. Bot. 54: 953-962. Johnson, L. A. S. & B. G. Briggs. 1963. Evolution in the Proteaceae. Austral. Jour. Bot. 11: 21-61. Lamb, H. H. 1959. The soutliern westerlies: a preliminary survey; main characteristics and apparent associations. Quart. Jour. Roy. Meteor. Soc. 85(363): 1-23. Mason, B, J. 1971. Global atmospheric research programme. Nature 233: 383-388. McClure, H. E- 1971. Some aspects of bird migration in Asia. Rec. Proc. XII Pacif. Sci. Congr. 1: 219-220. McKenna, M. C. 1972. Was Europe connected directly to North America prior to the middle Eocene? EvoL Biol. 6: (in press), Melville, R. 1966. Continental drift, Mesozoic continents and the migrations of the angio- sperms. Nature 211: 116-120. Moore, D. M, 1972. Connections between cool temperate floras, with particular reference to southern South America. Pp. 115-138, in D. H. Valentine (editor), "Taxonomy, Phyto- geography and Evolution. 99 & P. H. Raven. 1970. Cytogenetics, distribution and amphitropical affinities of South American Cammonf^ ( Onagraceae ) . Evolution 24: 816-823. Muller, J. 1970. Palynological evidence on early differentiation of angiosperms. Biol. Rev. Cambridge 45: 417-450. Miners, G. S. 1967. Zoogeographical evidence of the age of tlie south Atlantic Ocean. Stud. Trop. Oceanogr. Miami 5: 614-621. PrrMAN, W. C. & \L Talwani. 1972. Sea floor spreading in the North Adantic. Bull. Geol. Soc. Amer. 83: 619-646. Raven, P. H. 1963a. The generic position of "Boisduvalia iasmanica'* Aliso 5; 247-249. . 1963i». Amphitropical relationships in the floras of North and South America. Quart. Rev. Biol. 38: 151-177. — . 1971. The relationships between 'mediterranean' floras. In P, H Davis, P. C. Harper & I. C. Hedge (editors), "Plant Life of South-West Asia," pp. 119-134. Edinburgh. — . 1972. Evolution and endemism in the New Zealand species of Epilohium, Pp. 259- 274, in D. H. Valentine (editor), "Systematics, Ph>togeography and Evolution." — . 1973a. The evolution of mediterranean' floras. In H. Mooney & F. DiCastri (ed- itors), "Evolution of Mediterranean Ecosystems, In press. — . 1973b. The evolution of subalpinc and alpine plant groups in New Zealand. New Zealand Jour. Bot. (in press). & D. I. Axelrod. 1972. Plate tectonics and Australasian paleobiogcography. Science 176: 1379^1386. RzEDOWSKi, J. 1962. Contribuciones a la fitogeografia floristica e historia de Mexico. I. 246 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Algunas consideraciones acerca del elemento endemico en la flora mexicana. Bol. Soc. Bot. Mex. 27: 52-65. ScHALKE, H. J. W. G. & E. M. VAN ZiNDEREN Bakker, Sr. 1971. History of the vegetation. In E. M. van Zinderen Balcker, Sr., J. M. Winterbottoni & R. A. Dyer (editors), "Marion and Prince Edward Islands," p. 89-97. Cape Town. Schuster, R. M. 1969. Problems of antipodal distribution in lower plants. Taxon 18: 46- 91. Simpson, G. G. 1947. Holarctic mammalian faunas and continental relationships during the Cenozoic. Bull. Geol. Soc. Amer. 58: 613-688. . 1952. Probabilities of dispersal in geological time. Bull. Amer. Mus. Nat. Hist. 99: 163-176. Smiley, C. J. 1967. Paleoclimatic interpretations of some Mesozoic floral sequences. Amer. Assoc. Petrol. Geol. Bull. 51: 849-863. . 1969. Cretaceous floras of Chandler-Colville region, Alaska: Stratigraphy and pre- liminary floristics. Amer. Assoc, Petrol. Geol. Bull. 53: 482-502. Steabn, W. T. 1971. A survey of the tropical genera Oplonia and Tsilanihele ( Acanthaceae). Bull. Brit. Mus. Nat. Hist. Bot. 4: 261-323, pi. 38-47. Steenis, C. G. G. J. VAX. 1934a. On the origin of the Malaysian mountain flora, i. Facts and statement of the problem. Bull. Bot. Card. Buitenzorg III. 13: 139-262. , 1934&. On the origin of the Malaysian mountain flora. 2. Altitudinal zones, general considerations and a renewed statement of the problem. Bull. Bot. Card. Buitenzorg III. 13: 289-417. Sykes, W. R. & E. J. GODLEY, 1968. Transoceanic dispersal in Sophora and other genera. Nature 218: 495-496. . 1936. On the origin of the Malaysian mountain flora. 3. Analysis of floristic rela- tionships. Pt. i. The Sumatran track. Bull. Bot. Gard. Buitenzorg III. 14: 56-72. — & M. M. J. VAN Balgooy (editors). 1966. Pacific Plant Areas. Vol. 2. Blumea Suppl. Vol. 5: 1-312. Tabling, D. H. 1971. Gondwanaland, palaeomagnetism, and continental drift. Nature 229: 17-21. Tryon, R. 1970. Development and evolution of fern floras of oceanic islands. Biotropica 2: 76-84. Wegener, A. 1915. Die Entstehung der Kontinente und Ozeane. Sammlung Vieweg, Bruns- wick; No. 20. Wolfe, J. A. & E. B. Leopold. 1967. Neogene and early Quaternary vegetation of north- western Nortli America and northeastern Asia. In D. M. Hopkins (editor), "The Bering Land Bridge," pp. 193-206. Stanford. Young, S, B. & J.-R. Klay. 1971. Bryophytes in the 1969 crater of Deception Island, Ant- arctica: An apparent case of rapid long-distance dispersal. Ohio Jour. Sci. 71: 358-362. Zanten, B. O. van. 1971. Musci. In E. M. van Zinderen Bakker^ Sr., J. M. Winterbottom & R. A. Dyer (editors), "Marion and Prince Edward Islands," pp. 173-227. Cape Town. NOTES ON PANAMANIAN TREES AND SHRUBS COLLECTED IN 1971 BY L. R. HOLDRIDGE AND OTHERS John D. Dwyer^ The following is a list of trees and shrubs collected by L. B. Holdridge, E. A, Lao, L. Maasola, and A. Gentry during the spring and summer of 1971 in the Republic of Panama. All the collections were made at approximately sealevel; the few collected at or above 100 meters will be noted in the list. No collection was made above 650 meters. Collections of the following families seem especially important: Humariaceae, Lauraceae, Myristicaceae, and Sapotaceae. Some ma- terial was identified in the sterile state by Dr. Holdridge and possibly by his colleagues; the sterile collections are recognized in the list by the asterisk (*) placed before the collector's name. The author identified approximately one half of the collections which are in flower and fruit. The collections of the Burseraceae and Rutaceae were identified by Dr. Duncan Porter; the Boraginaceae were identified by Dr. Joan Nowicke. Dr. Thomas Croat gave valuable assistance in the critical identification of several species. Miss Mireya Correa A. of Panama University deserves a special note of thanks for forwarding some of the collections. Mr. Yow-Yuh Chen of the Department of Biology, St Louis University, prepared the figure of the Sacoglottis fruit. Among the more than 180 collections in the list one encounters 47 localities in Panama. For the sake of convenience the localities have been segregated along provincial lines and listed below, with each locality being given a number. The number is placed in parentheses folloA\dng the collector and his collection number. BOCAS DEL TORO 1. Isla Colon; 2. La Gruta, Isla Colon; 3. El Chumical. Canal Zone 4. Albrook; 5. Ancon; 6. Balboa; 7. Pipe Line Road; 8. Summit Garden. Colon 9. Between Colon and Portobclo; 10. Buena Vista; 11. Entrada a Sabanitas; 12. Gatiin; 13. Maria Chiqiiita; 14. Playa Langosta; 15. Portobelo; 16. Rio Indio; 17. Rio Piedras; 18. Rio Santa Izabel; 19. Rio Trapiche; 20. Road to Portobelo; 21. Salud; 22. Santa Rita; 23. Santa Rita Arrfl7a; 24. Villa Aloiidra, road to VovlohiAo; 25. West of Canal Zone. Herrera 26. Menchacha, Ocu; 27. Llano de Las Minas; 28. La Cabuya, Las Minas; 29. Cerro Colorado de Las Minas; 30. El Chunial; 31. Los Hatillos, Pese; 32. Quebrada El Canimaron; 33. Divisa. Los Santos 34. El Ejido de Los Santos. Panama 35. Arraijan, Cerro Silvestre; 36. Calzada Larga; 37. Canitos de Chepo; 38. Capira; 39. Cerro Azul; 40. Chichebre, Chepo; 41. Chiltepe; 42. La Chorrera; 43. La Cresta; 44. Nuevo Guarare; 45. Panama; 46. Rio Indio; 47. Sajalices, Capira. ^ St. Louis University and Missouri Botanical Garden, 2315 Tower Grove Avenue, St. Louis, Missouri 63110. Card. 59; 247-261. 1972. 248 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 In citing the collections the following format has been used. The families are arranged alphabetically. In parentheses^ following tlie name of the family, one finds the author, the part, the volume, pagination, and year of publication of the specific family as found in the Flora of Panama^ published periodically in the Annals of the Missouri Botanical Garden. Some families have not been treated in the Annals, e.g. Rubiaceae. Following the binomial name and its author, the collectors are abbreviated as follows: G — Gentry; H — Holdridge; L — Lao; M — ^Maasola. The collection site of each collection is indicated by the number in parentheses following the collection number. The DBH of the tree or shrub is given in centimeters and the height in meters, as for example 8 cm X 6 m. Common names, supplied by col- lectors, are appended. The notes which accompany many of the species in the list are those of the author, unless otherwise indicated. Special stress is placed upon range extensions within tlie Republic of Panama. Families recendy treated in the Flora of Panama may be considered reliable in determining geographical ranges of species in the Republic. All determinations made by the author were checked in the herbarium of the Missouri Botanical Garden. Dr. L. R. Holdridge also checked much of the material at the same institution. It is appropriate to point out that many species of trees and shrubs from Pan- ama are poorly represented in herbaria. The arboreal vegetation, and for that matter, all kinds of vegetation, along the Atlantic face of the Isthmus, is poorly known. This is especially true of the area extending from the Province of Bocas del Toro, especially its eastern part, to the western side of the Panama Canal and More tha om part of Gatun Lake in the Canal Zone. These collections serve to svipport the need for intensive collecting in Colon. Equally significant are the collections in the list, though fewer in number, made in Colon on the eastern side of the Canal. The Santa Rita area in Colon, which is currently attracting a large number of collectors, has yielded several new species and has proven important in extending the geo- graphical ranges of many others. While our collectors did not extend their efforts to the Province of Darien, it is appropriate to point out that there are few collections from the mountains of Darien. This makes it difficult to trace the geographical distribution of montane species at the eastern end of the Republic and in neighboring Colombia. By the same token the few specimens available from the Atlantic side of the western end of the Isthmus present a problem in defining phytogeographical relationships between Costa Rica and Panama. The reader is referred to my remarks concerning areas poorly collected in Panama (Dwyer,1964: 115). Annonaceae (Fries— Part 4. Vol. 49: 491-525. 1962). Annona glabra L. — L ir H 234 (21), 10 cm X 8 cm. This species grows in southern Florida, tropical America, and West Africa. It is common in Pan- ama. A. spraguei Saff. — L ir H 179 (25), 9 m. A strictly Panamanian species with wide distribution in the Repviblic. 1972J DWYER— PANAMANIAN TREES AND SHRUBS £49 Malmea ■L 96 (1), 30 cm X 20 m, "Yaya" This species treatment (MO)) Panama. The species extends from Mexico to y> Stelechocarpus burahol (Hook, f.) Thorns.—// 65/8 (8), 20 cm x 6 m. This exotic species, native of Malaya, was identified by Dr. Croat. The flowers are cauliflorous. Unoniopsis floribunda Diels — L 91 (1), 20 cm X 16 cm, "Yaya Amarilla. U. panamensis Fries — L ir H 161 (20), 4 cm X 7 m. This species heretofore has been known only in Panama from Cerro Campana, Province of Panama. U. pittieri Saff. — L i:r H 50 ( 16), 3 cm X 6 m. This species has been previously collected in the Canal Zone and in the Province of Colon. It has been re- ported from British Honduras, Honduras, and Panama. Xtjlopia macrantha Tr. & P. — L ir H 51 ( 16 ) . The fruits only were collected. The aggregate fruits are as large as a tennis ball. Dr. Croat reports that he has collected fruits twice this size on Barro Colorado Island, Canal Zone; the follicles, on splitting become brick-red within. Anacardiaceae (Blackwell & Dodson— Part 6. Vol 54: 351-379. 1969). Spondias mombin L. — L 62 (47), — 18 cm X 10 m. A common tree in Panama. Apocynaceae ( Nowicke— Part 8. Vol. 57: 59-130. 1970). Himatanthus sp.— L dr H 188 (21), "Calacuchillo," collected at 100 m. This sterile specimen is probably H, articulata ( Vahl) Woodson. It is known in Panama only from the Provinces of Darien and San Bias. Nowicke (p. 81) erroneously cites Williams 823 (NY), collected in Cana, Darien, as having been collected in the Province of Colon. Malouetia tamaguarina (Aubl.) DC — L 159 (20), 6 cm X 5 m. This species has not been reported north of the Guianas. Holdridge identified this col- lection, I note, however, that the fruits are not elongate and linear but rather measure about 1^/^ cm in diameter. The leaves of the Lao collection, on the other hand, certainly suggest M. tamaguarina. Aquifoliaceae( Edwin— Parte. Vol. 54: 381, 1967). Ilex guianensis (Aubl.)Ktze— L 6 H 229 (21), 10 cm X 8 m. This species has being the initial collection from Colon. Chiriq Araliaceae Dendropanax arboreus (L.) Dene. & Planch, — L, H ir G 15 (7), 15 cm X 10 m, collected at 350 m elevation. This species has been reported from prac- tically all Provinces of Panama. Bombacaceae ( Robyns— Part 6. Vol. 51: 37-68. 1964). Theobroma bernouUii Cuatr. — L & H 213 (21), 15 cm. The sole representative of this species is the type Pittier 4105 (US). It was also collected from the Province of Colon. The flowers are cauliforous. BoRAGiNACEAE ( Nowicke— Part 9. Vol. 56: 33-69. 1969). Cordia dentata Poir— L 45 (30), 30 cm x 10 m, "Cuguaro" and "Billulo." This is common on the Pacific slope, having been collected in the Provinces of Herrera, Los Santos, and Veraguas. 250 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 C. panamensis Riley — H 6444B (10), 30 cm X 10 m. A common species much in need of study for intraspecific variation. It ranges from El Salvador to Panama. C, porcata Nowicke — L, H 6- G 7 (22), shrub or treelet, collected at 100 m. Nowicke reports Dwijer ir Lallathin 8586 (MO) incorrectly as collected in the Province of Los Santos. BuRSERACEAE ( Porter— Part 6. Vol. 57: 5-27. 1970). Protium costaricense (Rose) Engl— L, H i^ G 16 (7), 20 cm X 10 m. This species is known from seven collections in the Province of Panama and one in the Canal Zone (Barro Colorado Island). P. glabrum (Rose) Engl— L I- H 172 (21), 3 ci While collections of this species from the Province of Colon, Pittier 3949^ 4190, 4191 (all US), and Tyson et al 450 (MO), the present collection from Salud represents the most westward collection of the taxon in Panama. P. tenuifolium subsp. sessiliflora (Rox.) Porter — L 123 (37), 15 cm X 6 m. A well collected species, known from numerous collections in the Canal Zone, as well as from the Provinces of Chiriqui, Darien, and Panama. Celastraceae Maytenus jamaicensis Kr. & Urb. — L h- H 23 (21), 25 cm X 10 m. This species, according to Dr. Croat, is present on Barro Colorado Island, Canal Zone. I suspect that it is common in the Provinces of Panama and Colon. Lao ir Holdriclge 23 is in fruit and flower, a combination rarely observed on her- barium specimens. The collectors describe the fruit as orange, Zinowieicia costaricensis Lundell— L ir H 49 (36). — This species, known from Costa Rica and from a solitary collection in Panama, Maxon 5109 (F), from El Boquete, Province of Chiriqui, has its range extended considerably eastward in the Republic. Coghlospermaceae ( Robyns— Part 6. Vol, 54: 61-64. 1967). Cochlospermum williamsii Macbride — * L ir H 178 (10), 15 cm X 9 m. While the collection is sterile, Holdridge has identified it as to species. This taxon, originally described from Peru, was not reported from Panama in Robyns' recent treatment. Combbetaceae ( Exell— Part 7. Vol. 45: 143-164. 1958). Terminalia bucioides Standley & Williams — L ir H 15 (24), 25 cm X 15 m; H ir M 6530 (13), 10 cm X 7 m. This species is not listed in the relatively recent treatment of the family in the Flora of Panama. T. edulis Blanco — H 6521 (8), 20 cm X 6 m. An exotic tree growing in Summit Garden. It is native to the Philippines. Dilleniaceae ( Hunter— Part 6. Vol. 52: 579-598. 1965). Saurauia laevigata Tr, & PI.— H ir M 6528 (13), 25 cm X 8 m, collected at 150 m; L ir H 182 (21), 10 cm X 7 m. This species is common in Panama. Erythroxylonaceae Erythroxylum ampliim Benth. — L ir H 226 (21), 8 cm X 6 m. I have seen three sheets of this species in the Herbarium of the Missouri Botanical Garden; all of these are from Barro Colorado Island, Canal Zone: Shattuck 1022 and Croat 8265, 12679. 1972] DWYER— PANAMANIAN TREES AND SHRUBS £51 EuPHORBiACEAE ( Wcbster & Burch— Part 6. Vol. 54: 211-350. 1967). Acalypha diversifolia Jacq.^ — H 6446 (19), 3 m. This shrub is common in Central America. Croton sp.—L ir H 224 (21), "Algodoncillo." This may be a new species. Tliis collection has 3-lobed leaves resembling in shape and size those of Gos- sypium harhar dense L. I have been unable to match it with any of our identified Central American crotons. Among the undetermined crotons I discovered a recent collection {Kirkhride & Duke 1665 (MO)) from nearby Cerro Trinidad, Province of Panama, with the same vernacular name, "Algodoncillo/' It compares favorably with the Lao and Holdridge collection. The field notes on the Kirkbride and Duke sheet are ilhmiinat- ing: "Tree without red latex, 20 m tall; bark whitish; wood malodorous, rank smelling; flowers scurfy, greenish; stamens greenish-yellow. Second growth and culled forest, SE slopes of Cerro Trinidad, yy C. hillhergianus MuelL-Arg.— L 148 (30); LirH 170 (21), 6 cm X 5 m; L ir // 202 (21 ), 8 m, "Vaguero.** A common species in Panama. C. panamensis (Kl.) Muell.-Arg. — L 148 (30), 15 cm X 7 m, collected at 140 m. i< Sangrillo." A common species in Panama. Mabea occidentalis Benth. — L, H 6- G 2 (22), 6 m, collected at 100 m. A com- mon species, collected on several occasions by the author on Cerro Azul and Cerro Jefe, Province of Panama. It is rather common arovmd the Canal Zone and adjacent Provinces, although readily confused with M. montana Muell.-Arg., the latter with leaves only subacuminate at the apex. Mabea occidentalis also occurs in Costa Rica. Margaritaria nobilis L. f. — L ir H 236 (21 ), 15 cm x 8 m. This has been widely collected in Panama but heretofore has not been reported for the Province of Colon. Pera arborea Mutis — L, H ir G 12 (23), 15 cm X 10 m. Known from previous collections in the Canal Zone and the Provinces of Colon, Darien and Pan- ama. It also occurs in Colombia. Phyllanthus acuminatus Vahl — H 6525 (13). A common species ranging from Mexico & Sapium caudatum Pittier — L 113 (44), 20 cm X 10 m, "Ohvo. yy rorchidium gorgonae Croizat subsp. robledoanum (Cuatr.) Webster — L 6- H 13 (23), 50 cm X 10 m; Lt-H 33 (39), 16 cm X 10 m, collected at 600 m. This species heretofore has been known only from Cen'o Jefe, Province of Panama, and west of Gatun Locks in the Canal Zone. The present collcc- Jefe Flacourtiaceae 1968) Casearia javitensis HBK — L ir H 230 (21), 15 cm X 6 m. Common in Panama. C. sylvestris Sw. — H 6458 (40), 10 cm x 7 m. Common in Panama. It ranges n Mexic thalmia 600 common I speciosa var. panamensis Monachino — H ir M 6532 (13), 5 m, collected 100 m. This species is known from the Provinces of Bocas del Toro, 252 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Canal Zone, Colon, and San Bias. The Holdridge collection was made only a few miles from Santa Rita Ridge, Province of Colon, where Dressier and Correa {647 (MO) ) recently collected material of this species. Zuelania guidonia (Sw.) Britt. & Millsp.— H 6445 (19), 20 cm X 10 m. This species, with a wide range from southern Texas to Panama, has heretofore been reported from Panama only once, Pittier 2710 from "the hills between Rio Grande and Pedro Vidal, Canal Zone." Dr. Croat has collected it sev- eral times on Barro Colorado Island, Canal Zone {Croat 4962, 4973, 7187, 8393 (all MO). Other collections from Barro Colorado Island, deposited in the Missouri Botanical Garden are Woochvorth ir Vestal 719; Zetek 3343, 3463). . GUTTIFERAE Calophyllum brasiliense Camb. — H b- M 6534 (13), 75 cm X 30 m. Tliis large tree was identified from dried leaves. The material matches well a fruiting collection of the author {Divyer et al 7301 (MO)) from the summit of Cerro Jefe, Province of Panama, a well-known collecting spot, less than 20 miles from Maria Chiquita, the location of the Holdridge and Maasola col- lection. This species ranges from southern Mexico to Panama. Marila verapazensis Donn. Sm. — L, H ir G 17 (7). This species is known from several collections in Panama. Holdridge notes that "this species grew in La Selva, Costa Rica." It ranges from British Honduras to Colombia. Rheedia madruno (HBK) Tt. & PI.— L ir H 190 (21), 40 cm X 20 m, "Satro." As this species has been collected on Barro Colorado Island ( Croat 10840 (MO) ), its collection in nearby Salud, Province of Colon, is not surprising. HUMARIACEAE Sacoglottis ovicarpa Cuatr.— L, H ^ G 5 (22), "Conocillo"; * L ir H i93 (21), 19 cm X 18 m, collected at 120 meters elevation; L b- H 195 (21), 60 cm X 25 m, "Corotu." L. R. Holdridge identified the three collections and in deference to his field experience with Sacoglottis I am giving his identifi- cations priority. As the discussion which follows indicates, however, I am prepared to challenge them. Of the three collections listed here, the one designated by an asterisk is sterile, while the others are in fruit. All three collections were made on the Atlantic side of the Isthmus. The fruits of Lao ir Holdridge 195 are oblong, measuring 3-4 cm in length and 2.2-2.8 cm in width. One drupe measuring 4 cm in length, was sectioned transversely with a handsaw. The surface (exocarp) is smooth and deep purple-red in color in the dry state. The exocarp and mesocarp are evident on the cut face, the mesocarp appearing as a thin line. In softening a slice of the fruit in aerosol and boiling in glycerine the mesocarp became even more evident. The exocarp is thin, measuring 0.1-0.2 cm in width; the ligneous endocarp is studded with resinous subrotund cysts which measvire 1-4 (-6) mm in diameter. There is no separation of the resinous cysts into large and small groups. The septal lines are faint on the cut face. The fruit of Lao, Holdridge ir Gentry 5 is oblong, measuring 3 cm in 1972] DWYER— PANAMANIAN TREES AND SHRUBS 253 .-•' re m n FiGUKE 1. Cross section of fruit of Sacoglottis sp. (or Schistostemon) showing cut surface natural size. En — endocarp; ex — exocarp; m — mesocaip; re — resinous cyst; sc — seminal cavity. Based on Lao ir Holdridge 195 (MO). X2. length and 2 cm in width. The surface is rugose and pitted (resinous cysts), the exocarp presumably having rotted away. The superficial pits of the endocarp are more or less circular, measuring 1-1.5 mm in diameter (c/.Fig.l). On being cut transversely with a band-saw, tlie ligneous endocarp of Lao, Holdridge h- Gentry 5 exhibited great resistance. Most noticeable on the cut surface are 3 large seminal cavities, measuring ± 6 mm in diameter. The resinous cysts, scattered throughout the valves of the endocarp, are small, with a diameter measurement of 0.5-1.5 mm and averaging about 6 per valve. The cut faces of tlie valves appear cuneiform. On cutting the fruit, the 5 parts (which I regard as valves) of the endocarp easily sepa- rated themselves ( or if they remained in position, could be pried loose with a fingernail) from 5 radially disposed septa of about equal length {of. Fig. 2). These septa may contain occasional resinous cysts or canals. Each arm of the "star" measures about 0.2 cm in diameter on the cut surface. Viewed laterally, the septum extends almost the full length of the endocarp (Fig. 1). The remnants of the disintegrated seeds are evident From a study of the illustrated fruits in Cuatrecasas (1961) revision of the Humariaceae, especially Fig. 31, n-p I decided that Lao, Holdridge 6- Gentry 5 either belongs to the genus Schistostemon, known from the Gui- anas and parts of the Amazon Basin, or is to be referred to Sacoglottis mattogrossemis Malme, found in Colombia, Venezuela, the Guianas, and Brazil. The genus Sacoglottis has been reported from Panama by Johnston (1949; 161) as occurring in the drift on the beaches of San Jose Island, Province of Panama, and by Holdridge (1970: 260). Holdridge in his manual does not designate the species of Sacoglottis which occurs in Pan- ama, however. At the same time he indicates that Sacoglottis occurs in 254 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 A Figure 2. Fruit of Sacoglottis sp. (or Schist ostemon sp.), endocarp only shown natural size. — A, External view, the resinous cysts visible as holes of varying diameter. — B. Endocarp shown in hemisection, one displaced valve shown to right; the seminal ca\ities are large, the resinous cysts smaller. — C. Lignified septa with all valves absent. Based on Lao, Holdridge ir Gentry 5 (MO), Xl%. Costa Rica. Dr. Holdridge says (personal communication) that S. ovicarpa Cuatr. is the forest dominant on Cocos Island, Costa Rica. In conclusion, one speculates whether Sacoglottis and Schistostemon may not be congeneric. The floral differences separating tlie two genera do not appear to be strong, judging from Cuatrecasas' work. What is espe- cially needed is intensive collection of the Humariaceae, particularly the fruits in various stages of development Lacistemaceae (Nevling— Part4. Vol 47: 124-127. 1960). Lacistema aggregatum ( Berg. ) Rusby — H 6451 ( 19 ) , 8 cm X 6 m, A common tree in the Republic. Lozania pedicelhta (Standley) L. B. Smith— f/ 6484 (15), 6 m. There are only three collections of L. pedicellata from Panama, von Wedel 2121 (MO) from Old Bank Island, Province of Bocas del Toro, Ebinger 348 (MO) from Cerro Campana, Province of Panama, and Shattnck 972 from Barro Colo- rado Island, Canal Zone. Lauraceae( Allen— Parts. Vol. 35: 1-68. 1948). Nectandra fuscobarhata (Mez) Allen — H 6403 (15), 40 cm X 15 m. Known previously from two widely separated localities in Panama: Isla Colon, Province of Bocas del Toro, and Pinas Bay, Province of Darien. N. gentlei Lundell — H 6407 (6), 30 cm X 10 m. Known in Panama from the lowlands of Chiriqui Province and from the Canal Zone. N. glohosa (Aubl.) Mez— L & H 32 (36), 20 cm X 10 m, collected at 100 meters elevation; H 6447 (18), 35 cm X 13 m. This species ranges from 1972] DWYER— PANAMANIAN TREES AND SHRUBS £55 Ci. Guatemala to Panama; it has been reported in Panama from the Province of Panama and the Canal Zone. N. martinicensis Mcz—H 6508 (21), 22 cm X 10 m. Our material in the Mis- souri Botanical Garden is from Tobago, Trinidad, and Granada in the West Indies. N. standleyi Allen— L 102 (2), 15 cm X 10 m, "Sigua Amarillo"; L 103 (2), Sigua Blanco." This species is known from Bocas del Toro and possibly from the Province of Panama, especially from Juan Diaz {Allen 942 (MO)). Persea caerulea (R. & P.) Mez — L 46 (27), 30 cm X 15 m, collected at 350 m elevation. There is a single collection of this well-known species in the Missouri Botanical Garden, Allen 1015 from Boquete, Province of Chiriqui. Phoebe johnstonii Allen— H 6433B (45), 10 cm X 5 m; H 6518 (5), 20 cm X 8 m. This has been previously reported from San Jose Island, Province of Panama (Johnston 697 (MO, the type collection) ) and from the Province of Chiriqui ( Stern et al 1143 ( MO ) ) . P. mexicana Meisn. — IL 6475 (40), 8 m. This species is well distributed in Mexico and Central America. This is probably the first report of the species in Panama, although Dr. Croat informs me that he has undistributed mate- rial of the species from Barro Colorado Island, Canal Zone. Lecythidaceae ( Woodson— Part 7. Vol 45: 115-136. 1958). Grias fendleri Seem. — L ir II 232 (21), 25 cm X 8 m. This species is known from the Provinces of Bocas del Toro, Canal Zone, and Colon. Dr. Croat reports its presence on Barro Colorado Island, Canal Zone. Eschweilera calyculata Pittier — L ir H 158 (20), 12 cm X 10 m. This is the first report of the species in the Province of Colon. It has been previously reported from the Provinces of Bocas del Toro and the Canal Zone. E. pittieri E. Kunth— L, H pe, Pittier 6601 (US), is from the vicinity of La Palma, Province of Darien, Panama. Literature Cited Cuatrecasas, J. 1961. A Revision of the Humariaceae. Contr. U.S. Natl Herb. 35: 25-214. DwYER, J. D. 1964. Panama, Plant Collection and the Missouri Botanical Garden. Ann. Missouri Bot. Card. 51: 108-117. 1968. Borojoa and Tocotjena (Rubiaceae) in Panama. Phytologia 18: 445-449 HoLDRiDGE L. R. 1970. Inventariacion y Demonstraciones Forestales Panama (Manual Dendrologico Para 1000 Especies Arboreas en la Republica de Panama (Programa de La.s Naciones Unidas) ). Panama. Johnston I. 1949. The Botany of San Jose Island (Gulf of Panama). Sargentia 8: 1-306. SOLANACEAE STUDIES II: TYPIFICATION OF SUBDIVISIONS OF SOLANUM W. G. D'Arcy^ A useful step in the study and identification of plants is the assembling of species into infrageneric groups, and this is especially so in the genus Solatium. In this genus the great number and diversity of species discourages study, yet many groups can readily be assembled witliin the genus which have every appear- ance of being natural and distinct from one another. In the past, many groups were formally recognised with a considerable redundance of names. To bring some order to the surfeit of names, names of subdivisions of the genus are reviewed here with respect to their rank, validity of publication and typification; and to clarify the relationships of taxa to one another, a provisional conspectus is pre- sented, not as a reworking of the taxonomy of the genus so much as a framework from which nomenclatural values can be seen. The desirable sequel, designation of the limits of taxa and keys to distinguish them, must await later efforts. The present paper is one necessary step in making the parts of this important genus more accessible to study. The literature search was careful and perhaps exhaus- tive, but there is the usual possibility of unexpected discoveries altering the nomenclature presented here, and in the future, especially when the flora of South America is better understood, there will be additions to the generic subdivisions noted here. Help was sought from other workers: Susan M. Coles, University of Birming- ham, chose the lectotype for section Cnjptocarpum; R. M. Polhill, Royal Botanic Gardens, Kew, chose lectotypes for African taxa, and Don Ugcnt, University of Southern Illinois, Carbondale, chose the lectotype for Regmandra, Roger Polhill also helped in checking bibliographic references, and R. K. Brummit, Royal Botanic Gardens, Kew, uncovered several names overlooked by the writer. In addition to the above who helped directly in the study, the late G. V. Morton, Smithsonian Institution, Washington, and W. D. Margadant, Biohistorisch Institut der Rijksuniversiteit, Utrecht, provided advice on particular aspects of the project. Assistance from these people is gratefully acknowledged: the writer at the same time takes full responsibility for all choices and decisions not clearly attributed to others. The alphabetical listing of names indicates the first rank, place, and date of the the This is followed by subsequent changes the type species, Holotypes and automatic choices under Art. 22 of the Code are indicated as "Type species." Other choices are "Lectotyp follows in square brackets. the writer, source of the ^ Missouri Botanical Garden, 2315 Tower Grove Avenue, St. Louis, Missouri 63110. Ann. Missouri Bot. Gard. 59: 262-278. 1972. 1972] D'ARCY— SOLANACEAE STUDIES II 963 In the provisional conspectus of the genus, synonymy based on the same type species is indicated by an equals sign ( = ). The nearly equals sign (^) is used to indicate taxonomic affinity which may in fact be taxonomic synonymy, but this paper does not present any new decisions as to taxonomic synonymy. The nearly equals sign and placement in the hierarchy either reflect a decision by a previous worker or merely suggest a close affinity. In most cases, taxa of equal rank are listed in alphabetical order. Names without clear indication of rank ( gradi ambigui ) are placed in the conspectus to reflect their presumed position in the taxonomic hierarchy witliout intention of assigning them to rank at this time. Changes in Art. 22 (on autonyms) in tlie 1972 Code require some different names from those called for under the 1966 Code. Thus section Herposolanum is now correct instead of the autonym Bassovia and sect. Acanthophora is correct instead of sect. Leptostemonum, This brings to mind a remark of Wm. T, Stearn (British Museum, N. H.) that there are more changes of names because of legis- lation than for any other cause. The ranks of names published by some authors are only indirectly evident. Dunal in 1813 {Hist, Sol) used twelve names for groups of Solanuin species, and r in 1816 (Sol. Syn.) he added two more. On page 43 of die 1813 work he referred to section Pteroidea. On page 172 he suggested that several species miglit be grouped into a distinct section, and in 1816 he gave these same species the name Leiodendra. As all names used by Dunal in these two pubHcations are of apparent equal rank, they are all treated as sections. More than 35 years later in the Sola- naceae of the De CandoUe Prodromus (Prodr. 13(1): 1-690. 1852), Dunal paid little attention to earlier names and erected a completely new framework for the genus. A host of categories between the rank of subsection and species were given names and diagnoses, and a few of these gradi ambigui have been assigned to rank by later workers. Lowe in 1868 (Man. FL Madeira) used new names for several infrageneric taxa in Solanum. Because he attributed two names {Morella and Tuberarium) to Dunal without comment, it is presumed he was accepting the names at Dunal's rank of subsection. Lowers names are all of apparent equal rank. The same rationale is used in accepting names in Solamim in Dumortier's Florula Belgica (1827), even though Dumortier changed tlie orthography of DunaFs Maurella to Morella. In 1913, Georg Bitter (Fedde Rep. 11: 381) equated the German word "Reihe" with the Latin "sectio," so his use of Reihe in subse- quent places is interpreted to mean section, Wessely (1961), in commenting on Pojarkova's ( 1955) series, also used die term Reihe in this sense. Bitters "Grossart" (Dunaliaiui) and Pojarkova's "Cycle" {Alata and Nigra) are considered to be gradi ambigui. The treatment of names published without clear indication of rank has long been the subject of debate, but all proposals to amend Art. 35 of the Code, which deals with this matter, were defeated at the Seattle Congress. Changes in Art. 22 (on autonyms) effected at the Seattle Congress do reduce the possibihties for negating the priority of names by assigning gradi ambigui to higher {e.g. sub- generic) rank and thus substituting an autonym for a name of lower rank {e.g. 264 ANNALS [Vol. 59 section), but the situation is by no means clear. An example is the taxon Lasio- carpa Dun. The present writer wishes to recognise a section with Solanum lasiocarpum Dun. as the type species. Introduction of a completely new name would leave open the possibility that another writer might at some time assign hasiocarpa Dun. to the rank of section and claim it had priority as a section from 1852 when first published by Dunal. The Dunal name is chosen for use here, and it appears in the form section hasiocarpa (Dun.) D'Arcy. But two other interpre- tations are possible: first that because the publication of the name by Dunal indi- cated no rank, it has no priority in any rank and the section should be treated as newly published, Le. sect, hasiocarpa D'Arcy; while a second alternative interpre- tation might be drawn from Art. 22 of the Code to say that because the section takes its name from its type species, prior use of the name at any rank other than section is immaterial and the name at sectional rank is sect, hasiocarpa D'Arcy. Fortunately most of the early names for subdivisions of Solanum are assignable to rank. The question of treatment of gradi ambigui has been left for future disputa- tion; but such names in Solanum have been assigned type species, and they have been placed in tlie generic conspectus in likely positions. Establishment of type species for many infrageneric taxa in Solanum is auto- matic, cither because the author included only one species in his original proto- logue (holotype species), or by application of Art. 22 of the Code which in part provides ; "When the epithet of a subdivision of a genus is identical with or derived from the epithet of one of its constituent species, this species is the type of the name of the subdivision of the genus unless the original author of that name designated another type. » This rule was not part of the Code when Seithe (1962) selected S. swartzianum rather than S. lepidotum as the lectotype of sect, hepidotum. Fortunately these iewpoint very By portion Petota and most series in subgenus heptostemonum are automatically typified. A future Botanical Congress might well consider a recommendation that new names for series and subseries be formed from the stem of the epithet of a con- stituent species. Lectotyp type for hijcianthes The lectotyp and living material of most taxa. In those few cases where material was not available, study of the literature and in one case of a type photograph preceded selection of the lectot}^e species. Lectot>pes for three series, Borealia, Glabres- centia, and Transaequatorialiu, were not selected, leaving the choice to someone more familiar with these groups than the present writer. Without some structmre, nomenclatural relationships cannot be understood, much less taxonomic relationships. The conspectus of the genus presented here is derived for the most part directly from the schemes of other workers, and the new elements of taxonomy are few. At the sectional level, the scheme used is 1972] D'ARCY— SOLANACEAE STUDIES II £65 roughly that of Seithe (1962) with corrections in nomenclature and the addition of several sections. The infrasectional arrangement in sect. Potatoe is adjusted from that of Hawl:es ( 1963 ) , that in sect. Basarthrum is taken from Correll ( 1962 ) and that for sect. Solatium from Pojarkova (1955). Infrasectional treatment for most sections in subgenus Leptostemonum are taken from various papers of Bitter (1911-1923; summarised by Seithe, 1962). Working mainly with the African species. Bitter distinguished several sections which may be considered taxonomi- cally synonymous when a full range of the genus in the New World is taken into account. In particular, sect. Melongena should probably include sects. Lathyro- carpum, Leprophora^ Oliganthes^ and Torva and perhaps also sects. Micracantha and Eriophyllum, The new sections Lasiocarpa and Extensum are proposed largely as a result of the author's recent studies of Central American Solanaccae. Dunal (1813, 1852), Bitter (1911-23) and Seithe (1962), among others, have tried to recognise a fundamental dichotomy on tlie genus which separates species with stellate hairs, tapering anthers, and spines from those species with simple hairs, stout anthers, and no spines. This simplistic approach has led to disagree- ment over placement of a number of groups. Rather than two lines of evolution, it would appear that there are several distinct lines of evolution in the genus. Seven subgenera are recognised here: subgenus Archaesolanum includes South Pacific species with distinctive leaf shape, axillary inflorescences, and sometimes aneuploidy; subgenus Bassovia includes groups with stout anthers, simple hairs, no spines, pinnate leaves without interstitial leaflets, and in many cases, axillary inflorescences and pointed fruits; subgenus Brevantherum includes groups with stout anthers, entire leaves, and dendritic or stellate hairs; subgenus Leptostem- onum includes groups with tapering anthers, stellate hairs, and often spines; sub- genus Lyciosolanum was recognised by Bitter and Seithe on the basis of the elongate filaments and local distribution (South Africa) of its sole species, S. ag- gregatum Jacq. (^ S. guienense L.), but its distinctiveness at the subgeneric level is not apparent to this writer and perhaps it should be considered a section of subgenus Solarium; subgenus Potatoe includes groups with scandent species, pinnate leaves often with interstitial leaflets, lateral pendulous inflorescences, and articulation of the pedicels above the base (sometimes only slightly so), which may indicate the site of an ancestral bracteole; and subgenus Sokinum which includes groups with stout anthers, simple hairs and no spines. Section Aculeigerum may also represent a distinctive line of evolution from the common ancestral Solarium stock and warrant recognition at the subgeneric level. For present purposes it is placed as a section of subgenus Potatoe. It is important that any omissions or errors in the following listing be brought to early notice so tliat nomenclature in the genus may soon achieve a good measure of stability. Names for Subdivision's of the Genus Solan um Acanthocalyx Series Bitt, Fedde Rep. Beih. 16: 175. 1923. Lectotype species: S, richardii Dun, Acantha})hora Section Dun., Hist. Sol. 131, 218. 1813. Subsection G. Don, Gen. Syst. 4: 434. 1838. Grad, ambig. Dun. in DC., Prodr. 13( 1 ): 30,235. 1852. Lectotype species: S. mammomm L. 2QQ ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Acaulia Series Juz. ex Buk. & Kameraz, Princ. Potato Breed. 21. 1959. Type species: S. acaule Bitt. Aculeastrum Series Bitt, Fedde Rep. Beih. 16: 165. 1923. Type species: S. aculeastrum Dun. Aculeata Grad. ambig. [Psubgenus] Dun., Hist. Sol. 125. 1813. Nomen nudum. Section G. Don, Gen. Syst. 4: 423. 1838. Grad. ambig. [^Division] Walp., Rep. Bot. Syst. 3: 67. 1844. Lectot>T)e species: S. mammosum L. Acttleigenim Section Seithe, Bot. Jahrb. 81: 291. 1962. Type species: S. wendJandii Hook. f. Acthiopica Series Bitt., Fedde Rep. Beih. 16: 43. 1923. Type species: S. aethiopicum^ L. Afrodecumbens Series Bitt., Fedde Rep. 16; 163. 1923. Type species: S. nigriviolaceum Bitt. Afroindica Series Bitt., Fedde Rep. Beih. 16: 4. 1923. Lectotype species: S. indicum L. Ajrosolanum Section Bitt., Bot. Jahrb. 54: 420-1, 440. 1917. Lectotype species: S. terminale Forsk. [Seithe, 1962], Alata Series Pojark., Not. Syst. 17: 336. 1955. Grad. ambig. ['Cycle ] Pojark., Fl. URSS. 22: 32. 1955. Type species: S. alata Mill. Albicaule Series Bitt., Fedde Rep. Beih. 16: 100. 1923. Type species: S. albicaule Dun. Alticola Series Buk., Prob. Bot. 2: 319. 1955. Nomen invalidum [no Latin description]. Type species: S. alticola Bitt. Anarrhichomenum, Section Bitt., Fedde Rep. 11: 247. 1912. Lectotype species: S. sodiroi Bitt. [Seithe, 1962], Andigena Series Buk. in Buk. & Kameraz, Princ. Potato Breed, 24, 1959. Type species: S, cmdigenum Juz. & Bxik. Andreana Series Hawkes, Bull. Imp. Bur. PI. Breed. 2: 50. 1944, Nomen nudum. Type species; S. andreanum Baker. Androceras Genus Nutt., Gen. Amer. 1: 129. 1818. Section Marzell in Hegi, Fl. Mitt. Eur. 2585. 1927. T>pe species: A. lobata Nutt. (=: S. rostratum Dun.). Andromonoecum Section Bitt., Fedde Rep. Beih. 16: 142, 157. 1923. (Based on Mogenoplum and Melongetia Dun.) Lectotype species; S. melongena L, Angustisegmentata Grad. ambig. Buk., Prob. Bot, 2: 325. 1955. Nomen invalidum [no Latin description] . Lectotype species: S. canasense Hawkes. Anisantherum Section Bitt., Bot. Jahrb. 54: 420, 503. 1917. Lectotype species: S. pnbescens Willd. [Seithe, 1962], Anomalum Series Bitt., Bot. Jahrb. 57: 276. 1922. Type species; S. anomalum Thonn, Anthopleuris Grad. ambig. Dun. in DC., Prodr. 13( 1): 29, 123. 1852. Section Bitt., Fedde Rep. 16: 10. 1920. Lectotype species: S. nudum H. & B. ex Dun. Anthoresis Grad. ambig. Dun. in DC., Prodr. 13( 1 ) : 29, 95. 1852. Section Bitt., Bot. Jahrb. 54: 489. 1917; 55: 66. 1919; Fedde Rep. 16: 79. 1919. Lectotype species: S. pulvemlentum Pers, Appendiculata Grad. ambig. Rydb., Bull. Torrey Bot. Club 51: 146, 174. 1924. Nomen nudum. Series Com, Potato & Wild Rel. 62. 1962. Type species: S. appendiculatum H. & B. ex Dun. Aquartia Genus Jacq., Enum. Pi. Carib. 1: 12, 1760. Grad. ambig. Dun. in DC, Prodr. 13( 1 ) ; 30, 193. 1852. Section D'Arcy, stat. nov. Type species: A. aculeatum Jacq. (= S. aquartia Dim.). Aracciana Grad. ambig. Buk., Prob. Bot. 2: 320. 1955. Nomen invalidum [no Latin descrip- tion]. 1972] D'ARCY— SOLANACEAE STUDIES II 267 Type species: S. aracc-papa Juz, Archaesolanum Subgenus Bitt. ex Marzell, Fl. Mitt. Eur. 2583. 1927. Type species: S. avicnlare Forst. f. Armutae Grad. ambig. [Psubgenus] C. H. Wright, Fl. Trop. Africa 4(2:2): 209. 1906. Lectotype species: S. aculeatisswium Jacq. (f^ S. capsicoides AH.). Articulata Series Corr., Potato & Wild Rel. 62. 1962. Type species: S. sanctae-marthae Bitt. Asterochlaena Subsection Lowe, Man. Fl. Madeira 2( 1): 80. 1868. Type species: S. auriculatum Ait. Asterotrichotum Subsection Dun. in DC, Prodr. 13( 1) : 30, 282. 1852. Lectotype species: S. carolinense L. Austroafricana Series Bitt., Fedde Rep. Beih. 16: 71. 1923. Lectotype species: S. tomentosum L. [Polliill]. Avicularia Series Herasim., Nov. Syst. PL Vase. 7: 270. 1970. Type species: S, avictilare Forst. £. Basarthrum Subsection Bitt., Fedde Rep. 11: 350. 1912. Section Bitt., Fedde Rep. 13: 101. 1913. Lectot>pe species: S. suaveolens Kunth & Bouche [Seithe, 1962]. Bassovia Genus Aubl., PL Guiane 1: 217, t. 85. 1775. Subgenus Bitt, Fedde Rep. 17: 329. 1920. Type species: B. sylvatica Aubl. ( = S. sp.). Bassovioides Gr^d. SLmhig. Dun m DC., Prodr. 13( 1) : 29, 154. 1852. Lectotype species: S. anceps R. & P. Benderanum Section Bitt., Bot. Jahrb. 54: 487. 1917. Type species: S. benderanum Schimp. ex C. H. Wright. Bifurca Series ['Reihe'] Bitt., Bot. Jahrb. 54: 436, 452. 1917. Lectot>pe si>ecies: S. hifurciim Hochst. ex Dun. (^ S. terminale Forsk.). Borealia Series Buk., Prob. Bot. 2: 323. 1955. Nomen invalidum [no Latin description]. Series Corr., Potato & Wild Rel. 388. 1962. Type species: Not selected. Brachylobus Grad. ambig. Dun. in DC, Prodr. 13(1): 31, 353. 1852. Lectotype species: S. esculentum Dun. (= S. melongena L.). Brevantherum Section Seithe, Bot. Jahrb. 81: 297. 1962. Subgenus D'Arcy, stat. nov. Type species: S. verhascifolium sensu Seithe (i= S. erianthtwi D. Don) non L. Bulbocastana Grad. ambig. Rydb., Bull. Torrey Bot. Club 51: 172. 1924. Nomen nudum. Series Corr., Potato & Wild Rel. 254. 1962. Nomen nudum. Type species: S. bulbocastanum Dun, Campanulisolanum Section Bitt., Fedde Rep. 11: 234. 1912. Lectotype species: S. fiebrigii Bitt. [Seithe, 1962]. C ampylacantha Subseries Bitt., Fedde Rep. Beih. 16: 202, 207. 1923. Type species: S. campijlacanthum Hochst. Canema Series Corr., Potato & Wild Rel. 99. 1962. Type species: S. canense Rydb. Capensiformia Series Bitt., Fedde Rep. Beih. 16: 62. 1923. Type species: S, capenseh, Cardiophylla Series Buk. in Buk. & Kameraz, Princ. Potato Breed. 26. 1959. Type species: S. cardiophyllum Lindl. Caripensa Series Corr., Potato & Wild Rel. 50. 1962. Type species: S. caripense H. & B. ex Dun. Cerasocarpum Subsection Lowe, Man. Fl. Madeira 2(1): 79. 1868. Type species: S. pseudocapsicum L. Chamaesarachidium Section Bitt., Fedde Rep. 15: 93. 1917. Type species: S. chamaesarachidium Bitt. Circaeifolia Series Hawkes, Ann. Mag. Nat. Hist. ser. 12(7): 702. 1954. Series Corr., Potato & Wild Rel. 245. 1962. Type species: S. circaeifolium Bitt. Chra Series Grab. & Dion, in Corr., Potato & Wild Rel. 243. 1962. Type species: S. cZaru/n Corr, 258 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Commersoniana Series Buk. in Buk. & Kameraz, Princ. Potato Breed. 19. 1959. Type species: S. commersonn Lam. Conicibaccata Series ['Reihe'] Bitt., Fedde Rep. 11: 381. 1912, Lectotype species: S. oxtjcarpum Schiede. Cuneolata Series Hawkes, Bull. Imp. Bur. PI. Breed. 118. 1944. Type species: S. infitndibuliforme Phil. [Hawkes & Hjerting, 1969]. Ctjphomandwpsis Section Bitt., Fedde Rep. 12: 461. 1913. Lectotype species: S. stuckertii Bitt. [Seithe, 1962]. {— Ctjphomandra stuckertii (Bitt.)) D'Arcy, comb, nov., based on S. stuckertii Bitt., Fedde Rep. 12: 461. 1913. Lectotype specimen: Argentina, Stuckert 215S9 (?B, not seen; photo NY)). Crypfocarpum Section Dun., Hist. Sol. 134, 232. 1813. Subsection G. Don, Gen. Syst. 4: 438. 1838. Grad. ambig. Dun. in DC., Prodr. 13(1): 30, 325. 1852. Lectotype species: S. balbisii Dun. [Coles]. Cypellocahjx Section Bitt., Bot. Jahrb. 55: 91. 1919. Lectotype species: S. sp. (^ Ltjcianthes sp.). Demissa Series Buk. in Buk. & Kameraz, Princ. Potato Breed. 27. 1959. Type species: S. demissum Lindl, Didcamara Genus Moench., Meth. 514. 1794. Section Dumort., Fl. Belg. 39. 1827. Subsection Dun. m DC, Prodr. 13( 1) : 28, 60. 1852. Grad. ambig. Dun. in DC, Prodr. 13( 1 ) : 28, 68. 1852. Section Bitt., Bot. Jahrb. 54: 428. 1917. Type species: D. flexuosa Moench. (=::: S. dulcamara L.). Dunahana Grad. ambig. ['Grossart'] Bitt., Bot. Jahrb. 55: 70. 1919. Type species: S. dunalianum Gaud. Durigibbosa Series Bitt., Fedde Rep. 16: 79. 1920. Lectotype species: S. cladotrichum Vandas [Polhill]. Endotricha Series Bitt., Fedde Rep. 16: 83. 1920. Type species: S. endotrichum Bitt. Eoafra Series Bitt., Fedde Rep. Beih. 16: 102. 1923. Lectotype species: S. zanzibarense Vatke [Polhill]. Episarcophyllum Section Bitt, Fedde Rep. 11: 241. 1912. Lectotype species: S. sinuatirecurvtim Bitt. [Seithe, 1962]. Eriophylla Section Dun., Hist. Sol. 127, 189. 1813. Grad. ambig. G. Don, Gen. Syst. 4: 426. 1838. Lectotype species: S. jamaicense Mill. Erythracanthum Subseries Bitt., Fedde Rep. Beih. 16: 105, 122. 1923. Type species: S. erythracanthum Boj. Etuberosa Series Juz. ex Buk. & Kameraz, Princ. Potato Breed. 18. 1959. Type species: S. etuberosttm Lindl. Euincana Subseries Bitt., Fedde Rep. Beih. 16: 206, 270. 1923. Type species: S. incanum L. Euleptostemonum Subsection Dun. in DC, Prodr. 13(1): 29, 183. 1852. Lectotype species: S. mammosum L. Euhjcianthes Grad. ambig. Dun. in DC, Prodr. 13( 1): 29, 161. 1852. Lectotype species: S. lycioides L. (rz: Lycianthes lycioides (L.) Hassl.). Eumelongena Grad. ambig. Dun. in DC, Prodr. 13(1); 31, 355. 1852. Lectotype species: S. esculentiim Dun. (= S. melongena L.). Eusolaniim Subgenus Bitt., Bot. Jahrb. 55: 62. 1919. Lectotype species: S. nigrum L. [Hitchcock & Green, 1929]. Eutorvum Series Bitt., Bot. Jahrb. 57: 251. 1922. Type species: S. iorvum Sw. F.xtensum Section D'Arcy, sect. nov. Arbores v. frutices scandentes, incrmes, folia tirones interdum adsunt; pilis stellatis porrectis gerentibus; calyce interdum accrescenti, foliaceo; floribus parvis; antheris aequalibus, crassis, poris terminalibus magnis aperientibus; acino globoso glabro v. pubescenti. Type species: S. cxtensum Bitt. Geminata Subsection G. Don, Gen. Syst. 4: 418. 1838. Section Walp., Rep. Bot. Syst. 3: 58. 1844. 1972] D'ARCY— SOLANACEAE STUDIES II £69 Lectotype species: S. nudum H. & B, ex Dun. Giganteiformia Series Bitt, Bot. Jahrb. 57: 255. 1922. Type species: S. giganteiim Jacq. Glahrescentia Series Buk., Prob. Bot. 2: 318. 1955, Nomen invalidum [no Latin description]. Series Buk. ex Buk. in Buk. & Kameraz, Princ. Potato Breed. 19. 1959. Type species: Not selected. Gonatotrichum Section Bitt., Fedde Rep. 11: 230. 1912. Type species: S. gonafotrichurn Bitt. Gonianthes Grad. ambig. Dun. in DC, Prodr. 13(1): 29, 163. 1852. Lectotype species: S. stellatum Jacq. (= Lycianthes stellata (Jacq.) Bitt.). Graciliflora Grad. ambig. Dun. m DC., Prodr. 13(1): 29, 183. 1852. Section Seithe, Bot. Jahrb. 81: 302. 1962. Type species: S. graciliflorum Dun. Grandiflorae Grad. ambig. ( Psubsection ) C. H. Wright, Fl. Trop. Africa 4(2: 2): 210. 1906. Lectotype species: S. melongena L. Herposolamim Section Bitt., Fedde Rep. 11: 250. 1912. Type species: S. reptans Bunb. Herpystichum Section Bitt., Fedde Rep. 17: 331. 1920. Lectotype species: S. trifolium Dun. [Seithe, 1962]. Heteracantha Grad. ambig. Dun. in DC., Prodr. 13(1): 30, 197. 1852. Lectotype species: S. jamaicense Mill. Holochlaina Subsection G. Don, Gen. Syst 4: 422. 1838. Section Walp., Rep. Bot. Syst. 3; 62. 1844. Lectotype species: S. bigeminatum Nees (= Lyciunthes bigeminafa (Nees) Bitt.). Holophylla Subsection G. Don, Gen. Syst. 4: 414. 1838, Section Walp., Rep. Bot. Syst. 3: 51. 1844. Lectotype species: S. pulverulentum Pers. Hijperbasarthrum Subsection Bitt., Fedde Rep. 11: 359. 1912. Lectotype species: S. tuberosum L. Incaniformia Series Bitt., Fedde Rep. Beih. 16: 201. 1923. Lectotype species: S. incanum L. Induhitaria Grad. ambig. Dun. in DC, Prodr. 13(1): 29, 123. 1852. Subsection Seithe, Bot. Jabrb. 81: 289. 1962. Lectotype species: S. brachystachys Dun. [Seithe, 1962]. Inevniiu Grad. ambig. [Psubgenus] L., Sp. Pi. 184. 1753. Nomen nudum. Grad. ambig. ['Division'] Walp., Rep. Bot. Syst. 3: 1844. Lectotype species: S. nigrum L. Inermes Grad. ambig. (Psubgenus) C H. Wright, Fl. Trop. Africa 4(2: 2): 208. 1906. Lectotype species: S. nigrum L. Inermis Section G. Don, Gen. Syst. 4: 400. 1838. Lectotype species: S. nigrum L. Ingaefolia Series Ochoa in Corr., Potato & Wild Rel. 129. 1962. Type species: S. ingaefoJium Ochoa. Irenosolanurn Section Seithe, Bot. Jahrb. 81: 301. 1962. Type species: S. tvoahense Dun. Ischyracanfhum Section Bitt., Fedde Rep. Beih. 16: 142. 1923. Lectotype species: S, ogadense Bitt. [Seithe, 1962]. Jasminosolanum Section Bitt., Fedde Rep. 17: 330. 1920. Nomen nudum. Section Seithe, Bot. Jahrb. 81: 291. 1962. Type species: S. jasminoides Paxt. Juciri Subsection Marzell m Hegi, Fl. Mitt. Eur. 2584. 1927. Type si>ecies: S. wendlandii Hook. f. Jughndifolia Grad, ambig. Rydb., Bull. Torrey Bot. Club 51: 173. 1924. Nomen nudum. Series Corr., Potato & Wild Rel. 102. 1962. Nomen invalidum [no Latin description]. Series D'Arcy, ser. nov. Frutices v. labruscae, stolones tiiberaque carentes, caulibus lig- nosis, floribus aureis. Type species: S. juglandifolia Bitt. Juripeha Grad. ambig. Dun. in DC, Prodr. 13( 1 ) : 30, 197. 1852. Type species: S. juripeba Rich. 270 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Kieseritzkianc Grad. ambig. Pojark., Fl. URSS. 22: 10. 1955. Nomen invalidum [no Latin de- scription]. Type species; S. kieseritzkii C. A. Mey. Laciniata Series Herasim., Nov. Syst. Pi. Vase. 7: 273. 1970. Type species: S. laciniattmi Ait. Lasiocarpa Grad. ambig. Dun. in DC., Prodr. 13( 1 ) : 30, 252. 1852. Section D'Arcy, stat. nov. Type species: S. lasiocarpum Dun. Lathy rocarpiwi Subsection G. Don, Gen. Syst. 4: 436. 1838. Section Wolp., Rep. Bot. Syst. 3: 88. 1844. Lectot>T)e species: S. carolinense L. Leiodendra Section Dun., Sol. Syn. 20. 1816. Grad. ambig. Dun. in DC., Prodr. 13( 1 ) : 29, 137. 1852. Lectotype species: S. nudum H. & B. ^x Dun. Lemtirisolanum Section Bitt., Bot. Jahrb. 54: 422, 436. 1917. Lectotype species: S. madagascanense Dun. [Seithe, 1962]. Lepidota Grad. ambig. Dun. in DC, Prodr. 13(1): 29, 131. 1852. Series [*Reihe'] Bitt., Fedde Rep. 16: 404. 1920. Section Seitlie, Bot. Jahrb. 81: 298. 1962. Type species: S. lepidotmn H. & B. ex Dun [S. sxvartzianum R. & S. design, err. Seithe]. Leprophora Section Dun., Hist. Sol. 125, 181. 1813. Grad. ambig. G. Don, Gen. Syst. 4: 423. 1838. Lectotype species: S. elaeagnifolium Cav. Leptostemonum Section Dun. in DC, Prodr. 13(1): 29, 183. 1852. Svibgenus Bitt, Bot. Jahrb. 55: 68. 1919; Fedde Rep. 16: 395, 405. 1920. Subgenus Marzell in Hegi, Fl Mitt. Eur. 2584. 1927. Lectotype species: S. mammosum L. Lohanthes Grad. ambig. Dun. in DC, Prodr. 13( 1 ) : 29, 174. 1852. Lectotype species: S. bigeminatum Nees (= Lycianthes bigeminata (Nees) Bitt.). LongipediceUata Series Buk. in Buk. & Kameraz, Princ. Potato Breed. 27. 1959. Type species; S. longipedicellatum Bitt. Lycianthes Subsection Dun. m DC, Prodr. 13(1): 29, 156. 1852. Section Wettst. in Engl. & Prantl, Nat. Pflanzenf . 4 ( 3b ) : 22. 1895. Subgenus Bitt, Bot Jahrb. 54: 424. 1917. Genus Hassl., Ann. Conser\^e. Jard. Geneve 20: 173. 1917. Lectotype species: S. lycioides L. [D'Arcy, 1972] (= Lycianthes lycioides (L.) Hassl.). Lycioides Section Walp., Rep. Bot. Syst. 3: 61. 1844. Type species: S. hjcioides L. ( = Lycianthes hjcioides ( L. ) Hassl. ) Ltjciosolanum Subgenus Bitt., Bot. Jahrb. 54: 421, 428. 1917. Type species: S. aggregatum Jacq. (^ S. guineense L.). Ly coper sicarpon Grad. ambig. Dun. in DC, Prodr. 13(1): 31, 350. 1852. Lectotype species: S. aeihiopicum L. Lycopersicon Genus Mill., Gard. Diet. ed. 4 abr. 1754. Section Wettst. in Engl. & Prantl, Nat. Pflanzenf. 4( 3b ) : 24. 1895. Section Bitt., Bot. Jahrb. 54: 500. 1917. Subgenus Seithe, Bot. Jahrb. 81: 204. 1962. Type species; L. esctdentuin Mill. (= S. lycopersiciim L.) Lysiphellos Subsection Bitt., Fedde Rep. 16: 90. 1919. Section Seithe, Bot Jahrb. 81: 288. 1962. Type species: S. decorticans Sendt. in Mart Macracanthum Series Bitt, Fedde Rep. Beih. 16: 99. 1923. Type species: S. macracanthum A. Rich. Macrocarpon Grad. ambig. Dun. in DC, Prodr. 13(1): 31, 353. 1852. Series Bitt., Fedde Rep. Beih. 16; 186. 1923. Type species: S. macrocarpon L. f. Macronesiotes Section Bitt., Bot. Jahrb. 54: 422, 432, 1917. Lectotype species: S. imamense Dun. [Seithe, 1962]. Macrophyllae Grad. ambig. (Psection) C. H. Wright, Fl. Trop. Africa 4(2; 2); 210. 1906. Lectotype species: S. aculeatissiimun Jacq. (^ S. capsicoides AIL). 1972] D'ARCY^SOLANACEAE STUDIES II £71 Madagascariensia Subseries Bitt., Fedde Rep. Beih. 16: 105, 124. 1923. Lectotype species: S. nossibeense Vatke [PoUiill], Maglia Series [*Roihe ] Bitt., Fedde Rep. 11: 360. 1912. Type species: S. maglia Schlecht. Maurella Section Dun., Hist. Sol. 119, 151. 1813. Lectotype species: S. nigrum L, Megistacroloba Series Card. & Hawkes, Jour. Linn. Soc. Bot, 53: 93. 1945. Type species: S. 7negistacrolobum Bitt. Lectotyp ambig Melongena Genus Mill, Gard. Diet. ed. 4 abr. 1754. : 29, 156. 1852. Ltjcianthcslycioides (L.) Hassl. ) Section Dun., Hist. Sol. 130, 208. 1813. Subsection G. Don, Gen. Syst. 4: 432. 1838. 350 Subseries Bitt., Fedde Rep. Beih. 16: 206, 292. 1923. Lectotype species: M. ovata Mill. (?=^ S. melongena L.). Micracaniha Section Dun., Hist. Sol. 128, 193. 1813. Grad. ainbig. G. Don, Gen. Syst. 4: 426. 1838. Grad. ambig. Dun. in DC., Prodr. 13( 1): 30, 216. 1852. Type species: S. micracanthos Lam. Micranthes Subsection Dun. in DC, Prodr. 13( 1 ) : 28, 95. 1852. [Section Marzell in Hegi, Fl. Mitt. Eur. 2584. 1927. err. = sect. Macracantha Dun.]. Type species: S. micranthum Willd. ex R. & S. in L. Microphyllae Grad. ambig. (Psection) C. H. Wright, Fl. Trop. Africa 4(2: 2): 209. 1906. Lectotype species: S. macracanthum A, Rich. Minutifoliola Series Corr., Potato & Wild Rel. 216. 1962. Type species: S. minutlfoliolum Corr. Mogenoplum Grad. ambig. Dun. in DC, Prodr. 13(1); 30, 335. 1852. Lectotype species: S. lanceolatum Cav, Monadelphoidea Series Bitt., Fedde Rep. 16: 87. 1920. Lectotype species: S. monodelphum Heurck & Muell.-Arg. Monodolichopus Section Bitt., Fedde Rep. 15: 94 nota. 1917. Nomen nudum. Section Bitt., Fedde Rep. Beih. 16: 297. 1923. Lectotype species: S. dubium Fresen non Dun. (^ S. thruppii C H. Wright) [Seithe, 1962]. Morella Section Dumort., Fl. Belg. 39. 1827. [Spelling changed from Maurella Dun.] Grad. ambig. G. Don, Gen. Syst. 4: 411. 1838. Dun Lectotyp Jahrb Morelliformia Series Hawlces, Scott. Pi. Breed. Sta. Ann. Rep. 156. 1956. Type species: S. morelliformia Bitt. & Muench. Muricata Series Corr., USDA Agr. Monogr. 11: 49. 1952. Type species: S. muricatum Ait. Nakuremia Series Bitt., Bot. Jahrb. 54: 447. 1917. Neolycopersicon Section Corr., Potato & Wild Rel. 39. 1962. Type species: S. pennellii Corr. Nigra Grad. ambig. ['Cycle'] Pojark., Fl. URSS 22: 25. 1955. Nomen invalidum [no Latin description] . Type species: S. nigrum L. Normania Genus Lowe, Man. Fl. Madeira 2: 70. 1868. Section Bitt., Fedde Rep. 11: 251. 1912. Type species: N. triphylla Lowe (=:: S. trisectum Dun., fide Bitt., loc. cit.). Nijcteriwn Genus Vent., Jard. Malm. t. 85. 1803. Section Dun., Sol. S>ti. 35. 1816. Subsection G. Don, Gen. Syst. 4: 439. 1838. Grad. ambig. Dun. in DC, Prodr. 13( 1 ) : 30, 331. 1852. Section Bitt., Fedde Rep. 16: 307. 1923. Type species: N. cordifolhim Vent. ( = S. vespertillio Ait.) 272 ANNALS [Vol. 59 Oliganthes Grad. ambig. Dun. m DC, Prodr. 13(1) : 30, 282. 1852. Section Bitt., Fedde Rep. Beih. 16: 1. 1923. Lectotype species: S, indicum L. [Seithe, 1962]. Oppositifolia Grad. ambig. Dim. in DC, Prodr. 13(1): 29, 123. 1852. Section Seithe, Bot Jahrb. 81: 288. 1962. Subsection Seithe, Bot. Jahrb. 81 : 289. 1962. Lectotype species: S. nudum H. & B. ex Dun. [Seithe, 1962]. Oxycarpa Grad. ambig. Rydb., Bull. Torrey Bot. Club 51: 172. 1924. Nomen nndum. Type species: S. oxtjcarpum Schiede. Pachtjphyllu Section Dun., Hist. Sol. 122, 168. 1813. Lectotype species: S. betaceum Cav. (= Cyphomandra betacea (Cav. ) Sendt. ). Pachystemonum Section Dun. in DC., Prodr, 13( 1 ) : 28, 31. 1852. Lectotype species: S. nigrimi L. Farviflorae Grad. ambig. (Psubsection) C. H. Wright, Fl. Trop. Africa 4(2: 2): 210. 1906. Lectotype species: S. aculeatissimum Jacq. (^^ S. capsicoides All.) Parvifolia Subseries Bitt., Fedde Rep. Beih. 16: 106,129. 1923. Lectotype species: S. hastifolium Hochst. ex Dun. [Polhill]. Persicaefolia Grad. ambig. Dun. in DC., Prodr. 13(10): 30, 183. 1852. Type species: S. persicaeifolium Dun. Persicariae Section Dun., Hist. Sol. 126, 183. 1813. Type species: S. persicaeifolium Dun. Petota Section Dumort., Fl. Belg. 39. 1827. Type species: S. tuberosum L. Pinnatisecta Grad. ambig. Rydb., Bull. Torrey Bot. Club 51: 146, 167. 1924, Nomen nudum. Series Corr., Potato & Wild Rel. 268. 1962. Nomen invalidum [no Latin description]. Type si)ecies: S. pinnatisectum Dun. Piurana Series Hawkes, Ann. Mag. Nat. Hist. ser. 12. 7: 693. 1954. Type species: S. piurae Bitt. Polyadenia Series Buk. in Buk. & Kameraz, Princ. Potato Breed. 26. 1959. Type species: S. polyadenium Greenm. Polybotryon Grad. ambig. G. Don, Gen. Syst. 4: 425. 1838. Section Bitt., Fedde Rep. 11: 469, 564. 1912, 1913. Lectotype species: S. mite R. & P. Polygama Grad. ambig. G. Don, Gen. Syst 4: 425. 1838. Grad. ambig. Dun. in DC, Prodr. 13(1): 30, 196. 1852. Type species: S. polygamum Vahl. Polynieris Section Dun., Hist. Sol. 123, 174. 1813. Grad. ambig. G. Don, Gen. Syst. 4: 420. 1838. Lectotyi)e species: S. stellatum Jacq. (= Lycianthes stellata (Jacq.) Bitt.). Potatoe Subsection G. Don, Gen. Syst. 4: 400. 1838. Section Walp., Rep. Bot. Syst. 3: 38. 1844. Grad. ambig. Dun. fn DC, Prodr. 13(1): 28, 31. 1852. Subgenus D'Arcy, stat. nov. Lectotype species: S. tuberosum L. Protocryptocarpum Section Marzell m Hegi, Fl. Mitt. Eur. 2585. 1927. Type species: S. sisymbriifolium Lam. Pseudo-cap sica Genus Moench., Meth. 476. 1794. Grad. ambig. G. Don, Gen Syst. 4: 410. 1838. Section Bitt., Bot. Jahrb. 54: 497. 1917. Type species: P. undulatifolium Moench. ( = S. pseudocapsicum L.). Pseudoflava Series Pojark., Not. Syst. 17: 338. 1955. Type species: S. pseudoflavum Pojark. Pseudolycianthes Grad. ambig. Dun. in DC, Prodr. 13( 1): 29, 156. 1852. Lectotype species: S. aggregatum Jacq. {^ S. guineense L.). Psihcarpa Grad. ambig. Dun. in DC, Prodr. 13 ( 1 ) : 30, 216. 1852. Lectotype species: S. lancaeifolium ]acq. Pterophyllum Grad. ambig. Dun. m DC, Prodr. 13(1): 28, 38. 1852. Lectotype species: S. fraxinifolium Dun. Pteroidea Section Dun., Hist. Sol. 43, 117, 136. 1813. Lectotype species: S. mite R. & P. 1972] D'ARCY— SOLANACEAE STUDIES II £73 Pijracanthum Series Bitt., Fedde Rep. Beih. 16: 139. 1923. Type species: S. pyracanthos Lam. Quadrangulare Section Bitt., Bot Jahrb. 54: 421, 428. 1917. Type species: S. quadrangulare Thunb. Regmandra Grad. ambig. Dun. in DC, Prodr. 13(1): 28, 60. 1852. Section Ugent, stat. nov. Lectotype species: S. montanum L. [Ugent]. Rhodacanthum Series Bitt, Fedde Rep. Beih. 16: 184. 1923. Type species; S. eickii Damm. Rhynchantherum Section Bitt., Fedde Rep. 12: 61. 1913. Type species: S. graveolens Bunb. Septemloba Series Pojark, Fl. URSS 20: 11. 1955. Nornen invalidum [no Latin description]. Type species: S. septemlohum Bunge. Silicisolanum Subsection Bitt., Fedde Rep. 16: 10. 1919, Type species: S. trachytrichum Bitt. Similia Series Herasim, Nov. Syst. PL Vase. 7: 274. 1970. Type species: S. simile F. Muell. Simplicipilwn Section Bitt, Fedde Rep. 18: 309. 1922. Nomen nudum. Section Bitt., Fedde Rep. Beih. 16: 147. 1923. Lectotype species: S. aculeatissimum Jacq. (^ S. capsicoides All.) [Seithe, 1962]. Sodomela Subsection Lowe, Man. Fl, Madeira 2( 1) : 81. 1868. Series Bitt., Fedde Rep. Beih. 16: 158. 1923. Type species: S. sodomeum L. Solanum Genus L., Sp, PL 184. 1753. Subgenus Seithe, Bot. Jahrb. 81: 285. 1962. Section Seithe, Bot. Jahrb. 81: 286. 1962. 'Chorus subgeneruin Seithe, Bot. Jahrb. 81; 285. 1962. Lectotype species: S. nigrum L. [Hitchcock & Green, 1929]. Somalanum Section Bitt., Bot. Jahrb. 54: 50O. 1917. Lectotype species: S. jubae Bitt. [Seithe, 1962]. Spinosa Grad. ambig. [Psubgenus] L., Sp. Pi. 186. 1753. Nomen nudum, StellatipUum Subgenus Seithe, Bot. Jahrb. 81 : 296. 1962. Section Seithe, Bot. Jahrb. 81: 297. 1962. 'Chorus subgcneruin Seithe, Bot. Jahrb. 81: 285. 1962. Type species: S. melongena L. Suaveolentia Grad. ambig. Rydb., Bull. Torrey Bot. Chib 51: 173. 1924. Nomen nudum. Series Corr., Potato & Wild Rel. 89. 1962. Type species: S. suaveolens Kunth & Bouche. Subcontinentalia Subseries Bitt., Fedde Rep. Beih. 16: 104, 107. 1923. Lectotype species: S. zanzibarense Vatke [Polhill]. Subdulcamara Grad. ambig. Dun. m DC., Prodr. 13(1): 28, 84. 1852. Lectotype species: S. ipomoea Sendt. Subifiermia Section Dun., Hist. Sol. 128, 198. 1813. Subsection G. Don, Gen. Syst. 4: 428. 1838. Type species: S. subinerme Jacq. Tarijema Series Corr., Potato & Wild Rel. 233. 1962. Type species; S. iarijense Hawkes. Torva Section Nees, Trans. Linn. Soc. London 17: 51. 1834. Subsection G, Don, Gen. Syst. 4: 430. 1838. Type species: S. torvum Sw, Torvaria Subsection Dun. in DC, Prodr. 13( 1 ) : 30, 258. 1852. Section Bitt., Bot. Jahrb. 57: 248, 250. 1922. Type species: S. torvum Sw. Transaequatorialia Series Buk. in Buk. & Kameraz, Princ. Potato Breed. 21. 1959. Type species: Not selected. Transcaucasica Series Pojark., Not. Syst. 17: 332. 1955. Type sipecies: S. iranscaiicasicum Pojark. Trifidu Series Corr., Texas Res. Found. Con tr. 1: 12. 1950. Type species: S, trifidum Corr. i 274 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Tuberarium Subsection Dun. m DC, Prodr. 13( 1 ); 28, 31, 1852. Section Bitt., Fedde Rep. 11: 349, 1912. Type species: S. tuberosum L, Tuberosa Grad. ambig. Rydb., Bull, Torrey Bot. Club 51: 146, 147. 1924. Nomen nudum. Series Rydb. ex Buk. & Kameraz, Princ. Potato Breed. 18. 1959. Type species: S. tuberosum L. Vaviloviaiia Series Buk. in Buk. & Kameraz, Princ. Potato Breed, 18. 1959. Type species: S. vavilovii Juz. & Buk. Yungasensa Series Corn, Potato & Wild Rel. 220. 1962. Type species: S. yungasense Hawkes. A Provisional Conspectus Genus Solarium L. Lectotype species: S. nigrum L. Chorus subgenerum Solanum (L.) Seithe Type species: S. nigrum L. Subgenus Solanum Subgenus Eusolanum Bitt. Lectotype species: S. nigrum L. Inermes C. H. Wright Lectotype species: S. nigrum L. Section Solanum Sect. Inermis G. Don Lectotype species: S. nigrum L. Sect. Maurella Dun, Lectotype species: S. nigrum L. Sect. Morella Dumort. Lectotype species: S. nigrum L. Sect. Pachystemonum Dun. Lectotype species: S. nigrum L. Sect. Campanulisolanum Bitt. Lectotype species; S. fiebrigii Bitt. Nigra Pojark. Type species: S. nigrum L. Series Transcaucasica Pojark. Type species: S. transcoucasicum Pojark. Alata Pojark. Type species: S. luiea Mill. Series Lutea Pojark. Type species; S. luteum Mill. Series Pseudoflava Pojark. Type species: S. pseudoflavum Pojark. Section Afrosolanum Bitt. TyjDe species: S. terminale Forsk. Series Bifurca Bitt. Type species: S. bifurcum Dun. Series Nakurensia Bitt. Type species: S. ndkurense C. H. Wright Section Benderanum Bitt. Type species; S. benderanum C. H. Wright Section Chamaesarachidium Bitt. Type species: S. chamaesarachidium Bitt. Section Episarcophyllum Bitt. Type species: S. sinuatirecurvum Bitt. Section Gonatotrichum Bitt. Type species: S. gonatotrichum Bitt. Section Leiodendra Dun, Lectotype species: S. nudum Dun. Sect. Anthopleuris (Dun.) Bitt. Lectotype species: S. nudum Dun. Sect. Geminata (G. Don) Walp. Lectotype species: S. nudum Dun = Sect. Oppositifolia (Dun.) Seithe Lectotype species: S. nudum Dun Subsection Micranthes Dun. Type species: S. micranihum R. & S. Subsection Siltcisolanum Bitt. Type species: S. trachytrichum Bitt. Section Lemurisohnum Bitt. Lectotype species: S. madagascariense Dun. Section LysipheUos (Bitt.) Seithe Type species: S. decorticans Sendt. Section Macronesiotes Bitt. Lectotype species: S. imamense Dun. Section Quadrangulare Bitt. Type species; S, quadranguJare L. f. Subgenus Archucsolanum Marzell Type species: S. avicuhre Forst. f. Series Avicularia Herasim. Type species: S. aviculare Forst. f. Series Laciniata Herasim. Type species: S. laciniatum Ait. Series Similia Herasim. Type species: S. simile F. Muell. Subgenus Bassovia (Aubl.) Bitt. Type species: B. sijlvatica Aubl. (= S. sp.) Section Herposolanum Bitt Type species: S, reptansBnnh. Section Herpysiichum Bitt. Lectotype species: S. trifolium Dun. Section P/ero/Jea Dun. Lectot>'pe species; S. ?nff^R. &P. Sect. Polybotrtjon (Dun.) Bitt. Lectotype species: S. mite R, & P. Bassovioides Dun, Lectotype species: S. anceps R. & P. Svihgcnus Brevantherum (Seithe) D'Arcy Section Brevantheru?n Seithe Type species: S. verbascijolhim auct. non L. SuhsecMonAstcrochlaena 'Lov/e Type species: S. auriculatum Ait, Section Extensum D'Arcy Type species: S. extensum Bitt. 1972] D'ARCY— SOLANACEAE STUDIES II 275 jphylla (G. Don) Walp. Lectotype species; S. piilvemlentum Pers. Sect. Anthoresis Bitt. Lectotype species: S. pulvendenhtm Pers. Series Durigibbosa Bitt. Lectotype species: S. cladotrichum Vandas Series Endotricha Bitt. Type species: S. endoirichum Bitt. Series MonadelpJwidea Bitt. Type species: S. monadelphinn Heurck. & Muell.-Arg. Subsection Indubitaria (Dun.) Seithe Lectotype species: S. brachystachys Dun. Section Lepzdofwm Seithe Type species: SAepidoium Dnn. Section Pseudocapsiciim Bitt. Type species: S. pseudocapsicum L. Subsection Cerasocarpwm Lowe Type species: S. pseudocapsicumh. Subgenus Lepiostemonum (Dun.) Bitt. Lectotype species: S. mammosxmi L. ^ Armatae C. H. Wright Lectotype species: S. aculeatisstmum Jacq. (^ S. capsicoides All. ) . Section Acanthophora Dun. Lectotype species: S. mammosum L. Sect. Aculeaia G. Don Lectotype species: S. mammosum L. Lepiostemonum Dun. Lectotype species: S. fnammosum L, Parviflorae C. H. Wright Lectotype species: S. aculeatissimum Jacq. (^ S. capsicoides All. ) . Simplicipihim Bitt. Lectotype species: S. actdeatissimum Jacq. (^ S. cap- sicoides All. ) . Subsection Euleptostemonum Dun. Lectotype species: S, mammosum L. Section Aculeigerum Seithe Type species: S. wendlandii Hook. f. Subsection Juciri Marzell Type species: S. wendlandii Hook, f. Section Androceras (Nutt.) Marzell Type species: A, lohata Nutt. (= S. rostratum Dun.) Willcl Type S. aquartia Dun.) Aquartia (Jacq.) Dun. Type species: A. aculeatiim Jacq. (= S. aquartia Dun.) Polygama G. Don T>pe species: S. polyga?num Vahl Section Cryptocarpum Dun. Lectotype species: S. balbisii Dun. Sect. Profocryptocarpum Marzell Type species: S. sisymbriifolium Lam. Section Eriophyllum Dun. Lectotype species: S. jamaicense Mill. Heteracantha Dun. Lectotype species: S. jamaicense Mill. Section Graciliflorum (Dun.) Seithe Type species; S. graciliflorum Dun. Section Irenosolanum Seithe Type species: S. woahense Dun. Section Ischijracanthum Bitt. Lectotype species: S. ogadense Bitt. Section Lasiocarpum (Dun.) D'Arcy Type species; S. lasiocarptim Dun. Section Lathyrocarpum (G. Don) Walp. Lectotype species: S. carolinense L. Asferotrichoiiim Dun. Lectotype species; S. carolinense L. Section Leprophora Dun. Lectotype species; S. elaeagnifolium Cav. Section Melongena Dun. Lectotype species: M. t?t'^?fl Mill. (= S. melongena L.). ^:z Sect. Andromonoecum Bitt. Lectotype species: S. melongena L. Sect. Stellatipilum Seithe Lectotype species: S. melongena L. Grandiflorae C. H. Wright Lectotype species: S. melongena L. Series Acanthocalyx Bitt. Lectotype species: S. richardii Dun. Series Aculeasirum Bitt. Type species: S. aculeastrum Dun. Series Afrodecumbens Bitt. Type species; S. nigriviolaceum Bitt. Series Incaniformia Bitt. Type species: S. incanum L. Subseries Campylacaniha Bitt. Type species; S. campylacanthum Hochst. Subseries Etiincana Bitt. Type species: S. incanum L. Typ Rhodacanthum Typ Brachylobus Dun. Lectotype species: S. esculentum Dun. (^ S. melongena L.) Eumelongena Dun. Lectotype species: S. esculentum Dun. ( ena Lectotype species; S. lanceolatum Cav. Section Micracantha Dun. Tyi>e species: S. micracanthos Lam, Psilocarpa Dun. Lectotype species: S. lancaeifolium Jacq. Section Monodolichopus Bitt. Type species: S. duhium Fresen Section Nijcterium (Vent.) Dun. Type species: N. cardifolium Vent. (=S. vespertillio Ait). - x_il_i ^ -*-■■ -— '.J-. .x^t^U^iL 276 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Section Oliganthes (Dun.) Bitt. Lectotype species: S. indicum L. ^ MicrophyUae C. H. Wright Lectotype species: S. macracanthum A. Rich. Series Aethiopica Bitt. Type species: S. aethiopicum L. Lycopersicarpon Dun. Lectotype species: S. aethiopicu7n L. Series Afroindica Bitt, Type species: S. indicuiii L. Series Albicaule Bitt. Type species: S. albicaule Dun. Series Ausiroafricana Bitt. Lectotype species: S. tomcniosum L. Series Capensiformia Bitt. Type species: S. capense L. Series Eoafra Bitt. Lectotype species: S. zanzibarense Vatke Subseries Erythracanthum Bitt. Type species: S. eryihracanthum Boj. Subseries Madagascariensia Bitt. Lectotype species: S. nossibeense Vatke Subseries Farvifolia Bitt. Lectotype species: S. hastifolium Dun. Subseries Subcontinentalia Bitt. Lectotype species: S. zanzibarense Vatke Series Macracanthum Bitt. Type species; S. macracanthum A. Rich. Series Pyracanthum Bitt. Type species: S. pyracanthos Lam. Section Persicariae Dun. Type species: S. persicaeifolium Dun. Persicaefolia Dun. Type species: S. persicaeifolium Dun. Section Subinermia Dun. Type species: S. subinerme Jacq. Jwripe&a Dun. Type species: S, /wnpefca Rich, Section Somalanum Bitt. Lectotype species: S. jubae Bitt. Section Torva Nees Type species: S. torvum Sw. Sect. Torvaria (Dun.) Bitt. Type species: S. torvum Sw. Series Anomalum Bitt. Type species: S. anomalum Thonn. Series Eutorvum Bitt. Type species: S. torvum Sw. Series Giganteiformia Bitt. Type species: S. giganteum Jacq. Dunaliana Bitt. Type species; S. dunalianum Gaud. Subgenus hyciosolanum Bitt. Type species: S. aggregatum Jacq. (?=* S. guineense L.). Pseudohjcianthes Dun. Lectotype species: S. aggregatum Jacq. (^ S. guineense L.) Subgenus Potaioe (G. Don) D'Arcy Section Petota Dumort. Section Potatoe (G. Don) Walp. Lectotype species: S. tuberosum L. Sect. Tuberarium (Dun.) Bitt. Type species; S. tuberosum L. Subsection Potaioe G. Don Lectotype species: S. tuberosum L. Subsection Hyperbasarthrum Bitt. Lectotype species: S. tuberosum L. Subsection Tuberarium Dun. Type species: S. tuberosum L. Series Acaulia Buk. & Kameraz Type species: S. acaule Bitt. Series Circaeifolia Hawkes Type species: S. circaeifolium Bitt. Series CZara Grah. & Dion. Type species: S. datum Corr. Series Bulbocastanum Corr. Lectotype species; S. bulbocastanum Dun. Series Conicibaccata Bitt. Lectotype species; S. oxycarpum Schiede Oxycarpa Rydb. Type species: S. oxycarpum Schiede Series Commersoniana Buk. Type species: S. commersonii Lam. s=^ Series Glabrescentia Buk, Type species: Not selected. Series Tarijensa Corr. Type species: S. tarijense Hawkes Series Yungasensa Corr. Type species: S. yungasense Hawkes Series Cuneolata Hawkes Type species: S. infundibuliforme Phil. Series Demissa Buk. Type species: S. demissum Lindl. Series Etuberosa Buk. & Kameraz Type species: S. etuberosum Lindl. Series Ingaefolia Ochoa Type species: S. ingaefolium Bitt. Series Juglandifolia D'Axcy Type species: S. juglandifoliurnBitt Series Longipedicellata Buk. Type species: S . longipedicellatum Bitt Series Borealia Corr. Type species: Not selected. Series Maglia Bitt. Type species: S. maglia Schlecht. Series Andigena Buk. Type species; S. aiidigenum Juz. & Buk. Series Andreana Hawkes Type species: S. andreanum Baker Series Minutifoliola Corr. Type species: S, minutifoliolum Corr. Series Transaequatorialia Buk. Type species: Not selected, Vaviloviana Buk. Type species: S. vavilovii Juz. & Buk. 1972] D'ARCY— SOLANACEAE STUDIES II £77 Series Megisiacroloba Card. & Hawkes Type species:. S. megisfacrohhnm Bitt. ^ Series A///coZa Buk. Type species: S. alticohiBitt. Series M or ell if arm la Hawkes Type species: S. morelliforme Bitt. & Mueneh. Series Pmrana Hawkes Type species: S, piuraeBitt. Series Polyadenia Corr. Type species: S. pohjadenium Greenm. Series Trifida Corr. Type species: S. trifidum Corr. ?== Series Cardiophylla Corr. Type species: S. cardlophyllum Lindl. Series Pinnatisecta Corr. Type species: S. pinnatisectum Dun. Series Tuberosa Buk. & Kameraz Type species: S. tuberosum L. Angustisegmentata Buk. Lectotype species: S. canasense Hawkes Aracciana Buk. Lectotype species: S. aracc-papa Juz. Section Anarrhichoinenum Bitt. Lectotype species: S. sodiroi Bitt. Section Basarthrum (Bitt.) Bitt. Lectotype species: S. suaveolens Knnth & Bouche Series Appendiculata Corr. Type species: S. appendiculaium Dun. Series Articulata Corr. Type species; S. sanctae-marthae Bitt. Series Canensa Corr. Type species; S. canense Rydb. Series Caripensa Corr. Type species: S. caripense Dun. Series Muricata Corr. Type species: S. muricatum Ait. Series Suaveolentia Corr. Type species: S. suaveolens Kunth & Bouche Pterophyllum Dun. Lectotype species: S, froxinifoUum Dun. Section Dulcamara Dumort. Type species: S. dxdcamara L. Kieseritzkiana Pojark. Type species: S. kieseritzkii C. A. Mey. Septemloba Pojark. Type species; S. sepiemlohum Bunge Subdulcamara Dun. Lectotype species: S. ipamoea Sendt. Section Jasminosolanum Seithe Type species: S. jasminoides Paxt. Section Neolycopersicon Corr. Type species: S. pennellii Corr. Section Normania (Lowe) Bitt. Type species: N. triphylla Lowe (=S. trisectum Dun.) Section Regmandra (Dun.) Ugent Lectotype species; S. montanum L. Section Rhy7}chantherum Bitt. Type species: S. graveolens Bunb. Nomina Solano Excludenda Ad Cyphomandra Sendt. Section Cyphomandropsis Bitt. Lectotyx>e sjx^cies: C. stuckeriii (Bitt.) D'Arcy Section Pachyphylla Dun. Lectotype species: Cyphomandra betacea (Cav.) Sendt. Ad Lycianthes ( Dun. ) Hassl. Section Cypellocahjx Bitt. Lectotype species: Lycianthes sp. Subsection Holochlaina G. Don Lectotype species: L, bigetniiuita (Nces) Bitt. Lobanthes Dun. Lectotype species: L. bigeminata (Nees) Bitt. Section Lycioides Walp. Type species: L. lycioides (L.) Hassl. Subgenus Lycianthes (Dun.) Bitt. Type species: L. lycioides (L.) Hassl. Section Lycianthes (Dun.) Wettst. Type species: L. lycioides (L.) Hassl. Subsection Lycianthes Dun, Lectotype species: L. lycioides (L.) Hassl. Eulycianthes Dun, Lectotype species; L. /f/ciaiV7(?^ (L.) Hassl. Meiomeris Dun. Lectotype species: L. hjciaides (L.) Hassl. Section Polymer is Dun. Lectotype species: L. stellata (Jacq.) Bitt. Qonianthes Dun. Lectotype species; L. stellata (Jacq.) Bitt. Ad Lycopersicon Mill. Subgenus Lycopersicon (Mill.) Wettst. Type species: L. esculentum Mill. Section Lycopersicon (Mill.) Bitt. Type species: L. esculentum Mill. Literature Cited Bitter, G. 1911-1923. [Series of many papers appearing in Fedde Rep. 10-19 and in Bot. Jahrb. 54-55. For a full bibliography see C. A. Weber, 1928. Ber, Deutsche Bot. Ges. 46 ( Generalvers. Heft 1 ) : 14&-156.] CoRRELL, D. S. 1962. The Potato and its Wild Relatives. Renner, Texas. D'Arcy, W. G. 1972. Proposal (328) to conserve the generic name Lycianthes (Dun.) Hassler (Solanaceae) vs. Otilix Rafinesque and vs. Parascopolia Baillon. Taxon 21: 211. Dumortier, B. C. 1827. Florula Belgica. Tornay. 278 ANNALS [Vol. 59 DuNAL, M. F. 1813. Histoire Naturelle, Medicinale et Economique des Solanum. Paris, Strasbourg, Montpellier. 1816. Solanorum generumque affinium synopsis, seu solanorum historiae, ed. 2. summarium. Montpellier. 1852. Solanaceae. In A. P, De CandoUe, "Prodromus systematis naturalis regni vegetabilis," 13(1): 1-690. Paris. Hawkes, J, G. 1963. A revision of the tuber-bearing solanums. Scott. PI. Breed. Sta. Rec. 1963: 76-181. & J, P. HjERTiNG. 1969. The Potatoes of Argentina, Brazil, Paraguay, and Uruguay. Oxford. Hitchcock, A. S. & M. L. Green-. 1929. Int. Bot. Congr. Cambridge, 1930, Nom. Prop. 111-199. Lowe, R. T. 1868. A Manual Flora of Madeira. London. PojAHKOVA, A. 1955. Solanaceae. In Flora U.R.S.S. 22: 1-117. Moscow, Leningrad. Seithe, a. 1962. Die Haararten der Gattung Solanum L. und ihre taxonomiscbe Verwertung. Bot. Jahrb. 81: 261-336. Wessely, I. 1961. Die mitteleuropaischen Sippen der Gattung Solanum Sektion Morella. Fedde Rep. 63: 290-321. rw NEW TAXA AND RECOMBINATIONS IN LOPEZIA (ONAGRACEAE) L Uzi Plitmann/ Peter H. Raven,^ and D. E. Breedlove^ the have led to the conclusion that these are best treated as a smgle genus. In the course of our investigations, three new taxa have been discovered and several new combinations have become necessary. These are published here so that they may be available for one or more papers on the group that will appear before our monograph, to be published in the Annals of the Missouri Botanical Garden in 1973. This work has been supported by a scries of grants from the U.S. National Science Foundation to Peter H. Raven. New Taxa Lopezia ciliatula Plitmann, Raven & Breedlove, sp. nov. Herbae annuae hirtellae, pilis rectis vel crispis. Rami 40-80 cm alti, erecti, plerumque ramosi; ramulosi tenues, teretes ad angulares. Folia 1-9 X 0.5-5 cm, alternantia ad raro opposita, ovata vel oblongo-ovata ad lanceolata, membranacea, longipetiolata, basi plerumque obliquo vel inaequale subcordata ad obtusa vel late cuneata, apice acuta ad acuminata, serrata vel serrulata, in quoque latere costae venulis 4-8 praedita, hispidula, pilis paucis strigulosis admixta, ciliata; petioli 0.3-6.5 cm longa, angustissime alati, pubescentes. Inflorescentia terminalis delicata infra foliosa; bracteae 0.3-2.5 X 0.05-0.7 cm, lanceolatae ad lineares vel subulatae, basi subcuneatae, apice acuminatae, subserrulatae vel integrae, petiolatae vel subsessiles; pedicelli 0.8-1.5 cm longi, subascendentes, filiformes atque angulares vel alati, dense et breviter hirsuti, Sepala 0.3-0.45 X 0.07-0.1 cm, plerumque linearo-lanceolata, acuta, subglabra, viridia ad purpurascentia. Fetala inferiora 0.25-0,4 X 0.15-0.2 cm, oblongo-ovata, in unquem brevem producta, subcrenulata, ad basin plus minusve breviter ciliata, rubescentes vel albesccntes; petala superiora 0.25-0.4 X 0.0.5-0.09 cm, anguste linearo-spatulata, ad basin parum angustata demum late obtriangularo-obovata auriculis alatis glandis 2 viridibus, sessilia, apice obtusa, ciliata, rubescentia ad purpurascentia. Stamen 0.3-0.4 cm; filamentum alato-dilatum apice exccpto; anthera 0.1-0.13 X 0.07 cm, plantis locarum umbrosarum pallidior, Staminodium 0.25-0.4 X 0.15-0.2 cm, suborbiculare vel obovato-spatulatum, apice subemarginatum, in unguem longum productum, reubescente. Stylus 0.25-0.35 cm longus; stigma parvum, capitato- obconideum; ovarium globosum ad ellipsoideum, pilosum. Capsula 0.15-0.35 X 0.15-0.35 cm, subglobosa, strigulosa. Semina ca 0.08 X 0.05 mm, oblongo-ovoidea. ^ Department of Botany, Hebrew University of Jerusalem, Israel. ^Missouri Botanical Garden, 2315 Tower Grove Ave., St. Louis, Missouri 63110. ^ Department of Botany, California Academy of Sciences, San Francisco, California 94118 Ann. Missouri Box. Card. 59: 279-281. 1972. 280 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 parum falcata, distincto papilloso-tuberculata, atrobrunnea. Numerus chromoso- maticus gameticus, n = 9, Type: Mexico, sinaloa: Sierra Surutato, Canon de Tarahumares between Arroyo Verde and Rancho Tarahumares, steep north-facing slope, 27 February 1968, D. E. Breedlove 15921 (DS, holotype). Distribution: Known only from a single population on a steep- walled canyon wi elevation ca. 1100 m. Lopezia ciliatula is easily distinguishable from the related species with two glands on each of the upper petals, by its small flowers, sessile and basally auriculate- winged upper petals, ciliation of the petals, pubescence of the petioles, pedicels, capsules, and inflorescence. It may have been derived from L. miniata Lag., w^hich it resembles most closely. Lopezia lacmiata (Rose) Jones subsp. ovata Plitmann, Raven & Breedlove, subsp. nov. Rami flexuosi, decumbentes vel ascendentes, interdum tenues, subglabra. Folia quam in subsp. lacimata plerumque latiora (ad 2.2 cm lata), ovata vel lanceolata, basi subcordata ad subacuto-cuneata, apice acuta, superiora distincte petiolata. Numerus chromosomaticus gameticus, n = 10. Type: Mexico, durango: 20 miles west of La Cicudad, steep moist cliff on edge of barranca, elevation 7,600 feet, 3 October 1966, D. E. Breedlove 15552 (DS, holotype). Distribution: Moist places, usually on wet cliffs and slopes, in rock crevices or near streams; altitude (800-) 1300-2800 m. In Durango and Sinaloa. Lopezia laciniata subsp, hciniata is found in similar habitats in the Sierra Madre Occidental of Jalisco. Lopezia nuevo-leonis Plitmann, Raven & Breedlove, sp. nov. iieimes 25- erecti, ramosi, basi sublignosi (ad 1 cm crassi), angulares, viridescentes ad purpureo-rubescentes. Folia 0.6-4 X 0.2-2,5 cm, oblongo-ovata ad lanceolata, basi rotundata ad late cuneata, apice subacuta vel acuminata, acute serrata, subcrassa, subglabra, quoque lateris costae venulis 2-6, basi opposita alia plerumque altemantia; petioli 0.1-3.8 X 0.05-0.2 cm, sparsissime strigulosi ad glabri, viridescentes vel rubrescentes; stipulae ca. 1 mm longae, acicularo-subu- latae, deciduae. Inflorescentia aperte paniculata interdum racemosa terminalis; 3-1 8-2 breve pubescentes ad lateram unam, subglabri ad lateram alteram, viridescentes ad purpurei. Sepala 0.45-0.7 X ca. 0.1 cm, linearia ad anguste lanceolata vel oblanceolata, acuta, viridescentes ad purpurascentes, glabra. Petala rubrescentes. 5-0.8 X 0.2-0 longum rotundata, interdum subcrenulata ad subretusa. snneriora 0.5-0 1-0 4-0 1972] PLITMANN, RAVEN & BREED LOVE— LOPEZ/ A £81 0.13 cm, caenileo-cinerea. Staminodium 0.35-0.45 X 0.2-0.3 cm, late obovatum, abrupte unguiculatum, apice emarginatum, rubescente atquc infra pallicliore. Stylus 0.3-0.4 cm longus, filiformis; stigma 0.5-1.1 mm diam., capitatum; ovarium glabrum atque infra breviter strigulosum. Capsula 0.3-0.5 X 0.2-0.4 cm, oblonga vel ovoidea-ellipsoidea, subglabra, nonnullam apice aperta. Semina 0.9-1.1 X 0.6-0.8 mm, oblongo-ovoidea, incurvata, crasse rugoso-tuberculata, subatrae. Numerus chromosomaticus gameticus, n = 8. Type: Mexico, nuevo leon: Sierra Madre Oriental, Santa Rosa Canyon, 21 miles west of Linares, cool north-facing talus and rock edges, elevation 2100 feet, 28 October 1964, IL D. Ripley and R. C, Barneby 13569 (DS, holotype; NY, isotype ) . Distribution: Rocky slopes in the Sierra Madre Oriental in Nuevo Leon and San Luis Potosi; rare and local. Lopezia nuevo4eonis differs from L. racemosa in its long, glabrescent stems, two (not one) glands on each upper petal, more or less oblong capsules, and other features. From L. miniata it can be distinguished by its angular reddish stems; acutely serrate leaves (generally glabrous, fleshy and with fewer veins); pinkish- red flowers; shortly and obtusely auricled upper petals; oblong or ovoid-ellipsoid and somewhat longer capsules; coarsely rugose-tubercled, blackish and slightly larger seeds. It likewise differs from both of these species, which have n = 10 (n ~ 20 in some populations of L. racemosa) in chromosome number. New Combinations Lopezia gentryi (Munz) Plitmann, Raven & Breedlove, comb. nov. Lopezia laciniata (Rose) M. E. Jones subsp. gentryi Munz, Brittonia 13: 84. 1961. Lopezia grandiflora Zuccar. subsp. macrophylla (Benth.) Plitmann, Raven & Breedlove, comb, nov. Lopezia macrophylla Benth., PL Hartw. 83, 92. 1841. Lopezia lopezioides (Hook. & Arn.) Plitmann, Raven & Breedlove, comb. nov. Diplandra lopezioides Hook. & Arn., Bot. Beech. Voy. 292, pi 60. 1838, Lopezia riesenbachia Plitmann, Raven & Breedlove, nom. nov. Eiesenbachia racemosa Presl, Rel. Haenk. 2: 36, t. 54. 1831. Lopezia semeiandra Plitmann, Raven & Breedlove, nom. nov. Semeiandra grandiflora Hook. & Am., Bot. Beech. Voy. 291, pi 5D. 1838. THE COMPARATIVE MORPHOLOGY OF THE COCHLOSPERMACEAE. HI. THE FLOWER AND POLLEN' Richard C. Keating^ Abstract The flowers' and pollen of Amoreuxia and Cochlospermum have been studied anatomically to contribute to determining trends of specialization and relationships of the Cochlospermaceae. Perianth vascularization consists of 10 traces in some species of Cochlospermum but is reduced to 6 or 5 in other species. Androecial development is centrifugal with the vascularization consisting of 5 or more trunk bundles. The ovary has a single, 3-5 carpellate, ramified locule in all species examined. Pollen is mostly intectate to semitectate and tricolporoidate. Floral morphology of Bixa is more speciahzed in some ways and less specialized in others and its placement in a separate family from Amoreuxia and Cochlospermum can be Justified, While the Cochlospermaceae show similarities to many parietalian families, the floral anatomy is also compatible witli that found in the Malvales. Previous papers in this series ( Keating, 1968, 1970 ) , discussed the relationships of the Cochlospermaceae using the evidence from vegetative anatomy. Although floral and pollen morphology have demonstrated value in systematics, there is no complete treatment of the flower of even a single species of Cochlospermum. The genus Amoreuxia is not know^n to have been anatomically investigated at all. Schnarf ( 1931 ) investigated the development of the seed of C. orinocense. His comments on the nature of the septa of the gynoecium made no mention of its vascular anatomy. In illustrating the vasculature of the perianth and androecium of C. vitifolium, Wilson (1937) pointed out that more species v^ould have to be investigated before drawing conclusions regarding the basic nature of the vascula- ture in this genus. The numerous stamens and reported parietal placentation have placed the family in the Bixales-Flacourtiaceae alliance (Cronquist, 1968; Takh- tajan, 1969), while vegetative anatomy (Keating, 1968, 1970) indicates malvahan affinities. Erdtman (1952) briefly described the pollen of three species of Cochlospermum and noted their resemblance to pollen of Bixa. Nair (1962) briefly described the Eiosum Floral anatomy and pollen of Rhopalocarpus, previously assigned to the Cochlospermaceae, has been investigated by Huard (1965^, b,c). My present purpose is to examine the floral anatomy and pollen of many species of Cochlospermaceae in order to determine more accurately the relation- ships of the family. ^ I am grateful to the office of Research and Projects, Southern Illinois University, for its support of this work. Valuable assistance was given by the staffs of the herbaria of the Field Museum of Natural History, Harvard University, Missouri Botanical Garden, New York Botan- ical Garden, and the Smithsonian Institution. Also, my thanks go to collectors and curators in many nations for their generous response to my requests for specimens. 2 Department of Biological Sciences, Southern Illinois University, Edwardsville, Illinois 62025. Ann. Missouri Box. Card. 59: 282-296. 1972. 1972] KEATING— THE FLOWER AND POLLEN OF COCHLOSPERMACEAE 283 Table 1. Specimens examined". Name Amoreuxia wrightii A. Gray A. palmatifida Mocino & Sesse Bixa orellana L. C ochlospermum fraseri Planch C. gillivraei Benth. C. gregorii F. Muell. C. orinocense Steud. C. parkeri Planch. C planchonii Hook. f. ex Planch. C. regium (Mart & Shrank) Pilger C. religiostim (L.) Alston ex Trimen C. tetraporum H. Hallier C. tinctorium A. Rich C. vitifolium (Willd.) Spreng C. williamsii Macbride Collector, source von Roztjnski s.n. ( 1932), F Webster 5^ Miller 13137, PUL Wooten s.n. (1919), US Axtell F -03484, SIUE Niks 425, ARIZ Tate 32, NY Keating 360, 414, SIUE Specht s.n. (1948), US Perry s.n. (1949), US Sprechts.n. (1948), US Perry s.n. (1948), US Kuhlmann h- Jimho 387, SP Holt l~ Gehringer s.n. (1930), US Piras s.n. (1947), US King s.n. (1933), US Wimhush s.n. (1962), SIUE Chevalier s.n. (1910), P Ihem s.n. (1963), SIUE Hoehne s.n. (1931), SP Flossier s.n. (1907-8), US Dahlgren s.n. ( 1935 ) , F Suhba Rao s.n. (1963), CAL Boh s.n. (1963), BLAT Yonghoonkird s.n. (1963), BK Pflanz 4034, US Chevalier 94, P Le Testu s.n. (1922), P Stern, Eyde ir Ayensu 1694, US Rowlee ir Mixter s.n. (1921), US Jack s.n. (1930), US Schipp 49, NY Williams (1929), F Material Bud Bud Bud Pollen Bud Bud pollen pollen Bud, pollen Bud Bud Bud Bud, pollen Bud, pollen Bud, Bud, Bud Bud Bud Pollen Bud Bud Pollen Bud Bud, pollen Bud Bud Bud Pollen Bud Bud Bud Pollen Bud ' Specimens cited according to the procedure recommended by Stem and Chambers (1960). * Southern Illinois University, Edwardsville, designated as SIUE. Materials and Methods Flowers and buds of Amoreuxia and C ochlospermum from a geographically diverse sampling of each genus were obtained for study ( Table 1 ) . Specimens of Bixa were available for comparison. Floral anatomy was studied from microtomed serial sections and from cleared thick sections of both dried and FPA preserved material. Thick sections were cleared using the NaOH-Chloral hydrate method of Amott (1939) or the peroxide-lactic acid technique of Sporne (1948), Occa- sionally 5% sodium hypochlorite (straight household bleach) was needed to remove dark deposits from the floral receptacle. Vascular structure of the clear- ings was readily visible without staining and drawings were made with the aid of a 50X dissecting microscope. Serial microtomed sections of most species were prepared using paraffin techniques from preserved material or from dried herbar- ium specimens. Staining was accomplished using Safranin O and Fast Green FCF for the preserved material while Methyl Violet 2B and Bismarck Brown Y proved to be the best combination for the restored material 284 ANNALS [Vol. 59 Pollen was prepared by KOH-acetolysis (Faegri & Iversen, 1964) and mounted in glycerin jelly containing safranin. Sections of some samples were made at 1 /a thickness with an ultramicrotome after embedding in Epon-Araldite. Measure- ments of polar and equatorial axes were based on at least 20 grains per sample from acetolyzed specimens. Descriptions of floral anatomy emphasize vascularization because of its demon- strated value in determining relationships (Moseley, 1967). Descriptions of vascu- lar "events" in the observations section are topographical and are designed for comparisons of the anatomj^ of mature flowers. Such terms as "diverge from the stele** refer to the appearance of vascular bundles viewed in successively higher (more distal) serial sections and yield the most efficient graphic descriptions. It should be understood that developmental interpretations of these terms will pro- duce nonsense. Observations General floral morphology and histology is similar for both Cochlospermum and Amoreuxia. The perianth is composed of five quincuncially imbricate sepals alternating with five rotate petals. Both genera show a tendency toward zygo- morphy, the condition being more pronounced in Amoreuxia. The bright orange- yellow petals in both genera have red glide ways on four of the five petals. These are only slightly visible on fresh flowers of C. vitifolium. Short unbranched tri- chomes range from dense to nearly absent on the receptacle and sepals. Vascu- larization enters the receptacle as a siphonostele and extraxylary fibers are absent at anthesis. In some specimens of Cochlospermum and Amoreuxia, vertical lysigenous canals are present in the ground tissue. Vascularization of the perianth is most discrete in Cochlospermum orinocense, C. gregori% and C. regium. In these species the vasculature of the calyx consists of five vascular bundles which diverge from successively higher serial sections in a % phyllotactic sequence. Each sepal is vascularized by one of these bundles (Fig. 29). At higher levels they trifurcate with the lateral branches fusing to form a ring (Fig, 10). At the base of the sepals, this ring of vasculature is used up in producing a series of parallel veinlets in each sepal. Alternating with the five sepal traces, five petal traces leave the siphonostele above them (Figs. 10, 29). At higher levels, the petal traces also appear to branch to form numerous parallel bundles in the petals. In Amoreuxia, C. vitifolium, and in most other species of Cochlospermum examined, the perianth traces are five to six instead of 10 in number. The traces diverge from the siphonostele in no discernible phyllotactic pattern. In higher sections, where the receptacle is considerably broadened, the traces trifurcate with the lateral branches fusing with neighboring laterals to form a circular vascular plexus (Figs. 2-3, 9, 31). Vascular bundles continue upward from the original traces and vascularize the sepals. Traces originating from the plexus between the original sepal traces vascularize the petals. The androecium.— The stamens have basifixed anthers which are four loculed. Pollen is shed through single introrse apical pores and a pair of lateral basal pores in most species of Cochlospermum (Figs. 4-6). In Amnreiiria the- nnthprs have a to to Figure 1. — Lougisection of flower of Cochlospennum x:ilijoUum^ X 6.5. Cut was made along plane of bilateral symmetry. Note differences in length of anther filaments^ the rccun*cd style, the curved pedimcle. 2 O > a O o c o in h3 W 53 to 00 ax 286 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 pair of obliquely arranged apical slits across the tops of lateral pairs of locules (Fig. 7), Septa between the pairs of locviles become thin and may disintegrate when pollen is shed. The elongate filaments have a single vascular trace which enters the base of the anthers and continues between the locules to the apex. Vascularization of the stamens originates above the origin of the perianth vasculature with 5-6 traces alternating with the previous series of bundles (Figs, 2-3). As many as 8-10 stamen bundles appear in some species due to early bi- or trifurcation, but five gaps in the stele are common even in these cases. These trunk bundles branch at higher levels to form a broken circle of vasculature which de- velops in a centrifugal pattern (Fig. 11). A broad zone of stamen traces is formed which at the surface of the receptacle appears to be a complete circle of 3-6 con- centric ranks of bundles. In Cochlospermum gregorii the most obscure pattern of vascularization is found. From the lowest level where perianth traces diverge from the stele^ there is an indistinct number of vascular traces, bifurcations, and fusions forming a discontinuous plexus which vascularizes all segments of the perianth. The remaining, slightly contracted, siphonostele provides androecial traces at a higher level. There are no distinct trunk bundles; a circular band of vasculature moves out centiifugally as higher sections are examined. Rows of stamen traces mature first nearest the center of the receptacle, the outermost being the most recently formed. The gynoecium. — The pistil of Cochlospennaceae consists of a globose ovary surmounted by an elongate, linear style, recurved at the summit in the plane of bilateral symmetry ( Fig. 1 ) . The ovary is three, four, or five carpellate, and all specimens showed the same anatomy with minor modifications. Transec- tions through the lower or upper parts of the ovary give the appearance of axile placentation while, transections through the middle part show parietal placenta- FiGUKES 2-18. — Floral anatomy of Cochlospennaceae. — 2-3. Cleared transverse thick sections through floral receptacle. — 2. Amoreuxia palmatifida, X 10. Note five perianth traces and six stamen trunk bundles. — 3. Cochlos]iermum vitifolium, X 5. Note six perianth traces and five stamen trunk bundles. — 4-7, Cleared whole anthers, X 6. — 4. Cochlospermum viti- folhim, adaxial view. Note small apical pore. — 5. Cochlospermiun religiosmn, lateral and adaxial vievi^s. Note basal pore in lateral view and elongate apical pore in adaxial view. — 6. Cochlospermum religiosum, adaxial view. Note large apical pore. — 7. Amoreuxia pal- matifida, lateral and adaxial views. Note pair of elongate apical slits. — 8. Cochlospermum religiosum, X 7. Longisection of ovary showing placental ridges. — ^9-10. Diagrammatic com- posite longisections showing perianth vasculature types. — 9, Perianth traces trifurcate with lateral branches fusing to form a plexus from which petal traces and lateral sepal traces arise, e.g. Amoreuxia, Cochlospermum vitifolium. — 10. Perianth vascularized by separate petal and sepal traces, e.g. Cochlospermum orinocense. —11. Cochlospermum religiosum, X9. Stamen trunk bundles and anastomosing branches which produce stamen traces. — 12-18. Cochlo- spermum regium, X 7. Serial transections through the pistil. — 12. View at level of receptacle. Note tiiree dorsal bundles and three ventral bundles. — 13. View of ovary above the receptacle, appears trilocular. — 14, View of ovary cut equidistant between base and upper surface. Note parietal placentation with pairs of ventral bundles in placental ridges. — 15. View above center of ovary. — 16. View at juncture of placentae to form a trilocular configuration. — 17. View at base of style. Note stylar canal at center surrounded by pairs of ventral traces. — 18. View through center of style. Note stylar canaL three dorsal traces and three ventral traces. 1972] KEATING — THE FLOWER AND POLLEN OF COCHLOSPERMACEAE 287 "-.-. ■•• '4* y" "■-.. i ■ * ******** \ I « t * \ --- '■— "-,-.«-''•' ^ • , , I : vv ' hi hI I* *1 i*: 1 i t tm ' #««■ Wj :** t » a ^i -.W'l I \ \ 3 6 10 / • \. ^*' 18 .ui^ mn^. 2gg AXXALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 tion (Figs. 8, 12-18). The ovary therefore has a smgle ramified locule with three, four, or five lobes corresponding to the number of carpels. The style is usually hollow and is lined with a glandular surface. The stylar canal is not usually con- tinuous with the locule of the ovary. The pistil is vascularized by what remains of the stele above the perianth and androecial vascularization. In Cochlospermum gregorii, C. frozen, C, religiosum, and C, vitifolium, the pistil is five carpellate. The stele divides into ten segments with five outer segments alternating with five inner segments. The five inner bundles represent fused lateral pairs of cai-pellary bundles each of which may divide into pairs at higher levels. The five inner bundles are inverted with the phloem on the adaxial side of the xylem (Fig. 30). The outer five bundles represent the dorsal carpellary traces. In other specimens of Cochlospermum orinocense, C. parkeri, C. reghim, C, tetraporum^ C. tinctorium, C. willianmi, and A. ivrightii, the pistil is three-carpel- late. Vascularization is as described above except that there are three dorsal and three pairs of ventral traces. Specimens of C. planchonii examined were four carpellate, and one specimen of C. religiosum had only four sets of ventral bundles but five carpels. The connate carpel walls are vascularized by pairs of bundles which supply traces to the placental ridges and ovules. The dorsal bundles may disappear at lower levels of the ovary wall by dividing into a number of minor parallel ovary wall bundles. w In a specimen of Cochlospermum regium, the ventral bundles are quite distinct in the base of the ovary with the phloem on the adaxial side of the three bundle pairs (Figs. 12-13). The dorsal bundles at that level branch into numerous traces to the ovary wall but are not used up in the process. At the top of the ovary, the three fused pairs of ventral bundles and the three dorsal bundles vascularize the length of the style as six evenly distributed traces (Fig. 18). The ventral bundles appear as free pairs only in the central portion of the ovary (Figs, 14^16). Pollen. — Grains of Amoreuxia and Cochlospermum are quite similar. Shape: mostly spheroidal to subprolate, occasionally suboblate. Structure: mostly semi- Figures 19-31. — Photomicrographs of pollen and floral anatomy.— 19-28. Pollen grains of Bixaceae, Coclilospermaceae, X 1000. — 19-20. Bixa orellana. — 19. Equatorial view. Note reticulate ektexine, sculpturing elements on colpus, pore with constricted lateral extensions. — 20. Equatorial view. Note oval pore. — 21-22. Cochlospermum religiosum, — 21. Equatorial view. No visible pore. — 22. Polar view, note finely reticulate ektexine. — 23-24. Cochlospermum tine- torium. — 23. Equatorial view. Note equatorial constriction in colpus. — 24. Polar view. Note intectate, rugulate sculpturing. — 25. Amoreuxia wrightii, equatorial view. Note constriction in colpus at equator. — 26. Cochlospermum parkeri, equatorial view. No pore is visible in colpus. — 27-28. Cochlospermum gregorii. — 27. Equatorial view. Note constricted colpus. — 28. Polar view. Note reticulate ektexine, sharply defined ends of colpi. — 29-31. Transections through the floral receptacle showing vasculature. — 29. Cochlospermum orinocense, x 10. Note five large sepal traces and five petal traces (arrows). — 30-31. Cochlospermum vitifolium. — 30. Gynoe- cium vasculature below surface of the receptacle, X 45. Note dorsal bundles (arrows) and fused ventral bundles (Vb) vdth xylem on the abaxial side. — 31. Section cut above level of origin of androecial vascularization, X 10. Note perianth traces; note three simple and two branched stamen tiTink bundles (arrows). 1972] KEATING FLOWER 289 ^-t^.-^ ->-r "_^ "J J ' j_ '_ it *'1f 'i 21 22 19 23 ! ■-:^^ '-Y::■W^?v^^ "t^^w-oL^-^: s- J- 26 -?■ 20 27 / y ^ ;.^. f*. ^^ **^; T * 1 --_- o 5^^ ^-^ wm^ ^< f'f;.^-^'4i 290 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 tectate or intectate. Exine thickness 1-1.5 ^. Ektexine equal to or greater than the endexine in thickness. Sculpture: psilate, scabrate, finely reticulate (Fig. 22) to rugulate (Fig. 24). Apertures: zoniaperturate, tricolporoidate, tricolporate, or tricolpate, Colpi are costate. Colpi often appear constricted at the equator or tliey may be divided in half by an equatorial bridge in the ektexine (Fig. 23). Pore zone often indistinct, appearing as a thin area in the endexine. When present, porae are usually lalongate. Delimitation of the porae is usually diffuse. Size: Cochlospermum averages for most species P — 17-23 ft; E — 16-27 ix (C gregorii P = 25 /x; E = 34 /t ) . Amoreuxia P = 33-34 [x; E =:=32-35 /x ( reported as 22 X 19.5 /x by Erdtman, 1952). (Figs. 21-28.) Pollen grains of Bixa examined in this study are easily distinguished from those of Cochlospermum and Amoreuxia. They are tectate, reticulate, tricol- porate with ektexine elements on the colpi membranes (Fig. 19) or around the porae (Fig. 20). The porae are oval, lalongate occasionally with ragged, indeter- minate ends. Bixa pollen is considerably larger with the longest axis extending to over 40 /x. Discussion The flower has been subject to a greater variety of interpretation as to its basic nature than any other structure of the plant. The basic assumption is made here that the flower is a modified shoot, i.e. a short stem with specialized lateral ap- pendages. Lengthy evidence has been presented (Eames, 1931, 1961; Moseley, 1961 ) that the vasculature of the receptacle is homologous with that of the vege- tative shoot. Ontogenetic studies by Tepfer (1953) uphold this view. Perianth parts, stamens, and carpels do not appear to be fvmdamentally different than the leaves of the shoot ( Sporne, 1958; Joshi, 1947) . Perianth. — The most primitive perianth vasculature is seen in Cochlospermum orinocense and C. regium in which five sepal traces alternate with five petal traces in the base of the receptacle. Quincuncial initiation is not usually apparent due to the lack of internodal elongation of the receptacle. Other species of Cochlospermum and all those of Amoreuxia examined show only five to six perianth traces which branch to form a plexus from which the traces are produced for the calyx and corolla. In all specimens examined, the outermost perianth traces trifurcate and form a plexus whether or not the petals may be supplied separately. The fusion of these traces is probably a natural result of the phylo- genetic shortening of the axis to where the phyllotactic spiral disappears com- pletely. Wilson (1937) accurately illustrated the perianth vasculature for Cochlosper- mum vitifolium as containing six perianth traces for its pentamerous perianth. It is difficult to understand the function of the sixth trace. It is possible that the basic number of perianth traces in the family is 10 and reduction to 5 may be a result of contracted space in the receptacle. Bilateral symmetry of the vasculature is well developed and the sixth trace is probably a vestige of separate petal vasculariza- tion. The sixth trace is produced on a side of the receptacle where perianth vascu- lature is least strongly developed. 1972] KEATING — THE FLOWER AND POLLEN OF COCHLOSPERMACEAE £91 Androecium. — In his discussion of the phylogeny of the stamen, Wilson (1937) used several lines of circumstantial evidence to support the hypothesis that stamens are derived from branching systems. He stressed that the Parietales and Malvales provide evidence upholding the telome theory. These two orders contain a number of families which have fascicled stamens or branched stamen trunk bundles in the receptacle, and the Cochlospermaceae were cited as an example. Wilson believed stamen trunk bundles to be relicts of highly branched clusters of sporangium-bearing branches or fertile telomes. In Cochlospermum vitifolium five gaps in the stele produce three single trunks, one double trunk, and one triple trunk (Fig. 31). The precocious branching in the latter two cases cannot be considered dichotomous branching, a condition cited as the anatomical basis for primitive telome formation. Eames (1961) agreed that the existence of trunk vascular supplies would seem to support the view that the fascicle is a unit organ but that actually the formation of fascicles is pi'obably an early step in androecium reduction. Basing his views on comparative anatomy of Dilleniaceae and Paeoniaceae, Eames held that the fascicle is an aggregation of simple stamens. Neither concept adequately explains the nature of stamen trunk bundles nor do they consider developmental data. In many families, floral vasculature reflects a history of separateness for many presently fused parts. Wilson s concept would require regarding malvalian genera with visible fascicles as primitive, those with numerous stamens and vascular trunk bundles in the receptacle (Cochlosperm- aceae, Bixaceae, Cistaceae) as intermediate, and laminar stamens with separate traces ( Magnoliaceae, Degeneriaceae ) as most specialized. Clearly, this concept does not agree with other lines of comparative anatomical evidence ( Bailey, 1944, 1951; Bailey & Swamy, 1951; Canright, 1952), Wilson (1965) concluded that reduction in the number of stamens per trvmk bundle and reduction in the number of trunk bundles has been the case in Hibbertia (Dilleniaceae). While this is probably the case for that genus, there is no reason to regard this as a unidirec- tional sequence. Selection pressure on the pollination system could prolong the activity of the androecial meristem. The genus B/.ta, many ways more specialized than Cochlospermaceae, including its anther morphology, has nearly twice as many stamens as any species of Amoreuxia or Cochlospermum. There is as great a difficulty in explaining the stamens as aggregates as there is with telomes. To consider the stamens as aggregates, it must be assumed that vascular fusion preceded the aggregation of anthers. Presently accepted tenets of floral evolution agree that fusion of parts is a specialized condition, and such specialization is almost universally reported to precede vascular fusion. The sta- men trunks are related to single gaps and show no evidence of any major reorga- nization at the stele of the receptacle. The genera showing stamen trunks (c/. Corner, 1946; Sporne, 1958) all have a shortened receptacle with a laterally expanded torus, a condition which removes the space where spiral phyllotaxis might otherwise occur. When shortening of the floral axis occurs in strobiloid flowers, crowding of the vasculature produces a complicated plexus of anasto- mosing traces as in Ntjmphaea (Moscley, 1961) and Nuphar (Moseley, 1965). 292 ANNALS [Vol. 59 Vascular fusion in these genera is quite different from those in which stamen trunk bundles are found. Difficulties presented by the telome and aggregation theories are eliminated by the centrifugal stamen concept as revived by Corner (1946) from original ob- servations by Payer (1857). Corner accumulated developmental evidence show- ing that the broad disc from which the stamens arise is the last portion of the floral apex to develop. Using the genera Wormia^ Tetracera, and Bixa, Corner showed tliat the stamen primordia arise radially from the base of the gynoecium separating it from the already well-developed perianth primordia. This disc interposes tissue in a centrifugal direction (basipetal with reference to the gynoecium). A pro- vascular zone, beginning as five traces, branches regularly to produce well spaced rows of traces and stamen primordia. One implication of this develop- mental pattern is that numerous stamens may be specialized over few, a trend opposite to that in strobiloid flowers. Centrifugal development probably arose early in angiosperm evolution, but it is not known if it developed out of or con- comitant with the strobiloid androecium. No indication of the origin of centrifugal stamens has appeared in genera thus far shown to have the condition. However, an intensive investigation from a developmental standpoint might prove to be highly informative. Gynoecium. — The ovary structure of Cochlospermaceae has been subject to misinterpretations of important systematic consequence which could have been avoided by serial sectioning. Eichler (1871) described the ovary of Cochlo- spermum as having three to five placentae with a septum dividing the ovary from base to apex into three to five locules. Pulle ( 1906) described a tiiree-chambered ovary in C. iventii. On the other hand, Pilger (1925) and Hutchinson (1967) separated Cochlospermum and Amoreuxia by the presence of a single chambered ovary in the former in contrast to a multichambered one in the latter. All material I examined had a single ramified locule in the ovary. Transections cut low or high appear axile while those at the center appear parietal. Both genera are identical in this respect. Description of the gynoecium of Cochlospermum orinocense by Schnarf (1931) was based on serial sections, but he made no mention of the position of vascular bundles. In speculating on the nature of the gynoecium, he felt that his material provided no evidence as to whether the septa were laminal ovitgrowths of the ovary wall or the united walls of carpels. He concluded that we can prob- ably assume they are not united carpel walls. Hutchinson (1959) illustrated the gynoecium of C. tinctorium as having marginally joined carpels with a single circular locule. No specimen studied here approximated his drawing. If we assume that the presence of pairs of vascular bundles running up the lamina of the ovary are ventral carpel bundles, then certainly the laminae are fused carpel walls and not secondary proliferations or septae. Several authors have discussed the evolutionary status of axile versus parietal placentation (Parkin, 1955; Purl, 1952; Eames, 1981). Arguments distinguishing between primitive parietal and advanced parietal placentation seem unconvincing. All such discussions seem to assume that carpels were completely closed or com- 1972] KEATING— THE FLOWER AND POLLEN OF COCHLOSPERMACEAE £93 pletely open when syncaipy took place. Gundcrsen (1939) has pointed out that Hypericum, Theohroma, Shortia, Feijoa, and Myrtus have ovaries which have axile placentation in the base of the ovary and parietal placentation in the upper part. There is no basis to assume that those ovaries were originally all closed or all open before syncarpy took place. Gundersen noted that wherever there is a change in placentation in ontogeny, it is always from parietal to axile. His opinion on the primitiveness of parietal placentation is thus based on recapitulationary arguments. There is no evidence for determining whether the partially parietal placentation in the Cochlospermaceae is primitively axile or primitively parietal rpels nature parietal condition actually exists. The condition is ht^ld by him to be derived from axile placentation in a number of families. In descriptive work, the terms axile and parietal should be used topographically without trends of specialization being necessarily implied. Level of specialization in the flower. — In both genera, primitive features in- r elude hypogenous structure, separate perianth parts, separate stamens, numerous ovules. Of intermediate specialization can be included open racemose inflores- cence, centrifugal stamen matviration, numerous stamens, expanded receptacle, incipient bilateral symmetry, poricidal dehiscence of the anther, and the syn- carpous gynoecium. The flowers of the family can be best characterized as between primitive and moderately specialized. Relationships among the Parietales and Malvales. — Bixa is undoubtedly related to the Cochlospermaceae but is more specialized in many ways including the more numerous stamens, folded anthers with apical (morphologically lateral) dehiscence (Venkatsh, 1956), and reduction of the gynoecium to two carpels. Bixa is more primitive in having no detectable bilateral symmetry and in having separate calyx and corolla vasculature. Trichomcs on the ovary wall are vascularized by tlie dorsal wall vasculature. Bixa has been separated biologically from Amoreuxia and C ochlospermum for a very long time, and its placement in a separate family seems justified. The evidence from vegetative anatomy supports this view. Flowers of Rhopalocarpus are readily distinguishable from the above genera by the four-merous perianth, fewer and larger ovules, an androgynophore, as well as numerous gross differences in vegetative anatomy. Its inclusion with Dialyceras in the Rhopalocarpaceae by Capuron ( 1962) is upheld by the studies of Huard ( 1965a, b,c). The Flacourtiaceae, thought to be the basal stock from which the Bixales (Cistales) series arose, are poorly known at present. Evolutionary tendencies in this family have produced perigyny, epigyny, loss of petals, unisexual flowers, reduction in stamen number, and reduction in carpel number (Cronquist, 1968). The genera of flacourts with which the Cochlospermaceae show closest affinities is yet to be identified. Many features found in the flowers of the Cochlospermaceae seem to be wide- spread in both parietalian and malvalian families. Trifurcating sepal traces with separate petal bundles are common in many families including the Cistaceae 294 ANNALS [Vol. 59 (Saunders, 1936), the Sterculiaccac (Rao, 1949), Tamaricaceae (Murty, 1954), and the Bixaceae. The Rhopalocarpaceae (Huard, 1965a) have separate petal and sepal traces with an indefinite number of parallel traces arising from a single gap for each appendage. The formation of a plexus or ring from which the petal traces arise is also found in the Onagraceae (Baehni & Bonner, 1948) and the Dilleniaceae ( Wilson, 1965 ) . Corner (1946), Sporne (1958), and Wilson (1937) have listed the famihes which have stamen trunk bundles as follows: Actinidiaceae, Aizoaceae, Bixaceae, Bombacaceae, Cactaceae, Capparidaccae, Crossosomataceae, Cochlospermaceae, Dilleniaceae, Flacourtiaceae, Hypericaceae, Lecythidaceae, Malvaceae, Onagra- ceae, Loasaccae, Paeoniaceae, Theaceae, and Tiliaceae. A characteristic of this type of stamen vasculature is almost complete separation of androecial and peri- anth vasculature. This is well illustrated by Sparmannia (Tiliaceae) figured by Sporne (1948). Of the families listed above, malvalian flowers have the most highly advanced centrifugal stamens with their clusters above the receptacle, staminal tubes, or fusions to the style. Cochlospermaceae have not reached that level nor have many other genera associated with malvalian families. The imbricate calyx of the Cochlospermaceae and Bixaceae might seem to preclude inclusion in the Malvales, but the valvate condition is easily derived from the imbricate. Ryania (Flacourti- aceae) shows the transition from an imbricate to a valvate calyx in each flower. Nor does parietal placentation really present a difficulty. In the Tiliaceae, the genera Belotia, Christiania, Colona, Entelia, Goethalsia, Mollia, Nettoa, Spar- mannia, Tilia, and Triumfetta all have parietal placentation appearing in at least some sections of the ovary (Weibel, 1945; Baehni, 1934). Several of the genera have ovary morphology almost identical to that of Amoreuxia and Cochlospermum, A comparison of pollen characters demonstrated that almost all families of the Bixales, Theales, and Malvales include tricolpor(oid)ate, zoniaperturate pollen. Their grains have mostly thin exines, psilate to reticulate sculpturing, range from oblate to prolate, and are intectate to tectate (Erdtman, 1952, 1969). Elaborations occur as in the highly sculptured large grains of many genera in the Malvales. Pollen of the Cochlospermaceae does not closely resemble that of Bixa but is distinguished by smaller size and less distinct porae. Many genera of the Fla- courtiaceae, Aizoaceae, Ochnaceae, Loasaceae, Tiliaceae, Sterculiaceae, and Theaceae have pollen close to that of the Cochlospermaceae, when all characters are considered. None of the other families in the Bixales or Theales have more than a few the rmaceae Their level of Malvales, yet they unquestionably allied to them. We can agree with Meeuse (1965 Flacourtiaceae and Dilleniaceae as basal families of this taxon the relationships of the many famiUcs will be difficult to clarify. Literature Cited Arnott, H. J. 1959. Leaf Clearings. Tiirtox News 37: 192-194. Baehni. C. 1934, Revision du ^cnrc Mollia Mnrt. f-t V.uon r.^nArAUr, t 403-426 1972] KEATING — THE FLOWER AND POLLEN OF COCHLOSPERMACEAE £95 — & C. E. B. Bonner. 1948. La vascularization des fleurs chez les Lopezieae (Ona- gracees). Candollea 11: 305-322. Bailey, I. W. 1944. The development of vessels in angiosperms and its significance in morphological research. Amer. Jour. Bot. 31; 421-428. , 1951. The use and abuse of anatomical data in the study of phylogeny and classifi- cation. Phytomorphology 1: 67-69. — & B. G. L. SwAMY. 1951. Conduplicate carpel of dicotyledons and its initial trends of specialization. Amer. Jour. Bot. 38: 373-379. Canright, J. E. 1952, The comparative morphology and relationships of the Magnoliaceae. L Trends of specialization in the stamens. Amer. Jour. Bot. 39: 484-497. Capuron, R. 1962. Revision des Rhopalocarpacees. Adansonia 2: 228-267. Corner, E. J. H. 1946. Centrifugal stamens. Jour. Arnold Arbor. 27: 423-437. Cronquist, a. 1968. The Evolution and Classification of Flowering Plants. New York. Eames, a. J. 1931. The vascular anatomy of the flower with refutation of the theory of carpel polymorphism. Amer. Jour. Bot. 18: 147-188. . 1961. Morphology of the Angiosperms. New York. EicHLER, A, W. 1871. Bixaceae. In: K. F. P, von Martins, "Flora Brasiliensis." 13(1): 427- 431. Munich. Erdtman, G. 1952. Pollen Morphology and Plant Taxonomy. Waltham, Mass. . 1969. Handbook of Palynology. Copenhagen. Faegri, K. & J. Iversen. 1964. Textbook of Pollen Analysis. New York. GuNDERSEN, A. 1939. Flower buds and phylogeny of dicotyledons. Bull. Torrey Bot. Club 66: 287-295. HuARD, J. 1965a. Anatomic des Rhopalocarpacees. Adansonia 5: 103-123. . 1965&. Remarques sur la position systematique des Rhopalocarpacees d'apres leur anatomic et leur morphologic pollinique. Bull. Soc. Bot. France 112; 252—254. — . 1965c. Palynologia Madagassica et Mascarenica. Fam. 127: Rhopalocarpaceae. Pollen et Spores 7: 303-312. Hutchinson, J. 1959. The Families of Flowering Plants. I. Dicotyledons. Ed. 2. London. , 1967. The genera of Flowering Plants. Dicotyledons. Vol. 2. Oxford. JosHi, A. C. 1947. Floral histogenesis and carpel morphology. Jour. Indian Bot. Soc. 26: 63-74. Keating, R. C. 1968. Comparative morphology of the Cochlospermaceae. I. Synopsis of the family and wood anatomy. Phytomorphology 18: 379-392. -. 1970. Comparative morphology of the Cochlospermaceae. II. Anatomy of the young vegetative shoot. Amer. Jour. Bot. 57: 889-898. Meeuse, a. D. J. 1965. Angiosperms — past and present. Phylogenetic botany and interpre- tative floral morphology of flowering plants. In L. Chandra (editor), "Advancing Fron- tiers of Plant Science 11/' New Delhi. Moseley, M. F., Jr. 1961. Morphological studies of the Nymphaeaceae. II. The flower of Ntjmphaea. Bot. Gaz. ( Crawfordsville ) 122: 233-259. •. 1965. Morphological studies of the Nymphaeaceae. III. The floral anatomy of Ntiphar. PhytomorphQlog>^ 15: 54-84. . 1967. The \'alue of the vascular system in the study of the flower. Phytomorphology 17: 159-164. MuRTY, Y. S. 1954. Studies in the order Parietales. IV. Vascular anatomy of the flower of the Tamaricaceae. Jour. Indian Bot. Soc. 33: 226-238. NAm, P. K. K. 1962. Pollen grains of Indian plants. II. Cochlospermaceae, Moringaceae. Phytolaccaceae, Portulacaceae, Tamaricaceae. Bull. Lucknow Natl. Bot. Card. 60: 1-8. Parkin, J. 1955. A plea for a simpler gynoecium. Phytomorphology 5: 46-57. Payer, J. B. 1857. Traite d'Organogenie Comparee de la Fleur. Paris. PiLGER, R. 1925. Cochlospermaceae. In A. Engler & K. Prantl, "Die natiirlichen Pflanzen- familien." Ed. 2. 21:316-320. Leipzig. PuLLE, A. A. 1906. An Enumeration of the Vascular Plants known from Surinam, together with their Distribution and Synonymy. Leiden. PuRi, V. 1952. Placentation in angiosperms. Bot. Rev. (Lancaster) 18: 608-651. Rao, C. V. 1949. Floral anatomy of some Sterculiaceae with special reference to the position 'of the stamen. Jour. Indian Bot. Soc. 28: 237-245. Saunders, E. R. 1936. The vascular ground plan as a gm'de to floral ground plan: illustrated from Cistaceae. New Phytol. 35: 47-67. 296 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 ScHXARF, K. 1931. Ein Beitrage zur Kenntnis der Samenenhvicklung der Cattung Cochlo- spernmm. Oesterr. Bot. Zeit. 80: 45-50. Sporne, K. R. 1948. A note on a rapid clearing technique of wide application. New Phytol. 47: 290-291. . 1958. Some aspects of floral vascular systems. Proc. Linn. Soc, London 169: 75-84. Stern, W. L. & K. L. Chambers. 1960. The citation of wood specimens and herbarium vouchers in anatomical research. Taxon 9: 7-13. Takhtajan, a. 1969. Flowering Plants — Origin and Dispersal, Washington, D. C. Tepfer, S. S. 1953. Floral anatomy and ontogeny in Aquilegia formosa var, truncata and Ranunculus repens. Univ. California Publ. Bot. 25: 513-576 + 81 plates. Venkatsh, C. S. 1956. The curious anther of Bixa. Its structure and dehiscence. Amer. Midi. Naturalist 55: 473^76. Weibel, R. 1945. La placentation chez les Tiliacees. Candollea 10: 155-177. Wilson, C. L. 1937. The phylogeny of the stamen. Amer. Jour. Bot. 24: 686-699. . 1965. The floral anatomy of the Dilleniaceae. I. Hibhertia Andr. Phytomorphol- ogy 15: 248-274. NEW WORLD JUGLANDACEAE, III. A NEW PERSPECTIVE OF THE TROPICAL MEMBERS WITH WINGED FRUITS^ Donald E. Stone^ Abstract Morphological and anatomical e\idence is presented for reco^ the two American species of Juglandaceae with winged fruits, endemic to Costa Rica, while O. mexicana ranges from Mexico ,„ „.,^™ . nicaragu crisis Molina is reduced to synonymy under O. mexicana subsp. mexicana subsp. costaricenm is described as new. Information is provided range, general ecology, morphology, and taxonomy of the American taxa. Thi 2n n 32 for O. mexicana subsp. mexicana is the first report for the genus. Oreomunnea pterocarpa is Costa Rica. Engelhardia The New World species of luglandaceae with winged fruits form an alliance that is recognized by some as the genus Oreomunnea (Hjelmqvist, 1948; Leroy, 1951, 1955) and by others as a section of the predominantly Old World genus Engelhardia (Candolle, 1914; Nagel, 1914; Standley, 1927a; Manning, 1949, 1959). While four Central American species have been described to date, the paucity of collections, particularly fertile ones, has handicapped critical taxonomic evalua- tions. Oreomunnea pterocarpa Oersted (1856), for example, is represented today by fewer than fifteen flowering and fruiting specimens. And until recently no fruits of Engelhardia mexicana Standley (1927a) had been collected since Rovi- rosa roamed Chiapas, Mexico, in 1891. Engelhardia nicaraguensis Molina (1968) is based on two fruiting specimens, but again no information was made available on the important floral characteristics. The fourth species was described by Standley (1940) as E. guatemalensis on the basis of several sterile collections, but as Manning (1959) has since pointed out some of the specimens have proved to be E. mexicana and Alfaroa costaricensis. The residue of Standley's JB. guatemalensis is by no means well understood. According to Manning (1959) it ". . . might not be distinct from Alfaroa manningii, A. hondurensis, or Engelhardtia pterocarpa^ Fortimately, this puzzle now seems to have been resolved by the discovery of fruiting material at the type locality. Engelhardtia guatemalensis Standley is in reahty an Alfaroa, apparently related most closely to A. hondurensis (Williams & Molina, 1970). ^ Supported by National Science Foundation grants GB-5233X and GB*28525X. The line drawings are the artistry of Susan Carlton Smith and were sponsored by the Duke University Council on Research. I am indebted to several biologists for their information and support in locating trees in the hinterlands of Mexico and Central America: Drs. Arturo Gomez Pompa and Ramon Riba, Mexico; Mr. Antonio Molina R., Nicaragua; and Drs. William H. Hathaway and Lester R. Holdridge, Costa Rica. The initial fieldwork that was conducted in the spring of 1966 was outstandingly successful, due in large part to the support of my research assistant, Mr. Louis F. Conde, and travelling companions, Drs. Marshall R. Crosby and James W. Walker. Appreciation is also'noted for services rendered by the curators of the many herbaria that have played such an important part in this study: A, CR, DUKE, EAP, F, GH, IICA, NO, NY, PH, US. Professor William Louis Culberson has kindly provided the Latin diagnosis, and Drs. R. L. Wilbur, B, G. Schubert, and W. E. Manning have been most helpful in suggesting improvements in the manuscript in form and content. - Department of Botany, Duke University, Durham, North Carolina 27706. Ann. Missouri Box. Gahd. 59: 297-321. 1972. 293 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Since Manning's last report (1959) on the status of the New World Engcl- hardias, several critical collections have become available for study. The purpose of this paper is to reassess the evidence for recognizing the genus Oreomunnea, detail new information on the cytology, morphology and distribution of the Amer- ican species, and discuss the patterns of variation in light of proposed taxonomic changes. r r Generic Considerations Oreomunnea Oersted, Vidensk. Meddel. Dansk Naturhist. Foren. K]0benhavn 3: 33-34. 1856. Engelhardtia sect. Oreomunea (Oersted) DC, Ann. Sci. Nat. Bot. IV. 18: 36. 1862. gelhardt 1864. Oreamunoa Oersted, Vidensk. Meddel. Dansk Naturhist. Foren. Kj0benhavn 1870: 166. 1870. j^ - -■ Trees, or rarely large shrubs; bark scaly or tight; pith solid; buds naked; leaves opposite, evergreen; leaflets 4-12, opposite to subopposite, symmetrical to asym- metrical at base, revolute on one or both margins, with or without auricles, with or without serrations on young trees and stump sprouts, entire on mature trees, adaxial surface glabrous, abaxial surface dotted with peltate scales; monoecious, with both staminate and androgynous panicles, in the latter case female spike flanked by 1-3 decussate pairs of staminate catkins; staminate flowers numerous, . alternately arranged, abaxial 3-lobed bract cupped or long and slender, receptacle round or elongate, inner floral envelope of 2 bracteoles and 2-4 sepals, stamens 8 in one series to 18 or so in two series, pollen tectate with fine scabrate sculpturing, diameter ca. 20/x; pistillate flowers numerous, alternately arranged, subtended by 3-lobed abaxial bract and adaxial bracteolar rim, calyx tube fused to bract-bracte- ole cup at base, forming distinct tube above, deeply parted, with 4 narrow or broad sepal lobes extending to summit of stigma or beyond, style short or long tapering, deep cleft separating two stylar arms, stigmas horseshoe shaped, carinal, capping the ends of the stylar arms; fruit 3-winged, medium to large (5-13 cm), adaxial bracteoles expanded to cover pistil, nut 8-celled at equator; germination hypogeal, first two aerial leaves opposite and simple or compound, succeeding several leaves alternate and simple or compound, shifting to opposite, compound leaves at maturity. w When Oersted described Oreomunnea^ (1856) as a new endemic genus from Costa Rica, he was quite aware of its close kinship to the Old World Engelhardias. However, he stressed particularly basic differences in the stigmas, fruit, and seed. Widi Htde additional information de CandoUe (1862) saw fit to reduce the New World taxon to sectional rank in the genus Engelhardia. Thus started the long- standing debate on the merits of Engelhardia sect. Oreomunnea versus Oreo- munnea. Once the Hues were drawn more opinions were added without shedding much additional light. Oersted ( 1870a, 1870fc ) restated the case for recognizing the genus Oreomunnea and was later given support by Hjelmqvist (1948) and ^ The name has been spelled in various ways since the genus was first described in honor of don Francisco Maria Oreamuno. Manning (1949: 200) discussed the tortured history of the spelhng in a footnote. 1972] STONE— NEW WORLD JUGLANDACEAE 299 Leroy (1951, 1955); however, de Candolle (1864, 1914), Standley (1927a), and Manning (1949, 1959) continued to classify New World Juglandaceae with winged fruits as a section of Engelhardia. As Manning ( 1949) has so aptly pointed out, the decision on what to call the New World group depends on which of many similarities and differences are emphasized. The issue took on new dimensions as a result of Standley's report (1927/;) of another new genus, Alfaroa, from Costa Rica, Though Standley mistook the superficial similarity of the fruit as a basis for close relationship with Jnglam, Manning (1938, 1940, 1948), Iljelmqvist (1948), and Leroy (1955) have since established that Alfaroa has its closest affinities with Oreomunnea, This appears to be a most remarkable assertion, because Alfaroa has a wingless fruit, whereas both Oreomunnea and Engelhardia have a conspicuous 3-winged bract. Manning (1949, 1959) summarized the many similarities of the New World members and presented substantial evidence that the fruit is about the only reliable basis for distinguishing between Alfaroa and Oreomunnea. The question then becomes one of assessing the relationship of Alfaroa and Oreomunnea as well as Oreomunnea and Engelhardia. On the basis "that the only reliable features are the position in flower and the condition in fruit of the bract and bracteoles (prophylls) . . " Manning (1949) concluded that Oreomunnea is best considered a section of Engelhardia and that Alfaroa should be recognized as a weak generic segregate. Since the last reappraisal, however, new data on leaf crystals (Stone, unpublished) and seedlings (Conde & Stone, 1970; Stone, 1970), when combined with previous findings on phyllotaxy, venation patterns, and fruit morphology, provide, in my view, persuasive evidence for conferring generic rank on Alfaroa^ Oreomunnea, and Engelhardia. The essence of these findings is summarized in Table 1 and discussed below. All World family. (1) Phyllotaxy is opposite at maturity (Figs. 2A, 7A; Stone, 1968: Fig. 1), though both genera pass through the seedHng stage with alternate leaves (Figs. 4F, 9G; Stone, 1968). Occasionally, however, the phyllotaxy may vary between Alf World Engelhardia and § Psilocarpeae ( Jacobs, 1960 (2) The intercostal and marginal leaf -venation patterns which permitted Wolfe Juglandaceae arc useful in recognizing two groups. A7/i which is characterized by an intercostal network of closed veinlets. Th group, in which Engelhardia is a member, has an intercostal mesh that network of free-ending veinlets. (3) Leaf crystals also serve to identify World and New World Metcalfe common of the juglandaceous plant. Leaf crystals in particular are common in certain genera and are often quite conspicuous. In hickories [Carija), for example, the in the palisade parenchyma. A survey of leaf crystals crys 300 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Tabi.f, 1. Diagnostic features of I Ufaroa, Oreomnnnea , and Engelhan lia. - Alfaroa Oreomumiea Engelhardia CHARACTiiK § Psilocarpeae alternate § Engelhardia Phyllotaxy opposite opposite alternate Leaflet venation, closed veinlets closed veinlets free-ending free-ending intercostal mesh \^einlets veinlets Leaflet crystals, present present absent absent large rhombic Fniit si/e large medium small small Fruit partitions 8-celled 8-celled 4-celled 2( 4) -celled Position of cotyle- separate separate tangled tangled dons in fi'iiit Testa of cotyledons tight tight loose loose Seedling germination hypogeal hypogeal epigeal epigeal Cotyledonary node 1-gap, 3-trace 1-gap, 3-trace 1-gap, 2-trace 1-gap, 2-trace Pollen diameter (/a) 24 (19-29) 21 (19-23) 15(13^17) 19(15-25) Pollen, polar view sub triangular subtriangular triangular subtriangular Pollen, equatorial suboblate suboblate oblate oblate-spher. view Pollen nexine thick thick thick thin Stigma orientation carina] carina! split-carinal commissural Stigma shape subglobose subglobose subglobose elongate Inflorescence terminal and terminal and terminal and lateral ■ lateral lateral lateral Fruit bract very small 3-winged 3-winged 3-winged Fruit bracteoles very small large small small Fruit wing venation absent triveined pinnate pinnate common in Canja, Juglans, Pterocarya, and Platycarija, but not in Alfaroa, Oreo- munnea, or Engelhardia. Instead, Alfaroa and Oreomunnea have numerous small crystals in the leaf struts formed by the bundle-sheath extensions and large rhom- bic crystals localized primarily in the palisade. Large rhombic crystals have not been detected in Engelhardia. (4) The internal anatomy of the fruit of Alfaroa and Oreomunnea is also distinctive. A median cross section of the fruit exposes eight chambers in these New World genera (Figs. 4E, 9F; Stone, 1968: Fig. 13) but only two or four in Engelhardia (Leroy, 1955; Manning, 1966). The basic partitioning of the fruit into two chambers is effected by the primary septum. Primary and secondary septa intersect to form the 4-chambered fruit, while the 8-chambered condition is the result of the elaboration of tertiary septa which bisect each of the four basic chambers. Leroy (1951, 1955) also discovered that tile vascularization of the fruit is identical in Alfaroa and Oreomunnea, Four placental strands fuse into one bundle in the plane of the primary partition prior to entering the funiculus. In Engelhardia, on the other hand, there are only two placental strands present to unite into a single vascular bundle, (5) The morphol- ogy of junglandaceous seeds and seedlings is a subject of considerable import and has been detailed elsewhere (Conde & Stone, 1970; Stone, 1970). The main findings confirm the separation of Oreomunnea from Engelhardia and its alliance with Alfaroa. Engelhardia has epigeal germination and one-gap, two-trace coty- ledonary nodes. Oreomunnea pterocarpa (Conde & Stone, 1970: Fig, 7) and O. 1972] STONE— NEW WORLD JUGLANDACEAE 301 mexicana (Fig. 9G) are hypogeal and have a one-gap, three-trace cotyledonary node as in Alfaroa (Conde & Stone, 1970: Fig. 20). While I was surprised to discover that the New World taxa with winged fruits have hypogeal germination, Hjelmqvist (1948: 58) shrewdly observed that the testa "attaches itself to all irregularities of the cotyledons, while in Engelhardtia it forms an even covering. Oreomunnea in this respect approaches the type that is found in Juglans and Carya, the character possibly being connected with the fact that the germination is hypogeal as in these; the germination of Oreomunnea is, however, not de- scribed." The similarities of Alfaroa and Oreomunnea as discussed above are subtle but quite substantial. The similarities of Oreomunnea and Engelhardia are more apparent than real. By this I mean that the conspicuous 3-winged fruits that are common to both taxa are, on closer examination, quite distinctive. As the paleon- tologists have long recognized and MacGinitie (1969) recently commented, Oreomunnea fruits have a 4-lobed involucre consisting of 3 wings (bract) and a well developed adaxial or basal lobe (bracteoles) that covers the pistil. Engel- hardia on the other hand has a 3-winged involucre with a highly reduced fourth lobe. Furthermore, the wings are triveined with looping laterals in Oreomunnea^ whereas Engelhardia wings have pinnate venation. The closest ties of Oreo- munnea to Engelhardia are through E. roxhurghiana (Leroy, 1955; Stone, 1970), a monotypic representative of sect. Psilocarpeae. Whitehead (1965), for example, viewed the small, triangular pollen grains of E. roxhurghiana as similar to an ancestral type from which were derived the larger subtriangular grains of Alfaroa, Oreomunnea, and Engelhardia sect. Engelhardia, The intermediate status of E. roxhurghiana is further supported by similarities to Oreomunnea in inflorescence position, stigma orientation, fruit anatomy, and staminate flower morphology. Both taxa, as well as Alfaroa, have terminal androgynous panicles from which specialized lateral inflorescences were derived (i.e. E. sect. Engel- hardia; Manning, 1938). Stigma development and orientation, a feature that has undergone considerable change in the evolution of the family (Manning, 1940), also characterizes £. roxhurghiana, Oreomunnea, and Alfaroa. The short style and stigma are bifurcate with the primary separation oriented parallel to the line of carpel fusion, i.e. carinal stigma (Manning, 1940; Leroy, 1955: Fig. 75). The secondary or minor separation that is oriented at right angles to the primary separation is quite pronounced in £. roxhurghiana and effectively separates the stigma into four lobes, hence giving rise to the term split-carinaL The more specialized members of Engelhardia have commissural stigmas with the elongate stigma halves positioned at right angles to the line of carpel fusion. As mentioned previously, Alfaroa and Oreomunnea have 8-celled flowers and fruits, whereas those of Engelhardia are either 2- or 4-cclled. The fruits of E. sect. Engelhardia are 2-celled at the equator, although a weakly developed secondary partition is frequently present at the base. Engelhardia roxhurghiana flowers are funda- mentally similar to Oreomunnea but have a much more simple organization (Leroy, 1955). They lack the tertiary partition altogether, and the secondary par- tition is only moderately well developed. In addition, however, Leroy notes that the development of a excroissance tabidaire at the plane of insertion of the ovule 302 ANNALS [Vol. 59 in E. roxhurghiana (Leroy, 1955: Fig. 72) establishes a relationship with Oreo- munnea (and Alfaroa), as well as with Pterocarija (and Flatijcarija). The staminate flowers are somewhat less revealing of relationship, althovigh E. roxhurghiana and Oreomunnea mexicana (Figs. 7D, E) are the only members of the complex with a regularly arranged floral envelope. Manning (1948) notes, for example, tliat "In all species of the genus except E, chnjsolepis [E. roxhurghiana Wall.] the bracteolcs and lateral sepals may be irregularly arranged so that it is often diffi- cult to determine the exact identity of a given organ." The male flowers of Oreomunnea pterocarpa more closely approximate the irregular floral envelope arrangement found inE. sect. Engelhardia, Even after the last character is assessed in the evaluation of relationships, the decision as to the appropriate taxonomic rank to assign a particular taxon remains in the subjective hands of the taxonomist. The evidence presented here leads me to conclude that the ties between Alfaroa and Oreomunnea are just as strong as those between Oreomunnea and Engelhardia. Consequently, it would seem logical that the three taxa should be given coordinate rank; whether it be at the sectional or generic level depends, of course, on their standing in relation to the other members in the family. In this regard Manning (1949) has stated that Alfaroa is as distinct from Engelhardia as ]uglans is from Pterocarya. Accepting this opinion I must conclude that Alfaroa, Oreomunnea, and Engelhardia deserve generic rank. The dilemma posed by these three taxa reflects a close and what appears to be almost linear relationship. Engelhardia sect. Psilocarpeae has intermediate char- acters linking the wind-dispersed members of E. sect. Engelhardia and Oreomun- nea. A shift in adaptive zones from wind to animal dispersed seeds, with the concomitant loss of wings and enlargement of seed, seems like a plausible explana- tion for the evolution of Alfaroa from Oreomunnea-\ike ancestors (Stone, 1970). Species CoNsmERATiONS Three taxa of New World Juglandaceae with winged fruits are herein re( nized. Oreomunnea pterocarpa is endemic to mid -elevations in the central cor leras of Costa Rica. Oreomunnea mexicana ( En^elhardtia mexicana M The trees from Mexico, Guatemala, and Nicaragua (E. nicaraguensis Molina) are treated as O. mexicana subsp. mexicana, while the Costa Rican plants are placed in a new subspecies, O. mexicana subsp. costaricensis. The following section is devoted to a taxonomic appraisal of these taxa. Key to Species of Oreomunnea Bark tight; shoot tip glaucous; petioles long (3.5-6 cm), hairy at base; petiolules long (5-15 mm); leaflets long (8.5-16 cm), and wide (3-6 em), sometimes revolute at base but not auriculate, the secondary veins not conspicuously forking; female flowers with elongate style and deep cleft separating stylar halves; male flowers with elongate receptacles bearing 16-19 stamens in two disorganized scries; fruits large, the lateral wing span wide (to 13 cm); first pair of aerial seedling leaves simple 1. O. pterocarpa Bark exfoliating; shoot tip bronzy-yellow; petioles short (1.3-3.3 cm), glabrous at base; petiolules short (to 3 mm); leaflets long (6-10 cm), and narrow (1.5-2.5 cm), revolute and at least some auriculate at base, the secondary veins conspicuously fork- ing; female flowers with shoit, nearly obsolete style, and shallow cleft demarcating 1972] STONE— NEW WORLD JUGLANDACEAE 303 ^^ Figure 1. Tuis, Costa Rica. Tree of Oreomunnea pterocarpa with 27 in dbh at Stone 2684 location east of stylar arms; male flowers with small oval to rectangular receptacles bearing 8-12 sta- mens in one or two series; fruits medium, the lateral wing span narrow (to 5 cm); first pair of aerial seedling leaves compound -2. O. mexicana 1. Oreomunnea pterocarpa Oersted, Vidensk. Meddel 33-34, 1856. (Tree: Figs. 1, 2A-C. Inflorescence: Fig. 3; 304 ANNALS [Vol. 59 Mannin j, 1938: Figs. 9, 20; Hjelmqvist, 1948: Fig. 8C. Staminate flower: Figs. 4A-C; CandoUe, 1914: Fig. I; Manning, 1948: Figs. 38-41, 64, 68, 72. Pistil- late flower and fruit: Figs. 3, 4D-E; Candolle, 1862: Fig. 50; Oersted, 1870fl: Plate II, Figs. 1-11; Eichler, 1878: Fig. 15M; Candolle, 1914: Fig. II 1-3; Manning, 1940: Figs. 32, 33, 61, 78, 102, 103; Hjelmqvist, 1948: Fig. 21a; Pittier, 1957: Fig. 20. Seedling: Fig. 4F; Conde & Stone, 1970: Fig. 7). Engelhardtia pterocarpa (Oersted) Standley, Trop. Woods 12: 15. 1927. Tree to 46 m tall, 73 cm dbh {Stone 2222)^ buttresses well developed; bark tight, outer surface grayish to reddish and verrucose, inner bark yellowish-orange; wood white throughout, with definite growth rings (Hcimsch & Wetmore, 1939), diffuse porous, the vessel members from trunk averaging 221/x X 937/x; pith solid; buds naked, protected by closely appressed, paired leaf-primordia, studded with butter-yellow peltate scales, without hairs, the axillary buds superposed; leaves decussate; petioles 3.5-6.0 cm, hairy at base in sapling and sucker sprouts; raches 7.0-10.0 cm; leaflets 6-8, opposite to subopposite; petiolules 5-15 mm; blades of leaflets 2.5-3 times longer than wide, 8.5-16.0 cm X 3.0-6.0 cm, symmetrical to asymmetrical at base, most revolute on one or both margins, most pronounced on basiscopic side, the margins entire, the secondary veins not conspicuously forked; yovmg leaflets pink to greenish yellow; mature leaflets corriaceous, the adaxial surface dark green and essentially glabrous, the abaxial surface light green to glaucous, dotted with peltate scales of two sizes, moderate number of large butter- yellow scales and numerous small transparent or brownish scales, no hairs. Inflorescences borne laterally on old wood or occasionally at junction of old and new wood; "androgynous panicle . . . , with usually four to six opposite staminate catkins, the inflorescence terminated by a fifteen- to twenty-flowered pistillate catkin" (Manning, 1938; diagrammed as terminal in Hjelmqvist, 1948: Fig. 8c). Staminate flowers alternately arranged; abaxial bract 3-lobed, long and narrow; mature flower with elongate receptacle; inner floral envelope of two bracteoles and 3-4 sepals; stamens ca. 19, in two disorganized series, filaments essentially obsolete, anthers glabrous, dehiscing by longitudinal slits; pollen tectate with fine scabrate sculpturing, suboblate, subtriangular in polar view, pores equatorial, averaging 21/jt in diameter. Pistillate flotvers small (6-7 mm), highly reduced, lacking petals; subtending hand-like cup formed from 3-lobed abaxial bract and adaxial bracteole rim; calyx tube tightly fused to bract and bracteoles, the four narrow sepal lobes appearing as distinct outgrowths of ovary, seemingly not united into calyx tube; ovary inferior, elongate tapering style extending to same height as divergent sepal lobes, with deep cleft separating two stylar arms; stigmas carinal, subglobose or horseshoe shaped, capping ends of stylar arms, verrucose Figure 2. Vegetative aspects of Oreomunnea. — ^A. Shoot of O. pterocarpa displaying decussate phyllotaxy {Stone 1907), X 0.4. — B. Pubescent node of sapling of O. pterocarpa (Stone 2169), X 1.7. — C. Leaflet of O. pterocarpa {Stone 1907) with secondary veins that cun^e upward toward tip and end without conspicuous branching, X 0.6. — D. Leaflet of O. mexicana subsp. costaricensis {Stone 2680) with auricles at base and secondary veins that curve upward toward tip, branch and then fuse near margin, X 0.8. 1972] STONE— NEW WORLD JUGLANDACEAE 305 306 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 stigmatic surface confined to rim and oviter surface. Fruits large, 3-winged, lateral wing spread to 13 cm, central wing to 11 cm long, strong central primary vein flanked by two subordinate primaries and interconnected by looping laterals; adaxial pair of fused bracteoles enclosing globose nut and stylar-stigma region; nut 8-celled at equator, 4-celled near poles; fruit splitting loculicidally along secondary partition on germination. Seedling with long, slender tap root; germi- nation hypogeal; first two aerial leaves opposite and simple, succeeding several leaves alternate and simple, followed by transition to compound, and later abrupt shift to opposite phyllotaxy; shoots pale to vivid pink; simple leaves with long petioles (20-30 mm) and large blades (40-80 X 75-250 mm), with conspicuous midrib and secondaries; transitional compound leaves (3-, 5-, and 7-foliolate) with extremely long petioles (30-60 mm), near-sessile lateral leaflets (to 2 mm), and an extended petiolule ( 15-20 mm ) supporting large terminal leaflet ( 50-100 X 100-200 mm); leaves with more than five leaflets have even-pinnate pattern char- acteristic of those of mature tree; simple leaves and leaflets of compound leaves entire, no evidence of inrolling of basal margins; adaxial leaflet surface dark green, glabrous; abaxial leaf surface glaucous, free of hairs, dotted with large butter- yellow peltate scales and numerous brown peltate scales. Oersted's original description ( 1856) and subsequent discussions ( 1870fZ, 1870Z>) were based on fruiting material only. It was nearly sixty years before collections by Adolfo Tonduz made vegetative and flowering specimens available for study (Candolle, 1914). The paucity of collections is as much a function of the difficulty of collecting specimens from these forest giants as it is locating them along the steep ravines of the valley and tributaries of the Rio Reventazon. Trees in the virgin rainforest around Platanillo and Tuis commonly have plank buttresses extending eight feet in diameter at the base and tapering to a 2 to 3 foot trunk at head height (Fig. 1). The bark is smooth and tight, and the trunk is often without branches for the first 40 to 50 feet. The mature trees may reach 46 m (Tonduz m de Candolle, 1914) to become upper story emergents. Young shoots are usually vivid pink, while mature leaves are dark green above and glaucous below. Perceptive native guides are remarkably accurate in spotting the broad crowns from distant hillsides. In the Platanillo area woodsmen readily differentiate between gavildn bianco {Oreomunnea pterocarpa) and gavildn Colorado ( Alfaroa manningii Leon ) . One of the most intriguing aspects about Oreomunnea pterocarpa is its highly restricted range. As both Standley (1927a, 1937) and Manning (1959) have pointed out, it is known only from the low to middle elevations on the Atlantic drainage of Costa Rica. All of the verified specimens are from the Rio Reventazon valley in Cartago Province, although Mr. Charles Lankester presented me with typical O. pterocarpa fruits, ostensibly collected at Laguna Hule in Alalueja Prov- ince. Collection data from herbarium sheets indicate that O. pterocarpa ranges from 200 m (Manning, 1959: locality not cited nor known to me) to 1500 m elevation (Tonduz in Candolle, 1914: Rincon del Indio). The large tree on the grounds of the IICA at Turrialba {Stone 1346) is situated at 550 m, while Leons topotype collection from La Gloria {Leon 1523) was located at 820 m. There is, 1972] STONE— NEW WORLD JUGLANDACEAE 307 r^TT_Fr ^ \ Figure 3. Inflorescence of Oreomunnea pterocarpa (Stone 1346) with laterally borne fniits, X 0.54. in addition, a cultivated tree in the Botanic Garden of the Universidad de Costa Rica in San Jose (1168 m) that appears to be doing quite well. Manning (1959) suggested that a specimen from Baja Verapaz, Guatemala {StancUey 90967, Gil) might be O. pferocarpa. However, the specimen is sterile and lacks the glaucous shoot tip and long petiolules characteristic of O. pterocarpa; it appears to be a member of the genus Alfaroa, probably A. guatemalensis (Standley) Williams & Molina. Oreomunnea pterocarpa^ like so many of the tropical fon^st trees, may be 308 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 locally abundant but not common. This is one of the puzzling aspects of tropical diversity that has led to speculation that many tropical plants are self pollinated (Baker, 1959, 1963). Oreomunnea has typical amentiferous catkins, and the large 3-winged fruits of O. pterocarpa (Fig. 3) show obvious adaptations to wind dispersal. Furthermore, the tree crown is favorably situated in the upper canopy. Fruit production on the tree in San Jose {Stone 1016) offers affirmative evidence for self pollination in O. pterocarpa. No other trees are known to occur within 20 miles of this specimen and yet fruit and seedlings are prodviced at least occasion- ally (Jan 1962, Nov 1967). Costa Rica, alajuela: Edge of Laguna Hule, ca. 12 km N of V. Poas, no date, Lankester (DUKE). CARTAGO: Juan Vifias, no date, Calveri 57 (PH); Cachi, Jun 1910, 4 Jul 1914, Lankester (NY); La Gloria, bordes del Rio Chiz, 820 m, 23 Apr 1949, Leon L523 (EAP, IICA); La Gloria, 700-800 m, 29 Apr 1914, Tondtiz 18000 (A, CR, EAP, F, GH, NY, PH, US); Peralta, 13 Jul 1923, Stevens 468a (US); grounds of Inst. Interamericano Ciencias Agricolas (IICA) near Rio Turrialba, 550 m, 7 Sep 1949, Leon 1819 (IICA); slope behind IICA station over- looking Rio Reventazon, 550 m, 22 Jan 1963, 28 Mar 1966, 26 Jan 1968, Stone 1346 (DUKE, NO), 8 Jul 1965, Stone 1908 (DUKE); on peninsula between Molina Quebrada and Rio Reventazon near Turrialba, 10 Jul 1965, Stone 1919, 1920, 1921 (DUKE), 23 Apr 1966, Stone 2169 (DUKE); Pejibaye, 2000 ft., 3 Jul 1928, Stork 2808 {¥); hills S of PlataniUo, 22 Jan 1963, Stone 1347, 1348 (DUKE); midslopes SE of PlataniUo, 25 Mar 1967, Stone 2222 (DUKE); Montana Baja Corono, E of Tuis and N of PlataniUo, 2 Sep 1968, Stone 2684 (DUKE); no locality, 1927 and 1964, Lankester (F, US), san jose: Botanic Garden, Univer- sidad de Costa Rica, San Jose, 3800 ft., 19 Jan 1962, 7 Jul 1965, Stone 1016 (Duke), 10 Nov 1967, Stone 1016 A, B, C (DUKE). 2. Oreomunnea mexicana (Standley) Leroy, Bull. Mus. Hist. Nat. (Paris) Ser. 2. 23: 127. 1951. (Tree: Figs. 5, 6A-C, 7A, 8. Inflorescence: Figs. 7B, 7C, 9A, 11. Staminate flower: Figs. 7D, 7E, lOA, lOB. Pistillate flower and fruit: Figs. 9A-F, lOC-J; Manning, 1940: Figs. 34, 35, 106. Seedling: 9G.) Engelhardtia mexicana Standley, Trop. Woods 12: 15. 1927. Engelhardtia nicaraguensis Molina, Fieldiana Bot. 31: 358. 1968. Tree to 32 m tall {Stone 2177B) or more, 150 cm dbh {Stone 2807 locality), buttresses developed on both downhill and uphill sides of trunk, occasionally extending to height of 3 m; bark exfoliating, outer surface reddish brown, inner bright orange; wood white throughout or heartwood sometimes pink, diffuse porous, the vessel members from trunk averaging 160;li X 693fi; pith solid; buds naked, protected by closely appressed paired Icaf-primordia, studded with bronzy- yellow peltate scales, without hairs, axillary buds superposed; leaves decussate, petioles 1.3-3.3 cm, with dense covering of short hairs (.04 mm) on young trees, raches 1.8-11.8 cm, with dense covering of short hairs on young trees; leaflets 4-12, opposite to subopposite; petiolules 0-3 mm; blades of leaflets 4-5 times longer than wide, 6.0-10.0 X 1.5-2.5 cm, symmetrical to asymmetrical at base, revolute on one or both margins, most pronounced on basiscopic side, auricles often present on one or both margins, the margins mainly entire, with coarse ser- rations on distal half of stump sprouts and some shoots, the secondary veins con- spicuously forked; young leaflets pink; mature leaflets coriaceous, the adaxial surface dark green and essentially glabrous, the abaxial surface light green to glaucous, dotted witli peltate scales of two sizes, moderate number of large bronze scales and numerous small brownish scales, short hairs on basal portion of midrib 1972] STONE— NEW WORLD JUGLANDACEAE 309 Figure 4. Flowers, fruit, and seedling of Oreomunnca pierocarpa. — A. Portion of staiiii- nate catkin {Stone 1346), X 4.3. — B. Abaxial view of staminate flower with 5 floral segments {Stone 1346), X 10.4. — C. Adaxial view of same flower as in C, exposing attachment points of 19 stamens, X 10.4.— D. Pistillate flowers {Siune 1346), x 3.4.— E. Transverse section of fruit at equator, oriented with primary partition in cast-west direction: 8-cclled, thin huslc and thin cartilaginous shell {Stone 1346), X 2.1. — F. Seedling with alternately arranged leaves, hypogeous cotyledons shed {Stone 2684), x 0.4. 310 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Figure 5. Oreomunnea mexicam {Stone 2718), 30.5 in. dbh, Tapanti, Costa Rica tucked Infl with three typ pairs of stammate catkins (Fig. 11); staminate panicle consisting of 1 to 3 de- 1972] STONE — NEW WORLD JUGLANDACEAE 311 % Figure 6. Habit and habitat of Oreomunnca mrxirana. — A. Moderate development of buttresses. Stone 2177B location. — B. Bark separating into long platelets on 14 in. dbh trunk, Stone 2181, — C. Fog-shrouded mountains south of Muncco, Co.sta Rica, ca. 1500 m elev., June 1970. 312 AN>; [Vol. 59 cussate pairs of catkins; pistillate panicle of one to several catkins, fruiting spikes elongate but fruits congested. Staminate flowers alternately arranged, small (2-3 mm), sessile or essentially so, outer surface sprinkled with butter-yellow and transparent peltate scales; abaxial 3-lobed bract cupped; mature flower with rounded receptacle; inner floral envelope of two lateral bracteoles and two (rarely three) sepals; stamens 8 in one series, ranging to 12 in two series, filaments essentially obsolete, anthers glabrous, dehiscing by longitudinal slits; pollen tectate with fine scabrate sculpturing, suboblate, subtriangular in polar view, pores equa- torial, averaging 19jli {Stone 2144) to 23/a {Stone 2181) in diameter. Pistillate flowers small (3-5 mm), highly reduced, lacking petals; subtending hand-like cup formed from 3-lobed abaxial bract and adaxial bacteole rim; calyx tube fused to bract and bracteoles at base, forming distinct calyx tube above, four broad sepal lobes incurved or spread, extending well beyond stigma; style short tapering, with deep cleft separating two stylar arms; stigmas carinal, subglobose or horse- shoe shaped, capping the ends of stylar arms, the verrucose stigmatic surface confined to rim and outer surface. Fruit medium-sized, 3-winged, with lateral wing spread to 5 cm, central wing to 4 cm long, with three primary veins and looping laterals; adaxial bracteoles enclosing globose nut and stylar-stigma region, to 12 mm long; nut 8-celled at equator, 4-celled near poles; fruit splitting loculi- cidally along secondary partition on germination. Seedling with long, slender tap root; germination hypogeal; first two aerial leaves opposite and pinnately com- pound, succeeding several leaves alternate and compound, abrupt shift to op- posite phyllotaxy in young sapling stage, 1-2 feet tall {Stone 1872); shoots pale to vivid pink; most leaves odd-pinnate, the 18-20 (14-22) leaflets coarsely serrate; petiole short (to 7 mm) in first formed leaves, longer (to 25 mm) in later ones; leaflets essentially sessile with petiolules rarely 2 mm long; later formed leaflets becoming entire, the last formed leaves of 3-4 foot saplings with entire margins on basal half of leaflets, coarse serrations on distal half; leaflet bases often asym- metrical, auricles usually present on leaflets of larger saplings; a coating of short ( .04 mm) hairs on stem tip, petiole, and rachis and a cluster of longer hairs (.6 mm) at junction of petiole and stem; adaxial leaflet surface dark green, with butter- yellow scales scarce; abaxial leaflet surface light green to glaucous, occasionally with a few long hairs at base, dotted with large butter-yellow peltate scales. Chromosome number, 2n = 32, based on counts from roottips of seedlings, green- house progeny of Stone 2141 { DUKE ) . Oreomunnea mexicana as here recognized constitutes a very natural assem- blage. The range from Oaxaca, Mexico, to Cartago, Costa Rica, may at first appear to be rather exceptional, but not when viewed in light of the similarity in environ- mental conditions at each locality. All trees that I have seen, including ones from Mexico, Guatemala, Nicaragua, and Costa Rica, are invariably situated on steep hillsides and fog-shrouded ridges. Miranda (1946) vividly described this type of habitat for the Orizaba population: ". . . emergian de la sombra de la neblina y cuyas altas copas volvian a perderse en ella.'* Present records indicate that the species ranges from 900 to 2600 m. There is reason to believe that the plants at higher elevations may not fare nearly so well as those growing in the 1000-1700 m 1972] STONE— NEW WORLD JUGLANDACEAE 313 A. FiGUBE 7. Vegetative and flowering features of Oreomunnea mexicana subsp. mexicana. Shoot displaying decussate phyllotaxy (Stone 2141), X 0.4. — B. Terminal staminate inflorescence {Stone 2141), X 0.4.~C. Portion of staminate catkin {Stone 2144), x 6.5. — ^D. Adaxial view of staminate flower {Stone 2144) with proximal sepal obscuring view of 3 stamens and receptacle, X 13. — E. Adaxial view of staminate flower exposing attachment points of 12 stamens in two series {Stone 2144), X 13. 314 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 ^B. r^-r d-KHi-jdriliqiMri^r -^_^__J Mji ^-^ m^ A M i : x^ r <.: j\ ki r. * f -f % ^ t ./ / J- f fr- > i: , , - ^..* I -■^.,V .^ ^^ ? . ^ / C ;^ r ^^' i /'^^V ■ *- ' '^ I' ^ 7 X ^y -?r FiGTTRE 8. Holotype of Oreomunnea mexicana subsp. mexicana {Roviwsa 1006, PH) mexicana subsD. mexicana have yielded flowering or fruiting specimens. These include: typ collection (fruits) from Chiapas (900 Martinez (fruits) and Stone (male and female flowers and fruits) specimens from Oaxaca (1700 m); female flowers by Stone from Alta Verapaz, Guatemala (1300-1500 m); collections by galpa, Nicaragua (1450 m). Jinotega and Mata mexicaiid 1972] STONE— NEW WORLD JUGLANDACEAE 315 IS a "shrub in elfin forest" at 1500 m suggests that unfavorable exposures at even moderate altitudes may handicap growth and reproduction. Perhaps marginal populations are maintained through asexual means. I have observed, for example, that saplings of O. mexicana are capable of vegetative propagation. If the stem is prostrated on the ground, adventitious shoots may be stimulated. Also, it is quite likely that stump sprouts arise here as in Alfaroa (personal observation). The geographical range of Oreomunnea mexicana has, to be sure, promoted varying degrees of population divergence, but the differences are subtle. Miranda (1946) thought he perceived a difference in the more numerous and sessile leaf- lets in the Orizaba specimens and indicated that he would later describe this element as a new species. However, no formal recognition was ever proposed, and, except for Molina's EngelJiarcltia nicaraguensis (1968), no other segregate or close relative has been recognized. The field characteristics of seedlings, saplings, and adult trees of Oreomunnea niexicana are remarkably distinctive and as far as I am aware uniform in all of the Middle American populations. Foliage of younger plants and sucker shoots on trees are typically reddish; the leaves are opposite, pinnately compound, and the leaflets are serrate. The larger trees can usually be spotted by the exfoliating bark that peels off in large strips (Fig. 6B). The leaflets on mature trees are mostly entire, but some serrate leaflets can always be found. Alfaroa costaricensis sensu lato (including A. wilUamsii Molina) is the only other species of tropical American Juglandaccae that approaches O. mexicana in this respect The leaflets and terminal shoots of O. mexicana tend to be more slender than those of O. pterocarpa and the Alfaroas. This combination of charac- ters unites the Mexican, Guatemalan, Nicaraguan, and Costa Rican populations of O. mexicana. Only in flower characters has it been possible to discern any signif- icant differences. The following key emphasizes these differences in recognizing two subspecies. Key to Subspecies of Oreomunnea mexicana Female flowers and fruits with conspicuous pedicels to 3 mm long; pistil oriented parallel to axis of catkin, sepals incurved; Mexico to Nicaragua 2a. subsp. mexicana Female flowers and fruits sessile or nearly so; pistil oriented at a 45° angle to axis of catkin, sepals spread; Costa Rica 2b. subsp. costaricensis 2a. Subspecies mexicana Holotype: Mexico, chiapas; Hab. regione temperata locis humidis Andium Chiapensium, 900-1000 m, Rovirosa 1006 (PH; photograph with two leaflets and sc\eral fruits, US). Engelliardtia nicaraguensis Molina, Fieldiana Bot. 31: 358. 1968. Type: Nicaragua, matagalpa; Bosque nebulosa vecindades de Santa Maria de Ostuma, 1300 m, Salas 2967 (F; isotype, EAP). There are no vegetative features that I have found that can he used to recog- nize the subspecies. Thus, when the sterile specimens are cited it should be noted that provisional identification was based solely on geographic location. The mexicana mexicana that curves to orient the pistil parallel to the axis of the catkin (Fig. lOH); in this respect it is reminiscent of the female flower of Engelhurdia roxhurghiana (sec Manning, 1940). 316 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 After careful examination of the type specimen of Engelhardtia nicaraguensis and a few subsequent collections, I can find no basis for recognizing this taxon. The female flowers (Fig. lOH) are indistinguishable from those of the Mexican {Stone 2807) and Guatemalan {Stone 3003) specimens. In the original descrip- tion Molina (1988) did not single out any diagnostic features, but rather relied on a host of meristic differences between the type specimens: E. mexicana E. nicaraguensis Pairs of leaflets 3-4 4-6 Petiole length (cm) 1.5-2 2.5-3.5 Rachis length (cm) 2^6 4-9 Petiolule length (mm) 2-3 1-2 Leaflet length (cm) 5-8.5 3-8.5 Nut length (mm) 6-7 6-8 The inclusion of additional samples in this comparison completely blurs the dis- tinctions noted above (see diagnostic descriptions under O. mexicana). Fertile specimens: Mexico, chiapas: Hab. regione temperata locis humidis Andiuin Chiapensium, 900^1000 m, 24 Jul 1891, Rovirosa 1006 (PH, holotype; US), oaxaca: Highway 175, 102.5 km S VaUe Nacional, 1700 m, 7 Nov 1962, Martinez 13^9 (MEXU, DUKE); highway 175, 42,1 km S Valle Nacional, 4900 feet, 16 Mar 1966, Stone 2141 (DUKE); high- way 175, 38.8 mi N Ixtlan de Juarez, 2 Feb 1971, Stone 2807 (DUKE). Nicaragua, matagalpa: Bosque nebuloso vecindades de Santa Maria de Ostuma, 1300 m. May 1957, Salas 2967 (F, holotype; EAP). jinotega: Cerca de Aranjuez, 1160 m, Aug 1957, Salas ir Taylor 2967 A (EAP); near Santa Maria de Ostuma, on road to Aranjuez, 0.9 mi E of highway 3, 1450 m, 6 May 1966, Stone 2181 (DUKE). Guatemala, alta verapaz: On highway 7E, 1.3 mi SE of highway 5 intersec. near Tactic, 25 May 1971, Stone 3003 (DUKE). Sterile specimens: Mexico: veracruz: Cerro de San Cristobal, near Orizaba, 4200- 5000 feet, 3 Aug 1953, Manning ir Manning 53746 (A), 4 Aug 1953, Manning b- Manning 53770 (NY, PH), Manning I- Manning 53775 (DUKE), 8 Mar 1966, Stone 2518, 2119, 2120, 2121, 2122, 2123, 2124 (DUKE); Santa Marta, 5000 feet, Los Tuxdas range, 17 Jun 1963, Ross 55 (US); near Ocotal Chico, 3000 feet, Los Tuxtlas, 24 Mar 1965, Ross 120 (US); Ejido San Fernando, vertiente SW Sierra de Santa Marta, 1000 m, 14 Mar 1968, Sousa 3551 (DUKE). oaxaca: Highway 175, 42.1 km S Valle Nacional, 4900 feet, 28 Jan 1965, Stone 1872, 1873, 1875 (DUKE); highway 175, 42.1 km S Valle Naciondl, 4900 feet, 2 Feb 1971, Stone 2808, 2809, 2810, 2811, 2812 (DUKE). Guatemala: alta verapaz: Region of Chelae, NE of Carcha, 1500 m, 2 Apr 1939, Standley 70379 (A, F); ridge SE of Tactic, 9 Nov 1969, Stone 2697 (DUKE); highway 7E SE of Tactic, 2 kni from junction with highway 5, 10 Nov 1969, Stone 2703, 2704, 2705 (DUKE). BAjA verapaz: Sierra de las Minas, E of Chilasco, 28 May 1950, Lamb 113 (F, EAP). el PROGRESSO: Montana Canahui, bet\\een Finca San Miguel and summit, 1600-2300 m, 10 Feb 1942, Steyermark 43810 (A, F). Guatemala; No locality or date, Aguilar 616 (F). huehue- : Cerro Huitz, Sierra de los Cuchumatanes, between Mimanjuitz and Yulhuitz, 1500- 2600 m, 14 Jul 1942, Steyermark 48619 (F). -> Figure 9. Fruits and seedling of Oreomunnea mexicana subsp. mexicana. — A, Inflores- cence with terminally-borne fruits (Stone 2141), X 0.8. — B. Adaxial view of mature fruit (Stone 2141) showing triveined wings and well-developed basal lobe, X 0.8. — C. Radial section of fruit along secondary partition exposing two stylar arms separated by major groove and enclosed by abaxial bract (b) and adaxial bracteoles (br) or basal lobe (Stone 2141), X 3.4. — D. Radial section of fruit along primary partition exposing bifurcation of one stylar 1972] STONE — NEW WORLD JUGLANDACEAE 317 arm E. Polar view oi horseshoe-shaped stigma on fruit, major groove oriented in north-south direction (Stone 2141), X 8.7. — F. Transverse with 8-celled, thin husk (outer dark band) and thin cartilaginous shell (stippled zone) {Stone 2141), X 3.4, G. hyp succeeding leaves alternate (Storie 2142), X 0.8. 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 v-» _ ■ * * ■ *' 5? --^ Y 5 -'?*.- VV --Via" ^ *^ "- ■•■*■' ' ^*H E>'j < - ■-_ * .... 1 'y J* ■?''-■. --. { ' ".t >> ■ ■>-> I ^» Figure 10. Inflorescence and flowers of Oreamunnea mexicana: A, C-G, subsp. costa- ricensis; B, H-J, subsp. mexicana (Nicaragua). — A. Adaxial view of staminate flower widi 8 stamens and aborted pistil {Stone 2680), X 15. — B. Adaxial view of staminate flower with 8 stamens in one series {Siofie 2181), X 15. — C. Pistillate flowers with spreading sepals and moderately-developed stigma (Stone 2177B), X 6. 18.— E. Carinal stigma (Stone 2177B), X Pistillate flowers with spreading sepals and well-developed stigma (Stone 2680), F. Side view of stigma sighting along major groove (Stone 2680), X 18. — G, Polar view of stigma (Stone 2680), X 18. — H. Pistillate flowers with incurved sepals (Stone 2181), X 5.1. — I-J. Side view of carinal stigma (Stone 2181 ), X 18. X 6.5.- 1972] STONE— NEW WORLD JUGLANDACEAE 319 r -* Figure 11, Androgynous panicle of Oreomwmea mexicana subsp. cosiariccnsis {Stone 2680), X2.L 320 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 2b. Subspecies costaricensis Stone, subsp. nov. Floras feminei pedicellis brevibus et calycum lobis rcflexis. Type: Costa Rica, cartaco: 4 km S of Muneco, ridge between Rio Patarra and Rio Som- brero, Stone 2680 (DUKE). This new subspecies is known only from Cartago Province, Costa Rica. While it grows in the same general areas as Oreomunnea pterocarpa^ Alfaroa costaric- ensis, and A. manningii, O. mexicana subsp. costaricensis is usually situated above the protected ravines on fog-shrouded mountain ridges in a premontane rainforest {sensu Holdridge, 1964). No good fruiting material has been collected, although remnants of the winged bracts were found attached to young seedlings. The female flowers are somewhat smaller than those of O. mexicana subsp. mexicana, in addition to the fact that the pedicel is short, the pistil is oriented at a 45° angle to the axis of the catkin, and the sepal lobes are spread and rather flat. There appears to be some variation in the extent of elaboration of the stylar arms. The material from Valle Escondido (Stone 2177B) has a deep stylar cleft, but only moderately developed minor grooves on the stigma lobes (Figs. lOC-D). Speci- mens from Muneco {Stone 2680) have deep minor grooves that give the stigma a pronounced 4-pronged appearance (Figs. lOE-G). The significance of these differ- ences cannot be evaluated until more collections are available. Fertile specimens: Costa Rica, cartago: Ridge above Valle Escondido, ca. 14 km NE of Tuis, 1100 m, 29 Apr 1966, Stone 2177B (DUKE); 4 km S of Muiieco, ridge between Rio Patarra and Rio Sombrero, 29 Aug 1968, Stone 2680 ( DUKE, holotype ) . Sterile specimens: Costa Riga: cartago: Valle Escondido, ca. 14 km NE Tuis, 1100 m, 1966, Hathewatj (DUKE), 29 Apr 1966, Stone 2177A (DUKE); Tapanti, 1300 m, 2 Sep 1966, Hatheway (DUKE), 6 Nov 1967, Stone 2316, 2317, 2318 (DUKE), 20 Jan 1968, Stone 2334A, 2335 (DUKE); ridge above Tapanti on precipitous road to Tausito, 4500 feet, 23 Mar 1968, Stone 2342, 2343 (DUKE), 13 Nov 1969, Stone 2718, 2718A (DUKE), 8 Jun 1970, Stone 2745, 2746 (DUKE); Muneco on Rio Navarro, 1400-1500 m, 6-7 Mar 1926, Standley j ' - ______ S: ^^,^«iSSgS^S^^sm ■ ■ X _ J J _ ■*"■ n r J _ _ ■;■:'^>:v : ^"XO!^! ■>:%':-:■ -%■ ^ V_» ^h »_l' _ i_i_rr-OCiiCj:.V'. _ -^^>>x-:^'.- ^^HBBp FtoMaslba 1 )i^ ;)^ I ^ Plast M^r 25 Tubm ■ ^ViV^WV^^ V ■ »^iH^>Tr*^'*vVfc.'-^^^^>:^ ■^O-'^^W B WqcWC^ CONTENTS I Rejoice in What He Affirmed 323 Edgar Anderson, 1897-1969 John J. Finan 325 The Pubhcations of Edgar Anderson Erna R. Eiscndrath 346 Student Days with Edgar Anderson or How I Came to Study Sunflowers Charles B. Heiser, Jr 302 Edgar Anderson: Recollections of a Long Friendship G. Ledyanl Stchhins 373 Hybridization, Evolution, and Systematics Duncan AL Porter 380 Cytological Studies of Natural Intergencric Hybrids and their Parental Species in the Moss Genera Astonuim and Wcissia Lewis E. Anderson 6 Betty E. Lemmon 382 Cytological Evidence of Introgression Between Drosophila Species in Panama Sarah Bedichek Pipkin 417 Levels of Confidence in the Analysis of Hybridization in Plants L. B. Gottlieb 435 Hybridization, Taxonomy and Avian Evolution Lester L. Short 447 Variation and Spcciation in the Genus Hudsonia Judith Troop Skog 6- Norton H. Nickerson 454 C-4 and C-3 Carboxylation Characteristics in the Genus Zygophyllum (Zygoph>llaceae) R, Kent Crookston & Dale N. Moss 465 New Combinations in Compositae R. P. Wunderlin , . . , 471 NOTES Capparis hypoleuca Presl: A Synonym of Solanum schlechtcndaUanum Walp. Hugh IL litis 474 The Chromosome Number of Utricularia dcnticulata Benjamin Katsuliiko Kondo . • 474 VOLUME 59 1972 NUMBER 3 ANNALS OF THE Missouri Botanical Garden The Annals contains papers, primarily in systematic botany, contributed from the Missouri Botanical Garden and the Department of Biology of Washington University. Papers originating outside the Garden or University will also be accepted. For information on preparation of manuscripts, see the inside back cover. The Annals appears three times a year, and three numbers, totaling about 400 pages, constitute a volume. For information concerning subscriptions, see the back cover. Matters regarding exchange of publications are handled by the Library, Missouri Botanical Garden, 2315 Tower Grove Avenue, St. Louis, Missouri 63110. Editorial Committee Marshall R. Crosby, Editor Missouri Botanical Garden John D. Dwyer Missouri Botanical Garden 6- St. Louis University Peter Goldblatt Missouri Botanical Garden Published by the Missouri Botanical Garden Press, St. Louis, Missouri 63110. © Missouri Botanical Garden 1973 ALLEN PRESS, INC, *'TJ'* LAWRENCE, KANSAS aj-' ' ?f-^ ^- SJ . ---^fe^. -! ^^- \ ■y- ^y .■-■^ "^. y^. \. ■■- .. X. 4 .« This publication is dedicated to the memory of Edgar Anderson ( November 9, 1897-June 18, 1969), lover of natm-al beauty and order; great scientist; good and gentle man. Tliis photograph was taken in 1966. VOLUME 59 1972 NUMBER 3 ANNALS OF THE Missouri Botanical Garden I REJOICE IN WHAT HE AFFIRMED^ ?> , rises Bear Mountain, climlx^d last Scptcuiber. I left the train at Becket and utilized the cool morning hours to good advantage, 330 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 making three miles an hour across hills via roads that were seldom more than cat tracks. Mountain laurel, fragrant pink azaleas, and a pair of oxen plowing, made the walk more interesting. At East Lea a truck picked me up for the short mile but at South Lea a benevolent, but slow-moving meat truck took me in and car- ried me thru the beautiful summer towns of Stockbridge etc. to Great Barrington where I turned off from the main road to follow this riv^er road. I took a three- minute dip in a forest-girt pool this morning and bathed my feet this afternoon. These rock out crops are interesting to me. I wish I knew the language in which they are written (Idea for boys' talk-diary) . Saturday, June 4th 192L 7:30 A.M. Canaan, Conn. After writing you [diaryl the other afternoon I sauntered on a mile or t\vo, passing pleasant comfortable farm houses where the cattle were first being brot up to be milked. I picked out a trim little house with a neat \a\vr\ and asked if I might sleep in their bam and if they would sell me some supper; that I was *'sane and respectable** in spite of my get up. Permission being granted I helped Mr. Cahill and his high-school boy son to put the cattle in the bam (note for father — they were grade Holsteins and the stable shone with white whitewash.) .... After supper I milked two cows, played mum- ble-ty-peg with Harold, and sat on the piazza vmtil dusk. Harold, finding that I was guilty of an Agricultural College education, in a mildly kidding way asked me what new farming wrinkles I had picked up in college. I told him we learned to plant potatoes and onions in alternate rows; this made the potatoes* eyes water and irrigated the potatoes. His Dad seemed to think I had the laugh on Harold and chaffed him about it. I departed for "bed" at nine, in spite of earnest entreaty to use a real bed in the house. . . It thunderstormed during the night but started to clear up w^hen I started at 5 and the air is now fresh and cool and the mountains in their new spring duds, are a bright light green. ... Althou fessor, Dr. East, he seems to have been fascinated by, but at the time not fully appreciative of, the Economic Botany course at the Bussey given by Oakes Ames, director of the Harvard Botanical Museum. It was only much later, as Anderson acknowledged in Plants, Man and Life (1952), that he gave importance to Ames's provocative questions about the origins of cultivated plants — questions that An- own and in that of many of his students on sunflowers, amaranth, tomatoes, rubber, beans etc. The person at the Bussey who turned out ultimately to be the most important to him was a young laboratory assistant to Dr. East, Dorothy Moore. She had been trained at Wellesley, and was completing Masters work in botany there while assisting at the Bussey. Edgar met Dorothy, as their daughter remembers the story, when Dorotiiy berated him for leaving a microscope dirty from staining. He courted her for two years, bringing her flowers (see diary entry of Feb. 12, 1920 above) and taking her on hikes. They were married in 1923. Totally uncon- trived, of quiet but keen intelligence, and always glowing with absolute inner goodness, Dorothy Anderson provided unfailing support in a union that v^as to last 46 years. After his degree completion at Harvard, Anderson was invited to join the staff of the Missouri Botanical Garden, which had, through the Henry Shaw School of Botany, a tie with Washington University. His adjustment to life in St. Louis seems to have been quick not only because he was reared a mid- Westerner but because St. Louis, like Boston, maintained cosmopolitan intellec- 1972] FINAN— EDGAR ANDERSON, 1897-1969 332 Itu The director espoiisible St. Louis, had, like Anderson, been trained at Hai'vard. George H. Pring, for many years superintendent of the Garden, has recalled Andersons arrival in St. Louis :^ Dr. Moore said to me that he was bringing a young geneticist from Boston for an interview. He asked me to show him about the Garden and to acquaint him with the city. After doing this, I took him to Riiggeri's for a meal. After eating, I asked if he had ever tasted our famous German cheesecake in St. Louis, and he said he hadn't. So he ordered a large piece of cheesecake and consumed it with gusto. Ever after he made reference to his first impressions of St. Louis centered around that cheesecake. Anderson held three posts during those first years in St. Louis: an assistant (later associate) to the Washington University, geneticist Emmet Layton, now in landscape architecture at the University of Wisconsin, Green Bay, who was study- the Even in the somewhat alien field of the Henry Shaw Scliool of Gardening, [Anderson] contributed as much to his students as did the more appropriate Land- scape Architect John Noyes. Where John taught techniques, Edgar taught scholar- ship and the love of learning, along with an empathy with the earth itself. He also had a genius for giving succor in serious situations which might other- wise have demolished the individual. He could rescue a failing student by sheer force of his own vast energy and a determination to teach. He also made sure there was no exaggerated sense of obligation. Where his gift was beyond price, he al- ways found a way for the recipient to contribute something, in return for which Edgar would express sincere gratitude. When I found advanced mathematics as impossible to learn as Sanscrit, Edgar not only resorted to absolutely Herculean methods to teach me so thoroughly that mathematics became a lifelong part of my personal vocabulary . . . but he arranged for me to drive him home to Webster Groves after those long night sessions and thanked me profusely for a performance I naturally regarded as a privilege. For Edgar opened the entire future for me, and led me through doors I could never have entered without him. Without his help I would never have been able to enter the School of Architecture at Washington Uni\ crsity, since I would have found the entrance examinations quite beyond my capacity. In turn, Ed^ar was careful to explain how much he appreciated my own skill in graphics presentations, in producing and interpreting plans which he always claimed he could never under- stand widiout a model — though he could take any blueprint and read it Hke a book, provided it applied to something in which he took a special interest. From the point of view of the history of science, these first years of Anderson in St. Louis are important because it was during them that there was developing the interest in *'biosystematics," as it came to be called, which was to pro\ade the framework for most of his later research. Unfortunately he did not keep diaries during most of this period, and it was a time when he published very little. ^ Letter to author from George H, Pring. * Letter to author from Emmet Layton. 332 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 He once said to me, although I did not reahze the full significance of his words until the preparation of this biographical essay, that it was during this period he very consciously decided not to publish but to think. Out of those months and years of reflection, and frequent visits, as Layton remembers, to "the vast fields of wild iris beyond the city of Saint Charles, in die Missouri River flood plain," emerged his now classic papers on speciation in Iris. The development of Anderson s thought leading to his discovery of the impor- tant role of baclccrossing ( "introgressive hybridization" as he termed it) in the evolution of natural populations and cultivated plants and his devising of means to anal}^ze and measure it, can be traced only superficially here. His going in 1922 from the Bussey Institution, where he had majored in genetics, to the Mis- souri Botanical Garden, which had an historic strength and distinction in tax- onomy, probably was fortuitously influential in stimulating his interest in the rich but largely unexplored area between the two fields. Too, his mentality was conducive to such a research direction; he frequently referred to his "summa- rizing mind" which was indeed highly classificatory but was at the same time dis- trustful of "labels" and was provoked by the analytical preclusions which he felt they often imposed. Also, his natural bent for turning associations with colleagues and students into mutually creative experiences must be taken into account. Tlie Garden staff and students whom he came to know in the years 1922 to 1931, when he returned to Harvard, contributed significantly to his scientific growth. The Director of the Garden, George Moore, administered the institution urbanely with a great tolerance for creative individuality, Anderson consulted with Moore frequently, as attested in the diaries, about professional and personal problems and always found understanding. The Curator of the Garden Herbar- ium, Jesse Greenman, austere and reserved but receptive to all serious inquiry, counseled Anderson as he explored the genetic significance of taxonomic group- ings. His debt to Greenman is acknowledged in a survey article published shortly before Anderson's death (Anderson, 1969): This review is dedicated to Dr. J. M. Greenman, former Curator of the Missouri Botanical Garden Herbarium and my colleague during much of the time these studies got underway. In answering my naive questions about the species prob- lem, he frequently quoted to me the statement "Species are but judgments." It was due to him that I became fascinated with the problem of finding factual evi- dence as to the ways in which such judgments are formed. tudents of this period who proved to be especiall)^ influential unking was a graduate student at the Garden. Robert Woodson, who had majored as an undergraduate at Washington University : Languages but shifted to Botany after taking courses with Greenman son. Only about six years Anderson's junior, and like him, finding joy in nature and the out-of-doors, Woodson established a friendship with Anderson which was more that of peer than student as they went on countless field trips together. Both intensely sensitive as well as highly intelligent, they were nevertheless in many ways unlike; and their association, which was to continue for four decades after Woodson joined the Garden staff, was, while mutually creative, often turbulent. It was perhaps only to be expected that the perfectionism which was character- istic of Woodson and the indi\ddualism which was a strone trait of Anderson 1972] FINAN — EDGAR ANDERSON, 1897-1969 333 February tensions. Anderson records one such clash in his diary Today I started in earnest to work out the taxonomy of the spongy-seeded Irises. A good part of the day was spent copying references out of Dykes and getting them smoothed out and on to a systematic arrangement of cards. Boh [Woodson] and I performed petulantly at lunch. He ended hy washing my face in my lettuce salad — a most disgusting sight. Two days later all had blown over and they were off agi Out of these trips and collaborative research was produced their monograph Woodson. 1935). More Woods 1969) ery as he has acknowledged (Anderson, In his Doctors thesis my colleague Dr. Robert E. Woodson published an inter- pretation of the phylogeny of the genus Apocynum which was decades ahead of its time. He envisaged it as a genus in which inter-specific hybridization has been so important that its evolutionary pattern is more like an anastomosing network than a branching tree. To me his ideas though stimulating and interesting, seemed rather in need of experimental confirmation by other than purely morphological criteria. After much friendly argument an experiment was planned, a simple progeny test of two common American species, Apocynum cannahinum and A. androsaemifolhim, strikingly different plants, and their putative hy- brid. It confirmed all his hypotheses, including some I had been skeptical about, . . . [It] also presents evidence for a phenomenon that I do not remember having talked, read or thought about up to that time, the various restrictions to free recombination of multiple-factor characters which operate in hybrid germ plasms. , , , It was these experiments with Apocynum which led me to examine the general restrictions to recombination in multiple factor characters and eventu- ally to describe, define and diagram introgressive hybridization. After the mid-1930's, the research interests of each became distinct and their relations, especially after Anderson became Director of the Garden in the 1950's, very strained. To the end, however, Anderson always spoke of his formcT stu- dent with affection and respect. Crucial to the development of Anderson's thought leading to his hypothesis of introgressive hybridization and to the establishment of its importance in evo- lution were the methods he \\'orked out to note and quantify character variation and associations in plants. Mildred E. Mathias, who was a student at the Garden in the late 1920*s and is now Professor of Botany and Director of the Botanical Gardens-Herbarium, University of California, Los Angeles, recalls Anderson's early measurement techniques:^ Edgar was my genetics teacher and at that time was completing the field studies on his treatment of Iris versicolor and J. virginica. The four of us in the class did many of the measurements of the flowers and night after night I carried home a quart jar of iris flowers in formaliii. From these measurements lie developed the "ideograms" and I can remember his delight at finally finding the right word ^ Letter to author from Mildrrd E, Mathias. 334 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 to describe the stylized iris flower parts. It was this same fertile mind which led to the many other pictorial methods which he devised to illustrate variation in plant populations, methods which are now so commonly used that many do not realize that it was Anderson who started it all. The "ideograms'' or ideographs were introduced in Anderson's seminal paper, "The problem of species in the northern blue flags" (Anderson, 1928). These early methods were later elaborated to his "whiskers" technique which allowed for measurements and depiction of four characters in a scatter diagram, as Ander- son (1952, p. 93-94) himself explained: •J m L By the time Dr. Woodson and I had finished monographing the American species of T rade scant ia, it was evident to both of us that hybridization between the var- ious species might well be of evolutionary importance in that group of plants. In those days there was a good deal of argument going on about this subject, much of it merely the voicing of opinions by biologists who had been studying other matters but had had some opportunity to observe hybridization or the lack of it. It seemed to me that instead of just arguing it would be better to set up some kind of a method for measuring the effects of hybridization. Postponing the fascinating question of whether or not hybridization was generally important in evolution as a whole, I began to work out ways for measuring what it was and was not doing in our spiderworts. . . . Eventually a simple and generally applicable technique was worked out. . , . One of my assistants always referred to it as "whiskers'* . . . and that term is still used informally around the laboratory. However, such sprightli- ness is not considered good form among scientists so when it was finally formally published it was described under the mouthfilling phrase of "pictorialized scatter diagrams." Fundamental, of course, to Anderson's success with these measuring tech- niques, however ingenious, was what David M. Gates, who became director of the Garden in the 1960's, called Andersons 'unique perception." As William Brown, Anderson's student and close associate for almost thi'ce decades in com research, has noted :^ One of the things I admired most about Edgar's scientific approach was his emphasis on what was biologically important. He had an excellent grasp of statis- tics and its use, yet his unique ability to search out critical traits with which to work and his emphasis upon the kinds of differences that delimit species and vari- eties, usually precluded the use of statistical methods to measure significance. . . - In Anderson s paper "The species problem in Iris' ( 1936) and the aforementioned ''Hybridization in Tradescantia III" (Anderson & Hubricht, 1938) are the fruit- ful hypotheses which are his greatest contributions to science and which guided almost all of his later research. As he summarized his work in a draft inventory of his professional interests written about 1957 and fortunately preserved among his papers: . . , . My basic professional discovery has been how to measure evolution in progress by analyzing variation between individuals so effectively that one could pre- * Letter to author from William L. Brown. 1972] FINAN— EDGAR ANDERSON. 1897-1969 335 cisely determine its source. From work on the species problem I have moved on to the classification and the origin of agriculture. This has led to studies on the importance of hybridization in evolution, particularly in a cryptic way through successive back-cross (introgression) to the original species. Gradually I learned to depart from conventional statistical methods, to analyze species differences ex- haustively at a morphological level before taking up quantitation of these differ- ences. Thus it was that during Anderson's first years at the Missouri Botanical Gar- den — ^in the 1920s — the scientific interests which became the basis for all his later major contributions were firmly formed. At the end of the decade, he re- ceived a National Research Fellowship for study in England. He w^orkcd during the academic year 1929-30 under the guidance of J. B. S. Haldane and studied cytology with C. D. Darlington at the John Innes Institution and statistics with R. A. Fisher at Rothampstead. There is recorded in his diary his first impressions of Haldane and Fisher: Oct. 16, 1929. . . . After dinner, I wrote letters to all the people suggested by Mur- rill, and a couple others stationed at Lake Baikal, begging for seed of Aquilegia. I was just finishing up on this job when J. B. Haldane came. We had a long con- ference and he was most encouraging in regard to the possibilities of my data and suggested extending them by something of the sort on Primula sinensis. Dr. East had said he was large — ^but I had failed to anticipate. He reminds me of a large St. Bernard or Newfoundland. Big and bulky, rough tweedy clothes, scriz- zly hair (such as there is) & general air of being difficult to comb. At tea Miss Fellow remarked that she wanted someone meek and industrious to score plants for her. J. B. S. H. said to apply to some Catholic father for a Catholic Botanist who needed to do penance. Nov. 5, 1929: Out to Rothamstead on the ten oVlock train to see Dr. Fisher. At the lab there had been some mLx-up and I had to cool my heels for two hours in the marble entrance hall. At length he appeared, short, slight, with greying hair, and a full beard of silky brown. All this one notices as he comes in and as c^uickly forgets as soon as talk lias started. He took me home to dinner. ... In criticizing Martin Arrotcsmith he said that S. Lewis had exaggerated the artist, in Martin the scientist — that a real scientist would have had more effect upon his assistant's habit of thought and habit of work. In the afternoon I went over my data with him and made plans to come up in February for the whole month. Until dien I shall work on Calculus and get my data into shape. He returned to the Missouri Botanical Garden in 1930 and remained there for one year when he accepted an appointment as Arborist at the Arnold Arboretum, where he remained on the staff until 1935. These were, on the whole, frustrating r ■ umber service obligations at the Arboretum his research interests. Indeed, the press of duties became so great that, as Doro- thy Anderson s diary records, he suffered severe exhaustion in the Spring of 1934. He went with his family to England in July, 1934 and he spent August and Sep- tember on a collecting trip to the Balkans. His interest in the Balkans stemmed from his belief that the U. S. climate is duplicated there more api)roximately than anyw^hore else in Europe. He brought back rooted cuttings of Balkan strains of ivy, yew, and boxwood, expressing the 336 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 hope, in a report to the U, S. Department of Agriculture, "that these strains, coming from a cHmate so similar to our own, will prove more adaptable to Amer- ican conditions than have those already imported from northern Europe." His hope was fulfilled far beyond expectations by the popular acceptance of his intro- ductions, especially the Balkan ivy, in many shady gardens in the United States. For this work in international botanical cooperation, he was awarded the Order of the Yugoslavian Crown in 1937. With his return in 1935 to the Missouri Botanical Garden, he continued, as mentioned above, his earlier work with Iris and from his studies of Apocynum enunciated his hypothesis concerning introgressive hybridization. In 1939, from studies of character recombination in the second generation of a semi-fertile cross between two very different species of flowering tobacco, he demonstrated the importance of linkage and the partial association in the progeny of all the mul- tiple factor characters going into the cross (Anderson, 1939). This study of hy- bridization of two very dissimilar species of Nicotiana coincided with the publi- cation, by his old Bussey roommate, Paul C. Mangelsdorf, and R. G. Reeves (1939), of a demonstration of another wide cross, Zea mays X Tripsacum. Their resulting hypothesis that teosinte, rather than being, as widely believed, the putative ancestor of Indian corn, was in fact a hybrid of maize and tripsacum, stimulated Anderson's interest in variation in these plants. A new focus of research thus began. For the next decade and a half, Anderson launched a series of studies toward a natural classification of maize. While he drew upon the important hypotheses and measuring techniques he had developed earlier, he knew relatively little about the com plant when he started this research. With typical lucidity and simplicity, he has described how he began his study of maize (Anderson, 1952, p. 211): . . . to understand maize one would have to know something about grasses in general. I knew next to nothing about them technically, so I began in the simplest, most fundamental ways I could put my hands on. The English morphologist, Mrs. Agnes Arber, had just published a general volume on grasses. I read it from cover to cover. It seemed so simple that I wandered at her writing so elementary a book, but as the months went on I began to realize that like a good sonnet the book had more than just words — it transmitted an attitude, and with new attitude one could look down whole new vistas of experience, and old facts took on new significance. After With microscope summer and learned how to identify all the wild grasses of a small area. I found, to my joy, that though most botanists consider grasses difficult to understand they are really quite simple. The trouble springs merely from the fact that they are specialized. Their leaves look rather like the leaves of other plants and their roots are obviously roots but the rest of the plant, including the tassels, is so very peculiar that a famiharity with ordinary plants is not much help in understanding it. With Mrs. Chase to guide me these troubles disappeared and by the end of the first sum- mer I could see that I was getting real insight into these curiously specialized plants, many of which man has used for so many thousands of years. Recognizing thus the close relation of maize to man, Anderson sought to broaden his knowledge and miderstanding of ethnobotany and historical geog- 1972] FINAN — EDGAR ANDERSON, 1897-1969 337 Edgar Anderson in 1948, in the laboratory where he accnmph'shod much of his important work on plant evolution and variation. raphy by using a Rockefeller grant in 1943 to study with Carl Sauer, the dis- tinguished geographer at the University of Cahfomia. This grant was followed by a Guggenheim award which allowed him to make a preliminaiy survey of maize in Mexico. It was typical of the man that he sought to undertake, essen- tially alone, the enormous task of a natural classification of maize in Mexico. That this goal could not be attained by any one man but ultimately by a team using the extensive resources at the disposal of the Rockcrfc^llcr Fimndiition is significant only to the extent that if Anderson had not begun the task, its com- 338 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 pletion would have been considerably delayed. His experiences in Mexico are touched upon in Plants, Man and Life. What is notable about tliem is that he manifested the same respect for and search for understanding of Indian behavior as he did for plant variation; his analysis of the logic behind an Indian garden, for example, is an ethnobotanical classic. The breadth of his concern in following years with almost all aspects of maize can be noted in the wide ranire of snecialists wth whom he collaborated: archae- ithropologists, geographers. It can articles on maize which he oroduced be as cited in the bibliographical article included with this issue of the Annals. He was aided in his com research by his informal association with the Pioneer Hi- Bred Corn Co. of Des Moines, Iowa. He spent parts of many summers at the Pioneer research center in Johnston, Iowa, where, with his former student, Wil- liam Brown, now Pioneer's vice-president for reseai^ch, he caiTied on a great part of his investigations. He summarized, on a scrap of paper — perhaps for Garden Dublic relations reauirements — the results of his research: He discovered that corn is made up of hundreds (but not thousands) of intergrad- ing races and worked out most of the methods for classifying them. He worked closely with professional hybrid com companies in demonstrating the value of these methods and discoveries in the continuing improvement of hybrid com. . , . In a larger sense his research on maize reiterated the validity, through his inter- disciplinary breadth of focus, and through his humanistic concern for showing the mutually influential relation of plants and man, of his oft-stated conviction that knowledge is all of a piece. Equaling his distinction and achievements in scientific research was his great talent in teaching. His fundamental educational philosophy was the ancient, Socratic one in which the teacher's role is maieutic. He loved to arouse and di-aw out latent creative intelligence, whether that of doctoral students or of garden club women; and his permissiveness (an excellent example is cited in Charles Heiser's essay in this issue of the Annals) must have frequently shocked col- leagues and irritated administrators. He gave particular attention to bright "undisciplinables'' — ^persons whose interest or curiosity had never been whetted enough or whose intelligence never appropriately challenged for creative achieve- ment. It seems to have mattered little to him whether a learner had much formal education or not; the formal degree-granting system he frequently characterized as "red tape" and having httle to do with real learning. It is significant in this regard that one of his most imp o M en O 2: 00 CD I CO 05 CD 4^ 342 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Emanuel Rudolph remembers another example of Anderson's generosity :^^ Andy was a person who could be embarrassingly generous. Although I was not his student, he took me under his wing as he did so many students. I remem- ber talking with him about my desire to take a particular course at the Scripps Oceanographic Listitute if I could afford to do so. He immediately insisted on giving me the money to take this summer course, I tried to refuse his offer to no avail. Fortunately for me, the course was filled and I did not have to make a decision about accepting his offer. He took special delight in working with those who loved plants but who had no specialized or professional training. While at the Arnold Arboretum in the early 1930's, he founded in Boston the Herb Society of America, "with a small group of Boston Back Bay dowagers" as he referred to them. He became the Society's president in 1935. He organized, with Mary E. Baer, a St. Louis branch in 1941. "He called us all liis herb ladies," she remembers,^^ "and dispensed his doses of herbal knowledge to his would-be herb savants." Another recollection is that of Mary A. Gamble, who joined the St. Louis group in the 1950's:^^ Dr. Anderson was simply tremendous with volunteers and amateurs. He cred- ited any reasonably intelligent and energetic person with the ability to do some pretty intricate work. . . . He kept a sharp eye on the herb garden given to the Missouri Botanical Garden by the Herb Society and it was not unusual to get a call from him saying that such and such a plant was labeled incorrectly. . . . He was pragmatic and inventive. I remember one spring, when as always, our savory and hyssop plants reached that point in their growth when they look like identical twins. We took one of each plant to tvvo botanists; the first sent for herbarium sheets; Dr. Anderson bit first into one plant and then the other. "This is winter savory," he said, handing us the one. . . . There were times when he was too busy to attend a meeting but he invariably came down the steps in time to join us in the kitchen back of the meeting room. He liked unusual foods and was an enthusiastic participant in the Herb Society's regular food tasting and testing session held in the kitchen after the regular meetings. He could always identify every herb in every dish. Method tliereafter. turned while in Graduate School to Quakerism, a faith Stefferud. 1969 Q I began to attend Friends Meetings when I was a graduate student at Harvard right after World War I. Since my lodgings were at Jamaica Plain, I attended the Dorchester Meeting not far away. It had a pastor and a piano but long periods of silence. I was greatly moved by the long group silences with which the Meeting began and by the handshaking and kindly interest of the old people who sat next to us. One Sunday I persuaded a fellow graduate student to accompany me. When meeting was over, the old lady nearest to him shook his hand and said, "Is tliee a Friend?" He cordially answered, "Not exacdy; just a room-mate." I joined the Cambridge Meeting shortly before it moved to the Brattle Street location. . . . n Letter to author from Emanuel Rudolph. ^- Letter of Mary E. Baer to author. ^^ Letter of Mary A. Gamble to author. 1972] FINAN— EDGAR ANDERSON, 1897-1969 343 C4 His faith was such an integral part of his life that he seems to have been spared those tensions in dealing with the trans-rational wliich afflict many intellectuals. I once attended with him the Friends Meeting in Washington and saw him gen- uinely overcome with emotion when a member reported a dying son's last words ; God is love." He told me afterwards, "You can't overestimate the power of love; it really makes the world go round/' At the same time, he could be coolly de- tached. In reply to a question I later posed concerning the role of religion, he said quite matter-of-factly, "It helps us to endure the disintegration of the body." In late 1959, during a visit to Chicago, he had an illuminating religious ex- perience which he described later to his daughter as similar to that of the con- version of St. Paul. Mrs. Anderson affirmed in her diary the impact of the experi- ence upon him: Nov 30, 1959. Edgar so very exhilarated from his trip to Chicago he really fright- ens me. He talks continually, mostly ahout "his religious experience." He's one of those rare geniuses — a twice born man. He has suddenly hecome more inte- grated. Whatever the significance of this experience his religion, in the broadest and most meaningful sense, was his firm belief in an orderly universe, his uncom- promising commifanent to intellectual honesty, and a hfestyle that paid quiet but clear respect to the Sermon on the Mount. An article hke this would be incomplete without some mention of Anderson^s distinguished prose style. He seemed to find great creative fulfillment in writing, and the volume of his output was so large as to give some credence to the sug- gestion of colleagues that as a writer he was "compulsive." His writings — even the most technical — are without exception models of clarity and readability. Despite an appearance of i-eady facility, he rewrote his material frequendy until his pieces attaiaed that succinctness and personally-informed simphcity which are as real a mark of his genius as his scientific achievements. Some of his ''re- writing" took place in his mind, while walking or driving, as he tested the ap- propriateness of words against his keen ear. He very consciously selected words of Anglo-Saxon derivation over those from the Latin, and he delighted in em- phasizing to students the few words of Latin derivation in the Lord's Prayer and the oft-quoted psalms of the King James Bible. I once asked him who in- fluenced liim most in his writing style and lie replied, "It's the utter transparency of Hazlitt that has been my model and should be a model for us all." TTie Mis- souri Botanical Garden Bulletin^ intended for a popular readership, became an important vehicle for many essays which ailiculate his extraordinaiy perceptions. Beginning the 1950's, he prepared several pieces for Landscape which also show his unusual ability to communicate to a \vider audience. The last decade and a half of Anderson's life were years of mixed blessings. When he assumed in 1954 the directorshio of the Missouri Botanical Garden. confronted the enormous postp decades. While seeking: to turn den in new directions, he found full-time administration frvistrating and cre- rcsumin lany important he National Academy 344 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 close to liis heart was a letter of bibute sent liim on the occasion by a group of wrote We craftsman of excellence, and we hope that we carry along with us, from our years with you, something of your great devotion to the search for truth, your intellectual honesty, magnanimity and very warm human understanding. ■Wallace Medal Linnean Society, London, and a Golden Jubilee Award of Merit from the Botan- ical Society of America. In the last yeai's, he suffered severe illness, but the "grandem- of the man," to use Emmet Layton's words, never faltered. He left as his legacy not only many fundamental conti-ibutions to our understanding of plant evolution but a whole veneration of students who because of him are lookin^r at the world about them his oath, seeking to teach others do the same.^^ Literature Cited Anderson, E. 1928. The problem of species in the northern blue flags, Iris versicolor L. and Iris virginica L. Ann. Missouri Bot. Card. 15: 241-332. . 1936. The species problem in Iris. Ann. Missouri Bot. Card. 23: 457-509. . 1939. Recombination in species crosses. Genetics 24: 668-698. . 1940. Edward Murray East (1979-1938). Proc. Amer. Acad. Arts 74: 117-118, — . 1952. Plants, Man and Life. Little, Brown & Co., Boston. . 1956. Natural history, statistics, and applied mathematics. Amer. Jour. Bot. 43: 882-889. . 1969. Experimental studies of the species concept. Ann. Missouri Bot. Card. 55: 179-192. ^* Letters were received from the following persons with information concerning Edgar Anderson. All correspondence is being deposited in the Missouri Botanical Garden. Henry N. Andrews, University of Connecticut; Mary E. Baer, St. Louis, Missouri; Clar- ence Barbre, Webster Groves, Missouri; Robert W. Baxter, University of Kansas, Lawrence; William L. Brown, Pioneer Hi-Bred Corn Company, Des Moines, Iowa; E. G. Cherbonnier, St. Louis; Mary Chomeau, St. Louis; Robert B. Clark, Department of Parks, County of Mon- roe, New York; Hugh C. Cutler, Missouri Botanical Garden, St. Louis; Gilbert S. Daniels, Hunt Botanical Library, Carnegie-Mellon University; Donald N. Duvick, Pioneer Hi-Bred Corn Company, Johnston, Iowa; F. Ehrendorfer, University of Vienna; Mary A. Gamble, St. Louis; Verne Grant, University of Texas, Austin; Charles B. Heiser, Jr., Indiana University, Bloomington; William M. Hiescy, Camino, California; Nell Homer, St. Louis; Harold J. Kidd, Pioneer Hi-Bred Corn Company, Plainview, Texas; Emmet Layton, University of Wisconsin, Green Bay; Mildred E. Mathias, University of California, Los Angeles; Frank L. Mercer, St. Louis College of Pharmacy; Lily N. Perry, Arnold Arboretum, Harvard University; George H. Pring, former Superintendent, Missouri Botanical Garden; Howard C. Reynolds, Fort Hays Kansas State College, Hays, Kansas; David J. Rogers, Gulf Universities Research Consortium, Bay St. Louis, Mississippi; Reed C. Rollins, Har\'ard University; Emanuel Rudolph, Ohio State University; Robert W. Schery, The Lawn Institute, Marysville, Ohio; Wellington F. Scott, Jr., St. Louis; Russell J. Seibert, Longwood Gardens, Kennett Square, Pennsylvania; Julian Steyennark, Instituto Botanico, Caracas, Venezuela; Yutaka Suda, University of Iwate, Ueda, Morioka, Japan; Thomas W. Whitaker, U. S. Horticultural Field Station, La JoUa, California; Louis O. Williams, Field Museum of Natural History, Chicago. Dr. Peter H. Raven and his staff at the Missouri Botanical Garden put at my disposal important archi\al material relating to Dr. Anderson. Their help in all aspects of the prepara- tion of this article was unfaiHng. Miss Phoebe Anderson was also especially helpful to me in my work. 1972] FINAN— EDGAR ANDERSON, 1897-1969 345 — & Robert E. Woodson. 1935. The species of Tradescantia indigenous to the United States. Contr. Arnold Arbor. 9: 1-132. — & Leslie Hxjbricht. 1938. Hybridization in Tradescantia III. The evidence for in- trogressive hybridization. Ainer. Jour. Bot. 25: 396-402. Mangelsdorf, p. C. & R. G. Reeves. 1939. The origin of Indian corn and its relatives. Texas Agric. Exp. Stat. Bull. 574. Roe, Anne. 1952. The making of a scientist. Dodd, Mead & Co. New York. Stefferud, a. 1969, I rejoice in what he af finned. Friends* Journal for Aug. 1, 1969, p. 419. THE PUBLICATIONS OF EDGAR ANDERSON Erna R. Eisendrath^ Although Edgar Anderson was bom in New Yorlc State, November 9, 1897, a good deal of his life was spent in the Middle West. He graduated, in 1918, from Michigan State University, where liis father was Professor of Dairy Hus- bandry, with a B.S. degree. Four years later he had received Master s and Doctor's degrees from Harvard, where he had served as an assistant in genetics. Immedi- ately after completing his studies, Dr. Anderson came to St. Louis, with the titles Geneticist at the Missouri Botanical Garden, Assistant Professor at the Henry Shaw School of Botany, Washington University. In 1923 Dr. Anderson married Dorothy Moore, who had a degree in Botany from Wellesley College. Dr. and Mrs. Anderson remained in St Louis until 1931; during this period Dr. Anderson was also Director of the School of Gardening run by the Missouri Botanical Garden, and travelled a good deal in connection with his research on the genus Iris. During the year 1929-30 he served, also, as a Fellow of the National Research Council (British), working at the John Inness Horticultural Institute, near London. During the last year of this first stay in St. Louis, Dr. Anderson was promoted to an Associate Professorship at the University. There are relatively few bibliographical listings from this period, probably because Dr. Anderson signed only a small number of the articles he wrote during that time for the Missouri Botanical Garden Bulletin; although the style of a number of unlisted Bulletin articles is very like his, there is no way of corroborating his authorship; consequently I have included only those to which he did at least append his initials.^ As is quite clear from the listings. Dr. Andersons activities shifted back to Harvard, in 1931, when he was appointed Arborist at the Arnold Arboretum. Four years later, however, he returned to St. Louis, appointed once again Geneticist at the Botanical Garden and, this time, a full Professor at the Uni- versity. In the same year, 1935, appeared one of his most important publications. In conjunction with Dr. Robert E. Woodson, Jr., Dr. Anderson had carried out extensive research on the American Tradescantias, much of it cytological. As a result, their monograph on this group of plants led him on to further experimental study of the role of interspecific hybridization in the evolution of both weeds and cultivated plants. During the first year after his return to St, Louis Dr. Anderson was elected to the presidency of the Herb Society of America, which he had helped establish. The following year, 1936, he was awarded the Order of the Yugoslavian Crown, ^ Department of Biology, Washington University, St. Louis, Missouri 63130. ^ I have, on the other hand, included some signed materials so slight as to raise the question whether they should not have been forgotten. I have Usted them primarily because they demonstate, often far more vividly than does his serious work, a very important aspect of Dr. Anderson's many-faceted personality. Ann. Missouri Box. Card, 59: 346-361, 1972] EISENDRATH— EDGAR ANDERSON: BOTANIST 347 in recognition of his introduction to this country of the hardy, vigorous Bul- garian ivy^ since 1934 increasingly popular in shady gardens. Further honors came his way as well: in 1937, the distinguished Engelmann Professorship at Washington University; in 1941, membership on the editorial board of the American Journal of Botany, And, in the same year, Dr. Anderson participated with Dr. Ernst Mayr in presenting the Jessup lectures at Cohimbia University. At about the same time, Dr. Anderson began to narrow his interest in genetics to a study of Zea mays and its relatives; this interest as well as his close associ- ation with Dr. Hugh Cutler in working out the problems of maize genetics, over a long period of years, are both evidenced by many entries in the bibli- ography. In 1943, a Rockefeller grant made it possible for Dr. Anderson to work with Dr. Carl Saner at the University of California, Berkeley; a Guggen- heim Fellowship permitted further study of the races of maize in the southwestern States and Mexico, In June of that year Dr. Anderson participated, as a member of its Committee on Common Problems of Genetics and Paleontology, in deliberations held by the National Research Council at Berkeley. His interest in Indian corn was further broadened when the Pioneer Hi-Bred Com Company offered him an opportunity to experiment with corn breeding on a scale pro- hibitive at the Missouri Botanical Garden. This association continued until F his death. The many references to various aspects of his maize studies during subse- quent years reflect not only Dr. Anderson's concentration in this area, but also the geographical spread of his visits elsewhere, to study living plants and herbaria. Dr. Anderson remained widely interested in all sorts of other subjects, as is made abundantly clear from the variety of topics listed in the bibliography. For still further evidence of the breadth of his study and activity one can find in the Annual Reports of the Garden references to a great number of lectures, travel in many parts of the world, and participation in institutes of various sorts. These include, for instance, the development of an experimental garden, devoted to growing tropical strains of important economic plants, in Honduras; a visit to India as the American delegate to a UNESCO s>Tiiposium; the winning of an alumni award, for distinguished service, from his college — all in 1951. In the following year Dr. Anderson was appointed Assistant Director of the Missouri Botanical Garden, but his important publications during that year prove that the added administrative duties in no way slowed his other work. He even managed to teach at Stanford University, during the spring quarter, and to serve, in 1952, as President of the Botanical Society of America. During the years 1954-56, Dr. Anderson was Director of the Garden, but again his productive activity was not slowed down. In the first of these years he was also elected to the National Academy of Sciences. As a member also of the American Academy of Arts and Sciences, he was one of the very few scientists ever to achieve both honors. In 1957 Dr. Anderson gave over his heavy administrative duties to assume the title at the Garden, Curator of Useful Plants. This title, however, behed the continued breadth of his activities: lie served as Visiting Fellow in Mathe- 348 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 matics at Princeton, as Visiting Fellow at the Center for Advanced Studies in the Behavioral Sciences at Stanford, as Senior Fellow of the Guggenheim Founda- tion, and as Visiting Fellow at the Center for Advanced Studies at Wesleyan, as well as President of the Society for the Study of Evolution, in 1959. It is no wonder that, at about the same time, Dr. Anderson received a number of dis- tinguished awards; a Certificate of Merit from the Botanical Society of America; the silver Darwin- Wallace medal of the Linnean Society (London); and gold medals from both the Men's Garden Clubs of America and the National Council of State Garden Clubs. In 1965-66 Dr. Anderson demonstrated his continued interest in the origin and evolution of cultivated plants by accepting the Presidency of the Society for Economic Botany; at the same time he continued to edit, as he had since 1963, the Garden^s popular publication, the Bulletin. When Edgar Anderson retired, in 1967, after a number of bouts of serious illness, he retained the University title, Engelmann Professor of Botany, emeritus; but at the Garden he was hsted until the time of his death (June 18, 1969) simply as **Botanist.'' This was at his own request because, as he wrote to the editor of The Herb Grower, "I am one, never wanted to be anything else, and (the title) covers all of my activities." Note: In the bibliography titles followed by an asterisk ( *) were signed "E.A. 5> 1924 Studies on self-sterility VI. The genetic basis of cross-sterility in Nicotiana. Genetics 9 13^0. January. 1927 Iris versicolor in northern Ontario. Bull. Amer. Iris Soc. 22: 47-48. January. 1928 The problem of species in the northern blue flags, Iris versicolor L. and Iris virginica L Ann. Missouri Bot. Gard. 15: 241-332. September. 1929 Variation in Aster anomalus. Ann. Missouri Bot. Gard. 16: 129-144. April. 1931 (Witli De Winton, Dorothea) The genetic analysis of an unusual relationship between self- sterility and self -fertility in Nicotiana. Ann. Missouri Bot. Gard. 18: 97-116. February. Interna] factors affecting discontinuity between species. Amer. Naturalist 65: 144-148. March-April. (Abstracted: Brooks & Chipp, editors, Report of Proceedings, 5th Inter- national Botanical Congress, Cambridge, 1930, 225 with additional note on 231.) Termites and ants. Missouri Bot. Gard. Bull. 19: 86-87. May. Henry Shawns Alma Mater and Peter Collinson. Ibid. 19: 88-94. May. (With Judd, W. H.) The art of budding. Bull. Popular Inform. Arnold Arbor. Ser. 3, 5: 49-52. August 21. The chromosome complements of Allium stellatum and Nothoscordum bivalve. Ann. Missouri Bot. Gard. 18: 465-468. September. (With Schafer, Brenhilda) Species hybrids in Aquilegia. Ann. Bot. (London) 45: 639- 646, October, Hybrid trees. Bull. Popular Inform. Arnold Arbor. Ser. 3, 5: 65-68. November 16. J-- .n-L -h-. Utf-.r- 1972] EISENDRATH— EDGAR ANDERSON: BOTANIST 349 427-429. June. 1932 Character recombination as a genetic tool. Proceedings 6th International Congress of Genetics, Ithaca, 1932. 2: 2. Baltic ivy. Bull, Popular Inform. Arnold Arbor. Ser. 3, 6: 1-7. March 29. (With Diehl, D. G.) Contributions to the Tradescantia problem. J. Arnold Arbor. 13: 213- 231. April. How big is a cell? Science 75: 359. April 1. Character recombination in Drosophila. Proc. Natl. Acad. U.S.A. 18: Fire! Bull. Popular Inform. Arnold Arbor. Ser. 3, 6: 21-24. June 4. (With Judd, W. H.) Rosa nigosa and its hybrids. Ibid. Ser. 3, 6: 29-35. June 24. (With Ames, Oliver) Botanizing from an airplane. Ibid. Ser. 3, 6: 37-44. August 15. Growing oranges in Boston. Ibid. Ser. 3, 6: 45-47. November 5. (With Judd, W. H.) The propagation of woody plants. Ibid. Ser. 3, 6: 53-56. December 28. 1933 (Witl:i Sax, Karl) Segmental interchange in chromosomes of Tradescantia. Genetics 18; 53-67. January. Trumpet-creepers. Bull. Popular Inform. Arnold Arbor. Ser. 4, 1: 1-5. January 15. (With Palmer, Ernest J.) Tlie river birch. Ibid. Ser. 4, 1: 6-8. January 15. Color variation in a Missouri colony of Hepatica aciitiloba. Rhodora 35: 66-67. February. (With Schafer, Brenhilda) Vicinism in Aqtiilegia vulgaris. Amer. Naturalist 67: 190-192. March-April. Variation in flower color in Hamamelis vernalis. J. Arnold Arbor. 14: 253-257. April. Shrubs attractive in late winter and early spring. Bull. Popular Inform. Arnold Arbor. Ser. 4, 1: 17-20. April 18. Leatherwood. Ibid. Ser. 4, 1: 25-28. April 29, Plants of current interest Ibid. Ser. 4, 1: 28, Apiil 29; 1: 29-32, May 12; I: 37, May 26; 1: 51-52, June 23. The distribution of Iris versicolor in relation to the post-glacial Great Lakes. Rhodora 35: 154- 160. May. Basswood bark and its use by the Indians. Bull. Popular Inform. Arnold Arbor. Ser. 4, 1: 33-37. May 26. Wanted. Ibid. Ser. 4, 1: 37. May 26. Jabberwocky. Ibid. Ser. 4, 1: 41-44, June 1. A simple device for exhibiting flowering shrubs. Ibid. Ser. 4, 1: 50-51. June 23, Hydrangea petiolaris and Schizophragma hydrangeoides. Ibid. Ser. 4, 1: 53-56. July 24. (With Abbe, Lucy B.) A comparative anatomical study of a mutant Aquilegia. Amer. Naturalist 67: 380-384. July-August. A convenient color chart for geneticists. Science 78: 150-151, August 18. Pterocarya Rehderiana. Bull. Popular Inform. Arnold Arbor, Ser. 4, 1: 57-59. September 30. (With Judd, W. H.) Fothergilla major. Ibid, Ser. 4, 1: 61-64. December 29. 1934 (With Whitaker, Tliomas W-) Spcciation in Uvularia. J. Arnold Arbor. 15: 28-42. January. (With Abbe, Ernst C.) A quantitative comparison of specific and generic differences in the Betulaceae. Ibid. 15: 43^9. January. (With Sax, Karl) Interlocking of bivalent chromosomes in Tradescantia. Genetics 19: 157- 166. March. Origin of the angiosperms. Nature 133: 462. March 24. Hamamelis vernalis Sarg. The Ozark witch-hazel. Bull. Popular Inform. Arnold Arbor. Ser. 4,2: 1-4. April 28. Hardy flowering cherries. Ibid. Ser. 4, 2: 5-8. May 5. Hardy forsytliias. Ibid. Ser. 4, 2; 9-14. May 14. (With Schmitt, Louis Victor) Plants of current interest. Ibid. Ser. 4, 2: 15-16, May 14. The genus Akebia. Ibid. Ser. 4, 2: 17-20. May 22. (With Sax, Karl) A cytological analysis of self-sterility in Tradescantia, Bot. Gaz. (Craw- fordsville) 95: 609-621. June. Rhododendrons. Bull. Popular Inform. Arnold Arbor. Ser. 4, 2: 21-24. July 9. 350 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 1935 Nature's bags of scents. Herbarist 1: 5-6. An endemic Sophora from Rumania. J. Arnold Arbor. 16: 76-80. January 25. (With Palmer, Ernest J.) Uvularia perfoliata in Louisiana. Rhodora 37: 58-59. February Science as a storehouse. Horticulture (Boston) 13: 124. March 15. A visit to the home of the lilac. Bull. Popular Inform. Arnold Arbor. Ser. 4, 3: 1-4. March 22, (With Sax, Karl) Chromosome numbers in the Hamamelidaceae and tlieir phylogenetic sig- nificance. J. Arnold Arbor. 16: 210-215. April 24. The last week in April. Bull. Popular Inform. Arnold Arbor. Ser. 4, 3: 3-8. April 27. Gametic elimination in crosses between self -sterile species. Amer. Naturalist 69: 282-283. May-June. The flowering quinces. Bull. Popular Inform. Arnold Arbor. Ser. 4, 3; 9-12. May 27. With the last week in May. Ibid. Ser, 4, 3: 13-16. June 4, (With Rehder, Alfred) New hybrids from the Arnold Arboretum, J. Arnold Arbor. 16: 358- 363. July 10. After months of abundant rain. Bull. Popular Inform. Arnold Arbor. Ser. 4, 3: 17-20. August 7. (With Woodson, Robert E.) The species of Tradescantia indigenous to the United States. Contr. Arnold Arbor. 9: 1-132. August 30. Brightening up the rose garden in late summer.* Missouri Bot. Gard. Bull. 23: 125-127. October. (Witli De Winton, Dorothea) The genetics of Primula sinensis. IV. Indications as to the ontogenetic relationship of leaf and inflorescence. Ann. Bot. (London) 49: 671-687. October. The irises of the Gaspe peninsula. Bull. Amer. Iris Soc. 59: 2-5. October. The oriental crab-apples. Bull. Popular Inform. Arnold Arbor. Ser. 4, 3: 21-24. October 7, Sporting of the chrysanthemum, Missouri Bot. Gard. Bull, 23: 161-163. December. (With Turrill, W. B.) Biometrical studies on herbariimi material. Nature 136: 986-987. December 21. 1936 The taragons, cultivated and wild. Herbarist 2: 9-11. A morphological comparison of triploid and tetraploid interspecific hybrids in Tradescantia. Genetics 21: 61-65. January. Are flower shows really helpful? House Beautiful 78: 89. February. (With Sax, Karl) A cytological monograph of the American species of Tradescantia, Bot- Gaz. ( Crawfordsville ) 97: 433^76. March. Viburnum fragrans, Gard. Chron. Ser. 3. 99: 197. March 28. An experimental study of hybridization in the genus Apocynum. Ann. Missouri Bot. Gard. 23: 159-168. April. Color variation in eastern Nortli American flowers as exemplified by Hepatica acutiloha, Rhodora 38: 301-304. September. Damage to forest trees at die Arboretum.* Missouri Bot. Gard. Bull. 24: 163-169. September. The species problem in Iris. Ann. Missouri Bot. Gard. 23: 457-509. September. Hybridization in American Tradescantias. Ibid. 23: 511-525. September. (With P., G. H.) Ground covers for St. Louis gardens.* Missouri Bot. Gard Bull. 24: 186- 193. October. An American pedigree for woolly hair. J. Heredit>^ 27: 444. November. (Witli Wlielden, C. M., Jr.) Studies in the genus Fraxinus II. Data on the flowering and fruiting habits of three American species of ash of possible economic importance in the production of rapid-growing forest trees. Ibid. 27: 473-474. December. 1937 Angelica archangelica, the great umbellifer, Herbarist 3: 21-24. Hybridization in Tradescantia. (Abstracted: Genetics 22: 183. January.) Some general features of the Ozark flora. Bull. Gard. Glub Amer. Ser. 6, 1: 20-22. January. Gardenworthy plants of the Ozarks. Ibid. Ser. 6, 2: 41-42. March. Sassafras leaves: a culinary possibility.* Missouri Bot. Gard. Bull 25: 58-60 March (Widi P., G. H.) Pigmy water-lilies.* Ibid. 25: 67-68. April. Supra-specific \ariation in nature and in classification from the viewpoint of botany Amer Naturalist 71: 223-235. May-June. The wild flower reservation.* Missouri Bot. Gard. Bull. 25: 79-82. May. „^ ^ -^'-^ji: _ ^^- ^ ^ --J- - x_\^ -A 1972] EISENDRATH— EDGAR ANDERSON: BOTANIST 35]^ Notes on the history of botanical illustration.* Ibid. 25: 83-87. May. (Reprinted: Card. Gossip 18: 3 & 8. January, 1943.) Cytology in its relation to taxonomy. Bot. Rev. (Lancaster) 3: 335-350. July. A white blazing star.* Missouri Bot. Gard. Bull. 25: 22. September. The development of the wild-flower gardens at the Arboretum. Ibid. 25: 129-139. October. 1938 Mints and microscopes. Herbarist4: 15—16. (Witli Turrill, W. B-) Statistical studies on two populations of Fraxinus. New Phytol. 37: 160-172. April. (With Hubrichtj Leslie) Hybridization in Tradescantia III. The evidence for introgressive hybridization. Amer. J. Bot. 25: 396-402. June. (With Hubricht, Leslie) The American sugar maples I. Phylogenetic relationships, as deduced from a study of leaf variation. Bot. Gaz. ( Crawf ordsville ) 100; 312-323. December. 1939 The hindrance to gene recombination imposed by linkage: an estimate of its total magni- tude. Amer. Naturalist 73: 185-188. March-April. Has this plant become extinct in Missouri?* Missouri Bot. Gard. Bull. 27: 102. April. Creating a flowering meadow.* Ibid. 27: 102-106. April. A classification of weeds and weed-like plants. Science 89: 364-365. April 21. John Kellogg, 1862-1939.* Missouri Bot. Gard. Bull. 27: 127-129. June. Recombination in species crosses. Genetics 24: 668-698. September. (With Ownbey, Ruth Peck) The genetic coefficients of s^eciiic difference. Ann. Missouri Bot. Gard. 26: 325-348. November. The rediscovery of Prenanthes crepidinea* Missouri Bot. Gard. Bull. 27: 189-191. November. (Witli Lehmann, John S.) An old-fashioned rose. Ibid. 27: 202-203, December. Watch out, pickle wonns! Real Gard. 2: 51-53. December. 1940 A cytological, taxonomic and genetic monograph of the genus Tripsacum with reference to its allies Zea and Eiichlaena. I A renort from recinient of irrant from the Penrose Yearb. Amer. Philos. Soc. 1940: 106-107. Sweet corn with a college education.* Missouri Bot. Gard. BiJl. 28: 73-75. April. Notes from the Missouri Botanical Garden Arboretum.* Ibid. 28: 106-107, May; 120-121, June. r A survey of modern opinion. In The concept of the genus. Bull. Torrey Bot. Club 67; 363- 369. May. William L.) Lawns and lawn grasses for St. Louis. Missouri Bot. Gard. Bull. 28: 125-170. September. Two Ozark wild flowers for the fall garden. Ibid. 28: 180-184. October. Edward Murray East (1879-1938). Proc. Amer. Acad. Arts 74; 117-118. November. (With Hubricht, Leslie) A method for describing and comparing blooming-seasons. Bull. Torrey Bot. Club 67: 639-648. November, (Witli Brown, William L.) Foa cuspidata of tire Appalachian plateau and Atlantic coastal 124-125 1941 228-^229 Extracts from a letter from Edgar Anderson. Herbarist 7: 38. Reports from cooperators. Maize Genetics Cooperation News Letter 16; 22-24. (Review of Huxley, editor, "The New Systematics." Amer. Midi. Naturalist 25; January. ) (Review of Goldschmidt, "The Material Basis of Evolution." Chron. Bot. 6: 238-239. February 10.) Binary variation in Tradescantia hracteata. Ann. Missouri Bot. Gard. 28: 147-163. April. The fragrant sumac. Garden Life, St. Louis Hort. Soc. 14: 51. July. (With Erickson, Ralph O.) Antithetical dominance in North American maize. Proc. Natl, Acad. U.S.A. 27: 436-440. September. 352 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 (With Cutler, Hugh C.) A preliminary survey of the genus Tripsacum. Ann. Missouri Bot. Gard. 28: 249-269. September. (Reprinted in translation: Revista Argent. Agron. 9: 249-250. ) The technique and use of mass collections in plant taxonomy. Ibid. 28: 287-292. September. (With Beihuann, A. P. & Clark, Robert B.) Evergreens for St. Louis. Missouri Bot. Card Bull. 29; 223-248. November. (With Hubricht, Leslie) Vicinism in Tradescantia. Amer. J. Bot. 28: 957. December. 1942 Naturalizing Narcissi in Missouri. Herbertia 9: 195-199. Reports from cooperators. Maize Genetics Cooperation News Letter 17: 17-19. The Missouri Botanical Garden Arboretum, a laboratory for the week-end gardener. Bull. Gard. Club Amer. Set. 7, 19: 94-98. January. (Reprinted: Missouri Bot. Gard. Bull. 30: 73-77.) (Review of Willis, "The Course of Evolution. Reflections on evolution." J. Heredity 33: 67-68. February.) Vegetable gardening in St. Louis. Garden Life, St. Louis Hort. Soc. 15: 13. February. Narcissi at the Missouri Botanical Garden. Missouri Bot. Gard. BvilL 30: 63-73. March. (With Cutler, Hugh C.) Races of Zea Mays I. Their recognition and classification. Ann. Missouri Bot. Gard. 29: 69-88. April. Veronica hederaefolia in Missouri.* Missouri Bot, Gard. Bull. 30: 110. May. (With Blanchard, Frederick D.) Prehistoric maize from Caiion del Muerto. Amer. J. Bot. 29: 832-835. December. 1943 Two varieties of Rosa aJha. Amer. Rose Annual. 28: 12-14. Races of Zea mays 11. A general survey of the problem. Acta Amer. 1: 58-68. January- March, The seeds of Tradescantia micrantha. Ann. Missouri Bot. Gard. 30: 69. February. Vegetable gardening in St. Louis. Missouri Bot. Gard. Bull. 31: 33-34. February. Two fragrant herbs. Ibid. 31: 73-75. March. (With Hubricht, Leslie) Histological basis of a specific difference in leaf texture. Amer. Naturalist 77: 285-287. May-June. Some Indian uses of corn. Bull. Gard. Club Amer. Ser. 8, 16: 18-19, October. The Tlacopintli. Tlalocan 1: 159-160. November. (With Kelly, Isabel) Sweet com in Jalisco, Ann. Missouri Bot. Gard. 30: 405-412. No- vember. (Witli Barlow, R. H.) The maize tribute of Moctezuma's empire. Ibid. 30: 413-420. No- vember . A variety of maize from the Rio Loa. Ibid. 30: 469-474. November. Mass collections. Chron. Bot. 7: 378—380. December. 1944 Reports from cooperators. Maize Genetics Cooperation News Letter 19: 32-33. "Maiz del pais" in California. J. Calif. Hort. Soc. 5: 1-6. April, Notes on variation in Tithonia tubaeformis, Ann. Missouri Bot. Gard. 31: 239-240. September. (With Schrcgardus, Dorothy) A method for recording and analyzing variations of internode pattern. Ibid. 31: 241-247. September. Maiz reventador. Ibid. 31: 301-315. November. Cytological observations on Tripmcum dactyloides. Ibid. 31: 317-323. November. Homologies of the ear and tassel in Zea mays. Ibid. 31: 325-343. November. Two collections of prehistoric com tassels from southern Utah. Ibid. 31: 345-353. November. The sources of effective germ-plasm in hybrid maize. Ibid. 31: 355-36L November. (Review of Cain, "Plant Geography." The bio-systematic no-mans land. J. Heredity 35: 349-351. November.) 1945 The Arizona." Amerind Foundation ^e in the New World. In Wilsoi Chanter 2. 27-42. W ?> w--—±,_, ' TT^'--Z-'"'-t^ ■-■■ -^ — - — T^ — 1972] EISENDRATH— EDGAR ANDERSON: BOTANIST 353 Pinole. Herbarist 11: 33-39. (With Snyder, Ray E.) Reports from cooperators. Maize Genetics Cooperation News Letter 20: 19-21. Two Balkan ivies for St. Louis gardens. Missouri Bot. Gard. Bull. 33: 59-63. March. An early- flowering Japanese cherry. Ibid. 33:113. May. The hardy orange. Ibid. 117-118. May. Rosa alba in St. Louis. Ibid. 33: 130-132. June. An early-blooming dark red day-lily. Ibid. 33: 132-133. June. What is Zea mays? A report o£ progress. Chron. Bot. 9: 88-92. Autumn. (With Carter, George F. ) A preliminary survey of maize in the southwestern United States, Ann. Missouri Bot. Gard. 32: 297-322. September. (With Finan, John Jay) Maize in tlie Yanhuitlan Codex. Ibid. 32: 361-368. September. Can we grow snow trillium in Missouri? Missouri Bot. Gard. Bull. 33: 227. December. 1946 Reports from cooperators. Maize Genetics Cooperation News Letter 21: 23-25. Report on maize from Cheran. In Beals, Ralph W., "Cheran: a Sierra Tarascan village." Smithsonian Institution Institute of Social ^\nthropology Publication 2: 219-223. What is a tetraploid? Missouri Bot. Gard. Bull. 34: 43-48. February. Who was who? Ibid. 34: 59. February. Daffodils for the garden. Ibid. 34: 63-66. March. The blue of the blue ash. Ibid. 34: 99-100. April. Maize in Mexico. A preliminary survey, Ann. Missouri Bot. Gard. 33: 147-247. April. (With Homback, Earl) A genetical analysis of pink daffodils. A preliminary attempt. J. Calif. Hort. Soc. 7: 334-344, Spring. More about Balkan ivies. Missouri Bot. Gard. Bull. 34: 142-143. June. Narcissus ''Moonshine" as a variety for forcing. Ibid. 34: 210-211. October. 1947 Corn before Columbus. Pioneer Hi-Bred Corn Co. 24 pgs. Correlated development of the tassel and the ear. In "Growth and development of tlie com plant." American Seed Trade Association, 25-31. ^^_ ^^ I ' Red seeds for butter coloring. Herbarist 13: 29-30. (With Brown, William L.) The northern flint corns. Ann. Missouri Bot. Gard, 34: 1-28. February. Snow trilHum, Missouri Bot. Gard. Bull. 35: 103. April. Plants for a St. Louis garden. Ibid. 35: 113-117. May. (Reprinted: Ibid. 42: 18-23. January, 1954.) Better Forsythias for St. Louis. Ibid. 35: 125. May. Popcorn. Natural History 51: 227-230. May. Redbud, cow pastures and beer bottles.* Missouri Bot. Gard. Bull. 35: 133. June. The trumpet-creeper in Missouri.* Ibid. 35: 141. June. Natural gardens of Amsonia* Ibid. 35; 145. Jime. The flower witli two smells.* Ibid. 35: 148. June. Narcissus gracilis * Ibid. 35: 149. June. Missouri gravel bars.* Ibid. 35: 166. September. You can enjoy daffodils all spring. House Beautiful 57; 117-119; 199-202. September. Field Studies of Guatemalan maize. Ann. Missouri Bot. Gard, 34: 433-467. November. V 1948 Further information on the genetics of pink daffodils, (a) Foreword. Herbertia 15: 83. (With Brown William L.) The history of corn belt maize in relation to the inbreds derived 5-6. January. r Plants." Amer from it. (Abstracted: Genetics 33: 98-99. January.) Missouri Asters for Missouri gardens. Garden Life, St. Louis Hort. Soc. 20: ( Review of Hanes & Hanes, "Flora of Kalamazoo County, Michigan, Vascular Plants." Midi. Naturalist 39: 250. January.) Hybridization of the habitat. Evolution 2: 1-9. March. Introduction to Agramont, et al.. The red cedar. Missouri Bot. Gard. Bull. 36: 86. May. 354 THE [Vol. 59 What will corn be like 50 years from now? Capper*s Farmer 59; 9; 81. May. Report on the ears of maize from Cottonwood Cave. In Hurst, C. T., "The Cottonwood Expedition 1947." Southwestern Lore 14: 17-18. June. (Review of Babcock, "Genus Crepis!* The definitive hawks-beard. J. Heredity 39: 243-244. August. ) (With Brown, William L.) The soudiem dent corns. Ann. Missouri Bot. Card. 35: 255-268. September. Racial identity in the corn from Castle Park. In Burgh, Robert F. & Scoggin, Charles R., "The archeology of Castle Park Dinosaur National Monument." University of Colorado Series in Anthropology, No. 2, Appendix 1, 91-92. October. Vegetable gardening in Hong Kong. Missouri Bot. Card. Bull. 36: 141. October. (With Brown, William L.) A morphological analysis of row number in maize. Ann. Missouri Bot. Card. 35: 323-336. November. Prince Maximilian's sunflower. Missouri Bot. Card. Bull. 36: 173. December. 1949 The corn plant of today. Pioneer Hi-Bred Co. 20 pgs. Introgressive hybridization. John Wiley and Sons. 109 pgs. (With Newlin, J. J. & Bressman, Earl N.) Revision of Wallace & Bressman, "Corn and Corn Growing." 5th ed. John Wiley and Sons. 424 pgs, (With Hurst, C. T.) A corn cache from western Colorado. Southwestern Lore 15: 3-11. January. (With Hurst, C. T. ) A corn cache from western Colorado. Amer. Antiquity. 14: 161-167. January. Gravel bars evolve their own flood control. Missouri Bot. Card. Bull. 37: 54-57. February. (With Stonor, G. R.) Maize among the hill peoples of Assam. Ann. Missouri Bot. Card. 36: 355-404. September. (With Rick, Charles M.) On some uses of maize in tlie Sierra of Ancash. Ibid. 36: 405-412. November. 1950 (With Brown, William L.) Reports from cooperators. Maize Genetics Cooperation News Letter 25: 17-18. ( Witli Agramont, Felix) Ibid. 25: 18-21. Reviews of herbal interest: more about corn. Herbarist 16: 54-58. 50-55 An Indian garden at Santa Lucia, Guatemala. Ceiba 1: 97-103. April. A rose for the herb garden. Missouri Bot. Card. Bull. 38: 97. May. (Review of Darlington & Mather, "The Elements of Genetics." An admittedly hazardous test. J. New York Bot. Card. 51: 145-146. June.) (With Cutler, Hugh C.) Methods of com popping and their historical significance. Southw. J. Anthropol. 6: 303-308. Autmnn. (With Morey, Dennison) The genetic basis and the origin of the varieties of Aquilegia canadensis, (Abstract: Amer. J. Bot. 37: 665-666. October.) Naturalizing flowers at Gray Summit Arboretum. Parks and Recreation 33: 345-349. October. (Reprinted: Missouri Bot. Gard. Bull. 39: 47-51, March, 1951.) (Review of Schmalhausen, "Factors of Evolution." Evolution in the light of modern genetics. J. New York Bot. Card. 51: 283-284. November.) 1951 The sacred plume. A description of the maize tassel with some indications of its importance. Pioneer Hi-Bred Com Co. 26 pgs, (Review of Stebbins, "Variation and Evolution in Plants." Bull Torrey Bot Club 78- 170- 171. March.) Ten Missouri vdld flowers for St. Louis gardens. Missouri Bot. Gard. Bull. 39: 52^5. March. The white rose of tire renaissance. National Gardner 22: 15-16. March- April. Concordant versus discordant variations in relation to introgression. Evolution 5: 133-141. June. Inclusive herbaria. Indian J. Genet. Pi. Breed. 11: 1-3. Tune 1^)72] EISENDRATH— EDGAR ANDERSON: BOTANIST 355 1952 (With Brown, William L.) Origin of corn belt maize and its genetic significance. Heterosis 124-148. Plants, Man and Life. Little, Brown & Co. 245 pgs. (Reissued, Univ. of California Press, 1967.) Two interesting books about trees. Missouri Hot. Card. Bull. 40: 46-47. February. Wild flower trails at the Missouri Botanical Garden Arboretum. Bull. Card. Club Amer. 40; 54-55. May. (With Brown, William L, ) The history of the common maize varieties of the United States corn belt. Agric. Hist. 26: 2-8. June. (With Gage, Amy) Introgressive hybridization in Phlox bifida. Amer. J. Bot. 39: 399-404. June. (With Brown, William L. & Tuchawena, Roy, Jr.) Observations of three varieties of Hopi maize. Ibid. 39: 597-609. October. Our common autumn wild flowers. Missouri Bot. Gard. Bull. 40: 111-126, September; 127-135, October. Summer flowers which linger into fall. Ibid. 40: 136-137. October. Foreword to McCue, "The history of the use of the tomato." Ann. Missouri Bot. Gard. 39: 289-290. November. (Review of Goldschmidt, "Understanding Heredity." J. Heredity 43: 291-292. November- December. ) 1953 (With Brown, WilHam L.) The popcorns of Turkey. Ann. Missouri Bot. Gard. 40: 33-48. February. Upland cress in tlie house and in the vegetable garden. Missouri Bot. Gard. Bull 41: 40. February. (Review of Davis & Steiner, "Philippine Orchids."* Ibid. 41: 41. February.) (Review of Gardner, "Flora of Western Australia, vol. I, pt. 1. Gramineae."* Ibid. 41; 41. February. ) Pepper and salt, our earliest wild flower. Ibid. 41: 45-46. March. Flora monacensis, a memorial to Dr. Hermann von Schrenk.* Ibid. 41; 53-54. March. An analysis of suspected hybrids, as illustrated by Berberis X gladwynensis, Ann. Missouri Bot. Gard. 40; 73-78. May. Lamartine lilacs for St. Louis. Missouri Bot. Gard. Bull. 41: 90-91. May. (Review of Peattie, "A Natural History of Western Trees."* Ibid. 41: 97. May.), (Review of Dice, "Natural Communities."* Ibid. 41: 98. May.) (Review of Benson, "The Cacti of Arizona."* Ibid. 41: 98-99. May.) (Review of Hitchcock, "A Revision of the North American Species of Lathyrusr"^ Ibid. 41: 99. May.) Spring and summer and the green lull between. Landscape 3: 25-27. Summer. Balkan ivy. Missouri Bot. Gard. Bull. 41: 101-106. June. Introgressive hybridization. Biol. Rev. Cambridge Philos. Soc. 28: 280-307. August, (Review of Hume, "Hollies."* Missouri Bot. Gard. Bull. 41: 130. September.) The Horticultural Council. Ibid. 41:144. October. (Review of Cornish, "Your Guide to a Greener Lawn."* Ibid. 41: 144-145. October.) (Review of Drummond, "Styling Corsages witli Garden Flowers."* Ibid. 41; 145. October.) (Review of Greene & Blomquist, "Flowers of the South, Native and Exotic.*'* Ibid. 41: 145- 146. October.) (Review of Reh & Reh, "The Reh Method of Greenhouse Culture under Plastic Fiber Glass."* Ibid. 41: 146. October.) (Review of Nelson, "Plants of Rocky Mountain National Park."* Ibid. 41: 146-147. October.) Daffodils, daffodil breeders, and daffodil breeding. Ibid. 41: 152-163. November. The best narcissi for naturalizing. Ibid, 41: 163. November. Maize of the southwest. Landscape 3: 26-27. Winter. 1954 Efficient and inefficient methods of measuring specific differences. In Kempthorne et al., editors, "Statistics and Mathematics in Biology." 93-106. Guar and jaguar. Herbarist 20: 9-10. Tower Grove in the 1870's- Missouri Bot. Gard. Bull. 42: 1. January. f, ■, ^r ^ ^ --^ ir*L-" -i-'jm- f 356 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Sunflowers in the west. Landscape 3: 9-10. Spring. (With WilUams, Louis O. ) Maize and sorghum as a mixed crop in Honduras. Ann. Missouri Bot. Card. 41: 213-221. May. The Missoiui Botanical Garden (the city garden). Bull. Gard. Club Amer. 42: 45-46. May. The Garden's most important crop. Missouri Bot. Gard. Bull. 42: 87. May. Introduction to Dr. Woodson's seven meditations.* Ibid. 42: 91. May. Harry Collins. Ibid. 42: 100-101. May. ( Review of Shepherd, '^History of the Rose."* Ibid. 42: 101-102. May. ) Mr. Shaw's Garden. National Gardener 25: 6-7; 29. May-June. Reflections on certain Honduran gardens. Landscape 4: 21-23. Summer. Henry Shaw — a pictorial biography.* Missouri Bot. Gard. Bull. 42: 103-118. June. (Review of Rickett, editor, "Wild flowers of America" and Jennings, "Wild flowers of Western Pennsylvania and tlie Upper Ohio Basin." Sci. Monthly 79: 59-60. July.) Analysis of introgression in a population of stemless white violets. Ann. Missouri Bot. Gard. 41: 263-269. September. (With Brown, William L.) Ladyfinger and Tom Thumb, t\vo old species of popcorn. Ibid. 41: 301-304. September. A field survey of chromosome numbers in the species of Tradescantia closely allied to Tradescanfia virginiana. Ibid. 41: 305-327. September. (With Anderson, Burton R.) Introgression of Salvia apianu and Salvia mellifera. Ibid. 41: 329-338. September. (Reprinted: Omduff, "Papers in Plant Systematics." Pp. 402-411. 1967.) Introgression in Adenostojna. Ibid. 41: 339-330. September. The importance of the African violet.* Missouri Bot. Gard. Bull. 42: 119. September, A botanist in the Christmas tree market. Ibid. 42: 196-198. December. Christmas trees around the world. Ibid. 42: 198. December. (With Stebbins, G. L., Jr.) Hybridization as an evolutionary stimulus. Evolution 8: 378- 388. December. (Review of Benson and Darrow, "The Trees and Shrubs of the Soutliwestern Deserts." Sci. Montlily 79: 416. December.) 1955 A series of bulletins about the Ozarks. Missouri Bot. Gard. Bull. 43: I. January. (Review of Christopher, "The Pruning Manual." Ibid. 43: 15. January.) The Henry Shaw papers. Ibid. 43: 29. February. Horticultural consultants.* Ibid. 43: 29. February. Horse-and-buggy countryside. Landscape 4: 34-35. Spring. A nice quiet evening with a potato. Missouri Bot. Gard. Bull. 43; 50-53. April. (Reprinted: Herb Grower Mag. 9: 41—44. Spring.) Some fine new Forsythias for St. Louis gardens.* Ibid. 43: 53. April. Is die tomato a fruit or a vegetable? Ibid. 43: 60-61. April. (Review* of Bower, "Winter-hardy Azaleas and Rliododendrons." Ibid. 43: 62. April.) (Review of Weatherwax, "Indian Corn in Old America." Sci. Monthly 80: 264. April.) (With Tryon, Rolla M.) Norman Carter Fassett. Bull. Torrey Bot. Club 82; 248-250. May- June. Confederate violets. Landscape 4: 7-11, Summer. The Missouri primrose. Missouri Bot. Gard. Bull. 43: 90. June. The Peruvian daffodil. Ibid. 43: 91-93. June. Learning a little sometliing about aroids. Ibid, 43: 111-113. October. A second harvest. Ibid, 43: 114-121. October. Memories of Alice Eastwood. Ibid. 43: 123-124. October. Go to the index first; don't use the key unless you have to. Ibid. 43: 125-126. October. (Review of Muenscher & Rice, "Garden Spice and Wild Pot Herbs." Ibid. 43: 126. October. Laelio-cattleya *'Ellen A. Ricker." Ibid. 43: 137. November. The star of Bethlehem orchid and its story. Ibid. 43: 143-145. December. 1956 Character association analysis as a tool for the plant breeder. In "Genetics in plant breeding." Brookhaven Symposia in Biology (No. 9), 123-140. Man as a maker of new plants and new plant communities. In Thomas, editor, "Man's Role in Changing the Face of the Earth," 763-777. (Reprinted: Smithsoilian Report for 1956: 461-479. 1957.) 1972] EISENDRATH^EDGAR ANDERSON: BOTANIST 357 Missouri vegetation and an English inatlieniatician. Missouri Bot. Gard. Bull. 44: 12-14. January. As the twig is bent: nature study from an automobile. Cranbrook Institute of Science News- letter 25: 66-67. February. Tlie country in the city. Landscape 5: 32-35. Spring. (Reprinted: Morton Arbor. Quart. 5: 17-21. Summer, 1969.) (With Rudolph, E. D.) An analysis of variation in a variable population of Cladoniu, Evo- lution 10: 147-156. June. (Review of Geddes, "Cities in Evolution." Missouri Bot. Gard. Bull. 44: 91-93. June.) Bald cypresses for St. Louis. Ibid. 44: 108-111. September. (Review of Li, "Chinese Flower Arrangement." Ibid. 44: 111-112. September.) In memoriam Benjamin Minge Duggar. Ibid. 44; 114-115. October. RoseofBrody: an outstanding new daffodil. Ibid. 44: 119. October. (With students in Field Botany course) Korean Lespedeza: Ozark gold. Ibid. 44: 124- 126. October, Plants for a hot place near a wall.* Ibid. 44: 128. October. Dr. George T. Moore, Director 1913-1953. Ibid. 44; 141-142. December. Why botanists visit matli departments. Ibid. 44: 148-151. December. Natural history, statistics, and apphed mathematics. Amer. J. Bot. 43: 882-889. December. (Reprinted: Steere, editor, "Fifty Years of Botany," 247-260. 1958.) (Review of Dobzhansky, "Evolution, Genetics and Man." Bull. Torrey Bot. Club 83: 441. December. ) The star of Bethlehem orchid. Amer. Orchid Soc. Bull. 25: 818-820. December. (Review of Eckbo, ''The Art of Home Landscaping." Landscape 6: 35-36. Winter, 1956- 1957.) 1957 Peruvian daffodils, good bulbs for the sunny herb garden. Herbarist 23: 21-22. (With Singh, B. H. & Pal, B. P.) Studies in the genetics of Triticiim vavilovi Jakub. Agronomy Journal 49: 4-11. January. The cornbelt farmer and the cornbelt landscape. Landscape 6: 3-4. Spring. An experimental investigation of judgments concerning genera and species. Evolution 11; 260-262. June. ( Review of Dansereau, "Biogeography." Science 126: 361. August 23.) A semigraphical method for the analysis of complex problems. Proc. Natl. Acad. U.S.A. 43: 923-927. October. ( Reprinted: Technometrics 2: 387-391. August, 1960. With appended note from Dr. Anderson. ) The city is a garden. Landscape 7: 3-5. Winter 1957-58. 1958 (With Dodson, Calaway H.) Introgressive hybridization in Oncidhnn. In "Proceedings of tfie Second World Orchid Conference," 209-213. (Reprinted: Amer. Orchid Soc. Bull. 29: 733-736. October, 1960.) Winter bouquet. Notes privately printed, as banquet souvenir. A Saturday afternoon witli Epidendrums. Missouri Bot. Gard. Bull. 46: 6-8, February. Anatolian mystery. Landscape 7: 14-16. Spring. Spring flowers of Missouri and how to know them. Missouri Bot. Gard. Bull. 46; 2 >-44 April. August (Review of Graf, "Exotica." Science 128: 296. August 8. [Reprinted: Missouri Bot. Card. Bull. 46: 82-83. November, 1958.]) Etiontjmous alattis, a good shrub for St. Louis.* Missouri Bot. Gard. Bull. 46: 67. September. (With Adams, Helen) Conspectus of hybridization in the Orchidaceae. Evolution 12: 512- 518. December. Winter 1959 Evolution of domestication. In Tax, editor, "Evolution after Darwin, vol. 2, The Evolution of Man," 67-84. Zapalote Chico; an important chapter in the history of maize and man. In "Congreso de Americanistas, San lose, Costa Rica," 230-237. 358 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Three lovely, tangled words: narcissus, jonquil, daffodil. Bull. Card. Club Amer. 47: 5-6. March. Daffodils: breeders and breeding. Ibid. 47: 7-12. Marcb. Old fashioned Cannas for St. Louis. Missouri Bot. Card. Bull. 47: 42-43. April. Plumbago and ivy: a good combination for late summer gardens in St, Louis. Ibid. 47: 43-44. April, Challenges in teaching adults. Adult Leadership 8: 184-185. December. College and the experience of nature. Landscape 9: 7-8. Winter, 1959-1960. (Review of Kieran, "A Natural History of New York." Ibid. 9: 33-34. Winter, 1959-1960.) 1960 (With Ramirez, Ricardo et aL and Calle, G, Edward Nicholson & Brown, William L.) Races of Maize in Bolivia. Natl. Res. Council Publ. 747. 159 pgs. (With O'Brien, Dorothy A.) Repulsive rhymes for ranch house residents. Privately printed. (With Mehra, K. L.) Introgression between Hyparrhenia cymbaria Stapf. and H. papillipies Anderss. in disturbed habitats of Ethiopia. Indian J. Genet. Pi. Breed. 20; 93-101. February. Research in useful plants. Missouri Bot. Gard. Bull. 48: 40. March. (Review of Grigson, "A Herbal of All Sorts." Landscape 9: 38. Spring.) Note appended to reprint of "A semigraphical method for the analysis of complex problems" (1957). Technometrics 2: 387-391. August. The considered landscape. Landscape 10: 8. Autumn. The pellitory problem I.* Missouri Bot. Gard. Bull. 48: 139-141. October. Trumpet creeper.* Ibid. 48; 141. October. Pink maids.* Ibid. 48: 157-158. November. 1961 (With Mehra, K. L.) Character association analysis; its use in grass taxonomy. In "Recent Advances in Botany." Vol. 2: 121-124. (Contribution to symposium on 'The natural classification of the Gramineae," 9th International Botanical Congress. Montreal. 1959.) (With Timothy, David H. et at and Brown, William L.) Races of Maize in Chile. Natl. Res. Council Publ. 847. 84 pgs. The pellitory problem II.* Missouri Bot, Card. Bull. 49: 30-31. February. The analysis of variation in cultivated plants with special reference to introgression. Euphytica 10: 79-86. March. The tree of heaven, Ailonthtis aUissima I. Missouri Bot. Gard. Bull. 49: 105-107. September. (Review* of Lawrence, "Gardens in Winter." Ibid. 49: 114-115. September.) (Review of Yoshimura & Halford, "The Japanese Art of Miniature Trees and Landscape, their Creation, Care and Enjoyment." Ibid, 49; 131-132. October.) (Review of Simoons, "Northwest Ethiopia, People and Economy" Landscape 11: 17. Winter, 1961-1962.) 1962 The role of hybridization in evolution. In Johnson & Steere, editors, "This is life. Essays in modern biology," 287-314. The tree of heaven, Ailanthus altissima II.* Missouri Bot. Gard. Bull. 50: 10. January. Wild flowers of the Ozarks. In "Gardening with native plants," Pi. & Gard. 18: 87-90. Spring. Knock away, gallant soldiers, and Aunt Lucy. Missouri Bot. Card. Bull. 50: 16. April. (Review of Ballard, "The Art of Training Plants." Ibid. 50; 17-19. April.) (With Peck, Kenneth O.) Late season vegetable gardens. Ibid. 50; 1-4. May. Warm 77 Ibid. 50: 9-13. May.) A tour of the Missouri Botanical Garden.* Ibid. 50: 2-16. June. Autumn on the Peninsula. Landscape 12: 13-14. Autumn. Feather geraniimi — Jerusalem oak. Missouri Bot. Gard. Bull. 50: 6. September. (Reprinted: Ibid. 51: 11-12. May, 1963.) (Review of Enari, "Ornamental Shrubs of California." Ibid. 50; 8. September.) (Review of Hughes, "Making an Orchard." Ibid. 50; 8-9. September.) (Review of Genders, "Miniature Chrysanthemums and Koreans." Ibid. 50:9. September.) (Review of Thomas, "Simple Practical Hybridizing for Beginners." Ibid. 50: 9. September.) ,- k_^VJ^-LJh V± * .J- ^-' 1972] EISENDRATH— EDGAR ANDERSON: BOTANIST 359 1963 Collecting boxwood in the Balkans. Boxwood Bull. 2: 26. January. (With Barbre, Clarence & Denison, Edgar) Revision of Meyer & Denison, *Thirty-one broad- leaved evergreens for the Central Midwest." Missouri Bot. Card. Bull. 51: 1-24. April, (Review of Haworth-Booth, "The Moutan or Tree Peony." Ibid. 51: 10-11. June.) Passion-flowers at the Garden. Ibid, 51:1-7. September. Our worst weed becomes a triple problem. Ibid. 51: 8-10. September. The Chinese parasol-tree in 1963.* Ibid. 51: 11. September. Moth mullein in a suburban lawn.* Ibid. 51:11, September. The Connecticut forests and what to do about them. Ibid. 51: 1-10. October. Early fall color as a danger sign. Ibid. 51: 11. October. The nutmeg trees and the colonial administrator. Ibid. 51: 13-14, October. (WithP., G. H.) Indian pipes.* Ibid. 51: 1-2. November. Introduction* to, and explanation* of accompanying photographs, Heiser, "A Trip to Tulcan." Ibid. 51: 2-3 & 4-9, November. Why should V/c^or/a leaves be stiff? Ibid. 51: 10-12. November. Sophora japonica. Ibid. 51; 1—4. December. A jujube tiee in the older part of St. Louis.* Ibid. 51: 5. December. The true Salvia coccinea, Indian fire. Ibid. 51: 5-7. December. 1964 Candelabra {Cassia alata)* Missouri Bot. Card. Bull. 52; 1-2. January. Robert Everard Woodson, Jr. Ibid. 52:2-4. January. Autumn color in 1963.* Ibid. 52: 11, January. For the naturalist, places and plants in Illinois. Ibid. 52: 1-12. February. A tree attacked by dogs. Ibid. 52: 3. March. Named varieties of American holly which are outstandingly hardy in St. Louis. Ibid. 52: 4. March. "Fair maids of February.'* Ibid. 52: 5. March. The old-fashioned flower garden and the new herb garden.* Ibid. 52:6. March. (With Van Schaack, George B.) Robert Everard Woodson, Jr., 1904-1963, Taxon 13: 45-48. March, Goldenrain-tree. Koelreuteria paniculata, Missouri Bot, Card. Bull. 52: 9-11. April. (Review of Stevens, "Kansas Wild Flowers." Ibid. 52: 12-14. April.) (Review of "A Lady Botanist of the American Wilderness," manuscript published by Garden Club of Orange & Duchess Counties, N. Y. Ibid. 52: 14-15. April.) A good early Crocus for St. Louis.* Ibid. 52: 16, April. The Garden as laid out by Henry Shaw.* Ibid. 52: 1, May. (Review* of Grigson, "A Herbal of All Sorts." Ibid. 52:6. September.) (Review* of Menninger, "Seaside Plants." Ibid. 52; 6-9. September.) (Review* of Li, "The Origin and Cultivation of Shade and Ornamental Trees/* Ibid. 52: 9-10. September.) Sacred trees and sacred forests. Ibid, 52: 5-7. October. A new staff member arrives.* Ibid. 52: 7-9. October. (Review of Rickett, "The New Field Book of American Wild Flowers." Ibid. 52: 9-10. October. ) "Jewels of Opar," T^/mam pa»/cuJflium.* Ibid. 52; 11. October. The Chinese witch hazel (Hamamelis mollis) * Ibid. 52: 11-12. October. False aloe, Agat;e ufrgmica.* Ibid. 52: 12. October. (Review of Grogger, "Soil Survey of Daviess County, Missouri." Ibid. 52; 6-7. November.) A bicentennial of gardening in St. Louis. Ibid, 52: 10-12. November. Modem developments of an ancient art at Tower Grove House.* Ibid. 52: 3. December. The European hornbeam, Carpinus hetulus. Ibid. 52: 13-14. December. Trifon von Schrenk, 1887-1964.* Ibid. 52: 14. December. Flowers and agriculture. Ibid. 52; 15. December. (Review of Steele, "Flowers for the King." Landscape 14: 43. Winter 1964-1965.) 1965 t Lad Cutak's two books about Cacti.* Missouri Bot. Card. Bull. 53: 2. January. Tree labeling by the Men's Garden Clubs of the Midwest Regional Council.* Ibid. 53: 7, January. ly - n H. 360 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Oa the bad habits of certain shade trees. Ibid. 53: 1-5. February. Lotus leaf candles at Chinese fiestas. Ibid. 53: 7, March. Standing in the shade of a cabbage. Ibid. 53: 8. March. The medlar {Mespilus germanica). Dr. Samuel Johnsons standard for wit. Ibid. 53: 11-12. March. A unicorn on tlie window-sill. Ibid. 53: 5-7. April. Victoria water lilies. Ibid. 53: 1-18. May. England and America; some quaking grasses and tlieir common names.* Ibid. 53: 22-23. May. How to smell a tree. Ibid. 53: 23-24. May. Foreword to "The first five years of the Climatron." 53; 1. June. A botanist looks at poetry. University of Michigan Quarterly Review 4: 177-184. July. (Review of Kingsbury, "Deadly Harvest." Missouri Bot. Gard. Bull. 53: 2-3. September.) Introduction to Shankland, "A Summer Afternoon with Henry Shaw and his Friends." Ibid. 53: 3-4. September. Yucca and biological research at the Missouri Botanical Garden. Ibid. 53; 6-7, September. Tapioca and the Climatron.* Ibid. 53: 1. October. (Witli Macrander, Dorothy) A charming new weed. Ibid. 53: 16. October. David M. Gates, the new Director. Ibid. 53: 1-7. November. Is your name correct in our roster? Ibid. 53; 16. November. 1966 Weeds, botanists and chemists. Missouri Bot. Gard. Bull. 54; 18-19. January. The golden larch, Pseudolarix amabilis. Ibid. 54: 3-11. February. The Alcohohaceae; a monographic account. Ibid. 54: 15. February. (Review of Thompson, "Creative Decorations with Dried Flowers." Ibid. 54; 16. February.) (With Baer, Mary E.) The Chelsea Physic Garden, yesterday and today. Ibid. 54: 20-23. March. Islands of tension. Landscape 15; 7-8. Spring, The stories behind two botanical names.* Missouri Bot. Gard. Bull. 54: 1. April. Springtime greens from square-stem. Ibid. 54: 15—16. April. Why aspen leaves flutter in the wind.* Ibid. 54: 8. May. Oil of broomsedge.* Ibid. 54:9. May. Answers to tough problems. Ibid. 54: 9-11. May. Poverty grass.* Ibid. 54: 12. May. A strange and edible Missouri grass.* Ibid. 54: 12. May. The new Park Building. Ibid. 54: 3-4. September. The death of Philip A. Conrath.* Ibid. 54: 4. September. The dynamics of landscape. Ibid. 54; 5, September. (WithB., M.) A welcome visitor,* Ibid. 54:10. September. The *lierb ladies" provide both gaiety and money,* Ibid. 54: 12-13. September. Ear-mioffs in spring, salt pills in summer,* Ibid. 54: 16. September. Green leaves in a hot summer.* Ibid. 54; 18. September. (With B., M.) Graduate students busy all summer.* Ibid. 54: 19-20. September. Old Japanese stone lantern on display. Ibid. 54: 20. September. Botanical gossip.* Ibid. 54: 20. October. The Garden's gift to American window sills. Ibid. 54: 2-4. November. Theobroma comes to the Climatron. Ibid. 54: 4-6. November, A glimpse of the Garden's staff picnic* Ibid. 54:11. November. Two of the Garden's alumni on a recent visit. Ibid. 54: 14-16. November, Learning from the bees: G. H. Pring and water-lily breeding. Ibid. 54; 8-10. December. Goethe and the Gm/cgo fci/oba. Ibid. 54: 15. December, 1967 The bearings of botanical evidence on African culture history. In Gabel & Bennett, editors, "Reconstructing African Culture History," 167-180. Perilla fmtescens: a multipurpose new vegetable. Bull. Gard. Club Amer. 55: 10-13. May. The landscape of the lower Meramec and its tributaries. Missouri Bot. Gard. Bull. 55: 11-16. May. (With Brockhoff, Dorothy) Henry Shaw — a pictorial biography. Ibid. 55; 1-16. June. Frank Steinberg and the Arboretum. Ibid. 55: 12-15. September-October. The glades of the Arboretmn. Ibid. 55: 16-18. September-October. - -^L -^ - 1972] EISENDRATH— EDGAR ANDERSON: BOTANIST 362 An unusual American shrub: salt bush {Baccharis halmifolia) . Ibid, 55: 32-33, September- October. The botany of Christmas. Ibid. 55: 12-18. Nov^ember-December. (Review of Sauer, "Plants and Man on the Seychelles Coast." Ibid. 55: 19. November- December. ) Ginkgo biloba, the Ginkgo or maidenhair tree. Ibid. 55: 20-25. November- December. Living with a GmTcgo. Landscape 17: 33-34, Winter 1967-1968. 1968 (Review of Sehwanitz, "The Origin of Cultivated Plants." Landscape 17: 33-34. Spring.) The story of the white redbud. Missouri Bot. Card, Bull. 56: 5-7. March-April. (With Denison, Edgar) Ginger {Zingiber officinale). Ibid. 56: 17-19. May-June. Evening primrose I. Ibid. 56: 20-21. May-June. Individual insight and irreducible facts. Friends' Journal: 320. July 1. Sunflowers in the fields. Missouri Bot. Gard. Bull. 56: 5. July- August. Experimental studies of the species concept. Ann. Missouri Bot. Gard. 55: 179-192. September. (Review of Jeffers, "The Friends of John Gerard [1545-1612], surgeon and botanist," Missouri Bot. Gard. Bull. 56: 18-19. November-December.) (Review of Kenfield, "The Wild Gardener in the Wild Landscape." Ibid. 56; 20. November- December. ) (Review of Arnold, "Poisonous Plants of Hawaii." Ibid. 56: 21. November-December.) (Review of Menninger, "Fantastic Trees." Ibid. 56: 22. November-December.) What we do not know about Zea maijs. Trans. Kansas Acad. Sci. 71: 373-378. Winter. 1969 A message from the midwest. In "How to Grow Herbs in the Midwest/' 3. St. Louis Herb Society and Missouri Botanical Garden. Reports from cooperators. Maize Genetics Cooperation News Letter 43: 117. Godfather of the Garden (George Engelmann). Washington University Magazine 39: 38—43. Spring. Calceolaria, Missouri Bot. Gard. Bull. 57: 6. March-April. Winter visitors in the Garden. Ibid. 57:7. March- April. The partnership of hollies and birds.* Ibid. 57: 32. March-April. (Reprinted: Ibid. 57: 4. November-December, 1969.) The horsechestnut — Aesculus hippocastanum. Ibid. 57: 4-9. May-Jime. The music of a scientific name. Ibid. 57: 18. May-June. Original edition, Steyermark book. The Garden Today. June. (Dr. Anderson died on June 18, 1969; all subsequent listings were published posthumously.) Chicory, botany by the wayside. Missouri Bot. Gard. Bull. 57: 7. July- August. Wliat I found out about tiie com plant. Ibid. 57: 6-9. September-October. Missouri cedars.* Ibid. 57:9. September-October. Ecology and the duck pond. Ibid. 57: 10-11. September-October. The red hibiscus I. Ibid. 57: 12-13. September-October. The purple and copper beeches. Ibid. 57; 15-19. September-October. (Review of Barlow, editor), "Darwin and Henslow. The growtli of an idea. Letters 1831- 1860." Ibid. 57: 28-30. November-December. 1970 ' A letter. Herb Grower Mag. 23; 27. Linnaeus' description of the first pineapple he ever tasted.* Missouri Bot. Gard. Bull. 58: 17. January- Fe b ru ar y . Our soybean climate. Ibid. 58: 18-21. January-February. (Review of Stout, "How to have a Green Thumb mtlaout an Aching Back." Ibid. 58: 31-32. January-February. ) A few Anderson gems. Ibid. 58: 27-28. May-June. Knot gardens in Parkinson's Paradisi in Sole (1629 A.D. ). Boxwood Bull. 10; 9, July. •■-■ -> |i-T^^ ■- ,I^S- STUDENT DAYS WITH EDGAR ANDERSON OR HOW I CAME TO STUDY SUNFLOWERS' Charles B. Heiser, Jr.^ When Dr. Porter invited me to give this address in a symposium honoring Dr. Edgar Anderson, I v^as, of course, deUghted to accept. In my reply to him I suggested several possible topics — the origin of Solatium quitoense, the current status of introgression, variation in the bottle gourd, why did man J plant seeds, and the title that he selected. I really didn't expect him to choose the one he did, for at the time I suggested it I thought it a rather clever title, but I did not have the vaguest idea as to what I would say, whereas for most of the others I did have some hard data. In a sense it is a most appropriate title, and Dr. Anderson would have enjoyed seeing that I still get into the kind of predicaments I did 30 years ago. It is also appropriate, for after a day of data and hypotheses, something different is called for, and I promise you that this will be different — if nothing else. I hesitate to give this talk for two reasons. First, I do not feel that I am old enough to give a talk of reminiscences. But the other day my son said to me, "Dad, do you realize that you are a half century old." Somehow when put this way it makes me seem old enough. Secondly, I hesitate giving this talk since it involves talking a lot about myself and exposing a brash young student still wet behind the ears, and it becomes necessary to say a lot about him to develop the theme that the title implies. By way of background I should say that I was a student at Washington University and that sunflowers have been my principal research interest for a number of years. In a sense the story began in a room in Rebstock Hall in 1939 when I took Bob Woodson's Botany 101 course. I took it as my science requirement, for at the time I had listed myself as an English major. Before the course was over I had changed my mind. However at Bob's suggestion I took no botany the next year, for he thought diat I shouldn't change my major too rapidly. So I didnt really become acquainted with Edgar Anderson until my senior year and that was somewhat of an accident. That year I had planned to take an advanced taxonomy course with Wood- son, but that was all changed when he was asked to give a geography course for students the Army had sent to campus. Anderson volunteered to give Woodson's course. That in itself now appears unusual to me. How many of you have colleagues who will volunteer to give your courses? There were two students in Botany 550 that year— myself and another botany major, Dorothy Gaebler (I wonder what became of her?). We soon found that this Missouri Botanical Garden. ^ Department of Botany, Indiana Unh Ann. Missouri Bot, Gard. 59: 362-372. [uni 1972] HEISER— STUDENT DAYS WITH EDGAR ANDERSON 363 Figure 1. Dr. Anderson with Ada Hayden, Richard Hohii, LiHian Nagel, and Dorodiy Gaebler Heiser. — Edgar was way ahead of his time in many ways, including fashion as this photograph of a field trip to the Arboretmii shows (about 1945). I don't think he was ever happier than when he was conducting students on field trips. was no ordinary course. At the time I assumed it was because of the small stude ordinary He was a most unorthodox but a most effective teacher. I don t think that anyone else should try to teach a course the way he did, for only an Edgar Anderson could do it that way. One thing, I remember, was that I wanted to look up answers in books. This he did not encourage — one looked at plants. The plants that fall were composites, and the entire semester was devoted to them. We started with sunflowers and then nrnnpf^dfid to Asfprs ^Droceeded to Asters, Solidagos, and many others. Weedy composites occupied most of our time. Gas rationing limited our trips outside of St. Louis, and, of course, as is well known, Anderson had a special affection for weeds. Of these, none held a greater fascination for him than the common sunflower, and these plants served as the primary basis for his classes' studies year after year. St Louis with all of its slums, railroad yards, and dump heaps had plenty of sunflowers in those days. grew. In fact, when he met visitors at Union Station he took great delight in taking them to the Garden on one of his *'short-cuts." The visitor would see nothing but slums and factories and lots of sunflowers. In addition to the beautiful scenery the Edgar knew all the places where sunflowers ■■■- 1"^ -"- >- ' . ' ^ALf_"-^_ z _.C 354 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 short cut afforded, it didn't take more than 10 minutes longer than the normal route. Edgar would also delight m telling about the time he was in the hospital and some of his students brought him a bouquet of weedy sunflowers. It was on that occasion he told me that he first understood the marvelous events that took place from hour to hour in a sunflower head. I recall that it was on one of the field trips to Dago Hill with his class of two students that he said, "Charley, I think that if someone looked into sunflowers carefully, he would discover something very significant/' That remark stuck with me. He had a singular ability to make even the most insignificant plant a thing of great interest. He always conveyed the im- pression that there was something very mysterious going on in plants about which scientists knew very little and, moreover, that they didn't realize how little they knew. One had to look at plants in a new way. One way for people to do this was to use a ruler and to look at populations, not single plants. This, of course, we now take for granted. It was also at this time that I became aware that he could see things in plants that others couldn't, or a least that I couldn t. Some years later I had my doubts and decided that maybe this was partly an act, for there is no doubt about it, he was a showman! But he was a natural showman not an intentional one, and I now realize that he could look at a field of plants and actually see things that escaped others. In trying to analyze this ability I have decided that it involved two things: an ability to cor- relate a large number of independent observations in a very short time — he was a numerical taxonomist with his own built-in computer — and an ability to analyze patterns. For example, inflorescences of plants which meant little to the ordinary taxonomist revealed a great deal to him. But back to our course. I recall that he had us get some colored modeling clay and asked us to make models of disk flowers. There were no regular lectures, although he would frequently come into the lab and pick up some plant or our clay model and ask what we had learned. He would hsten a while (and for a great talker one of the amazing things about him was that he was also a good hstener) and then he might turn around and walk — maybe stalk is a better word — out of the room, or he might hold forth on some subject for 15 minutes or so. I don't recall taking any notes— maybe be- cause I knew there was to be no exam. Our grade was to be based on two term papers which the students were to do jointly. It takes great courage — or perhaps foolishness — on my part but I am going to read you some from the first term paper. I ask your indulgence and patience but I assure you that I shall not read it all to you, for which you should be tliankful. C SEX LIFE OF COMPOSITES AND OTHER INTERESTING FACTS GLEANED FROM BOTANY 550 by dorothy m. gaebler and diaries b. heiser, jr. "To name all of the species of Solidago or the asters at sight is a feat probably no one living can perform." — Nature's Garden (It's also hard to do with Grays Manudl^-Ueiser) 1972] HEISER— STUDENT DAYS WITH EDGAR ANDERSON 355 In the Ode, "Intmiations of Immortality from Recollections of Early Childliood, rt Wordsworth says: "To me the meanest flower that blows can give Thoughts that do often lie too deep for tears." The Romanticists all found depth of meaning, beauty, and expression m the Wordsworth didn't find any such meanings "A iDrimrose by a river's brim A yellow primrose was to him, And it was nothing more." to us a sunflower was nothing Well, it was that way with us until receni but a sunflower. Tlien for a brief period it became "that damn thing," and now what it is is simply amazing! "A fool sees not the same tree that a wise man sees" — Blake- This article seems very useless in that w^e won't tell Dr. Anderson very much about the composite family that he doesn^t already know, but the point of this is that we get something out of writing the article and that poor Dr. Anderson has to read it. But we're going to try to tell him some things that he doesn*t know; for instance, how if he had entered the morphology lab late on a certain afternoon last fall he would have had a wad of clay thrown in his face, pappus and aU. At the beginning of our course we were given several sunflowers. We were told to examine them, which we did. There was nothing extraordinary about them — or so we thought — but we learned fast as the result of some very em- barrassing questions asked us about the sex life of Helianthus, Our next line of study consisted of making a clay model of a disc flower of Helianthus^ an assignment that sounded ludicrous at the time and was even called worse things before we finished the model. "Wee, modest, crimson-tipped flovvV, Thou's met me in an evil hour/' — To a Mountain Daisy y Bums A very great friendship was almost broken up over the question of whether to put hairs on the stigma or not to put hairs on the stigma. However, at last we finished and we decided that it really hadn*t been so bad, that it was rather fun, and that we did know the disc flower inside out — and we do mean inside out. That was done. We would start on the other composites next time. We were through with tire clay (we thought). So imagine our surprise and consternation when our professor told us that he wanted us to make models of the pappi of as many different composites as we could obtain. We considered hiring some first grade pupil who had some experience with modeling clay, but as we couldn't afford paying union rates we decided to do the things ourselves. The arguments about the hairs on the stigma were now forgotten for far greater and more important controversies. In fact, when Heiser took Gaebler's model of Gaillardia and finished it to his satisfaction, sparks flew; but when he criticised and tried to change her model of the marigold pappus, it was practically the last pappus straw. No more work was accompUshed that day. A long and beautiful friendship was tottering on the brink, but alas, love conquers all and finally the models were finished. We really began to learn things now. We learned that one tenth of all of the flowering plants of the world belong to the composite family. In Missouri according to Palmer and Steyermark there are 65 native and 15 introduced genera, 243 native and 71 introduced species, five native forms and two introduced ones, and 15 hybrids of composites. ... From here on it begins to sound a little more like an ordinary term paper. th again until 1960, when it arrived '_■« '--^JF'T, _|_ 366 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 by mail along with a letter from Edgar. After a first paragraph dealing with variation in Claytonia, he went on to write: Topic No. 2, HeliantJius annuus. I am returning herewith the paper by Heiser and Gaebler, You will note that it is still in its original cover. The writing on it is Dr. Moore's [G. T, Moore, then Director of the Garden]. I told Dr. Moore what I was doing in the course and that I had leaned over backwards farther, in giving a good mark to tliis term paper than I ever had before. His eyes twinkled and he said he would like to see the paper, so I turned it over to him. He took it home and read it at his leisure and spoke to me about it some days afterwards. He handed it back witli the remark "Anderson, I must say in this case I doubt your judgement. You have had a number of unusual students and have done well with them, but in this case it seems to me that tliis young man ought to have his ears boxed." However, you will notice that it still bears the pencilled grade of "A" which I put on at that time. I hope you will have the courage and the intelligence to keep this paper. If I thought you would destroy it, I would not let you have it back again, but it is a personal thing and therefore it belongs to you and your wife. However, it has a scientific career of its own and is a demonstration of something which precious few teachers of elementary students seem to get into their heads. One is not turning out finished botanists at the end of the first or second or third course in botany. One is leading students on to become better and better botanists. I still refer to the year (when I just had the two of you) as the most successful year my class ever had, and I feel that very deeply. I hope someday you will have the courage and intelligence to show this to young teachers of botany in mingled pride and humility. It might help them to understand the lengths to which I was willing to go in following William James Beal's method of having promising students write up their own discoveries from every-day examination of plants brought in for them to study. Here perhaps is a missed opportunity, for this would be a good stopping place, but I agreed to tell why I came to study sunflowers so I must return once more to the forties. Our course in composites was supposed to be followed by one in grasses, but Dr, Anderson had the opportunity to go to Mexico the next semester so the course wasn't given. I wasn't particularly enthusiastic about studying grasses at the time, but I now look upon it as a great opportunity missed. After receiving my A.B. I continued on at Washington for a Master's under Dr; Greenmans direction. I suggested to him that sunflowers would make a good problem, but he discouraged me. However, since by now I was most interested in composites he suggested another genus in the family, Psilostrophe. The thesis was eventually published and forgotten. At the same time, however, my interest in sunflowers continued, probably largely because Edgar gave me little pep talks from time to time. I began to assemble various varieties of cultivated sunflowers which I grew in the greenhouse. I made morphological comparisons and had httle thought of doing anything more until Edgar asked if I minded if he attempted to make hybrids between some of the varieties. I said, "No, of course not," but he didnt get around to it immediately so I started making hybrids— which I am sure is exactly what he had hoped I would do. Edgar decided that I should go to Arizona the next summer to learn about the Hopi sunflower. He managed to get $50 from the Garden for the ti'ip and arranged with the Museum of Northern Arizona for me to spend a week there. I went by Greyhound bus— and I had a wonderful week althoudi I had to live 1972] HEISER— STUDENT DAYS WITH EDGAR ANDERSON 367 FiGUBE 2. Dale Smith and Dr. Anderson. Edgar examining a sunflower on a visit to the Experimental Garden at Indiana University (about 1955). "Nehru hat*' which he had obtained on a visit to India. He was very fond of his on cheese, since I had forgotten to bring any meat rationing coupons with me. It was my first contact with a flora outside of the Middle West and, as a good taxonomist should, I spent a lot of time filling presses with plants — or bailing hay as it is known to the non-taxonomist. I didn't learn very much about the Hopi sunflower. Because of gas rationing I never did get to the Reservation, although I did get to talk to the Hopis at the Museum about its uses. But fortunately weedy sunflowers — both the common sunflower, Helianthus annuus^ and H. petiolaris — grew near the Museum. I was struck by some unusual variants, but I didn't analyze them fully at the time. When I returned to St. Louis, I decided to take a closer look at H. petiolaris — a population of which grew at the streetcar stop next to the University. On the same day I looked at some plants of H. annuus which were growing about a quarter of a mile away where the streetcar tracks crossed Skinker Blvd. My attention was drawn to three very depauperate plants which grew a few yards away. The more I examined them the more excited I became, but I decided to look at them again the next day before telling Edgar that I had found some sunflower hybrids. Needless to say, he was delighted, and so was I, for the first time I had discovered something about sunflowers that he didn't already know. I then recalled the unusual plants that I had seen in Arizona and pulling out my dried specimens I realized that they included plants that were probably both Fi's and backcrosses of H. annuus and H. petiolaris. Today tlie finding of another hybrid 368 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Figure 3. Robert McClary, Vernita Neher, Jack Humbles, Emma Fraysur, Ralph Jacqmain, A. Isabella Hamilton, Dr. Anderson, Jane Haskett. — A botany class visits the Arboretum in 1959, Photograph taken in front of "Woodson's cabin/' is hardly cause for a celebration, but this was more than a quarter of a century ago when many taxonomists were still very skeptical about Anderson's claims for hybrids. I remember one saying that calling a plant a hybrid was an excuse for not being able to identify it. After receiving my Master's I stayed on the next year at Washington University as an instructor, as a replacement for Bob Schery, who had gone to Brazil in the wartime rubber program. At the same time I was asked to coach the basketball team at tlie University. I was offered $100 for the season. I accepted immediately, for I thought of it as pure gravy, for after all, I was already being paid $1500 for the year as a botany instructor, and ;ht of myself as well off. I won t dwell at length on my career as a basketball coach which lasted only the one season. I kept losing players to the draft and I ended up the season with two war veterans, a couple of 4-F's and some 17 year old boys. In fact, the player situation became so desperate that tlie manager, Mas Yamada,^ who may be here tonight, had to suit up for the final games. As I recall, Mas distinguished himself by picking up two personal fouls in one minute's playing time in our final game. The reason for the inclusion of basketball here is that it leads up to one of Edgar's thou =*Mas, who is still employed in the Department of Biology, was brought to Washington University by Dr. Anderson. 1972] HEISER— STUDENT DAYS WITH EDGAR ANDERSON 3gg favorite stories, or if not his favorite, one that I heard him tell on numerous occasions. My office in Rebstock Hall had the only phone for the Department of Botany; and since most of the staff members spent most of their time at the Garden, it fell upon me to handle incoming calls at certain times of the day. Edgar delighted to say that I would always answer, "Botany Department, Coach Heiser speaking/' I don t recall that it ever happened and I never thought it particularly funny, but I do know that both Edgar and Bob Woodson were quite pleased tliat a botany instructor was also a basketball coach, for they thought that it gave a new image to botany. It was in the spring of the year 1945 that I learned that I was going to the University of California. Actually it wasn't my decision, for I had no particular desire to leave Washington University. But both Edgar and Bob decided that I should go away for a year of seasoning and that Berkeley would be the appropriate place. Bob arranged an assistantship, and Edgar called me in to tell me about the people that I should see immediately — these were W. L. Jepson, whom I never did meet, Carl O. Sauer, and G. Ledyard Stebbins, Jr. I was then quite excited about making chromosozne smears, and he told me that Ledyard had a dropper bottle of aceto-carminc within arm's reach anyplace in his lab. Shortly after my arrival I presented myself to Ledyard. I recall noting that there was only one bottle of aceto -carmine in his lab, but that was my only disappointment. I believe that we discussed possible Ph.D. problems for me on this first meeting. I brought up the possibilities of Stephanomeria, a composite that I had become acquainted with in the herbarium of the Missouri Botanical Garden, and sunflowers. Helianthus annuuSy of course, was common in California, and I knew that a closely related annual, //. holanderi, also grew in California. Ledyard told me he knew where it grew — I was later to learn that he knew where nearly everything grew in California — and we set up a field trip for a couple of weeks later. In the meantime I wrote Anderson about my reception in Berkeley and about the possibilities for my research topic. He wrote back almost immediately (Oct. 16, 1945): Personally, I have a very different opinion about the Stephanomeria problem. It is a very nice problem, you already have your teeth into it and so does Stebbins, and working under his direction you would finish it in record time. It would be a good piece of work and widely acclaimed, BUT it would be just another monograph done in cyto-taxonomy. You wouldn't have learned very much, merely practiced. On die Helianthus problem you are not merely working out a problem, you are uncovering an entire new field of work. There is no question in my mind that for your eventual growth and development you would get the most out of the sunflowers. Up to the present, you have worked on the problem largely with me, and my ideas are pretty violently warped. It would be very helpful to you to work on Helianthus under Stebbins, who has another set of ideas and who undoubtedly is not mad about the genus. It is your decision, however, and not mine and I shall continue to vote for you even though you do nothing but Stephanomeria. I won t go into detail concerning the trip I made with Ledyard except to 370 ANNALS OF THE MISSOUKI BOTANICAL GARDEN [Vol. 59 say that we found Helianthus annuus and H. bolanderi growing together in several places and there were hybrids in nearly every mixed population. I was excited and so was Dr. Stebbins. My research problem for my Ph.D. was determined that day. I wrote back to Anderson to tell him that I had decided to work on Helian- thus which delighted him^ of course. I also took the opportunity to ask him to explain a remark in his previous letter. If I am uncovering an entirely new field of ^^^ork in Helianthus I wish you would let me in on what it is. Is it . , . tracing the origin of the cultivated sunflower or the hybrid work? Stebbins, I tliink, thinks you mean the latter. He is very interested in the hint that I dropped that California annuus may be a new sub- species . . . derived from hybridization. To which he replied: The taxonomy of cultivated plants is an unsolved and neglected problem. It will require new attitudes, new techniques, and new horizons. What you are doing with Helianthus annuus as a prehuman and post human weed, and as a series of cultivated entities is a contribution to a discipline which is yet a-borning. Having read tliis much go and talk the whole biss over with COS and see if he doesn't agree with me. It is a field which abuts upon the following fields: Genetics, Taxonomy, Agronomy, History, Archaeology, Anthropology. Its major techniques I imagine have still to be invented though one or two are suggested in Anderson and Cutler and in Carter and Anderson. After seeing COS and thinking this over please write me about it again. I can go into horribly complete details if necessary. Have you been to Davis and seen Beetle? Try to find out ever>thing you can about what Agronomists are like and what they know and don't know. My best to you both. Don't kill yourselves just because U. of C. will give you the chance. There is no-one there to be an old mother hen on the sidelines to cluck at you when you are not sitting back calmly from time to time to enjoy life. The days I shamelessly played hookey when I was a Graduate Student have paid better dividends than tlie times I worked too late in the lab. Of course there are limits. And then on Nov. 12 I wrote: Stebbins is quite excited (hardly a strong enough word) by sunflowers by now. If you are an authority on taxonomists I am becoming one on geneticists. I see what you mean when you say it would never do for Stebbins and me to work in the same lab. Dot [Dorothy Gaebler Heiser] says that we literally become make us both nervous wrecks. influence I should add here that although I had been sent to California for a year and was expected to return I never did so. I took my degree there in Botany, not Genetics, so although Stebbins was not my major adviser in name, he was in spirit and deed. I continued to exchange letters with Edgar while I was in Berkeley and in fact, ever since. I have a file of correspondence with him over 2 inches thick. At the time it didnt seem unusual to receive a two page letter from him, but looking back I realize that it is something that an ordinary professor doesn t find time for. Over the years there was no telling from where I would get a postcard from him with some observations about a sunflower he had seen in his travels. This past summer I spent a most delightful afternoon reading all his letters, and I would like to quote just a few comments from some of them. 1972J HEISER— STUDENT DAYS WITH EDGAR ANDERSON 371 Figure 4. Dr. Anderson with potatoes. Here we are spending a nice quiet evening with potatoes in 1959. Edgar's essay, "How to spend a Nice Quiet Evening with a Potato" (Missouri Botanical Garden Bulletin 43: 50-53. 1955), is my favorite among the many delightful popular articles he wrote. The table where many a student and visitor has eaten is situated behind the "Barn" at the Arboretum. Edgar and Mrs. Anderson used to live at the "Barn*' during the summer in his earlier years at the Garden, while he grew his experimental plants in nearby fields. On February 6, 1946 — on being frustrated at not findmg something he wanted in the herbarium — he wrote, Oh God, oh stamp c6ltecting, when will taxonomists ever take any interest in being biologists. Once when I traveled with E. J, Palmer I went to a good deal of trouble to get a whole sheet of Lily pods and he threw it away because it made specnnen anyway Later the same year (Nov. 18): Jon Sauer is starting in on several minor cultivated plants. I've learned a lot as usual. What an incredible gift good students are. In 1950 (Feb. 2) when I wrote him that I was looking for some new research problems. Butt. What's this about ditching Helianthtis? I suspect you have been traveling with Job's comforters. If you are tired of the damned weeds and don't want to look at em any more for a while, why by all means put them aside. Don't let anybody's advice, including mine, keep you from what you are happiest doing. This was followed by some compelling reasons why I should continue. 372 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 Once I obtained roots of a hybrid of Helianthus grosseserratus X maximiliani which he grew as an ornamental in his backyard but I failed to get information on its parentage. So I wrote him and he sent back this very brief message (May 26, 1960): The hybrid plants came up in my garden where I was growing bodi parent plants. I ask you, Mr. Taxonomist, is this a natural or an artificial hybrid? After I came to Indiana University in 1947 he continued to be not only my teacher, but my students' teacher as well. For many years I used to take my class to the Missouri Botanical Gardens Arboretum at Gray Summit for a weekend field trip. Whenever possible he would come out to join us. He delighted in teUing my students stories about me when I was a student and loved to try to embarrass the proper young professor. He also came to Bloomington frequently to see my sunflower garden. It's time to stop and I haven't told you anything about his music sessions He played the recorder. Nor about the square dances at the "Barn." Nor about his cooking. I think one of the worst dishes I have ever eaten was his spam covered with bread crumbs soaked in loganberry juice — ^perhaps because he raved so about it. I have talked way too much about myself, but I hope by doing so it has given you some insight into the character of Edgar Anderson, teacher and botanist, The latter is the title he chose for himself in his later years at the Garden. EDGAR ANDERSON: RECOLLECTIONS OF A LONG FRIENDSHIP G, Ledyard Stebbins^ Edgar Anderson was one of the most extraordinary men whom I or anyone else has ever known. As a scientist, he was remarkable for the precision with which he recorded what he saw, as well as his distaste for excessive detail, which sometimes led him to gloss over valid objections to his theories as he stated them. Long before other geneticists, he perceived almost intuitively the enormous complexity of genetic variation in natural populations. Most of his scientific career was devoted to a search for techniques by which this complexity could be expressed in such a way that any biologist, regardless of his mathematical background, could understand and appreciate it. In achieving this goal, he was largely successful. He was a true naturalist — a keen, analytical observer of plants in their natural surroundings. Perhaps his interpretations were sometimes too facile. Nevertheless, no plant scientist known to me was more able than Edgar Anderson to compare in their natural habitat the leaves, flowers, and wonderfully symmetrical architecture of plants, and to express their inter- relationships in a meaningful way. In his personality as in his scientific achievements, Edgar was a man of almost unbelievable contrasts. He could be rudely abrupt, if the occasion arose, with either his scientific opponents, his acquaintances, or even his closest friends. On the other hand, he could never hold a grudge, or hurt people intentionally. He could be equally warm and friendly to scientific colleagues, older scientists whose opinions he respected, to the grandes dames of Phil- adelphia's main line, or the high society of St. Louis; as well as to his sub- ordinates in the Garden, and to the country people of Missouri, California, Mexico, or any other country that he might be visiting. His warmth, friendliness, and encovu'agement was, above all, showered lavishly upon younger scientists and students who, in his opinion, were sufficiently intelligent and ambitious to warrant his attention. He set great store by his unconventionality. He could never resist the temptation of exercising his keen wit at someone else's expense. Nevertheless, he was even more concerned with improving his ability to understand the needs and motives of others. In this, he was usually successful. In these pages I shall illustrate the above remarks as well as I can by telling a few anecdotes about my 38-year friendship with Edgar Anderson. In the summer of 1930, as a graduate student on my first trip to Europe, I looked forward with particular anticipation to the International Botanical Congress in Cambridge, England. In many ways, the Congress lived up to expectations. I had a chance to meet, even if only casually, the renowned botanists whose works I had been reading: A. C. Seward, Otto Rosenberg, Georg Tischler, Agnes Arber, and many others. At other times, however, the ^ Department of Genetics, University of California, Davis, California 95616. Ann. Missouri Box. Card. 59: 373-379. 374 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 sessions were long, tiring, and, to me, excessively occupied with meaningless trivia. At the end of one of these sessions an expansive young man, in his broad Michigan accent, gave a clear, incisive description of differences between species of Iris that had been well known to me already in my childhood spent on the coast of Maine. Tying his account to chromosome numbers and to the probable evolution of the species with relation to the Pleistocene glaciation, Dr. Anderson immediately set going in my mind a train of thoughts about plant evolution that have been with me ever since. With some trepidation, I went up to him after the session and asked a question about his talk. His response was immediate, cordial, and enthusiastic. In a long session in his room that evening, we discussed his research on Iris, Aquilegia and Traclescantia, my own Ph.D. problem in Antennaria, and many other items, relevant and irrelevant. During the rest of the Congress, I spent as much time with him as I could. I gained a new insight into our field of common interest and learned to look at botanists in a new light. "Look at Edith Saunders in her square hat, doesn't she look just like a female bishop?" "There's that bore Ruggles Gates, trying to hide his Canadian origin by being more British than the British." "The trouble with C. D. Darlington is that after Newton's death, he came up too fast. He doesn't realize that to write intelligently about plant chromosomes, you at least ought to know what the plants look like that they came from." "There's Agnes Arber; why don't you go up and talk to her abovit plant anatomy and evolution. I'm sure she'll give you some good ideas." A few years later, when I was a young instructor at Colgate University, Edgar paid me his first visit at my home. His visit was only partly on my account. He had also been attracted by the incomparable collection of peony species and hybrids being grown by a Professor of Chemistry, A. P. Saunders, at nearby Hamilton College. His letter to Professor Saunders included a characteristic question: "Do you have in your garden any peonies except tliose big puffy ones that aren't good for anything except a horse's funeral?" We assured him that he had, and the visit went off most successfully. Among other things we had a musical session at which Professor Saunders displayed his superb artistry on the violin, and I was persuaded, I don't know why, to present a sample of my woefully inadequate piano playing. Its sequel came a few weeks later when I had an occasion to visit Harvard and the Bussey Institution. After dinner at his home, Edgar addressed a characteristic remark to Mrs. Anderson: "Dorothy, I want you to hear Ledyard play 'Joshua fit de battle of Jericho,' not hitting any note right, but getting the spirit of it better than you do!" That was the best compliment on my piano playing that I have ever received! After going to California in 1935 I saw less of Edgar, but we renewed our friendship at the A.A.A.S. meetings in Dallas at the end of December 1941, Three weeks after Pearl Harbor, these meetings were noted chiefly for the difficulty that all of us had in paying attention to science. At one session, a paper was presented by the father of hybrid vigor in plants, G. H. Shull. Al- though all of us revered Dr. Shull, his current research was rather hard to take. It consisted of a detailed account of gene transmission in shepherd's 1972J STEBBINS— EDGAR ANDERSOX: RECOLLECTIONS 375 purse, for which he used the then current generic name Bursa, and was il- lustrated by an endless succession of impossibly detailed typewritten charts. In the middle of this performance I felt a hand on my shoulder, took a note from Edgar, who was sitting behind me, and opened it up. It read: "Times are bad and getting worsa, But G. H. Shull still diddles with Bursa.'' At the end of the talk, I turned around and remarked: "You ought to be made poet laureate of this session." At which, Edgar leaned back in his chair, spread out his arms, and let out a characteristic, resounding guffaw. It coincided exactly with the end of the modest applause that Dr. Shull received for his paper. Nearly everyone looked around at us, and I could only imagine what they were thinking. The Chairman, Dr, Castetter, a tall man with impressive dignity and enormous bushy eyebrows, transfixed me with a stare that made me feel like crawling out of the room under the seats. After the session, I went up to Dr. Shull to apologize and explain, but found him more interested in discussing with me our mutual interest in genetics. A few years later, when I gave lectures on plant evolution at Columbia University in New York, Dr. Shull came up from Princeton to hear me, and we developed a most pleasant acquaintance. The longest continuous association that I had with Edgar Anderson was in 1943, when he came to Berkeley on a Guggenheim Fellowship, in order to be associated with my then chief. Professor E. B. Babcock, as well as with C. O. Sauer in the Geography Department. One incident that occurred during this period is best related in Edgar s words, as I heard him tell it many times to mutual acquaintances. "When I was in Berkeley during the war, Ledyard and I used to have lunch together in the warm sunny alcove outside of his laboratory on the top floor of Hilgard Hall. We had many good talks together, but I'll never forget his reaction to something that I did a little while after I arrived in California. You see, almost all of the men were gone, and I ])ad heard that the sororities were having trouble in finding help. So I went to one of them, offered my services, and they hired me as a waiter. When I told Ledyard this, he looked at me, all of the blood of his strict New England ancestors curdled in his veins, and he blurted out: *Edgar, you'd do anything to be bizarre, wouldn't ?' your Evening seminars with Edgar, Professor Sauer, and the few students who were around during that period were memorable occasions. I learned to look for and find evidences of man's impact upon his environment far greater than I had previously suspected. Theories about the conditions under which agri- culture developed were new to me, and although I could not accept all of these that Anderson and Sauer suggested, they stimulated all of us to a lot of careful thinking. The same can be said about their theories regarding pre- Columbian exchanges of crop plants between the Old World and the New. That spring, we had a day together in the field that remains as a high point in my long experience of plant hunting excursions. Our objective was to obtain precise data on variation in the flowers of Iris longipetala, an endemic of the San Francisco Bay Region, which at that time was abundant on the -t I J ' -^ 376 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 lower slopes of San Bruno Ridge, on the outskirts of South San Francisco. The March sun was warm and inviting, and the green hills were flushed with the pale blue of hundreds of Iris flowers. We sat down in the middle of one of the larger patches, and Edgar inspected carefully the delicate, fragile petals of the nearest flowers. He could see immediately that flowers on each plant had their own intricate and distinctive design of lines and spots on the outer petals and asked for my help in reducing the patterns to a diagrammatic order that could be expressed in quantitative terms. Then in the middle of the most intense observation and recording of data, he leaned back, spread himself out on the soft grass, looked up at the blue sky and remarked: "Aren't we just the two luckiest people on earth to be out here in this gorgeous sun- shine, looking at the incredible beauty of design in these flowers, and calling it ^Work. After Western ?> I made a tour of several Mi 3efore settling down to my J Lectureship at Columbia. Not unnaturally, the tour began when I left my westbound family at St. Louis, and visited for several days with Edgar, who was then occupying the home on Flora Place that was his before he became Director of the Garden. We spent evenings in his garden, looking at the native perennial species of sunflower that he was cultivating there, and discussing the origin of temperate North America's most distinctive cultigens, the Annual J A memorable trip to Gray Summit, including my first introduction to the fascinating vegetation of limestone glades, was a high point of this visit. Another interesting conversation occurred when he showed me paintings done by an artist friend of his, w^ho had been captivated by the beauty of design that one can see in histological, sectioned and stained preparations of human viscera. "You know," said Edgar, "she came to me before she put together a one-woman show of these paintings, asking permission to borrow a portrait that she had made of me the previous year. Here it is. You see, I was working out at Gray Summit on a warm summer day, and had taken off my shirt and undershirt so that I would be more comfortable. She thought that the people who came to see the exhibit, and who might be repelled by nothing but human viscera, would feel better if a more conventional portrait were included. But tlic result turned out to be exactly the opposite. When she asked her friends how they liked the exhibit they replied: 'Well, the livers and kidneys were all right and rather fascinating, but why did you have to include that huge, monstrous bare torso!'" At the end of this visit, Edgar drove me to Urbana for my next visit, at the University of Illinois. "You know," he said, "that your host in the Herbarium will be George Neville Jones. I have known Jones for a long time. He first came from the state of Washington, where he was rather a lone wolf and little known. He went to Harvard, where he worked under Elmer Merrill at the Arnold Arboretum. He was very much under Merrill's thumb, shy and very subdued. But then he came to Illinois, where he was the only taxonomist, could pose as an authority from Harvard, and had the job of building up the 1972] STEBBINS— EDGAR ANDERSON; RECOLLECTIONS 377 Herbarium. In this connection, he built himself up, too. Every time I have visited him he has displayed a little more self confidence, until now, he greets me as one African potentate to another." The simile was so apt when we actually arrived that I had difficulty in restraining my laughter. A few years later, in 1951, Edgar came out to the meeting of the Society for the Study of Evolution in Berkeley. Both of us were on the Executive Covmcil of this Society, which had its annual meeting at the home of Vice President Ralph Chaney. The meeting was preceded by a lunch in the patio, but before we sat down we were given a tour of Dr. Chaney's unique garden. As a paleontologist, he had assembled all that he could of the "living fossil" plants that are now relics of past geological periods and had arranged them into assemblages representing the forest communities that dominated past geological epochs. As we were shown the distinctive characteristics of each species, 1 lingered behind the group in order to inspect more carefully a shrub with which I had been previously unacquainted. In doing so, I failed to notice that the branch that was closest to me for inspection contained at its other end a paper wasps' nest. The wasps, however, did notice me, and I was impelled to make the most rapid trip through the Eocene, Oligocene, Pliocene, and Pleistocene up to the recent security of the lunch table that any traveller has ever made. As soon as I had related my troubles, Edgar came out with one of his immortal couplets: "The day was hot, our host was gracious, But Ledyard got stung in the early Cretaceous.'* The following year, Edgar was elected President of the Botanical Society of America. This honor led to his being upstaged in the deftest manner possible. Professor Katherine Esau is not only one of America's most eminent botanists: she also has a dry, rapier-shaip wit. At the annual banquet, the President has, among his other duties, that of introducing the speaker of the evening, who is always the Society's President for the previous year. This was Edgar's task for Dr. Esau at the banquet held at Cornell As might have been expected, he was not thoroughly at ease before such a large gathering, and he showed this self consciousness by preceding his introduction with a long peroration about his being confused with Dr. E. G. Anderson of Cal Tech, about his aversion to college deans, of which the one at Cornell was a happy exception, and various other irrelevant matters. When he finally called attention to Dr. Esau and her botanical eminence, his audience had become somewhat fidgety and impatient. She got up, and in her quiet, matter-of-fact tone of voice replied with this succinct remark: "I have read in a book of etiquette that the function of the master of ceremonies is to draw attention away from himself and toward the speaker of the evening. » During the same year, I received a copy of his new book, Phnts, Man and Life, characteristically inscribed: "To Ledyard from Edgar with many thanks." I, of course, felt that the thanks should go in the other direction. Rarely has the first reading of a new book given me such delight. It is Edgar Anderson, tlirough and through. His chief technical contribution to science, the picto- 378 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 59 rialized scatter diagram, appears there in its final and most useable form, in a chapter, entitled characteristically, "How to Measure an Avocado." Another gem is his diagrammatic map of an orchard-garden in Guatemala, showing how a community of cultivated plants can be made compact and harmonious. His devotion to the lore of the origin of cultivated plants, and his appreciation of the complexity that apparently simple problems concerning them may present, is set forth in his chapter on sunflowers. Sprinkled through the book are thumbnail sketches of personalities, living, historical, and partly fictional or anonymous, like the field taxonomist who rushes past the fascinating, unsolved problems of tropical crop plants and their associated weeds, in order to concentrate on the "tiny patches of cloud forest" that cover the remote ridges rising above tlie plains. Then oblivious of the awe-inspiring beauty and solemnity of the scene, he 'rushes about with a great excess of energy, throwing the plants into presses, searching here and there for something yet uncollected." The characterization that appealed to me the most, however, was that of the medieval herbalist, Leonhard Fuchs, whom he visualizes as "a big, broad- shouldered Honry-the-Eighth sort of man, with handsome clothes and a general air of getting things done.'' Couldn't there be something of a self portrait included, either consciously or unconsciously, in this characterization? When Edgar came to Berkeley in 1953 for his second year in Sauer's Department, I was already in Davis, and we saw each other only occasionally. He did, however, come out to Davis to give a seminar to the Agronomy Department about his work with corn and preceded it with one of his favorite stories. According to him, in an unnamed agronomy department of a Middle Western university, an unimaginative professor assigned to an equally unimaginative student the task of distinguishing between oranges and apples. Both of them were first rate statisticians, and so they went about their task in a businesslike, objective, and thoroughly scientific fashion. The student measured heights, diameters, and circumferences. He obtained weights and specific gravities. Then he used chi square tests, t-tests, correlation coefficients, and all of the other accepted statistical methods for combining data. At the end of six months' work, the student reached the firm, statistically irrefutable conclusion that, while one could always distinguish populations of oranges from populations of apples, it is impossible to distinguish a single orange from a single apple! The story impressed me in two ways. First, Edgar had the nerve to tell it to an audience of agronomists, and he got away with it. Second, the kernel of ti'uth that it illustrates impressed me then and still impresses me. If one is to use statistical methods with biological material, one cannot succumb to the temptation to select parameters of measurement because they are the easiest ones to handle statistically. Usually, the best distinctions between two different kinds of objects involve such subtle non-metric characters as (in oranges and apples) shades of color, roughness versus smoothness of surface, chemical content, etc. My personal experience has been many published statements to the effect that, for instance, a particular chromosomal race cannot be 1972] STEBBINS— EDGAR ANDERSON: RECOLLECTIONS 379 distinguished from another are based only upon an inadequate survey of a few easily measured characters. During his visit of 1953, Edgar and I decided to record in print the outcome of our numerous discussions on the subject of hybridization between plants in nature. We agreed that the important fact about this phenomenon is not its taxonomic but its ecological and evolutionary consequences. Our observations of many plant groups had convinced us that, almost regardless of the degree of fertility of an Fj hybrid, the chances that such a hybrid will contribute significantly to future gene pools and so to evolutionary change depends almost entirely upon the environment in which the hybrid is growing. If this en- vironment presents an array of new and unoccupied or poorly filled ecological niches, some of the hybrid progeny are likely to be capable of filling them. On the other hand, if the hybrid is formed in a closed habitat, which presents no ecological opportunities for the establishment and spread of its descendants, it will have no effect on evolution, regardless of its vigor or fertility. The purpose of our sole joint contribution, published in the American Naturalist in 1954, was to point out that mans disturbance of natural habitats, although it is at present the chief source of evolutionary opportunity, is by no means the only kind of significant disturbance that exists and has existed in the past. Glaciations, volcanic eruptions, advance and retreat of epicontinental seas, the rise and extinction of large herbivorous animals: all of these, at one time or another, have favored the evolutionary success of hybrid progeny. Neither he nor I have seen any reason since then for deviating from this belief. My last extensive visits with Edgar were during 1958, when he was a Fellow of the Center for Advanced Studies in the Behavioral Sciences at Stanford. Our discussions during this time were largely reminiscences and reviews of ground that we had already covered. After he left Stanford, I saw him only a few times. He had made his impression upon me and had given me a great deal to live by. Not all of our discussions consisted of science and witticisms. I cannot con- clude without referring to Edgar Anderson's great faith in mankind, which led him to adopt and follow zealously the Quaker religion and way of life. He accepted family tragedies calmly and resolutely. His inner conflict with him- self was never wholly resolved, but he never wavered in his belief that he could make life better for others by his kindness toward them, and his ability to share with them his extraordinary perception of the wonders of plant life, and what plants could mean to people. - ij- ^ ' ^ *F-i '_-»'jj— i-f - I I ^ ' HYBRIDIZATION, EVOLUTION, AND SYSTEMATICS Duncan M. Porter to stud It was appropriate that the Missouri Botanical Garden's Seventeenth Annual Systcmatics Symposium be dedicated to the late Dr. Edgar Anderson, who was associated with the Garden for almost 45 years .^ Dr. Andersons long series of researches on certain aspects of hybridization, and especially his book Intro- gressive Hijhridization (Wiley, New York. 1949), have profoundly influenced ridization under natural conditions in the 1930's, most investigations of the phenomenon were made in the garden and the greenhouse, Anderson took the investigator out of the laboratory and into the field, successfully integrating ecology and genetics. He also pioneered in using simple methods to solve complex infraspecific variation, and the effects of hybridization thereon. The Symposium was held in the Museum building of the Missouri Botanical Garden on Friday and Saturday, 16-17 October 1970. About 225 botanical and zoological systematists, representing 65 institutions in the United States and Canada, attended. The theme of the Symposium was the importance of hy- bridization in evolution and its effects on systematics. Six papers were presented at the Symposium, with the resulting discussion directed by the Symposium moderator. Dr. G. Ledyard Stebbins, University of California, Davis. Speakers and their topics were Dr. Lewis E, Anderson, Duke University, "Population cytology, hybridization, and systematic relationships in mosses"; Dr, Leslie D. Gottlieb, University of California, Davis, ''Levels of Dr. Norton Nickerson, the ?? Tufts University, "Introgression in Hudsonia (Cistaceae)"; Dr. Sarah B. Pipkin, Howard University, "Introgression between Drosophila sibling species in Pan- ama"; Dr. Robert K. Selander, University of Texas, "Biochemical genetics of hybridization in European house mice"; and Dr. Lester L. Short, American Museum of Natural History, "The significance of hybridization in avian evo- lution and systematics." J versity, presented a colorful evening talk on "Student days with Edgar Anderson ^Or how I came to study sunflowers" at Wilson Auditorium, Washington Uni- versity. The papers of Drs. Anderson, Gottheb, Nickerson, Pipkin, and Short, plus the talk of Dr. Heiser, and reminiscences of Edgar Anderson by Dr. Stebbins follow in this issue of the Annals of the Missouri Botanical Garden. All contributed to a successful Symposium. Dr. John Finan, The American University, and Erna R. Eisendrath, Wash- ington University, have contributed, respectively, a biography and bibhography 1 Missouri Botanical Garden, 2315 Tower Grove Avenue, St. Louis, Missouri 63110. Present address: Department of Botany, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. 2 The Symposium was supported by National Science Foundation Grant GB-22057. Ann. Missouri Box. Card. 59: 380-381. 1972] PORTER— HYBRIDIZATION, EVOLUTION, AND SYSTEMATICS 381 of Dr. Anderson. These two papers, truly labors of lovc^ have been added to those from the Symposium to complete the section of this issue of the Annals dedicated to Edgar Anderson. My acquaintance with Edgar Anderson was all too short. I was struck by his eager willingness to help others to understand problems of variation. His ability to quickly grasp a complex problem and to explain it clearly to others is sorely missed. _Hi X ^ - ^—■--..ajt^ ^'T . _t . ■_: ^_^' i_i _ra_ .V. 1^1- ^^-^-_, - CYTOLOGICAL STUDIES OF NATURAL INTERGENERIC HYBRIDS AND THEIR PARENTAL SPECIES IN THE MOSS GENERA, ASTOMUM AND WEJSSIA^ Lewis E. Anderson^ axd Betty E. Lemmon^ The relationships of a complex of species within or near the moss genus, Weissia Hedw., have puzzled bryologists for more than a hundred years. The main problem is the uncertain status of two segregate genera, Hijmenostomiim R. Br, and Astomum Hampe. Recurring reports of intergeneric hybrids between species of all three genera have tended to weaken the case for maintaining separate genera, yet the segregate genera are based on characters which usually are considered strong. An opportunity to study meiosis in two natural inter- generic hybrids, Astomum ludovicianum Sull. X Weissia controversa Hedw. and A. muhlenhergianum (Sw.) Grout X W. controversa, initially prompted this study. Taxonomic relationships within the Weissia complex and arguments for and against maintaining one or more of the segregate genera have been discussed by Lindberg (1879), Andrews (1920, 1922, 1924, 1933), Hilpert (1933), Grout (1938), Jensen (1939), Chen (1941), Steere, Anderson and Bryan (1954), Podpera (1954), Nyholm (1956), Demaret and Castagne (1964), Reese and Lemmon (1965), and Williams (1966). Tlie distinguishing characters of the three genera, as presently understood, are summarized in Table 1. Astomum includes species with immersed capsules in which an operculum is lacking or very poorly differentiated and non-functional. Capsules are cleistocarpous, breaking open irregularly at matvuity. Peristome is lacking, and there is no membranous covering or "hymenium" at the mouth of the capsule. Hymenostomiim embraces species with exserted capsules, a functional oper- culum, and a thin membrane which covers all or a portion of the mouth of the capsu