i W ANNALS OF THE MISSOURI BOTANICAL GARDEN ^ ^ 'r.. f N Annals of the Missouri Botanical Garden Volume XVII 1930 With Fifty-one Plates and Twenty-two Figures Published quarterly at 8 West King Street, Lancaster, Pa., by the Board of Trustees of the Missouri Botanical Garden, St. Louis, Mo. Entered as second-class matter at the post-office at Lancaster. Pennsylvania, under the Act of Mardb 3, 1879. 54304 Annals of the Missouri Botanical Garden A Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Graduate Labora- tory of the Henry Shaw School of Botany of Washington Uni- versity in affiliation with the Missouri Botanical Garden. Editorial Committee George T. Moore Benjamin M. Duggar Information The Annals of the Missouri Botanical Garden appears four times during the calendar year: February, April, September, and November. Four numbers constitute a volume. Subscription Price $6. 00 per volume Single Numbers 1 .50 each Contents of previous issues of the Annals of the Missouri Botanical Garden are listed in the Agricultural Index, pubhshod by the H. W. ^^^ilson Co. Jh r STAFF OF THE MISSOURI BOTANICAL GARDEN Director, George T. Moork. Assistant to the Director, Katiierine H. Leigh. Hermann von Schrknk, Pathologist. Jesse M. Greenman, Curator of the Herbarium. (On leave of absence) .Adele L. Grant, Acting Curator of Herbarium. Edgar Anderson, Geneticist. Ernest S. Reynolur, Physiologist. David H. Linder, Mycologist. Anton Hogstad, Jr., Pharmacognosist. Roland V. La Garde, Research Assistant. Robert E. Woodson, Jr., Research Assistant. Nell C. Horner, Librarian and Editor of Publications BOARD OF trustees OF THE MISSOURI BOTANICAL GARDEN President, George C. Hitchcock Vice-President, Samuel C. Davis. Second Vice-President, Daniel K. Catlin. L. Ray Carter. Thomas S. Maffitt. George T. Moore. Albert T. Perkins. Philip C. Scanlan. Ethan A. F. Shepley. Fred G. Zeibig. EX-OFFICIO MEMBERS: George R. Throop, Victor J. Miller, Chancellor of Washington University. Mayor of the City of St. Loida Frederick F. Johnson, Bishop of the Diocejse of Missouri. # A. F. Satterthwait, President of The Academy of Sci- ence of St. Louis. Emil J. Barth President of the Board of Education of St. Louis. Daniel Brbck, Secretary, TABLE OF CONTENTS PAGE Studies in the Apocynaceae. I. A Critical Study of the Apocynoideae (With Special Reference to the Genus Apocy- num) Robert E. Woodson, Jr. 1-212 Studies in the Umbelliferae. III. A Mon- ograph of Cymopterus Including a Criti- cal Study of Related Genera... .Mildred E. Mathias 213-474 General Index to Volume XVII* * Since Volume XVII contains but two papers, and both have complete indices^ the General Index to the volume has been omitted. Annals of the Missouri Botanical Garden Vol. 17 FEBRUARY-APRIL, 1930 Nos. 1-2 STUDIES IN THE APOCYNACEAE. I^ A Ceitical Study of the Apocynoideae (With Special Reference to the Genus Apocynum) ROBERT E. WOODSON, JR, Research Assistant^ Missouri Botanical Garden Formerly Rufus /. Lackland Research Fellow in the Henry Shaw School of Botany of Washington University Table of Contents I. The status of the Apocynoideae 2 Introduction 2 Systematic history of the Apocynaceae 2 Morphological and phylogenetical discussion 10 The genera of Apocynoideae 39 Key to the genera of Apocynoideae 40 Summary 40 II. A monograph of the genus Apocynum 41 Introduction 41 History of the genus 44 General morphology 47 Intra-generic relationships 69 Geographical distribution 80 Economic uses 81 Taxonomy 83 Excluded species 144 Abbreviations 149 List of exsiccatae 150 III. Trachomitum: a new genus of Apocynoideae 156 Key to the species 158 Abbreviations 164 List of exsiccatae 164 An investigation carried out at the Missouri Botanical Garden in the Graduate 1 Laboratory of the Henry Shaw School of Botany of Washington University, and submitted as a thesis in partial fulfillment of the requirements for the degree of doctor of philosophy in the Henry Shaw School of Botany of Washington University, Issued May 14, 1930. Ann. Mo. Bot. Gard., Vol. 17, 1930 (1) 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN (Vol. 17 IV. A revision of the genus Poacynum 164 Key to the species 166 Abbreviations 168 V, Index to species 168 1. The Status of the Apocynoideae INTRODUCTION Somewhat less than four years ago, when the writer commenced a revision of the genus Apoeynum, the possibility of further studies in the family Apocynaceae was reserved for the future. However, it soon became apparent, as the study progressed, that the problem was so complicated and had been neglected so long that it could be dealt with satisfactorily only in intimate conjunction with a thorough investigation and readjustment of the entire family. The situation was complicated, moreover, not only with numerous Apocynaceae, but with Asclepiadaceae as well. A glance at the synonymy of Apoeynum, for instance, will show that that genus alone has been entangled with no less than thir- teen genera of Apocynaceae, and, even more surprising, twenty- two genera of Asclepiadaceae. Observation of other genera of Apocynaceae has shown that the situation of Apoeynum is repre- sentative of the group as a whole. With such conspicuous need of an extensive revision, the present study has been prepared as the forerunner of a group of similar papers which will be concerned with the comparative morphology, taxonomy, distribution, and system of Apocynaceae, and, it is hoped, later of Asclepiadaceae as well. SYSTEMATIC HISTORY OF THE APOCYNACEAE Previous to Tournefort, practically all of the plants then known which are now distributed between the families Apocynaceae and Asclepiadaceae were included in one monstrous group Apoeynum, a Latinized form of a name of Dioscorides, 'A'j:6kuvov, applied to a plant of Greece used as a poison for wild dogs and other animal pests. This plant, according to Dioscorides, was distin- guished above all by its milky juice, and consequently any plant which happened to yield latex was apt to be named Apoeynum, 19301 WOODSON — STUDIES IN APOCYNACEAE. I 3 first by the followers of Pliny, who evidently deserves credit for the composition of the name, and later by the disciples of the revered physician Galen. One generic name, therefore, charac- terized multitudes of pre-Tournefortian polynomials applied to milk-yielding plants which were indiscriminately Euphorbiaceae, Asclepiadaceae, or Apocynaceae. By the time of Tournefort, students of botany had become sufficiently keen-sighted to eliminate the various Euphorbiaceae (chiefly species of Euphorbia) from the melee which was the genus Apocynum, but most of the genera then known, which are now understood to be Apocynaceae or Asclepiadaceae, were yet in- cluded under that single name. The generic splitting of Apocynum began with Tournefort,* who recognized three generic entities in the old group. These genera, Apocynum, Periploca, and Asclepias, were deftly dis- tinguished both verbally and pictorially and heralded the dis- tinction of the families Apocineae and Asclepiadeae which was to be made by Robert Brown over one hundred years later after the prestige of the artificial system of Linnaeus had been overthrown by Jussieu and his contemporary naturalists. Seven other genera now placed among the Apocynaceae were also recognized by Tournefort, which, however, were not understood by him to have any direct affinity with Apocynum. These genera are Rauvolfia, Cerbera, Viiica, Nerium, Plumeria, Cameraria and Tahernaemontana, The sexual system of Linnaeus proved artificial indeed in the manner with which it treated the genera of Apocynaceae and Asclepiadaceae. The 'Genera Plantarum'^ split the aggregate of the two families between the "pentandria monogynia," including the genera Rauvolfia, Cerhera, Vinca, Nerium, Plumeria, Camer- aria, and Tahernaemontana of Apocynaceae and Ceropegia of Asclepiadaceae, and the "pentandria digynia," including the genera Periploca, Cynanchum, Asclepias, and Stapelia of Asclepi- adaceae and the genus Apocynum of Apocynaceae. The natural system of Jussieu* proved its worth by the man- * Tourn. Inst. ed. 2, 2: 91-94. 1700. 3 Linnaeus, Gen. PI. ed. 5, 98-102. 1754. Bartl. Ord. Nat. PI. 203-205. 1830. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 5 men; and Rauwolfea, genera with follicles connivent into a bi- or mono-spermous drupe at maturity and seeds with little albumen. The system of Endlicher^^ divided the Apocynaceae into three tribes and four sub tribes. These divisions were enunciated as follows: Carisseae: ovary simple, bilocular or unilocular, fruit baccate, or rarely capsular; Ophioxyleae: ovary double, fruit drupaceous; and Euapocyneae: ovary double, fruit bifollicular, occasionally pulpy. Under the last tribe, there were four sub- tribes as follows: Plumerieae: seeds ecomose, peltate; Alstonieae: seeds peltate, ciliate-comose; Echiteae: seeds basally comose; and Wrighteae: seeds apically comose. This system was accepted by Spach^2 and others. Alphonse de Candolle^^ in accepting Lindley's family Apocy- naceae, redistributed the genera into original tribes much re- sembling those into which they are at present found. The family as a whole was defined almost exactly as Brown had outlined it in 1809. The seven tribes recognized by de Candolle are as fol- lows: Tribe I. Willughheiae: ovary simple, unilocular, two placentae, ovules amphitropous, seeds naked, calyx eglandular; Tribe II. Carisseae: ovary simple, ovules amphitropous, fruit a drupe or a berry, seeds naked, calyx eglandular; Tribe III. Plumerieae: ovary double, ovules amphitropous, fruit a drupe, a berry, or a true follicle, seeds naked; Tribe IV. Parsonsieae: ovary single, bilocular, seeds apically comose, cotyledons con- volute; Tribe V. Wrightieae: ovary double, ovules amphitropous, folUcles two, seeds basally comose; Tribe VI. Alstonieae: ovary double, ovules amphitropous, follicles two, seeds comose through- out; Tribe VII. EcMteae: ovary double, ovules amphitropous or occasionally anatropous, follicles two, seeds apically comose. Mueller^* in 1860 divided the family into six tribes: Tribe I. Allamandeae: anthers completely fertile, ovary single and uni- locular, fruit a capsule or a drupe; Tribe II. Carisseae: anthers completely fertile, ovary single, uni- or bi-locular, fruit a berry or a drupe; Tribe III. Ophioxyleae: anthers completely fertile, "Endl. Gen. PL 577-586. 1836-1840 (1841). 12 Spach, Hist. Nat. V6g. 8 : 499-503. 1839. "A. DC. in DC. Prodr. 8: 317-489. 1844. » Muell.-Arg. in Mart. Fl. Bras. 6': 6-7. 1860. 6 ANNALS OF THE MISSOUBI BOTANICAL GARDEN [Vol. 17 ovary double, fruit drupaceous; Tribe IV. Plumerieae: anthers completely fertile, ovary double, fruit follicular; Tribe V. Ma- louetieae: anthersj_fertile Jonly towards the apex, seeds comose over the entire surface; Tribe VI. Echitieae: anthers fertile only towards the apex, seeds apically comose. In 1876 the tribes of Apocynaceae were reduced by Bentham^* to three: Carisseae: including de Candolle's tribes I and II; Plumerieae^ including de Candolle's tribes III and VI; and Echi- tideae^ corresponding to de Candolle's tribes IV, V, and VIL I Miers,^® in 1878, divided the family into an extremely elaborate system, but one that appears highly artificial and ambiguous : i 'Class L HAPLANTHEREAE adnate, or often dorsally attached to a soft connective. "A. Ovule and seed anatropous. "Tribe 1. Ophioxyleae: fruit drupiform and indehiscent, either single or double; cells monospermoua .... "B. Ovule and seed heterotropous, when the hilum is on the middle of one of the faces, equidistant from the radicular and cotyledonary extremities. ''Tribe 2. Carisseae: fruit drupaceous, indehiscent, 2-locular, or 1-ceIled by abortion; seeds imbedded in pulp, . . . embryo straight, in albumen, ''Tribe 3. Willxighbeieae: fruit drupaceous, indehiscent, 1-2-celled; seeds compressed, imbedded in pulp; embryo straight, without albumen . . . : thus differing from the preceding in the absence of albumen. ^'Tribe 4. Thevetieae: 1 or 2 indehiscent drupes, 1- or 2-celled, cells or pseudo- cells monospermous; seeds without pulp, oval, fleshy, peltately affixed, exal- buminous; embryo with a short radicle. "Tribe 6. Hunterieae: 1 to 5 oval dry indehiscent drupes, monospermous: seed laterally attached by a central hil with *' Tribe 6. Aspidospernwae: 2 follicles, rarely combined into a 2-locular fruit, or generally only 1 by abortion . . . ; seeds large, parallely adjacent, extremely compressed, broadly winged all round a central embryoniferous scutcheon . . . '* Tribe 7. Allamarideae: a single orbicular subcompressed dry capsule, unilocular , . . ; seeds not very numerous; testa oval, with a broad thick callous margin (instead of a wing) surrounding a flat embryoniferous scutcheon "Tribe 8. Plumerieae: 2 large, divaricated, thick, Unear-oblong folUcles, opening along their ventral suture, which expands internally into 2 flat septi- form placentae: seeds many, sub-compressed, furnished below with a lacerated wing, and affixed peltately and imbricately upon each semiseptum , . . '* Tribe 9. Alyxieae: 2 lomentaceous flat indehiscent follicles, transversely articulated into several dry, monospermous cells • . • » Benth, in Benth. & Hook. Gen. PI. 2: 681-728. 1876. " Miers, Apoc. S. Am. 6-10. 1878. 19301 WOODSON — STUDIES IN APOCYNACEAE. I 7 "Tribe 10. Crasjiidospenneae: an elongated subcompressed, 2-locular capsule, resolvable into 2 follicles by the splitting of its bilamellar dissepiment . . . "Class II. symphyanthereae: stamens connivent, each with 2 parallel anther-cells introrsely fixed upon a much longer horny connective, usually membranaceous or cuspidate at the apex, and terminating below in 2 longer or shorter forks, the pollen-cells adhering to the clavuncle of the style, all thus held together in a cone. "Tribe 11. Tabemaemontaneae: 2 follicles, ovoid or oblong, often pointed, dehiscing along their ventral suture, whose introflexed margins are semini- ferous ... "Tribe 12. Malouetieae: 2 follicles, Unear, oblong, or terete, dehiscing along their ventral suture, the margins of which are thickened introrsely into a solid, resilient placenta, seminigerous on both sides . . . "C. Seeds anatropous without an apical coma. "Tribe 13. Robbieae: 2 long follicles dehiscing along their ventral suture, the margins of which expand internally into 2 septiform membranaceous placentas bearing many imbricated seeds, which are oblong, . . • generally clothed with many long, soft hairs . . . "Tribe 14. Odontadenieae: 2 elongated follicles (or 1 by abortion) dehiscing along their ventral suture, the margins of which are invariably expanded into 2 broad septiform placentae . . . ; seed long, terete, erect, narrow at its two extremities, everywhere bare . . . "D. Seeds anatropous with an apical coma. "Class III. echiteae: seeds linear oblong compressed or terete, often with an elongated rostrum terminated by the micropyle, which is surrounded by a cup-shaped ring, bearing a crown of 1 or 2 series of long hairs, usually called a COMA. "Tribe 15. Macrosiphonieae: 2 very long subtorulose follicles, dehiscing along the ventral suture, the narrow margins of which are inflected and semini- ferous . . . Low erect, or prostrate plants, with a few axillary handsome flowers, having an extremely long narrow tube with a broad rotate border. "Tribe 16. Stipecomeae: 2 follicles with very thick pericarp, sometimes very long, rugous or verrucose, dehiscing along the ventral suture, either with 2 placentas conjoined by a keel attached to the suture and then detaching itself, or else narrower remaining separately attached to the suture . • . "Tribe 17. Wrighiieae: an oblong 2-celled capsule, sphtting septicidally through a thick bilamellar dissepiment (becoming like 2 follicles) . . • "Tribe 18. Preslonieae: an oblong 2-celled capsule, sphtting septicidally, as in the preceding tribe - . - : seeds many, imbricated, oblong, furnished at the apex with a long coma, near which they are suspended . . . "Tribe 19. Dipladenieae: 2 long terete folUcles dehiscing along the ventral suture . - . The chief peculiarity consists in a disk of 2 flat opposite lobes alternating with the ovaries. "Tribe 20. Prosechiteae: 2 oblong or terete folUcles . . . disk urceolate, entire, or more often partly cleft into 5 or 10 lobes. "Tribe 21, Mesechiteae: 2 long terete or torulose follicles, dehiscing along the ventral suture . . . Corolla small, or of moderate size, with a short tube, and rotate segments simply convoluted . . ." 8 ANNALS OF THE MISSOURI BOTANICAL GARDEN lVOL.t Asa Gray,^'' in systematizing the Apocynaceae of North America in 1878, reduced the subdivisions of the family to two, namely Plumerieae, characterized by free stamens (unconnected with the clavuncle), the cells of the anthers polliniferous to the base; and Echitideae, characterized by stamens closely connivent about, and appressed to, the clavuncle or stigmatic-head; and anthers largely sterile and appendiculate, polliniferous only near the apex. In thus basing his classification, Gray plainly followed Mueller in using the stamen as the important criterion of cleavage. In 1895 Schumann' 8 raised Gray's two tribes to the rank of subfamilies, calling them Plumeroideae and Echitoideae respec- tively, and subdivided them into tribes as follows: Subfam. I. plumeroideae: Tribe 1. Arduineae: ovary syncarpous, not divided to the base; Tribe 2. Pleiocarpeae: ovary divided to the base, follicles more than two ; Tribe 3. Plumiereae: ovary syncarp- ous, divided to the base, follicles two; Subfam. II. echitoideae; Tribe 1. Echitideae: anthers inserted within the corolla-tube; Tribe 2. Parsonsieae: anthers exserted beyond the corolla-tube. Since 1895, systematic works have dealt more generally with phylogenetic speculations among groups of families rather than with the detailed system within the individual family, and in the case of the Apocynaceae, Schumann's system has remained the most recent until the present. This historical sketch of the system of the family Apocynaceae, although incomplete and lacking in detail, is sufficient to illustrate the statement that the problem has never been studied by a con- sistent monographer who was willing to give undivided attention to the group. The work of Miers, while a careful compendium, is limited in its scope, as the title suggests, and frequently is labored and ambigu- ously composed. It is always lacking the proper contrast to make a monograph usable. The monograph is moreover entirely with- out keys, except for a key to the genera of the tribe Tabernae- montaneae. In careful examination of the work, one is struck with the unfortunate fact that although the author makes in- numerable new combinations, species, and even genera, actually "A. Gray, Syn. Fl. N. Am. 2^: 79-85. 1878. " K. Sch. in Engl. & Prantl. Nat. Pflanzenfam. 4': 109-189. 1895. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 9 sixty per cent of the species and genera (frequently new) were known to the author only at second hand. Further reference to the work of Miers will perforce frequently be made throughout these studies. Since the last intensive work upon the system of the Apocy- naceae was published over a half century ago, it may not be sur- prising that further study of the group should result in an attempt to readjust its divisions. Recent studies during several years have resulted in the following new classifications of the family: Fam. Apogynaceae Lindl. Nat. Syst. ed. 2, 299. 1836; Endl. Gen. PI. 577. 1838; A DC. Ann. Sci. Nat. Bot. 3^: 235. 1844; in DC. Prodr. 8 : 317 1844; Benth. in Benth. & Hook. Gen. PI. 2: 681. 1876; K Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4^ : 109. 1895 Apocineae Juss. Gen. PI. 143-151. 1789, in pari.; R. Br. Mem Wern. Soc. 1:12-58. 1809. Subfam. I. echitoideae K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4^: 160. 1895, r in part. Stamens largely sterile, basally appendaged, the anthers 4-loc- ular, connivent about the clavuncle, pollen tetrads separating into individual grains; ovary apocarpous; corolla variously ap- pendiculate within, or naked; calyx bearing various glandular appendages within, or rarely (f) naked. Woody vines or clamber- ing shrubs, rarely rhizomatous herbs. Subfam. II. apocynoideae, n. subfam. Stamens largely sterile, basally appendaged, the anthers 4-loc- ular, becoming 2-locular at maturity, connivent about the clav- uncle, pollen in persistent tetrads; ovary apocarpous; corolla appendiculate within by 5 individual or coalesced flanges opposite the corolla-lobes and alternate with the stamens; calyx naked within. Rhizomatous herbs. Subfam. III. plumeroideae K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4^: 122. 1895 10 ANNALS OF THE MISSOURI BOTANICAL GARDEN |VoL. 17 Anthers completely fertile, not basally appendaged, 4-loc- ular, free, not connivent about the clavuncle, pollen tetrads sepa- rating into individual grains; ovary apocarpous or syncarpous, bi- or occasionally uni-locular; corolla variously appendiculate within, or naked; calyx naked, or bearing various glandular appendages within. Trees, woody vines, and shrubs, rarely rhizomatous herbs. Tribe 1. PLUMIEREAE K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4^: 122. 1895. Ovary apocarpous, follicles two. Sub tribe A. T abernaemontaninae K. Sch. I. e. 1895. Calyx bearing numerous glandular appendages within, fruit pulpy. Subtribe B. Alstoniinae K. Sch. l. c. 1895. Calyx unappendiculate within, fruit dry. Tribe 2. pleiocarpeae K. Sch. I. c. 1895. Ovary apocarpous, follicles more than 2. Tribe 3. rauwolfieae, n. tribe. Ovary syncarpous, bilocular, placentation axile. Tribe 4. arduineae K. Sch. I. c. 1895. n Ovary syncarpous, unilocular, placentation parietal. In the present studies, investigation was commenced about the genus Apocynum, so an attempt to express the natural relation- ship in the Apocynaceae naturally centers about Schumann's inclusive subfamily Echitoideae (including the new sub-family Apocynoideae) to which that genus pertains. MORPHOLOGICAL AND PHYLOGENETICAL DISCUSSION Before undertaking a discussion of the differences separating the Apocynoideae and the Echitoideae, a few words justifying the foregoing system of Apocynaceae as a whole may be a pwpos. It will quickly be perceived that the new system proposed is an almost complete reversal of Schumann's system elaborated in 'Die natiirlichen Pflanzenfamilien.' The general theme in Schumann's system is elaboration. In separating the two subfamilies Plumeroideae and Echitoideae 19301 WOODSON — STUDIES IN APOCYNACEAE. I 11 Schumann bears in mind the popular conception that sterili- zation of tissue is an "advanced" condition, and so places the divisions in the above sequence, since the former group includes genera with completely fertile anthers in contrast to the latter which includes genera with largely sterile, appendiculate anthers. Such a view is generally logical, but may be superseded by ad- ditional considerations. With the exception of the genera of Apocynoideae, excluded in the foregoing section, the Echitoideae as visualized by Schumann presents a group of genera which are relatively homogeneous from a morphological and phylogenetical point of view. All genera are distinctly apocarpous. All genera, with few exceptions, as Nerium, and with some modifications, such as coalescence (cf. Odontadenia, etc.) and reduction of the number of units (cf. Dipladenia, etc.), display a cycle of vestigial nectaries sur- rounding the carpels. All genera, as far as is known at the present, display the appendiculate condition of the calyx. The group as a whole is an exceedingly natural one. Such, however, is not the case of the subfamily Plumeroideae. The phylogenetical range is wide. Some genera, it is true, are so similar to those of the Echitoideae that only the complete fer- tility of the anthers bars them from that classification. Such genera, of course, are apocarpous and display a squamelliferous calyx and a cycle of nectaries about the gynoecium. However, gradations appear from that condition, through degeneration of 4 the calyx squamellae and vestigial nectaries with gradual coales- cence of the apocarpous carpels, to a condition of a syncarpous (occasionally even unilocular) gynoecium and a completely naked calyx and receptacle. In some genera, as in Stemmadenia and T abernaemontana respectively, the coalescence and subsequent disappearance of the nectaries are accomplished through adnation to the gynoecium, thereby exerting an unmistakable influence towards syncarpy^^; in others the nectaries disorganize without visible trace, as in Amsonia. The Plumeroideae, then, are a very heterogeneous group, and one which shows the effects of evolu- tion clearly. The tendencies manifest in it, moreover, are more Woodson Stemmadenia. Ann. Mo. 347^49. fig. 1. 1928 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 varied, and of a more ''advanced" nature than those in the Echitoideae. It appears unfortunate, therefore, that Schumann considered the mere sterilization of the anthers of such importance that he was led to interpret the Pluineroideae as demonstrating a "primitive" condition. In his tribal divisions of the Plumeroideae Schumann is rather difficult to follow. In his system lower genera have a syncarpous, unilocular ovary with two lateral parietal placentae, while more advanced genera pass through a syncarpous, bilocular stage with axile placentation to complete apocarpy. Moreover, lower genera have unappendaged, eglandular calyces, while higher genera gradually develop distinct, orderly calycine appendages. Such a system has philosophical justification, perhaps, but should be superseded by more natural considerations. In the system which is here proposed, therefore, the order of Schumann in the Plumeroideae has been reversed, and we pre- sume that lower, or more primitive, genera have an apocarpous ovary and progress through bilocular syncarpy with axile placen- tation to higher, or more advanced, genera with unilocular and An illustration of this furnished in text-fig. 1, which presents diagrams of ovary cross- sections of three representative genera of Plumeroideae, namely, Atnsonia, Amhelania and Allarnanda. It will be seen that in Allamanda the irregular, binate, parietal placentations are sterile in the central region, hinting strongly that a transverse fission of the septum separating a bilocular ovary has resulted in a more or less unilocular condition. Such a view of the advanced con- dition of the syncarpous unilocular ovaries is strengthened by recapitulatory evidence in the very young stages where they appear as bilocular with axile placentation, much resembling the diagram of Amhelania in text-fig. 1. We may be fairly sure, then, that as a matter of fact the bilocular axile type is the forerunner of the unilocular parietal type. Certain Gentianaceous ovaries, as those of Frasera and Halenia, unilocular with parietal placentation when mature, also display axile placentation when immature. Anatomically, an explanation of such a process of ovarian evolution appears relatively simple, since the "margins" of the apocarpous carpels forming the binate 1930] WOODSON — STUDIES IN APOCYNACEAE. I 13 axile placentations are never completely coalesced, and in mi- croscopical preparations may be seen to be anatomically separate (cf. text-fig. 4, diagram 6, and text-fig. 7, diagrams 10-14). Therefore, in order to obtain a unilocular parietal ovary from a bilocular axile ovary, it is necessary only to project the coales- cence of the carpels of the ovary and the widening of the breach between the halves of the binate axile placentae in either carpel, thus forming an unilocular parietal condition. As a matter of fact, the stages of such tendencies are frequently found in transverse sections of the ovary of Nerium, especially towards the apex of the ovary cavities. Fig. 1. Diagrammatic transverse sections of ovaries of various Plumeroideae ; Armonia ciliata; B: Ainbelania tenuiflora; C: Allamanda cathartica. The separation of the Ap(3cynoideae and the Echitoideae upon the basis outlined just preceding is supported by a morphological and anatomical study of the floral mechanism of those subfamilies. For this study, fresh flowering material was available only in the case of the genera Apocynur.i and Nerium. For an understanding of the exotic genera, herbarium specimens were employed. All material was embedded in paraffin, sectioned at 5-10 ;jl, and stained with anilin safranin used in combination with either licht gruen or gentian violet. The combination of Delafield's haematoxylin with safranin also proved very satisfactory. Herbarium material requires special technique in order to pre- pare it for the microtome, and to meet that need the following procedure is followed. The specimens are carefully chosen, care being taken to detach only fragments affected at a minimum [Vol. 17 14 ANNALS OF THE MISSOUKI BOTANICAL GARDEN by pressing. The fragments are placed in a vessel containing distilled water and heated to about 90 °C. The material is then placed in a Stender dish containing the heated distilled water, covered and placed in an oven thermostatically heated at about 60°C. Although the time of soaking in the oven varies with the condition of the fragments, twenty-four hours is probably a minimum and a week a maximum. After the material appears properly soaked, it is removed to a vial containing Farmer's Fluid (six parts absolute alcohol ; one part glacial acetic acid) . After at least twenty-four hours in that fluid, it is washed in absolute alcohol where it may be left for an additional twenty-four hours. Xylol can then be added gradually to the absolute alcohol, eventually attaining purity. The xylol is in turn gradually replaced by paraffin. It is well to leave the material in pure paraffin for some time before imbedding and sectioning, — a week proved favorable in most cases. While the schedule given above is purely tentative, and lacks refinements in many details, notably the clearing reagent employed, very satisfactory serial sections were cut of over forty genera of Apo- cynaceae by that method. Flowers treated in that manner proved satisfactory for morphological and anatomical study even after fifty to seventy years of desiccation. As outlined in the preceding section, the chief characters of cleavage between the Apocynoideae and the Echitoideae are the presence or absence of calycine squamellae, and the state of the mature pollen. A study of the pollen of the three genera of the subfamily Apocynoideae has demonstrated that the pollen of those genera always develops and matures in persistent tetrads, the grains never normally becoming separated. On the other hand, the mature pollen of the genera of Echitoideae apparently always occurs as single grains, the tetrads breaking up relatively early in the life of the anthers. These observations have been made after microscopic studies of the more important genera selected at random throughout the subfamily, including inhabitants of all the major land divisions of the earth. The Echitoideae share with the Plumeroideae the individual condition of the pollen grains at maturity. 1930] WOODSON — STUDCES IN APOCYNACEAE. I 15 The cal3rx of the Apocynoideae is naked within; while the calyx of the Echitoideae is glandular-appendiculate within, although the individual appendages, or squamellae, may be extremely reduced and inconspicuous. The squamellae are found most frequently in groups opposite the caljrx-lobes. In such genera as Prestonia and Echites there may be only a single squamella for each lobe of the calyx. The genus Cycladenia probably exhibits the most reduced form of tlie calycine appendages, where they appear merely as an almost, microscopic fringe. In Prestonia, however, they are very conspicuous, and frequently equal in size the lobes of the calyx. The squamellae are thought to be of phylogenetic interest because of their histology and teratology. For an histological study cf the squamellae of the Echitoideae, fresh material was available only for the genus Nerium. However, preparations were also made from herbarium specimens of Pres- tonia mexicana, Echites umbellata, Echites n. sp., Odontadenia speciosa, Macrosiphonia hyfoleuca, Mandevilla tuhiflora, Cycla- denia humilis, Neohracea valenzuelana, Elytropus chilensis, Wright- ia tinctoria, and Anisolohus Andrieuxii. Two species of Nerium were studied, namely, N. odcrum and N. Oleander, but since they proved practically identical from an histological standpoint, at least in regard to the details under observation, only the generic name will be used when reference to either species is made. Text-fig, 2, diagram 7, illustrates the floral diagram of Nerium. The flower is seen to consist cf a bicarpellate center (c) surrounded by cycles of fives. Between the calyx (a) and the corolla (h) and opposite the lobes of the former are situated the squamellae (d) in groups of various numbBrs of individuals. Plate 2, fig. 1, represents a longitudinal sec bion of an individual appendage sub- tending a lobe of the calyx. A central core (a) is composed of vertical, elongate cells which are tapered at either end, somewhat resembling vascular cells. No spirally thickened cells are dis- cernible in the organ itself, but one is occasionally able to find immediately below it, in the 1 hickened receptacle, traces composed of spirally thickened cells (c) leading directly to the squamella. At right angles to the central conducting core of the squamella are found horizontally compTessed cells rich in cytoplasmic con- tent, forming a conspicuous glandular tissue (b). The nuclei of 16 [Vol. 17 ANNALS or THE MISSOURI BOTANICAL GARDEN the glandular cells are found to be chiefly in the region of the proximal half near the transverse wall adjoining the ''conductive" or core tissue, apparently indicating a physiological relationship between the two tissues. Floral anatomy in recent years has endeavored more and more & 2 h 4 5 6 7 Fig. 2. Nerium Oleander. 1: habit; 2: ventral surface of corolla- lobe, illustrating the tridentate flange, a; 3 : ventral surface of calyx-lobe, illustrating the calycine squamellae, 6; 4: ventral surface of petalode, illustrating the staminal anthers, h, and rostrum, a; 5: connate mass of petalodes, illustrating the staminal anthers, h, and rostrum, a; 6: stamen, illustrating rostrum, a, and anthers, 6; 7: floral diagram illustrating the position of the calyx, a, corolla, 6, gynoecium, c, squamellae, d, and androecium, e. 1930] WOODSON — STUDI]]:S IN APOCYNACEAE. I 17 to explain structures casually described by systematists, and has thus contributed greatly to phylogenetical study. The morphol- ogist and the anatomist have frequently been aided or even inspired by teratological objects. The elder de CandoUe depended frequently upon teratological observations for his morphological interpretations, and even declared that it was his belief that the morphology of Begonia could probably best be clarified by the appearance of a monstrosity. Robert Brown interpreted several teratological structures as reversionary, and recently the theory of "Carpel Polymorphism "^o has had its inception after the ap- pearance of teratological carpel's of Matthiola. Generally speaking, the evolutionary theory has been immensely helped by the dis- covery and careful interpretation of monstrosities. With the exception of Vinca, Nerium is probably the most widely cultivated genus of Apocynaceae. The intensive cultiva- tion to which it has been subjected has produced several races of double-flowered forms. While the petalody is occasionally of the stamens, producing supernumsrary petals within the true corolla, by far the most frequent condition is for the supernumerary petals to be produced between the true corolla and the calyx. If such a double flower is dissected it will be perceived that the supernumerary petals are o:: very unequal size (text-fig. 3), grading from petals equallinj? the lobes of the true corolla to inconspicuous and malformed filaments much smaller than the lobes of the calyx. If the calyx is examined for squamellae in such a case, it will be found thj^t they are either absent, or extreme- ly reduced in numbers. That the double flowers of Nerium are produced by petalody of the sciuamellae has long been appreciated and the occurrence has been called ''adenopetaly" by Morren," who also noted it under somewhat similar circumstances in the flower of Lopezia. Among the Apocynaceae, adenopetaly has also been found to contributes to the double flower of Tabernae- montana Coronaria. If the supernumerary petals of Nerium are examined carefully (text-fig. 3), it will be seen that they are almost exactly similar, from the largest and most pei'fect to the most inconspicuous and "Saunders, E. R. Ann. Bot. 38: 451-482. 1923. *i Morren, C. 51&-524. 1850. 18 [Vol, 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN malformed, to the lobes of the true corolla if those organs were polypetalous. As a matter of fact, in the double flowers the hint of reversion is frequently heightened by the partial dialysis of the gamopetalous corolla. This similarity is strikingly illustrated by the presence of the conspicuous tridentate flange which the corolla lobes bear upon their ventral surface. Similar flanges are borne upon each of the supernumerary petals except when those bodies are reduced to mere filaments, which occasionally occurs. c b a A e f g k I. k Fig. 3. Adenopetaly in Nerium Oleander: A, lobe of normal corolla; a-m, super- numerary petals, illustrating gradation to squamellae. Histological preparations of the double flowers disclose the fact that the vascular traces which were seen occasionally to lead to the squamellae in the single flowers actually traverse the supernumerary petals in the double flowers. Plate 2, fig. 5, represents a cross-section of a supernumerary petal of Nerium illustrating the vascular system which is exactly similar in number of traces and cellular constitution to that of the corolla-lobes. Fig. 2, of plate 2, illustrates the longitudinal section of a super- numerary petal, to contrast with the longitudinal section of the squamellae in fig. 1 of the same plate. Text-fig. 2 illustrates the gross morphology of the flower of Nerium. Diagram 1 represents 1930] WOODSON — STUDIES IN APOCYNACEAE. I 19 the habit of the flower, diagram 2, the ventral surface of the corol- la-lobe indicating the tridentate flange (a), and diagram 3, the ventral face of the calyx-lobe indicating the group of squamellae (b). Text-fig. 3 illustrates t\w stages of adenopetaly in a double flower from a normal corolla-lobe (A) to the smallest of the supernumerary petals (w) . The significance of the tridentate flange upon the ventral face of the corolla-lobes and supernumerary petals is also interpreted through teratology, it is believed. Text-fig. 2, diagram 5, illus- trates petalody of the stamens of a double flower of Nerium. The connate stamens have iri this case formed a gamopetalous pseudo-corolla composed of five petaloid stamens. If this tera- tological structure is dissected, it is seen that each of the five lobes (diagram 4) is composed in a manner similar to those of the corolla (diagram 2) and the supernumerary petals of the adeno- petalous flowers (text-fig. 3 a--m). In some cases the staminode is completely sterile, while in others the ventral face of the organ bears two loculae of an anth(;r (6) similar to those of a normal stamen (diagram 6), although conspicuously reduced in size. If the diagrams are examined carefully and comparatively, it will be seen that the blade of the staminodes (and consequently the corolla-lobes and the adenopetalous supernumerary petals) is equivalent to the filament and connective of the stamen (dia- gram 6), while the lateral and median lobes of the flange are equivalent to the anther-lobes (b) and the elongate rostrum (a) of the stamen, respectively. Somewhat more precise ini'ormation concerning the vascular relationships of the single and double flowers of the Oleander may be obtained by an examination of microscopic preparations of serial transverse sections. Text-fig, 4 presents a series of diagrams of transverse sections of a single flower of Nerium to illustrate the origin of the vascular supply of the floral organs. Diagram 1 represents the vascular constitution of the pedicel jus' below the gradual enlargement of the receptacle. The stele is in the form of a hollow cylinder, and although circular in cross-section immediately after leaving the peduncle, it soon assumes the triangular form indicated in the diagram. Gaps are soon apparent in the triangular cylinder, as [Vol. 17 20 THE MISSOURI BOTANICAL GARDEN higher examined, and eventually the (presented by diagram 2 is reached, where the 1 2 3 c f S -■--■— 6 Fig. 4. Diagrammatic serial transverse sections of the receptacle of a single- flowered variety of Nerium Oleander. Explanation in the text. broken in two opposite places, and is rather conspicuously and regularly lobed throughout. 19301 WOODSON — STUDIES IN APOCYNACEAE. I 21 ' *■ I When the stage represented by diagram 3 is reached, the caly- cine traces (c') have in all but one case partially detached them- selves from the central cylinclerj and the coroUine traces (c") are evident as conspicuous alternate lobes. The calycine traces almost immediately resolve into a stout dorsal trace and two finer ventral traces, evidently illustrating a case of the fusion of three traces at the plat^e of departure from the central cylinder. After the complete departure of the calycine traces, the isolated corolline lobes unite into a fsirly well-defined five-lobed cylinder (diagram 4). This cylinder quickly becomes of relatively uniform thickness, after which, in the place of the five original lobes, ten lobes appear opposite and alternate with the calycine traces. These lobes eventually disengage into ten traces, five larger alter- nating with the calycine traces, and five smaller opposite traces. At the same time either cycle gives rise to numerous residual frag- ments, or bundles which remain towards the center of the axis (diagram 5). Diagram 6 illustrates the receptacle just prior to the full elaboration of the floral organs. Of the three cycles of traces other than the calycine, the five larger are found to give rise to the corolla-lobes, the iive smaller to the stamen-filaments (s) which are adnate to the corolla-tube, and the numerous resid- ual traces (r), frequently as many as forty or sixty, to the gynoe- cium. Although the reunited cylinder of the corolline and stam- inal traces regularly breaks up into ten traces, frequently extra traces originate (sq), and it is these traces (text-fig. 4, diagrams 5sq-(jsq) which, in longitudinal sections, are found directly be- the squamellar gnifi cance of these extra traces may be appreciated from the phenom- ena described in the following paragraphs. The early stages in the br(;aking up of the pedicellar stele are similar in both the double and the single flowers of Nerium, The stages diagrammed in t(3xt-fig. 4, diagrams 1-4, are almost exactly similar for bo th the sir igle and the double forms. However, after the reorganization of the central cylinder after the departure of the calycine traces, the stele breaks up into a larger number of traces than was observed in the single flowers. In almost all cases, this number was twenty, in the manner indicated in text- . \ ^ t I /■ ^f >V 22 [Vol. 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN figure 5, diagram 4. Subsequent sections (text-fig. 5, diagrams 5-6) proved that these twenty traces were destined for two -c r 1 C 1 2 8 Fig. 5. c* s r p H ... ff - — i — . s «1 6 agrammatic serial transverse sections of a double flower of Nerium Oleander. Explanation in the text. cycles of exterior supernumerary petals in addition to the cycle of gamopetalous corolla and the cycle of epipetalous stamens. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 23 The vascular traces traversing these supernumerary petals are similar to those of the true corolla in remaining unbranched until entrance into their respecti^'e organs. Both petals (including the supernumerary and the corolla-lobes) and stamen-filaments contrast vnth. the calyx-lobes in possessing simple traces while within the receptacle. While the supernumerary ])etals of Nerium were found in the vast majority of cases to be provided with vascular traces in the manner described in the preceding paragraphs, occasionally very reduced supernumerary petals were detected which contained no evident vascular supply. Thus in text-fig. 3, in addition to the corolla-lobe {A), the dissection of the particular double flower displays ten supernumerary petals supplied with conspicuous vascular traces {a-j) and four extremely attenuate appendages, larger than the normal squamellae, and evidently also greatly reduced supernumerary petals, but not vascularly supplied. Such extremely reduced supernumerary petals retain to a certain degree the glandular nature of the true calycine squamellae. In view of the evidence presented, it is thought likely that in the possession of calycine squamellae the Apocynaceae, as well as certain other families of the Contortae, as the Gentianaceae, suggest a rather striking relationship with polyandrous poly- petalous families of plants. Because of the extreme variability of the reversion of these organs in teratological varieties, due no doubt to very great reduction, the exact nature of a hypothetical polyandrous type cannot be described. However, it might be supposed that (from the evidence of the two cycles of vascularly supplied supernumerary petals) such an ancestor in all proba- bility displayed at least two additional cycles of floral organs to those which occur in the modern flower (single), and possibly two cycles besides those, evidenced by the non-vascular appendages frequently detectable. From what has been said in the preceding paragraphs, it will be perceived that the stamens, corolla-lobes, and supernumerary petals are believed to be homologous. It is also believed that in the single flowers the supernumerary petals are represented by the squamellae. In other words, the squamellae are held to re-p- resent staminodia. The case of staminody, however, is considered 24 ANNALS OF THE MISSOURI BOTANICAL GAEDEN IVoL. 17 SO extreme that constancy in numbers and arrangement has dis- appeared. The importance of the squameUiferous calyx, however tenuous the explanation, must be accounted for nevertheless be(!ause of its almost universality in the Apocynaceae and its frequency in other families of the Contortae, as in the Gentiana- ceae. Of course to clinch the theory that the squamellae represent r staminodia it would be necessary to find, perhaps, a teratological specimen upon which the squamellae had become not only petal- oid but antheriferous. However, this condition is virtually obtained by the presence of the dorsal flange, which we have seen in the foregoing paragraphs and accompanying figure to be the homologue of the anther. Reasons accounting for the presence of the ^^staminodiar^ squamellae outside the gamopetalous corolla will follow in a succeeding paragraph. The mechanics of such centripetal sterihzation as is projected to have occurred in some polyandrous form to produce a structure similar to the modern single flower of Nerium is evidently at work in many of the polyandrous Polypetalae of the world to-day. Such genera are Nymphaea^ Cactus^ Mentzclia^ and many another, each of which displays the effects of progressive centripetal ster- ilization of the multiseriate androecium, from exterior completely sterile petaloid laminae to completely fertile interior stamens. The suggested explanation of the single flower of Nerium merely requires a continuation of sterilization in addition to the oft- invoked processes of reduction, coalescence and adnation. An interesting and significant morphological feature which the Apocynoideae have in common with most Echitoideae is a cycle of usually five, fleshy, ovate-cylindrical nectaries surrounding the carpels (cf. text-fig. 10/). In the Apocynoideae the genera Apocynum and Trachomitum exhibit a cycle of five individual and distinct bodies, while in Poacynum the nectaries are more or less united into a fleshy annulus about the gynoecium. In the Echitoideae the nectaries may be either present or absent. The genus Nerium is a representative of the comparatively few genera in which they are lacking. When present, the nectaries usually number five, but may be reduced to three or two, as in Dipladenia. Although usually individual and separate, the nec- taries are sometimes coalesced into an annulus closely investing the gynoecium, as in Odontadenia. 1930] WOODSON— STUDIES IN APOCYNACEAE. I 25 Examined histologically, the nectaries piovide a fascinating study. Plate 3, fig. 2, represents diagrammatically a cross-section of the flower of Apocynum androsaemifolium. It will be seen that the nectaries (e) form a circle about the carpels (d), alternating with the filaments of the stamens (a). According to the diagram the carpels contain three traces each, as do also the nectaries. Plate 6, fig. 2, represents in more detail a longitudinal section through a nectary of the same species of Apocynum. It will be seen that the vascular element is practically identical with the traces found in the gynoecium and figured in the clavuncle in fig. 3 of the same plate. For a more detailed knowledge concerning the interpretation of the nectaries, the vascular system of the flower of Apocynum was studied by means of serial transverse sections in a manner similar to that by which the squamellae of Nerium were studied. The diagrams in text-figures 6 and 7 illustrate this phase of the inves- tigation, being numbered consecutively throughout the two figures. In the pedicel of a flower of Apocynum cannahinum the stele is at first found to form a continuous bicoUateral cylinder which is nearly circular in cross-section (diagram 1). Very soon, however, the triangular cross-section noted in the pedicel of Nerium re- places the circular (diagram 2). Still further towards the distal end of the pedicel the stele becomes still more angled, eventually assuming the five-angled appearance illustrated in diagram 3. This stage indicates the transformation of the pedicel into the receptacle. In diagram 4, the five angles noted in the preceding diagram have become very prominent, and alternating with them are discernible five additional, lesser lobes. In diagram 5, the condi- tion noted in diagram 4 has almost resulted in the disruption of the stele into ten discrete segments. These consist of a series of five prominently angled sectors, identifiable as the bases of the caly- cine traces (c'), and a series of five alternating broader segments destined for the corolla lobes {c"). In diagram 6, the condition in the preceding diagram has become modified bj'- the conspicuous bilobed appearance of the five prominently angled, or calycine, lobes of the receptacular 26 IVoL. 17 ANNALS OF THE MISSOUEI BOTANICAL GARDEN % c'- c*. 4 c c* re 6 c re' s * 8 Fig. 6. Diagrammatic serial transverse sections of the flowe cannabinum. Explanation in the text. s n V n c* Apocyn stele. As subsequent sections ealed. the dorsal lobe was destined for the lobe, whereas the ventral lobe would the staminal This condition 19301 WOODSON — STUDIES IN APOCYNACEAE. I 27 evidently explained by the adnation of the calyx traces, or rather by their failure to depart promptly from the receptacular stele. 10 cl n s c f- n n cl 14 Fig. 7. Continuation of fig. 6. This is appreciated by comparing text-fig. 6, diagram 6, with text-fig. 3, diagram 4, and text-fig. 4, diagram 3, of the genus Neriumy in which the calycine traces depart from the central 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 cylinder before the appearance of the staminal lobes. This situation, then, is an adnation of the calycine traces, rather than adnation of the staminal traces, as it might appear in longitudinal sections. In the same diagram (diagram 6), residual traces (re) appear within the central cylinder, thus appearing in Apocynum at about the same stage in the development of the receptacle as was observed in Nerium. In diagram 7, the central cylinder of the receptacle has at last broken up into calycine traces (c'), corolline traces (c"), and staminal traces (s). The residual traces {re') have formed them- selves into two irregular, lunate semicircles. This action of the residual traces may be visualized as a breaking up of the roughly circular form in diagram 6 {re), and a very decided ventral devel- opment, producing the in-rolled semicircles illustrated in diagram 7 {re'). In diagram 8, the first differentiation of the receptacle into floral organs is perceived in the appearance of the ovarian cavities. In the lower of the two cavities illustrated, the placentation has not yet formed, and the only vascular element is the semicircular formation of the residual traces. However, in the upper of the two cavities, the first evidence of placentation is perceptible, and coincidently it is found that the vascular equipment is nearly, although not quite, completed. The ventral, or ovuliferous, traces {v) of either carpel, as may be seen from this figure, are formed by the two in-rolled margins of the residual semicircular cylinders, while the lateral veinlets (r) usually found towards the base of the carpels also arise from the residual traces. In diagram 8, it will quickly be seen that midribs, or dorsal traces, are absent from the region of the ovarian cavities. However, the staminal trace nearest the region of the dorsal wall of the ovarian cavity where the dorsal trace should be found is discovered to be con- spicuously bi-lobed. The ventral lobe of this compound trace eventually proves to be the dorsal trace of the carpel {d), which is adnate to the dorsal lobe, or proper staminal trace (s). In the upper ovarian cavity illustrated in diagram 8, the lobing of the compoimd staminal-carpellary trace is almost complete, while in the lower cavity the lobing has just begun to be noticeable. In diagram 9, the elaboration of the ovarian cavities is completed. 1980] WOODSON — STUDIES IN APOCYNACEAE. I 29 The vascular constitution of either carpel is seen to consist of a dorsal trace, or midrib {d) ; two ventral, or ovulif erous traces {v) ; and several smaller lateral traces, usually completely unlignified, numbering about twelve. These lateral traces, formed after the disorganization of the semicircular cylinder of residual traces to which attention has been called in the preceding paragraphs, progressively reduce in number as additional sections are ex- amined. In diagram 10 of text-fig. 7, either carpel is seen to contain four lateral traces. In diagram 13 of the same figure, the lateral traces have completely disappeared. In diagram 9, also, the cycles of calycine (c'), coroUine (c") and staminal (s) traces are conspicuous, and in addition another cycle of traces, alternate with the staminal, and opposite the coroUine, are evident. These are the traces to the cycle of nectaries (n), and they arise from the lobing of the coroUine traces {c") in exactly the same manner as did the d<3rsal traces of the carpels from the compound staminal traces. The manner in which the traces of the nec- taries (n) emerge from adnation with the coroUine traces (c") is illustrated in diagram 9, where a large coroUine trace is found constricting to free the trace of the nectary to which it is ad- nate. Text-fig. 7, diagram 10, discloses for the first time the floral traces assuming their final position. Furrowing has begun to distinguish the carpels from the surrounding receptacle. In diagram 11, the calyx-lobes are becoming free of the receptacle, whereupon the single calycine strands separate into a large dorsal and two small ventral branches. Furrowing has completed the differentiation of the ovary walls. In diagram 12, the calyx-lobes are entirely free from the re- ceptacle, and each contains three traces. The nectary traces have likewise divided into three traces, and the shapes of the nectary- bodies have begun to be visible due to the inauguration of furrows. Diagram 13 proves that, even in the genus Apocynum which is described as possessing discrete nectaries, two of the bodies may develop together, evidently providing a transition to the coalesced annular nectaries of Poacynum. In diagram 13, also, furrowing, which has differentiated the nectaries, has served to demark the bases of the eoipetalous stamen-filaments. The coroUine traces 30 [Vol. 17 GARDEN (c") soon branch, forming a strong dorsal trace and two finer ventral branches. In diagram 14 the corolla-tube, adnate stamen-filaments, and nectaries are fully differentiated, as they are viewed in ordinary dissections. The lobes of the corolla each contain three traces, but the lateral traces of the calyx-lobes have disappeared, leaving only the midrib. The lateral traces of the nectaries have also disorganized, for the most part. The foregoing discussions of the floral anatomy of Nerium and Apocynum are incomplete and lacking in detail, but they are in- cluded in the discussion to aid in the interpretation of the vestigial structures believed to be present in those genera. A detailed account of the floral anatomy of many representative genera of Apocynaceae is in the course of preparation. From the immediately preceding paragraphs it has been ex- plained with regard to the anatomy of Apocynum that (1) the vascular supplies of the carpels and the nectaries are of similar origin; (2) the carpels and the nectaries are similar in containing dorsal and ventral traces; (3) the bodies of the carpels and the nectaries differentiate from the receptacle in a similar manner and at about the same time. An additional similarity of the carpels and the nectaries is the fact that both are made up of highly glandular cells, and thus assume a similar staining reaction in microscopical preparation. For these reasons, in addition to certain others, as similarity in shape {cf. text-fig. 10), the nectaries are interpreted in this study as sterile carpels, and are considered phylogenetical the Apocynaceae. Of course if one is merely inclined to take the view that the nectaries are vestigial, and consider their number and alternate position with the stamens, they might be interpreted as stami- nodia. Such an interpretation of an inner cycle of staminodia has been adopted by systematists with regard to the nectaries of Parnassia, and recently Mrs. Arber^s has expressed her approval of this view. On the other hand, subsequent studies may lead to the interpretation of these vestigial organs of Parnassia as sterile carpels, as in the case of Apocynum. " Arber, Agnes. On the structure of the androecium in Parnassia, and its bearing on the affinities of the genus. Ann. Bot. 27: 491-510. j^. S6. 4 figs. 1913. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 31 In a systematic arrangement of taxonomic units, one is forced to grant a prior position to some and a subsequent position to others. It is very difficult for most taxonomic systems to depict accurately the probable evolution and phylogeny of the particular group. In arranging the three subfamilies of Apocynaceae, one is confronted with the problem of which should be first enumerated, since that position is tacitly interpreted as the '^ most primitive." And although one realizes that evolution is so complex and phy- logeny so incomplete that the true situation is inscrutable, one is nevertheless forced to a conclusion, and must reconstruct his personal conception of what the whole edifice of nature might have been were all the elaborate units of structure still extant, with only the meagre materials which he has at hand. In deciding the question regarding the probable relation be- tween the three subfamilies of Apocynaceae, one is confronted primarily by considerations of distribution, habit and structure. Geographically, the Apocynoideae are sharply distinct from the other subfamilies of Apocynaceae in having a completely north temperate habitat. The genus Apocynum is found scatteringly in northern Mexico as far south as central Chihuahua, but there is apparently escaped from ballast. It is as nearly cosmopolitan over the United States and southern Canada, including Newfound- land, as almost any other genus native to this continent. Tra- chomitum is a more southerly genus, and is found in Adriatic Italy, in Asia Minor, and in Asia as far south as the Persian Gulf and as far north as the vicinity of Peiping. Poacynum is appar- ently confined to a small district in central Asia. Since the overbearing majority of all Apocynaceae is dis- tinctly tropical (on!}'- seven or eight genera out of a total of over two hundred are characteristically temperate), it has been pre- sumed that the group has probably had a tropical origin. Hence, with such an assumption, the Apocynoideae might be taken as representing an advance from the tropical condition of the Echitoideae and the Plumeroideae. Closely coinciding with the geographical conclusions are con- siderations of habit of the subfamilies. The Apocynoideae are rhizomatous herbs entirely, while the Echitoideae and Plumeroi- deae are woody vines, clambering or procumbent under-shrubs 32 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 or trees, primarily, since the herbaceous habit is only rarely found in those groups. However, of the two subfamilies the latter contains the more numerous and the more typical herbs. Am- sonia, Rhazya, and Vinca are rhizomatous perennials of the Plumeroideae, the former genus frequently developing a woody condition of the rhizome. Lochnera, of the same group, contains the only annuals of the entire family. In the Echitoideae the herbaceous habit is much more rare. Cycladenia is a small sub-alpine rhizomatous perennial. Several genera, as Macrosiphonia, Dipladenia, and certain species of the inclusive genus Echites (cf. E. stans A. Gray) have become diminutive under-shrubs, apparently on the verge of the her- baceous habit. The herbaceous condition is now almost uni- versally viewed as an adaptation from a woody condition, and the fact that the Apocynoideae contain only herbaceous genera in contrast to the predominately woody genera of the other sub- families should lend support to the hypothesis that that subfamily demonstrates a relatively "advanced" condition. From the standpoint of structure, consideration naturally cen- ters about the condition of the mature pollen and the interpre- tation of the vestigial organs. That the tetradenous pollen of the Apocynoideae should be considered as phylogenetically more advanced than the granular pollen of the other groups can scarcely be gainsaid. In the condition of the mature pollen, the Apo- cynoideae share with the Asclepiadaceae an interesting analogy with the Orchidaceae. Since the Apocynoideae are interpreted as phylogenetically more advanced than the Echitoideae and Plumeroideae in other particulars, it might be reasonable to suppose that in the former subfamily the calycine squamellae have been totally effaced. Another view might be that the group is descended from an entirely different phylum than are the Plumeroideae and Echitoi- deae. Additional anatomical studies, it is hoped, will contribute to the elucidation of this problem. However, since it is known that the squamellae are eventually obliterated in certain genera of the Plumeroideae, the former view appears the more plausible. Two recent memoirs concerning the phylogeny of the families of Contortae are of especial interest with reference to the family 1930] WOODSON — STUDIES IN APOCYNACEAE. I 33 Apocynaceae. The first is the portion of Wernham's^s treatises on "Floral evolution" concerning the Contortae. In his delib- erations, Wernham comes to the conclusion that the "apocynal stock," represented by the present family Apocynaceae, should be considered the most primitive group of the order Contortae and the Sympetalae as a whole. This conclusion is reached be- cause that family, in general, demonstrates the simplest condition of all the families in its apocarpous, bicarpellate gynoecium. Although agreeing perfectly with Wernham 's main thesis concern- ing the probable relationship between the Apocynaceae and the other families of the Contortae, several of his statements should be supplemented or modified in the light of the foregoing observations regarding the morphology of the family. In conjecturing a group of Archichlamydeae which could pos- sibly be regarded as offering a stock from which the Contortae (through the "apocynal stock") could be obtained, Wernham" exclaims: "A gulf is thus discovered between the Contortae and the Archichlamydeae which at first sight seems difficult to bridge ; for the combination of an isomerous alternating androecium with a bicarpellary and superior ovary is extremely rare in the latter series — the sole group in w^hich this combination occurs at all to any extent being the highly evolved Umbelliflorae, in which the ovary is inferior." Again he states: ''In the Archichlamydeae, however, we must contemplate the tendency to economy from the aspect of its progress rather than of its realization; and the rarity of forms which have fully worked out this tendency before adopting sympetaly is scarcely matter for surprise. For the ancestry of the Contortae we must look for a group in which a tendency to isomery of the androecium is definitely traceable, together with a tendency to a bicarpellary gynoecium; at the same time any tendency to epigyny must he absent, or practically absent. These conditions are satisfied by the stock represented in the Geraniales-Sapindales plexus, and by no other." Still again he remarks: ''Unfortunately, however, the Contortae have left no traces of their progress from polypetaly to sympetaly in the " Wernham, H. F. Floral evolution: with particular reference to the sympetalous dicotyledons. New Phytologist 10: 217-266. 1911. "/. c. 220. 1911. [Vol. 17 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN shape of pentacyclic forms; neither a second staminal whorl nor any hint of it ever occurs." The itahcs are ours. The treatises of Wernham, it may be said, are quite generally stimulating to the study of phylogeny, which, as some critics have remarked, is the only phase of taxonomy which at present appears to have any vitality. However, as is evident from the few quo- tations which have been included in the preceding paragraph, his opinions, like those of many another phylogeneticist, were very evidently formulated from faulty morphological observa- tions. Although by no means representing all points of objection and exception which might be made to Wernham's phylogenetical ideas, several of his misleading morphological remarks should be recalled, namely: (1) that the ovary of the Apocynaceae is superior, and hence that "any tendency to epigyny must be absent, or practically absent" in a prospective hypothetical group of ancestors for the family; (2) that in the Apocynaceae there are "no traces of their progress from polypetaly in the shape of pentacyclic forms"; and (3) that "neither a second staminal whorl nor any hint of it ever occurs." As early as 1857, Payer^" referred to the genus Apocynum as an excellent example of the ontogenetical adnation and coalescence of floral parts. In his morphological experiments Payer found that in very young flower buds the corolla of Apocynum is poly- petalous, the pistil completely apocarpous, and the stamens free from the corolla. To these observations of Payer, all of which have been independently confirmed during the course of these studies, additional ontogenetical details have been added. Diagr the development of the flower of Apocynum hypericifolium are presented in text-fig. 8. In diagram A, the ovary appears inferior and the young stamens epigynous and organically free from the corolla. The receptacle is adnate to the whole outer surface of the carpels except at the apex where the clavuncle has not yet been formed by the fusion of the carpellary apices. In diagram B, the flower has assumed its mature condition. The stamens are now perigynous, for the carpels have developed above the surface of the receptacle. The " Payer, J. B. 564-565 1857. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 35 bases of the stamen filaments, meanwhile, have fused in their superficial tissue with the base of the corolla-tube, thus developing their epipetalous position. However, as pi. 3, fig. 4 clearly demonstrates, this adnation of the filament to the corolla is scarcely complete adnation, for the vascular element of the stamen fuses with that of the corolla-tube only below the surface of the receptacle. The surrounding cells of the filament, moreover, never completely lose their identity in the adnation with the corolla. A B C Fig. 8. Diagrammatic longitudinal sections of the flowers of species of Apocynwn and Sarifraga: A, young bud of A. hypericifolium; B, mature flower of same; C, mature flower of S. nivalis. As text-fig. 8, diagram B, illustrates, the ovary of Apocynum never becomes completely free of the surrounding receptacle, for the loculae are seen to penetrate into it. This penetration of the loculae is also conspicuous in the transverse sections of Nerium (text-figs. 4r-5) and Apocynum cannabinum (text-figs. 6-7). This condition has been called ^ ^ subinf erior ' ' by Dr. A. M. Johnson^^ and certain other specialists who note its occurrence in the genus Saxifraga and throughout the family Saxifragaceae in varying 26 Johnson, A. M. A revision of the North American species of the section Bora- phila of the genus Saxifraga (Tourn.) L. Univ. Minn. Biol, Studies 4: 6. pL 3, 1923. [Vol. 17 36 ANNALS OF THE MISSOURI BOTANICAL GARDEN degrees. A diagram of the mature floral structure of Saxifraga nivalis in longitudinal section is presented in text-fig. 8, diagram C. That the floral structure of the Apocynaceae is fundamentally pentacyclic there can be little room for doubt. The floral en- velopes, the androecium, the groups of squamellae and nectaries, are invariably in fives. The gynoecium alone being bicarpellate, typically, would appear to detract from the 5-merous symmetry. However, a 5-8-carp ellate gynoecium actually occurs in the fam- ily in the genera of the tribe Pleiocarpeae. The case of the Pleiocarpeae can scarcely be viewed as other than an indication of the prhnitive condition of the family, in the light of what phylogenetic studies in general, and the floral anatomy of Apocy- num in particular, have taught. The anatomy and probable significance of the calycine append- ages have been elaborated in some detail in previous paragraphs. However, it might be appropriate to add in this place reasons accounting for the existence of staminal vestiges outside the gamopetalous corolla. If gamopetaly is considered as an adapta- tion from polyretaly, and there are few indeed who do not accept such an interpretation, we have in the five-lobed corolla of the Apocynaceae the homologue of five individual petals. Within such a polyj etalous flower, let us conjecture an androe- cium of two or mere cycles of fives surrounding the partially infericr, bicarpellate gynoecium. Such a flower is more or less typical of the polyi etalous Saxifragaceae. In the Saxifragaceae, m( reover, the ruter cycle of stamens, alternating with the petals, frequently becomes adnate to the tube of the calyx. In such a case, if the outer cycle of stamens should become sterile during the process of centri] etal sterilization, a very common occurrence, the assumption of gamopetaly by the corolla would certainly exclude the staminal vestiges from the confines of the corolla in what might be considered a manner precisely similar to the exclusion of the calycine squamellae by the gamopetalous corolla of the Apocynaceae. On the other hand, let us imagine the remote ancestors of the Ajjocynaceae, as many another angiosperm, as possessing no sharply defined corolla, but only a many-seriate androecium illustrating centripetal sterilization, as in the flowers of Mentzelia, 1930] WOODSON — STUDIES IN APOCYNACEAE. I 37 etc. In such a flower could it not be conceived that very extreme sterilization^ adnation, and finally coalescence could produce such a flower as that of Nerium: the squamellae (which in teratology develop into two to several flanged petals similar to those of the gamopetalous corolla) representing the outer cycles of the androe- cium which have become extinguished due to the infinite contin- uation of centripetal sterilization and adnation to the calyx-tube; the gamopetalous corolla representing an inner cycle of sterilized micro-sporophylls which have coalesced; and the cycle of highly evolved and largely sterile stamens representing the innermost and last cycle of the once many-seriate androecium? Such an explanation of the condition of the single flower of Nerium is forcibly suggested by the appearance of the adenopetalous, or double, flowers, where the adenopetaly is frequently accompanied by dialysis of the gamopetalous corolla. From what has been explained concerning the anatomy and histology of the Apocynaceae in the foregoing paragraphs, we find that (1) the ovary of the Apocynaceae is not superior, but '^ semi-inferior, '^ as in certain families of the Rosales, especially the Saxifragaceae; (2) therefore the stamens should be considered as perigynous, and in young stages they are practically epigynous; (3) the ontogeny of such a flower as that of Apocynum indicates very clearly its progress from polypetaly in the late coalescence and adnation of floral parts; (4) pentacyclic gynoecia are present in the tribe Pleiocarpeae of the Plumeroideae ; (5) although the theory of the calycine squamellae as staminal vestiges has not been absolutely proven, such an interpretation is both attractive and plausible. In view of the preceding considerations, it appears rather un- necessary for Wernham to assign the ancestry of the Contortae to the ''Geraniales-Sapindales plexus, and ... no other/^ The Rosales, and in particular a more primitive condition of the present Saxifragaceous condition, provide a more accurate and a far more plausible aflSnity. Instead of appearing as a ^^gulf which at first sight appears difficult to bridge,'^ the hypothetical connection between the Archichlamydeae and the Sympetalae via the Saxifragaceae- and Apocynaceae-plexus appears astonish- ingly close and clear. The relationships among the families of Vol. 17 38 ANNALS OF THE MISSOURI BOTANICAL GARDEN Contortae are indeed fascinating, and it is intended that further studies will concern that problem. In 1922 there appeared a very significant paper by Demeter^' upon the morphology of the Asclepiadaceae. Although without the general consideration of the present studies, attention should be called to certain conclusions which were drawn. Briefly stated, the author found that the only real difference between the Apo- cynaceae and the Asclepiadaceae lies in the translators of the latter. The two families are also closest linked by the genera Apocynum and Periploca of either family through the pollen characteristics (persistent tetrads). Hence the author considered it proper to merge the two families into a single large group as Jussieu^^ conceived them. He also called attention to the fact that the tetradenous pollen of Apocynum forms a mass which may be transported in a single bulk of many tetrads by visiting insects, somewhat resembling the pollinia of the Asclepiads. Finally, the author reunites the two families and revises the bulk into three subfamilies: 1. Apocyneae; 2. Periploceae (including the genus Apocynum) ; and 3. A sclepiadeae . Demeter's reasons for reuniting the Apocynaceae and the As- clepiadaceae are scarcely original. Adanson, Giseke, and A. - L. de Jussieu advanced very similar opinions in the early years of the nineteenth century. A. de Candolle^^ retained the separation by Brown, explaining that he did so although aware that the only clear-cut distinction is the organization of the pollen. Like de CandoUe, Baillon^" also was forced to admit that: "Les Ascl4piad6es ont tous les caracteres des Apocyn6es, sauf ceux de leur pollen qui est reuni en masse." However, as Garcin^^ states: "... nous ferons remarquer que toute une tribu d'As- cl^piad^es, les P^riploc^es, qui comprennent de nombreux genres, ont un pollen libre et forment ainsi une transition insensible aux Apocyn^es. D'autre part le Vinca Rosea a un pollen agglu- " Demeter, K. Vergleichende Asclepiadeenstudien. Flora 115: 130-176. 1922. ••Jussieu, Gen. PI. 143-151. 1789. "A. DC. M6tQoire sur les Apocyndes. Ann. Sci. Nat. Bot. III. 1: 255-256. 1844. " Baill. Trait6 de botanique mddicale phan^rogamique. 1277. 1884. " Garcin, A. G. Recherches sur les Apocyn6e8. Ann. Soc. Bot. Lyon 15: 220- 221. 1888. 1030] WOODSON — STUDIES IN APOCYNACEAE. I 39 tin^." The same author, moreover, found that the histology of the two groups is practically identical. Although they certainly bear an extremely close affinity to one another, it is here considered better to continue the Apocynaceae and the Asclepiadaceae as separate families, even allowing the translator of the latter group, an extremely significant and specialized structure, to be the chief difference. However, the connection between the Apocynoideae and the Periplocoideae is unmistakable. At the conclusion of this morphological and phylogenetical discussion, the writer takes particular pleasure in acknowledging his gratitude to Dr. A. J. Eames, of Cornell University, who read the manuscript and encouraged its publication. THE GENERA OF APOCYNOIDEAE Among the three genera of the subfamily Apocynoideae several significant and easily distinguishable indicators of phylogeny appear. Of these perhaps the most conspicuous, as well as one of the most significant, concerns the composition of the inflores- cence. The inflorescence of Apocynum is a trichasium while that of Poacynum and Trachomitum is a monochasium. Phylo- genetically, the trichasium is almost universally conceded to be a more primitive structure than the monochasium, and the justice of that view is shown by additional characteristics of the three genera . The corollar appendages which have already been described with reference to Apocynum and Nerium are another source of valuable distinction. In the American genus Apocynum, as has already been explained, they are individual (c/. text-fig. 10, diagram 2g), while in the Eurasian genera Poacynum and Tra- chomitum they are coalesced into the form of a ring which appears more reduced in the former genus than in the latter (pi. 20). An additional character of undoubted phylogenetic significance is the state of the nectaries, which are distinct and separate in Apocynum and Trachomitum and more or less coalesced into an m Additional characters in the the two latter genera are the tubular-campanulate corolla and opposite leaves of the former, and the pelviform corolla and al- ternate leaves of the latter. 40 ANNALS OF THE MISSOURI BOTANICAL GARDEN rvoL. 17 KEY TO THE GENERA Inflorescence trichasial; corollar appendages distinct and separate, not coalesced into a ring; corolla glaljroiis or very rarely irregularly hirtellous; auricles of the anthers parallel; genus of North America I, Apocynum Inflorescence monochasial; corollar appendages coalesced into a ring; corolla uniformly pulverulent-papillose without; auricles of the anthers con- vergent; genera of Eurasia. Corolla tubulo-campanulate; nectaries distinct, not coalesced into a ring; pedicels single, not conspicuously paired; leaves opposite.. II. Trachomitdm Corolla pelviform, without a definite tube; nectaries more or less coalesced into a fleshy annulus about the carpels; pedicels conspicuously paired; leaves alternate III. Poacynttm Sufficient has been indicated of the relationship of the three genera of Apocynoideae in the preceding paragraphs ; details con- cerning each will be reported in the revisions which follow. SUMMARY 1 . An historical account is given of the systematic arrangement of the Apocynaceous genera from the time of Tournefort to the present. 2. A new systematic arrangement of the family Apocynaceae is presented, based partially upon the system of K. Schumann in Engler & Prantl's 'Die natiirlichen Pflanzenfamilien' and partially upon the results of original morphological investigations. 3. The genera Apocynum, Trachornitum, and Poacynum are segregated into a new subfamily based upon tetradenous pollen and an eglandular calyx. 4. The floral morphology and teratology of the genus Nerium are found to suggest explanations for several features of the floral organs of the Apocynaceae in general. 5. After a comparison of the calycine appendages or "squamel- lae'* of representative Apocynaceae with the morphology and teratology of the single and double blossoms of Nerium it is con- jectured that the calycine squamellae of numerous Apocynaceae may be interpreted as staminal vestiges which were excluded from the interior of the flower by gamopetaly. 6. The fleshy nectaries surrounding the carpels of numerous genera of all groups of Apocj^naceae are perceived to be anatom- ically and ontogenetically similar to the carpels and are inter- preted as a cycle of sterile carpels. 1930) WOODSON — STUDIES IN APOCYNACEAE. I 41 7. Close relationship of the Apocynaceae with the Saxifraga- ceae is indicated by (1) the bicarpellate and '^semi-inferior^' gynoecium; (2) the interpretation of the calycine squamellae as homologous to stamina! whorls; (3) the polypetalous condition of very young flower-buds; and numerous other details of the morphology, anatomy and ontogeny. 8. Two recent and important treatises dealing with the phy- logeny of the Apocynaceae are reviewed. 9. A key is provided to the genera of Apocynoideae, which are briefly described and contrasted. IL A Monograph of the Genus Apocynum INTRODUCTION ^^ When the Synoptical Flora, Vol. 2, part 1, was issued in 1878> only two species of Apocynum were recognized in North America? viz., A. androsaemifolium and A. cannabinum , and the distinction between them was none too marked. At the present time there are some thirty species recognized in North America, which as a rule are referred to one or the other of the two groups based on these original species. The result is that a sort of ^apocynal chaos' obtains such as we find existing in only too many of the so-called 'tough groups'.'^ Thus the author^^ of a new species of Apocynum prefaces his description. Upon bringing his observations to a close, he further remarks in none too reassuring a vein: ''The plant . . . cannot be referred to any recognized species, and as its characters are fully as distinctive as any of the genus, I have felt justified in proposing it as the type of a new species." Since the time of the above writing, this ''chaos" has become more and more perplexing, for in the period between 1910 and 1928 more than eighty additional North American species of Apocynum have been described, sixty of these having been pro- posed by the late Edward L. Greene. Beside these have accu- mulated many new varieties and forms. In spite of the many additions to the genus, however, many botanists still refer the various Apocyna to two or three species in identification, un- «MacGregor, E. A. Bull. Torr. Bot. Club 37: 261-262. 1910. [Vol. 17 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN doubtedly because of the tenuous differentia to be encountered in an attempt to segregate them more accurately. In 1913 a monograph of Apocynum was published by two Italian botanists/' which, however, for several cogent reasons is unsatisfactory to American botanists. In the first place, direct familiarity of the genus by the authors was confined to the few Eurasian representatives, constituting only about 10 per cent of the specific aggregate; knowledge concerning the North Amer- ican species, constituting fully 90 per cent of the genus as conceived by the authors, was limited apparently to specimens borrowed from the United States National Herbarium, Secondly, analytical drawings are not present save for a very few of a gen- eral and elementary nature, which are diagrammatic to the point of obscuring morphological details of considerable importance. Thirdly, the citations of the authors are practically useless and frequently absolutely unintelligible, — evidently merely guesses of what the English inscriptions upon the labels of the American specimens might likely be, producing page upon page of what Prof. Fernald,''* while referring to the paper in a recent review, has aptly called "Jabberwocky." All in all, the monograph is more of a curiosity than a scientific work. The condition of the genus at the present time not only war- rants another attempt at an adequate treatment of the group, but renders imperative a careful morphological study to substantiate the taxonomy of such a treatment. The present revision is an attempt to portray the evolutionary complexity of the genus as it appears to all who study it in the field; and at the same time to confine the specific concept applied to it within the bounds of practicality. To this end, extensive field studies have been made during five summers, embracing practically all of the species and varieties recognized. The collections of the genus in all the larger herbaria of the United States have been personally visited, and from several of the lesser herbaria abundant material for study has been borrowed through the courtesy of the various curators. Upon a smaller scale, breeding observations including both ^ B^guinot, A., & Belosersky, N. Revisione monografica del genere Apocynum Linn. R. Accad. Lincei Atti, Mem. CI. Sci. Fis. V. 9: 695-734. IB pi 12 figs. 1913. " Feruald, M. L. Science N. S. 68: 146. 1928. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 43 Apocynum and the closely related genus Trachomitum have been undertaken in the experunental gardens and greenhouses of the Missouri Botanical Garden. During the course of this study of the Apocynoideae, and in particular the genus Apocynum, two years were spent at the Missouri Botanical Garden, and one year at the Gray Herbarium of Harvard University. For the use of the herbarium, library, and laboratory facilities, as well as for the other conveniences of the Missouri Botanical Garden, the author is indebted to Dr. George T. Moore, the Director of the Garden. For the use of the excellent facilities for study at the Gray Herbarium, as well as for valuable advice and innumerable courtesies, he is likewise happy to acknowledge his obligations to Dr. B. L. Robinson, Curator of the Gray Herbarium, and also to the other members of the staff of that institution, particularly Prof. M. L. Fernald, and Dr. I. M. Johnston. In connection with his visits to various herbaria of the country, the author desires to acknowledge the courtesy and hospitality of Dr. N. L. Britton, Dr. P. A. Rydberg, and Mr. Percy Wilson, of the New York Botanical Garden; Dr. F. W. Pennell, of the Philadelphia Academy of Natural Sciences; Dr. W. A. Maxon, Mr. T. H. Kearney, and Mr. E. P. Killip, of the United States National Herbarium; Dr. J. A. Nieuwland, in charge of the herbarium of Dr. Edward L. Greene at the University of Notre Dame; and Messrs. P. C. Standley and J. F. Macbride, of the herbarium of the Field Museum. For the use of borrowed specimens of the genus Apocynum, the author is indebted to Dr. P. A. Munz, of the herbarium of Pomona College, to Prof. Harold St. John, of the State College of Washington, and to Mr. C.C. Deam, of Bluffton, Indiana, whose unexcelled private collection of Indiana plants proved an invalu- able addition to the specimens of the various endowed institutions. In the course of study of such a problem, material aid is fre- quently given by numerous friends, and for much assistance in the prosecution of these studies the author desires to thank the following: Dr. Edgar Anderson, Dr. D. H. Binder, Miss Nell Horner, Mr. Julian A. Steyermark, and Mr. Harry Fuller, of the Missouri Botanical Garden; Mr. Harry Piers, of the Provincial 44 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 Museum of Halifax, Nova Scotia; Mr. W. E. Saunders, of London, Ontario; and his father, Mr. Robert E. Woodson, Sr. Finally, the author wishes to express his gratitude for aid and courteous criticism to Dr. J. M. Greenman, Curator of the her- barium of the Missouri Botanical Garden. HISTORY OF THE GENUS It is generally admitted that Dioscorides was the originator of the name Apocynum, and the purgative potentialities of the 'Ax6kuvov, as his Greek original existed, were fully known to him. The medicinal use of the rhizomes of the plant were also appreci- ated by Pliny and Galen, who derived the word Apocynum. from the Greek. The plant was so named from a composition of the Greek words d^i, from, far from, hence meaning '^bane," and kuv6<;, of a dog, implying the use of a decoction of the plant as a poison for wild dogs and other animal pests. The identity of the plants called 'Axokuvov by Dioscorides, and Apocynum by Pliny and Galen, as true Apocyna in the modern sense, however, has been brought into question by some anti- quarians,^* who contend that the plants mentioned by the ancient writers were more likely referable to the genera Cynanchum and Marsdenia, both members of the Asclepiadaceae occurring in Mediterranean, Adriatic, and Aegean lands. Sibthorp'^ upholds such a view, and confidently refers the Apocyna of the ancients to Cynanchum erectum L. Recently the late Dr. Lunell/^ pursuing his views of absolute priority, has changed the name of the North American Apocyna to Cynopaema, upon the testimony of Sibthorp that the Apocynum of the ancients was in fact Cynanchum. Although the name Apocynum appears consistently throughout the herbals from the time of Pliny to the middle of the seven- teenth century, no direct proof that the name was then applied to a rightful member of that present genus can be found until 1669, when Robert Morison's Traeludia Botanica'" appeared with the *' cf . Bartolozzi, F. Memoria . . . sopra le quality che hanno i fiori della piante, &c. in Opusc. scelt. sulle scienze e sulla arti. p. 293, Milano, 1779. "Sibthorp, J. & Smith, J. E. Flora Graeca 1: 178. 1806. »^Lunell, J. Am. Mid. Nat. 4: 508. 1916. »• Morison, R. Prael. Bet. 12. 1669. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 45 first recognizable plate of that genus and with polynomial phrases which describe the plants fairly well for the first time. Seven Apocyna were listed and illustrated by Morison as representing all of the species known to him. With the aid of his plates and polynomials, it is fairly safe to state that of these seven various so-called Apocyna, three were undoubtedly true members of that present genus, and that all three were natives of North America. Thus, as early as 1(169 the American element of that genus was evidently the one associated with the name Apocynum. This is accentuated by the fact that the Linnaean Apocynum venetum, taken by some to be the type species of the genus, was yet in- cluded in the same work as " Tithymalus maritimus, venetus, . ." The three plants of Morison, taken to be referable to North American species of the present genus Apocynum, are his "Apo- cynum Canadense foliis Androsaemi," "Apocynum Canadense angustifolium flo. aureo," and Apocynum Canadense angusti- folium maximum flore minimo herbaceo." After each of these phrase names, the sign 3& appears, denoting that the combina- tion had originated with Morison himself. Tournefort,^^ in 1700, provided Apocynum the most competent treatment with which it had ever before been accorded. He recognized in his generic description two main divisions, the first coinciding well with our present genus Apocynum, and the second equally well with the existing genus Asclepias and its immediately neighboring genera. This was the first well-defined break in the solidity of the great amorphous aggregate then called by the name Apocynum. Tournefort made the distinction of his un- named divisions still more evident by referring each to a well- prepared analytical plate, presenting in themselves indisputable evidence of their individuality. In Tournefort's treatment of the genus for the first time appears the European element of the Linnaean genus Apocynum, A. vene- tum L. as "Apocynum maritimum, Venetum, Salicis foliis, siliqua longissima." Preceding the "Apocynum maritimum, . . .", however, appear Apocynum cannabinum L. as "Apocynum Vir- ginianum, flore herbaceo, siliqua longissima," and A. androsaemi- folium L. as "Apocynum Indicum, foliis Androsaemi majoris, »»Tourn. Inst. 2: 91-94. 1700. 46 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 flore Lilii Convalliuni serrarubentis," implying that the author was more familiar with the two preceding than with the following (( Apocynum maritimum, . . ." and desired the reader to be conscious of that fact. The fifth edition of Linnaeus' 'Genera Plantarum'*° was no improvement over Tournefort's 'Institutiones,' as far as the genus Apocynum is concerned. Although restricting the genus within smaller bounds than had Tournefort, the group was scarcely more homogeneous. Linnaeus' 'Species Plantarum'^' recognized five species of Apocynum, as follows: (1) A. fol. androsaemij*^ (2) A . cannahiniim, (3) A . venetum, (4) A . frutescens, and (5) A . reticuJalum. The first three species have remained until now as the traditional species of the genus, while the fourth and fifth have been transferred to other genera. It is not thought expedient here to devote further space to the history of the specific concept of the genus Apocynum. Suffice to limit the present remarks to less obvious features of the problem. In 1748 Heister''^ listed during the course of a synopsis the genus Apocynastrum, corresponding to the "Apocynum Canadense foliis Androsaemi" of Morison; separating it from the typical Apocynum as represented by Morison's "Apocynum Canadense angustifolium flo. aureo." Fabricius" in 1759 retained and am- plified the description of Apocynastrum, although he referred no species to it. Jussieu^* in 1789 altered the name of the genus to Apocinum, without transferring any species to the new generic spelling. Reference has already been made to the work of Robert Brown, which did so much to clarify the specific complex of the genus *" Linnaeus, Gen. PI. ed. 5. 101. 1754. « Linnaeus, Sp. PL ed. 1. 1: 213. 1753. « It is a little-appreciated fact that the first edition of the 'Species Plantarum' published the familiar species of eastern North America as "Apocynum fol. andro- saemi," and not as "Apocynum androsaemifoHum," which appeared in L. Sp. PI. ed. 2. 1 : 311. 1762; and several floras, overlooking this, have referred to the species as "Apocynum androsaemifolium L. Sp. PI. 1: 213. 1753." This latter familiar form Is used throughout this monograph, although the rights of the former are duly acknowledged. « Heistr. Syst. PI. Gen. 8. 1748. « Fabric. Enum. Meth. PI. 256. 1759. "Jusa. Gen. PL 146. 1789. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 47 Apocynum. Since the time of Linnaeus, species had been relegated to the genus carelessly and volubly, until a confusion had re- sulted which only the understanding of genius could clarify. The classic work of Brown"^ in ordering the chaos into two well- defined families, and finally into properly proportioned genera, will remain a monument for generations to come. Alphonse de Candolle^^ accepted the conclusions of Brown in his synopsis of the genus in the Trodromus,' and his revision remained the last until 1913 when the monograph by B6guinot and Belosersky appeared, to which reference has previously been made. Between 1844 and 1913, the only notable event in the history of the genus was in 1888, when Baillon,*^ after a study of the peculiar Asiatic element of the genus, segregated it from the mother genus, giving it for a name the anagram Poacynum. GENERAL MORPHOLOGY Roots. — The root system of all species of the genus Apocynum is extensive, filling the soil to a depth of three to four feet.^^ The root system is dual, comprising purely absorptive roots which are relatively fine and fibrous, ramifying the soil in the usual manner; and so-called ''gemmiferous" roots, which are relatively stout and woody and assume a more or less horizontal position, pre- senting the general appearance of a rhizome. Plate 5, fig. 4, illustrates the subterranean axes of Apocynum canndbinum. As indicated in the figure, the fibrous absorptive roots (/) are pro- duced freely along both the vertical rhizome (a) and the horizon- gemmiferous root (e), but The gemmiferous roots produce lateral buds (c) which at length de- velop into new vertical rhizomes. These new rhizomes are typical root-shoots, '*° such as those occurring in many arboreal genera as PopuJus and Rohinia, and correspond to the "additional root- « R. Br. I. c. 67-68. 1809, *' A. DC. in DC. Prodr. 8: 439-440. 1844. <» Baill. Bull. Soc. Linn. Paris 1: 757. 1888. « cf. Weaver, J. E. Root development of field crops, pp. 83-84. fig. 28. 1026. 5° Holm, Theo. On the development of buds upon roots and leaves. Ann. Dot. 39: 867-881. 1925. 48 ANNALS OF THE MISSOURI BOTANICAL GARDEN IVoL. 17 shoots" of Wittrock." The young gemmiferous roots {d) are in turn produced laterally from the vertical rhizomes. A fully mature gemmiferous root of A. cannabinum is usually .5-1.0 cm. in diameter, and may be several meters in length, giving rise to a dozen or more root shoots at one time. A single well-developed plant in a cultivated field may become spread by the plow-share to form colonies all over the field, the blade cutting the gemmiferous root into pieces and forming a new plant at each lateral bud. In this manner yI. cannabinum and its varieties, especially, have become common pests of agriculture. Plate 8 consists of two photographs of A. medium var. lividum growing in a cultivated field near Bayfield, Colorado. The photographs are unfortunately rather insufficient in detail, since an attempt was made to include as much in the plate as possible. Inunediately adjoining the areas photographed were several areas of equal or greater extent. The part which vegetative propagation has played in obtaining the infested condition of the field may be appreciated when it is understood that the variety in question is evidently absolutely sterile, since no fruiting specimens have ever been found either in herbaria or in the field. The field photo- graphed was visited in July, at which time not a single foUicle or any indication of one, either young or of previous years, was found by the writer among the several thousand plants which infested it (see also p. 72). A. androsaemi folium and its near relatives, although possessing the same general form of rhizome and gemmiferous root, are more frequently found in woodlands, and are much less apt to invade cultivated land than A. cannabinum or A. medium. A cross-section of a young absorptive root of A. hypericifolium (b) con- is presented in pi. 4, fig. 1. The epidermis (a) is seen t of a single row of cells, from which numerous root hairs duced. Within the epidermis lies an extensive cortex sisting of about ten rows of rather spongy parenchymatous cells. In the center of the section, and occupying about one-third of the diameter, is found the stele, bounded by a rather poorly defined endodermis (c). The vascular structure is tetrarch in all of the " Wittrock, v. Br. Om rotskott hos ortartade vaxter, med sarskild hansyn till deras olika biologiska betydelse. Bot. Notis. 1884: 21. 1884. 1980] WOODSON — STUDIES IN APOCYNACEAE. I 49 cases examined, but has been found in a single instance to be pentarch. The four alternating masses of phloem (e) and xylem (d) are embedded in a conspicuous pericycle (/) composed of small and closely packed parenchyma. Larger and more loosely ar- ranged parenchymatous elements occupy the center of the stele, and may be regarded as an elementary pith. A sector of a cross-section of a fully mature gemmiferous root is presented in pi. 4, fig. 2. The secondary structure in such a root is very conspicuous. The epidermal cells have ceased to exist as such, and constitute a dead outer sheath, beneath which is a conspicuous periderm. The periderm consists of a several- layered phellem (a), a conspicuous and active phellogen (6), and a relatively inconspicuous phelloderm (c). Within the periderm lies an extensive cortical zone (/), traversed by latex tubes (/i). The cells of the cortex contain much storage starch, which will be separately considered subsequently. Within the cortex Hes a broad cylinder representing a coincidence of secondary phloem and peribycle. Isolated areas of crushed and half-absorbed cells are conspicuous, which are thought to be the remains of the primary phloem and endodermis. A functional endodermis is evidently lacking. Occupying the interior of the stele is a broad cylinder of sec- ondary xylem (d), composed of parenchyma and conspicuous and very numerous vessels. Traversing the xylem are found conspic- uous storage rays (l) which extend through the phloem and cortex to the vicinity of lenticels {k) in the bark. A well-defined and active cambium (g) separates the phloem and the xylem. Stem. — The stem system of the genus Apocynum, like the root system, is dual, comprising a more or less upright aerial stem and a vertical subterranean rhizome, passing reference to which has abeady been made in the preceding paragraphs with regard to the root system. The rhizome appears to be developed almost entirely from the hypocotyl, in plants propagated from seed, stem-buds appearing upon that axis either laterally or in the axils of the cotyledons. In producing lateral stem-buds upon the hypocotyl, which in seedlings of all herbaceous Apocynaceae is a very conspicuous organ, the genus Apocynum differs from the genera Amsonia and Vol. 17 50 ANNALS OF THE MISSOURI BOTANICAL GARDEN im Rhazya, and is similar to the nearly related genus Trachomitum. The rhizome may be regarded as merely an underground portion of the true aerial stem, with which it is more or less perfectly con- fluent, the foliage leaves of the aerial portion grading uninter- ruptedly to the inconspicuous cataphylls of the rhizome. Since Apocynum is a north temperate genus, the chief distinction be- tween the proper stem and the rhizome is that the former is annual, while the latter is perennial. The first aerial stem of a young rhizome is terminal, but following ones, appearing after the destruction of the first, are always lateral. Since the internal anatomy of both rhizome and aerial stem is fundamentally identi- cal, as we shall presently see, the differences between rhizome and stem are general habit and longevity, production of roots by the rhizome, and the production of foliage leaves by the stem. An jortant additional difference is the method of branching, which is according to a definite plan in the stem, unlike the rhizome in which branching is evidently haphazard. Taxonomically and phylogenetically the stem of the various species of Apocynum offers characters of considerable value. The stem of A. androsaemifolium and A. pumHum and their varieties has a relatively lax, spreading habit, and the branches are pre- ponderantly alternate. The stem of A. cannahinum and its allied species and varieties differs in its opposite or sub-opposite, some- what fastigiate branching, and is chiefly erect in habit. The intermediate character of A. medium and its varieties is notice- able in the mode of branching, which is intermediate between that oiA. androsaemifolium and A. cannahinum. The branching is alternate to subalternate by the abortion or deferred devel- opment of an opposite axillary bud. The habit is also inter- mediate, since, although it is erect and not lax as in ^1. androsaemi- folium, it has not the fastigiate appearance of A, cannahinum. Ecological conditions easily influence the height of the stem. Plants of A. pumilum may be only two or three centimeters in height when growing at very high altitudes. A plant of ^. andro- saemifolium growing in an exposed, dry or sandy environment may produce stems only a few centimeters tall; while a plant of the same species growing in a shady, moist woodland may exceed a meter in height. Although in every other respect quite typical, 1930] WOODSON — STUDIES IN APOCYNACEAE. I 51 unusual developments of the stem due to ecological conditions have frequently been described as new species. The anatomy of the stem and rhizome presents interesting features for study. Although a detailed account of the develop- mental anatomy is without the scope of the present paper, it is thought that a general account of the cellular constitution may be helpful in securing a relatively broad conception of the genus. Plate 5, figs. 1-2, represent sectors of cross-sections of the stem of A. cannabinum and A. androsaemifolium, respectively, the former at a somewhat earlier development than the latter. In both certain anatomical characteristics are conspicuous. The epidermis proper (a) consists of a single layer of cells in which stomata are occasionally to be found. Immediately beneath the epidermis is a hypodermis (6) composed of cells somewhat similar to those of the epidermis, but appearing laterally more compressed. The cortex (c) is extensive, occupying from approximately one- eighth to one-third of the cross section. The cells of the cortex are characteristically spongy and parenchymatous; and contain resinous secretions occasionally, and stored starch grains almost invariably. A cross-section of the stem stained with a combination of safranin and licht gruen or gentian violet detects the presence of starch in the cortex. The individual grains are roughly el- liptical-discoid in shape, with a conspicuous and highly refractive hilum. Conspicuous and irregular fissures in the hila are almost always perceptible. Illustrations of the included starch from the cortex of A, cannabinum are presented in pi. 5, fig. 5. Within the cortex proper the coUenchyma differs from the strictly parenchy- ma by a lack or paucity of included starch. Within the cortex the pericycle (j) appears as prominent and isolated groups of fibers of great length and strength, the use of which by the aborigines of North America as a constituent of rope and thread has caused the popular name "Indian Hemp" to be given A . cannabinum. Associated with the pericyclic fibers are groups of latex tubes. The family Apocynaceae shares with other families of the order Contortae a bicollateral development of the stele. In the genus Apocynum the intraxylary phloem forms a conspicuous vascular rvoL. 17 52 ANNALS OF THE MISSOURI BOTANICAL GARDEN feature. The external phloem (e) forms a more or less continuous cylinder within the pericycle, and consists largely of fibrous elements. The irregular masses of pericycle, however, sometimes invade the zone of phloem, causing its disruption into discon- tinuous segments (fig. 2j) . The xylem (g) consists of a continuous hollow cylinder enclosing the internal phloem and pith. The woody elements are conspicu- ous and relatively few-angled. Vessels are not as numerous or as conspicuous as in the gemmiferous roots. The woody cylinder becomes greatly enlarged during the period of growth of the stem. In the young stem represented by fig. 1, the diameter is ap- proximately 1 mm. and there are about seven rows of cells in the xylem. In a stem of about 1 cc. in diameter, the rows of xylem cells number about one hundred. The internal phloem (/) in the genus Apocynum forms a rather loose, interrupted cylinder, separated from the xylem by a row or two of parenchymatous cells. The elements of the internal phloem are more parenchymatous than those of the external phloem. The origin of the internal phloem in dicotyledons has been a disputable subject. According to Scott and Brebner,^^' who devoted extensive research to the question, the internal phloem arises from the procambium. On the other hand, H^raiP^ and Lamounette" are of the opinion that the internal phloem origi- nates from the pith. In the course of the present studies it was not considered feasible to investigate this technical problem. The author, however, is inclined to accept the view of Scott and Brebner. The reader is referred to the literature. The center of the stele is invariably occupied by a conspicuous pith {h). In young stems certain cells display included starch grains (fig. li), but that condition is rather rare. Frequently isolated latex tubes (m) appear in the region of the pith bordering the internal phloem. " Scott, D. H., & Brebner, G. On internal phloem in the root and stem of dicoty- ledons. Ann. Bot. 5: 259-300. pi 18-20. 1891. " H6rail, J. fitude de la tige des dicotyl^dones. Ann. Sci. Nat. Bot. VIL 2 : 267. 1885. " Lamounette, B. Recherches sur I'origine morphologique du liber interne. Ann. Sci. Nat. Bot. VII, 11: 193-282. 1890. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 53 The internal anatomical structure of the rhizome (pi. 5, fig. 3) is practically identical with that of the stem, differing chiefly in the possession of a distinct periderm (6), and conspicuous vascu- lar rays (k). The vessels of the metaxylem of both the rhizome and the stem contain many-seriate, alternate bordered pits, closely pressed together and thus assuming a roughly hexagonal outline. Plate 4, fig. 6, illustrates the pitting of a tracheid of secondary xylem of A . cannabinum. According to Scott and Brebner" internal centripetal wood occurs in the stem of A . cannabinum, associated with the internal phloem. According to those authors internal cambium makes its appearance all around the pith, forming on its outer margin fresh cells of internal phloem, and on its inner isolated masses of xylem amounting to a maximum of twenty-five elements, equal to about one-eighth of the normal wood. No internal wood was discovered in the stems of Apocynum which were studied in the course of this study. This failure may be accounted for by the fact that Scott and Brebner examined wood of greater diameter than that studied by the present writer. Another source of error may be fluctuation of this phenomenon. The interpretation of the horizontal subterranean axis of the genus Apocynum as a gemmiferous root rather than as a horizontal rhizome is not of very common occurrence. Even anatomical studies have occasionally identified that axis as a rhizome.'* From the stem and rhizome, however, the gemmiferous root differs in several conspicuous and concise anatomical features. As will be seen from a comparison of pi. 5, figs. 1-3, and pi. 4, fig. 2, the stem and the rhizome possess pericyclic fibers, internal phloem, and central pith, as well as resinous cortical cells, which the gem- miferous root does not. The latex tubes of the stem and rhizome occur in association with the pericycle, whereas those of the gemmiferous root occur in the cortex. In addition, the xylem of the gemmiferous root differs from that of the stem and the rhizome in the greater size and number of vessels and the more parenchymatous nature of the cells. " Scott, D. H., & Brebner, G. I c. 283-285. pi 19, figs. IS-I4. 1891. ^ cj. Garcin, A. G. Recherches sur les Apocyn^es. Ann. Soc. Bot. Lyon 15: 210-221. 1888. [Vol. 17 54 ANNALS OF THE MISSOURI BOTANICAL GARDEN Leaves. — The leaves of the genus Apocynum are opposite, membranaceous; entire, and either distinctly petiolate or sessile. Superficially they may be either glabrous or variously pubescent. The outline and surface of the leaves are extremely polymorphic, which fact has been the cause of much of the taxonomic confusion regarding the specific complex of the genus. Leaf size, shape, and pubescence or glabrity are characters evidently to be con- sidered with caution. Two foliar characters are evidently reliable; however. These characters are the relative shape upon the pri- mary axis of the plant, and the position which the leaf takes with regard to the axis of the stem. Illustrating the former category, it can be shown that the majority of the species possess leaves which are nearly uniform throughout the plant, at least in shape; while others, as A. hyper-- icifolium and its varieties, have leaves which are relatively broad, blunt, sessile, and cordate or amplexicaul upon the primary axis, while those upon the secondary axes may be petiolate and rela- tively narrow, giving the plant a somewhat heterophyllous foliage. The second category is illustrated by the leaves of A. andro- saemifolium and its allies, which are drooping and somewhat pendulous; by those of A. medium and its allies, which spread at nearly right angles to the axis of the stem; and by those of A. cannahinum and its allies, which ascend from the main axis. Connecting the base of the petioles of each pair of leaves occurs an inconspicuous ring of triangular, membranaceous, pec- tinate glands 1 mm. or less in greatest dimension. These append- ages are superficial in origin, and whatever their origin or function is difficult to determine, since they present no distinct anatomical clue to their identity. They are probably best regarded as Lindley" regarded them, as vestiges of stipules, for a critical examination of them shows that they are connected by a shallow annular lamina. In studying the morphology and anatomy of the foliage of Apocynum, A. androsaemifoliu7n and A. cannahinum were selected " Lindley, J. Nat. Syst. ed. 2. 300. 1836. Lindley regarded the stipular glands wliich he observed in several genera of the Echitoideae as evidence of relationship with his family Cinchoniaceae, the stipules of which are conspicuous. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 55 as contrasting objectives. Representative sectors of cross- sections of leaves of either species are represented in pi. 6, figs. 5 and 4, respectively. The leaf in either case is a typical dorsi- ventral mesophyll. The epidermal layers consist of only one layer, in either case, and are composed of relatively even and regular cells upon the upper surface (a) , and relatively irregular and loose cells upon the lower surface (6). The mesophyll of either species is highly de- veloped. In A. cannahinum (fig. 4) the ratio of palisade to spongy parenchyma is about 3:2. The palisade cells (c) in that species are rather compact oblong-ovoid cells, and are arranged in two to several layers. The spongy parenchyma {d) is composed of larger, flabby cells, rather distantly and irregularly arranged, forming numerous air-chambers. The ratio of palisade to spongy parenchyma in leaves of A. androsaemifoUum is rather higher, about 3:1. The palisade cells (c) are more uniform than in the preceding species, in shape, size, and arrangement, constituting a single layer of cells. The spongy parenchymatous cells {d) are similar to those of A. cannahinum. The difference in the number of layers of palisade cells between A. androsaemifoUum and A. cannahinum, noted in the preceding paragraph, forms a tangible and evidently constant internal distinction between the two species. The number of layers is not dependent upon ecological conditions, as experiment has proven. Both in the studies of Mrs, Clements^* with the former species, and of those of the present investigation with the allied species of the genus, the number of layers of palisade cells has appeared constant. In the species intermediate between A . andro- saemifoUum and A. cannahinum the layers of palisade cells vary from one to several. Evidently stomata occur only on the lower surface of the foliage. These organs are most frequently accompanied by two or more subsidiary cells which are placed parallel to the pore, simulating the ''Rubiaceous type" of Solereder.^^ " Clements, E. S. The relation of leaf-structure to physical factors. Trans. Am. Microsc. Soc. 26: 19-102. 9 pis. 1905. "Solereder, H. Syst. Anat. Dicotyledonen, 911. 1899. 56 [Vol. 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN In the leaf laticiferous tubes occur only in the veins, and in no case were they observed among the cells of the parenchyma. Inflorescence. — The inflorescence of Apocynum is a trichasial cyme, varying greatly in size, shape, and construction. A cyme may produce but one to four flowers in some cases, while in others the flowers may be extremely numerous. In certain species, as in A. cannabinum, where the inflorescence is usually dense, each 1 a. 3 Fig. 9. Diagrams illustrating three characteristic inflorescences among species of the genus Apocynum: andros pedicel and peduncle is subtended by a distinct bract which may be scarious, as in A. cannabinum and its varieties, or semifoliace- ous, as in yl . hypericifolium and its varieties. In the case of others, as A. androsaemifolium and its allies, where the cymes are fewer- flowered, pedicellar bracts may be evident or virtually lacking. Text-fig. 9 illustrates very diagrammatically the three most X)rominent forms of inflorescence from which all other forms in the genus appear to be derived. Diagram 1, representing the 1930] WOODSON — STUDIES IN APOCYNACEAE. I 57 inflorescence of A. pumilum, shows the inflorescence to be chiefly terminal, but evidently axillary also in decreasing quantities along the upper nodes of the stem, receding from a multiparous cyme, usually composed of three fully developed cymes, to uni- parous axillary cymes in the axils of the upper leaves (which are usually somewhat reduced in size), and finally to single axillary flowers subtended by leaves of normal size. The first diagram demonstrates that the central cyme of the pleiochasium is the most extensive and perfect in form, branching regularly and determinately. The ventral terminal branches, however, are less extensive, and evidently less regular than the dorsal, demonstrating a suppression of the determinate flower or cymule. Finally, the small axillary cymes are found to con- tinue this suppression until the inflorescence is reduced to a single flower, and later disappears altogether. Diagram 2, representing an inflorescence of A. androsaemi- folium, illustrates the subsequent continuation of the tendencies noticed in the inflorescence of A. pumilum. The central cyme of the terminal pleiochasium is here also seen to be the most elaborate of the three branches of the trichasium, and nearer to the determinate type; but here, in its ramifications, is again noticed the suppression of the determinate member of the cyme, the central cymules being reduced as a rule to a single flower, and the central flower of the cymule frequently suppressed. The tendency toward the development of the paniculate type is shown by the two ventral cymes of the trichasium, where the determinate character has practically disappeared. Axillary flowers or cymules appear only abnormally. Diagram 3 represents an inflorescence of A. cannahinum, where the tendencies noted in the preceding paragraphs find final amplification. Here the inflorescence has become a panicu- late cyme, and axillary flowers are completely suppressed. The determinate character is also wanting, except theoretically in the triple multiplication of the terminal pleiochasium. Here also it is found that one of the lateral cymes of the trichasium, or not infrequently both of them, has acquired the power to continue vegetatively to form a new and complete stem with a terminal trichasium, and thus to continue and prolong the re- [Vol. 17 58 ANNALS OF THE MISSOURI BOTANICAL GAEDEN productive potentialities of the plant. In this manner a plant of A. cannahinum and its allies may continue in bloom and vege- tative growth from spring until late in the autumn, a distinct advantage over the much shorter periods of growth of the other species. A condensed view of the tendencies of the inflorescences il- lustrated appears to be to regard the entire process as the building up of a compact, specialized form from a branch of axillary- flowers through the reduction of foliage and the aggregation of flowers. From such a viewpoint, it would be reasonable to regard A . pumilum, with a nearly perfect determinate cyme and normal 1 axillary flowers, as the most primitive of the types discussed; A . androsaeniifolium, with a less determinate cyme and abnormal axillary flowers, and A. hypericifolium, with paniculate cymes, foliaceous bracts, and no axillary flowers, as intermediate types; and A. cannahinum, with a paniculate cyme, scarious bracts, and no axillarj^ flowers, as the most advanced. Flowers. — The floral anatomy of the genus Apocynum has al- ready been described in some detail (pp. 25-30). In this section the separate organs will be discussed in a general manner, only. The calyx of Apocynum is synsepalous. The five regular lobes are separate almost to the base, and may be greenish, reddish, or whitish in color, and pubescent or glabrous. It appears that the ecological situation of the plant has much to do with the pigmentation of the vegetative parts, including the calyx; an exposed sunny situation almost invariably imparting a ruddy tint to those parts. The lobes of the calyx vary considerably in outline; and the margins are entire, save in one variety, A. medium var. litddum, where they are minutely laclniate or toothed. The corolla is gamopetalous and regularly five-lobed. Contrary to the opinion of several students of the genus, the shape of the lobes appears of little taxonomic im f the lobe with reference to the corolla-tube, however, appears dependable. In A. androsaemifolium and its allies, the corolla- lobes are sharply recurved, differing distinctly from those of A. cannahinum and its allies, which are nearly, or quite, erect. The lobes of A. medium and its varieties add an additional inter- 1930] WOODSON — STUDIES IN APOCYNACEAE. I 59 mediate character to their striking affinities to both the andro- saemifolioid and the cannabinoid groups by assuming a spreading position. The color of the corolla varies from greenish white to pinkish or purplish. The relative length of the corolla and the calyx-lobes has proved to be an important taxonomic criterion. In A. andro- saemifolium and its allies, the ratio of the length of the corolla (including both tube and lobes) to the length of the calyx lobes is at least 3:1. In ^, cannahinum and its allies, the same organs have a ratio of less than 2:1, and the calyx-lobes usually equal the corolla in length. In A. medium and its varieties, the ratio of corolla to calyx-lobes is almost exactly 2:1. The comparative width of the corolla at base and orifice is an additional character which is frequently used. In A. androsaemifolium^ the tube of the corolla dilates considerably, and the orifice is much wider than the base, producing a campanulate flower. In A. pumilum, the corolla is cylindrical, and the base and the orifice are nearly or quite equal. So also the corolla-tube of the typical variety of A, hypericifolium is somewhat campanulate, while that of A. hypericifolium var. salignum is nearly cylindrical. The corolla is preponderantly glabrous throughout the genus, but very rarely, as in A. medium vars. sarniense and vesiitum^ microscopic tufts of tomentum are developed upon the exterior. The cellular constitution of the corolla is illustrated by a sector of a cross-section in pi. 6, fig. 1. As illustrated in the figure, the most characteristic feature of the corollar anatomy is the external epidermis of a single layer of bulbous, or papillate cells. It is a further elongation of these protuberant cells which causes the peculiar and characteristic pulverulent-papillate condition of the corollas of Poacynum and Trachomitumy as well as the micro- scopic tomentum of the corollas of Apocynum medium vars. sarniense and vestitum. The mesophyll of the corolla is relatively thick, and is com- posed of large, irregular, loosely connected cells. Latex tubes are numerous, but do not appear to be associated with the conductive system as in the leaves and to some extent in the calyx-lobes. Fifteen vascular strands enter the corolla, after it is fully differ- entiated from the receptacle, assuming five groups of three 60 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 strands each. These vascular groups correspond to the lobes of the gamopetalous corolla, and are homologous to petals. In the neighborhood of the median vascular strand of each lobe, the cells of the mesophyll are smaller and more compact than else- where. This region is occupied by the peculiar coroUar flanges characteristic of the genus. These flanges are composed of three internal ridges at the base of the corolla-tube which gradually converge into a tiny apiculate tip, forming a flat cordate-sagittate process about 2 mm. tall and 1.5 mm. broad, which alternates with the stamens, the apex fitting closely into the spaces between the auricles of the anthers. Conjectures concerning the signifi- cance of these corollar flanges were offered in a previous section of this study; their function will be discussed subsequently. Plate 6, fig. 1, illustrates the transverse constitution of the three corollar flanges at slightly different levels. Text-fig. 10, diagram 2 {g), presents their superficial habit. The stamens are very highly developed, and sterilization of tissue is apparent. They are rather loosely attached to the base of the corolla- tube, alternating with the lobes. The filaments are short and thick, and frequently very pubescent. The anthers are introrse, connate, and appressed about the pistil in the form of what has been described as a " gynostegiura " by Demeter, who wished to emphasize an analogy with a somewhat similar condition in the Asclepiadaceae. The anthers consist largely of a greatly elaborated, sterile, pel- tately affixed, narrowly sagittate connective, with a deeply auricu- late base. The four small sporangia are borne above the insertion of the filament. The basal, or auriculate, half of the anther is com- pletely sterile. At the maturity of the pollen, or shortly before, the partition between each of the dorsal and ventral loculae is ruptured, giving the appearance of a two-celled anther. Finally the ventral cells dehisce longitudinally and shed the pollen. The false bisporangia of Apocynum are characteristic of the genera of Apocynoideae, and appear to constitute an additional affinity with the Asclepiadaceae. The pollen grains are relatively small (diam. about 10 [x), spherical, and smooth-walled. The pollen tetrads, as has already » Demeter, K. I c. 1922. 60 \ 19301 WOODSON — STUDIES IN APOCYNACEAE. I 61 been explained, never normally separate into individual grains. The pollen of practically all the species of Apocynum is largely abortive, and polysporous tetrads are frequent, apparently in- dicating frequent hybridization. Upon the receptacle are situated the two carpels comprising the ovary and the cycle of five fleshy nectaries which encircles them by their common immersion in the receptacle, and at the tips, where they are united by a common fleshy style called the clavuncle. The clavuncle is thickened and truncate, and bears the terminal stigma. Plate 6, fig. 3, illustrates the cellular condition of a medi- an longitudinal section of the clavuncle of Apocynum androsaemi- folium. The organ should evidently be construed as the product of the fusion of two bodies representing the individual style and stigma of either of the two carpels, as the two vascular strands (the dorsal strands of either carpel) present in it are entirely separate, and the body cells themselves retain an individual ap- 3specially in the central region. The i the clavuncle, is highly glandular, and stigmat During the earlier phases of the present study, the shape of the clavuncle of the various species of Apocynum was considered to be of taxonomic importance. Extensive studies of the clavuncle of all the described species and varieties were accordingly made. Many vials were filled with dissected clavuncles, and correctly labelled with the appropriate taxonomic designations. Micro- photographs were also made in the hope that they would provide a concise criterion for difl"erentiation. After several months of concentration upon that organ alone, however, the conclusion was reached that the clavuncle is apparently quite variable, and of doubtful taxonomic assistance. Other technical characters have likewise been exhaustively studied to aid in the delimitation of species, but their value has appeared so dubious that in the taxonomic treatment which follows, macroscopic characters have been employed in entirety. Dr. Ballard," however, has employed the character of the clavuncle in providing a survey of the species of the eastern United States for pharmaceutical use. " Ballard, C. W. Taxonomy and pharmacognosy of the genus Apocynum. Proc. Tntftrnat. Cons. Plant Sci. 2: 1406-1412. 1929. 62 [Vol. 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN Either carpel of the gynoecium contains a single loculus bearing numerous anatropous ovules upon an axile binate placenta, placenta contains two main longitudinal vascular bundles which the traces to the ovules depart. A common misconception regarding the Apocynaceae h the nectaries which frequently a "disc." Such, however, is not the case, unless the term "d The surround the carpels represent is to become an entirely meaningless shibboleth. The nectaries, as has already been observed, are probably developments of a vestigial cycle of carpels. Detailed evidence has already been presented to support the assertion that they should be viewed e 1 3 Fig. 10. Floral morphology of Apocynum androsaemifolium. 1: longitudinal view, with portion of corolla removed, a- corolla; b- clavuncle; c- stamen; d- calyx; e- ovary;/- nectary; g- corollar appendage. 2: interior of dissected corolla, indicating the corollar appendages, g. as equivalent to a cycle of carpels. Plowever, the nectaries occasionally demonstrate a coalescence with the receptacle (cf. Odontadenia) , or even with the carpels (cf. T ahcrnaemoniana) , when their status naturally becomes obscure. Text-fig. 10 il- lustrates the organography of the flower of A. androsaemifolium. Floral Mechanism, — The floral organs of the genus Apocynum, which have been separately described in the preceding paragraphs, compose a reproductive mechanism of extreme interest, but one which has never yet been satisfactorily explained. As will be seen from a glance at text-fig, 10, diagram 1, the four 1930] WOODSON — STUDIES IN APOCYNACEAE. I 63 cycles of reproductive organs form an intricate and comimct mass within the center of the perianth. Without careful dissec- tion and keen observation the parts and composition of the complex mechanism are scarcely easy to comprehend. The center, of course, is the bicarpellary gynoecium (e) surmounted by the barrel-shaped clavuncle (b). As will be perceived from the diagram, the clavuncle is divided equatorially into an upper and a lower region. The upper region is slightly bilobate, and was considered by Linnaeus and Tournefort to bear the stigma. Pressed tightly about the clavuncle and gynoecium is the cycle of five stamens (c) alternating with the ovoid nectaries (/) and the sagittate corollar appendages (g). The anther of the stamens is fertile only toward the apex (i. e., the portion adjacent to the upper half of the clavuncle). At the point of attachment of the filament to the anther, elongate glandular hairs arise from the epidermis and meet with similar epidermal outgrowths from the equatorial region of the clavuncle. These elongate cells secrete viscin, which binds the androecium to the gynoecium. So tightly are the two cycles of sex organs cemented by these epidermal appendages and their secretions that one can scarcely remove the stamens without dislodging the clavuncle from the apex of the carpels. The true position of the stigmatic surface and the procedure the been much and discussed. As has been stated above, the stigma was called terminal by Linnaeus and Tournefort, neither of whom attempted an explanation of its pollination. In 1759 Fabricius^^ wrote of the attraction which A. androsaemifoUum exerts upon nectar- loving insects and observed that frequently the visitors are found trapped within the flower. In 1783 Lamarck" described the insect relations of the same species as follows: "The name gobe-mouche [fly-trap] has been given to it, because the flies, in their greed of the honeyed juice which is found at the bottom of its flowers, insinuate their proboscides by the narrow passage which is found between the small bodies [nectaries] which surround the ovaries, and the ovaries themselves, and when the insects would withdraw their proboscis, it is found to be held the faster, the more efforts they make to withdraw it. Thus these insects, half buried in the flowers, are caught as in a trap, and perish there without the power to escape." li^ Fabric. Enum. Meth. PL 256. 1759. "Lam. Encycl. 1: art. Apocynum. 1783. 64 ANNALS OF THE MISSOUBI BOTANICAL GARDEN [Vol. 17 In Europe, A. androsaemifolium was cultivated since early days, the exact date of its introduction being unknown.^^ How- ever, by the last quarter of the eighteenth century it was evidently widely cultivated and viewed as a curiosity because of its pro- pensity of capturing insects. In 1794 Moench^^ proposed the name A. muscipulum for A. androsaemifolium. Although it is now fully understood that insects trapped in flowers of Apocynum are probably always detained by the viscous floral secretions, some fantastic explanations have been advanced. Lamarck^® was of the impression that mechanical pressure ex- erted by the nectaries is responsible for detaining the proboscis. A somewhat more current view was that the anthers actually grasped the insect as do the leaves of Dionaea. Erasmus Darwin®^ appears to have been one of the first to sponsor this view, and wrote as follows : "In the Apocynum Androsaemifolium, one kind of Dogsbane, the anthers converge over the nectaries, which consist of five glandular oval corpuscles surrounding the germs; and at the same time admit air to the nectaries at the interstice between each anther. But when a fly inserts its proboscis between these anthers to plunder the honey, they converge closer, and with such violence as to detain the fly which thus generally perishes." Although accounts of other authors regarding the muscipulous habit of Apocynum would not be suitable in this general account, the description of the plant by the anonymous author of ^The Journal of a Naturalist'®^ should be included as perhaps the most fabulous of all : "But we have one plant in our gardens, a native of North America, than which none can be more cruelly destructive of animal life, the dogsbane (apocy- num androsaemifolium) which is generally conducive to the death of every fly that settles upon it. Allured by the honey on the nectary of the expanded blos- som, the instant the trunk is protruded to feed on it, the filaments close, and catching the fly by the extremity of its proboscis, detain the poor prisoner writhing in protracted struggles until released by death, a death apparently occasioned by exhaustion alone; the filaments then relax, and the body falls to the ground. The plant will at times be dusky from the numbers of imprisoned wretches. This elastic action of the filaments may be conducive to the fertiliza- "Ray (Hist. PL p. 1089) mentions "Apocynum flore Lilii convallium purpuras- cente" as appearing in the famous physic garden at Chelsea in 1688, « Moench, Meth. 464. 1794. " Lam. /. c. 1783. " Darwin, E. The botanic garden. Part 2: The loves of the plants. 126. 1791. " Anon. The journal of a naturalist. 80-81. 1829. 1930] WOODSON — STUDIES IN APOCYNACEAE, I 65 tion of the seed by scattering the pollen from the anthers, as is the case of the berberry; but we are not sensible, that the destruction of the creatures which excite the action is in any way essential to the wants or perfection of the plant, and our ignorance favours the idea of wanton cruelty in the herb; but how little of the causes and motives of action of created things do we know!" This highly fictitious narrative is illustrated by a sketch show- ing the flower of A. androsaemifolium before: with outspread anthers awaiting the visit of the victim; and after: with the an- thers pressed tightly about a fly. Apparently the first man to interpret the visits of insects to Apocynum as a means of pollination was Bartolozzi/^ who ad- vanced the opinion that the visiting insects inject the proboscis between the connate mass of the clavuncle and anthers both to obtain the pollen and to apply the pollen to the stigma. The account of Bartolozzi regarding the insect visits and the mor- phology of the flower was an exceptionally full and accurate one for its day. In 1809, as is well known, Robert Brown^" described the stig- matic surface of Asclepias as occupying the lower portion of the stylar shaft, and succeeded in tracing pollen tubes from that organ to the placenta and o\aile. Contemporaneous botanists were quick to accept Brown's observations, and in 1849 Schleiden^^ also drew the conclusion that the stigma of Apocynum is situated likewise on the basal half of the clavuncle. In 1867 Delpino^2 largely recapitulated the conclusions of Bartolozzi, favoring the view that pollen is transmitted to the basal stigma by the efforts of the proboscis to reach the nectaries surrounding the ovary. During the years 1872-73 W. H. Leggett^^ became interested in the biology of Apocynum^ devoting several short but highly interesting articles to the subject in the ^Bulletin of the Torrey Botanical Club.' Leggett followed Delpino^s de- °' Bartolozzi, F. Memoria . . . sopra le quality che hanno i fiori della pi ante detta Apocynum androsaemifolium di prender le mosche, etc. Opusculi scelti sulle scienze e sulle arti, 193-200. 1779. 70 Brown. K Mem. Worn. Soc. 1: 12-18. 1809. '^Schleiden, M. J. Botanik. 494. 1849. 72 Delpino, F. Sugll apparechi della fecondazione nelle piante antocarpee . , . 19-21. 1867. 73 L[eggett], W, H. Bull. Torr. Bot. Club 3: 46, 49-50, 53-55. 1872; 4: 1-2, 23. 1S73. [Vol. 17 66 ANNALS OF THE MISSOUEI BOTANICAL GARDEN scription of the pollination of the genuS; adding many original observations on the manner in which insects are frequently trapped within the flower. The manner in which pollen can be transferred to the stigma by the proboscis is easily understood, but the method of acquiring the pollen itself is not so easy to explain, since the insects send the proboscis straight to the receptacle of the flower by way of the exterior of the stamens and not usually through the connate mass of the anthers. This was neatly explained by Leggett as follows : '*The Qy probably dips its head down to get at the nectaries in the bottom of the cup, and, in drawing it back, sometimes gets its proboscis caught in the groove between the anthers, to which it is guided by their diverging bases. As the proboscis is drawn up the groove, it passes the glutinous glands [i. e., the viscous secretions from the epithelial gland-hairs of the stamen and clav- unclej; which are easily detached when the flower is mature, and, being thus charged with one or both of these glands, draws out the pollen from one or both of the adjacent anther cells. Perhaps alarmed by this rude entertainment, the insect flies off to a more distant flower, where it goes through the same process, first, however, leaving the stranger pollen on the stigmatic surface." It is evident that all who have dealt with the subject have been put to some pains to explain the pollination of Apocynum. By no means a help to the situation is the reassignment of the stigmatic surface to the upper half of the clavuncle by Demeter/* who investigated the cellular constitution of that organ and found the typically stigmatic cells located in that position. Plate 6; fig. 3, represents a median longitudinal section of the clavuncle of A. androsaemifolium. It will be seen that the upper portion bears typically glandular stigmatic cells (d), while the lower region is scarcely equipped to carry on the function of a stigma. Pollen tubes have also been found by the writer to penetrate through the upper region of the clavuncle. As all of the investigators dealing with the subject in recent years have believed, the flowers of Apocynum are apparently unable to fertilize themselves. At the present, however, the process of pollination appears almost as much in the dark as ever. It would appear that normal insect visits do not succeed in pol- lination, and that it is only when a visitor becomes trapped and struggles to escape that the floral mechanism is disarranged " Demeter, K. I c. 1 922. 1930] WOODSON — STUDIES IN APOCYNACEAE, I 67 sufficiently to allow the reception of the pollen by the stigma. At any rate, fertilization is truly a phenomenon in the genus, for follicles are relatively rare. In 1891 Robertson ^5 compiled an extensive series of notes regarding the visits of insects to native flowers. His list of insect visitors to Apocynum cannahinum includes nineteen species of Hymenoptera, seventeen species of Diptera, two species of Lepidoptera, one species of Coleoptera, and two species of Hemip- tera. Fruit. — The fruit of the genus Apocynum is a pair of elongate, terete follicles. The distinctness of the two is marked, save in very early stages when they may be united at the tips. The short union at the tips, however, is responsible for a taxonomic char- acter which has been used occasionally in the delimitation of taxonomic units. This character is a falcate appearance caused by unequal growth of the follicles when united at the tips, pro- ducing a gibbous condition. The follicles are usually pendulous, but are erect in A . pumilum and its variety rhomhoideum. This character, in the light of what has been previously written concerning the inflorescence of A. pumilum, might lead one to the thought that possibly the pendulous follicles of the other species represent a modification from a primitive erect condition. The seeds are abundant, linear-terete, and abundantly comose at the micropylar end. The coma develops from the superficial cells of the outer integument which elongate into a sessile whorl. The seeds possess but a single integument, and the nucellus, com- posed of but endosperm likewise, is absent from the mature seed. The length of the seed and coma is occasionally used as a taxonomic criterion in the revision which follows. At maturity the placenta withers away, leaving the follicle filled with free seeds which are easily dispersed by the wind upon dehiscence. The embryo is straight, and typically dicotyledonous. Plate 4, fig, 5, represents the habit of the embryo of Apocynum hyperid- folium magnified about twenty-five diameters. The embryo itself "Robertson, C. Bot. Gaz. 16: 70-71. 1891. " Frye, T. C, & Blodgett, E. B. Bot. Gaz. 40 : 61-52. 1905. [Vol. 17 68 ANNALS is shuttle-shaped, the two fleshy cotyledons equalhng about one- half the length of the body. The radicle is differentiated from the hypocotyl as an abruptly bevelled point. The plumule is not differentiated in the seed. The vascular element of the embryo in the dormant seed is limited to procambium. It is not until the plumule is formed in the seedling that the normal vascular constitution is evident. The cells of the embryo are rich in protein and starch, although the latter is concentrated chiefly in the cotyledons. Latex in- clusions are present only within cells similar in all other respects to the normal parenchyma. Such latex cells are infrequent. Laticijerous System. — The Apocynaceae as a family are popu- larly notable chiefly because of the internal secretory system which occurs in all genera. The internal secretions, or latex, may be colorless, greenish, most frequently whitish, and rarely somewhat ochraceous. In the genus Apoqjnum the latex is chalky white, viscid, and coagulates spontaneously. The latex of the genus Apocynum, as evidently throughout the family Apocynaceae, is contained in extensive cells which ramify nearly the entire plant. A few organs, notably the young ab- sorptive roots and the anthers, are evidently without latex cells. In the embryo, the initials of the latex cells first appear in the plane which coincides with the node of the cotyledons." In the embryo of the fully mature seed, they are observed in the cotyle- dons which resemble the fundamental cells except in their con- tents. At such a stage, the hypocotyl does not usually appear to contain latex cells. Later, during the seedling stage, they extend to the hypocotyl as well as to the plumule. In young absorptive roots, as has already been stated, latex tubes are lacking, or rare (pi. 4, fig. 1). In older roots, especially the gemmiferous roots, they occur freely, apparently being con- fined to the cortex in the majority of cases (pi. 4, fig. 2h). In the rhizome and stem they are apparently confined to the pericycle, where they are intimately associated with the pericyclic fibers (pi. 5, figs. 1-3 f). In some cases, particularly in young stems, however, isolated tubes are found in the outer layers of the pith (pi. 5, figs. 1-2 m). In the leaves, the laticiferous tubes are aDDarentlv restricted to the vascular bundles. " cf. Chauveaud, G. 108-109. 1891. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 69 In the calyx, as in the leaves, the latex tubes are evidently associated with the vascular bundles. In the other floral parts, the corolla, the stamen filaments, and the gynoecium, including the nectaries, they ramify chiefly the non-vascular tissue (pi. 6, figs. Ic, 2c?, and 3c; also pi. 3, fig. 4). The latex of Apocynum contains suspended proteins, starch, and crystalloids, and yields over 2 per cent caoutchouc. The economic possibilities of the caoutchouc will be discussed in a subsequent section of the present study. The latex cells are evidently coenocytic. INTRA-GENERIC RELATIONSHIPS In the case of a genus so complex and difficult as Apocynum, it is manifestly impossible to submit a taxonomic revision immune to much justifiable criticism. In the opinion of Asa Gray, the total number of the species of Apocynum was two. His younger contemporary, Edward L. Greene, of an entirely different view- point in matters taxonomic, himself published over fifty species of the genus, and left herbarium names upon nearly a score of specimens in the United States National Herbarium and the Greene Herbarium at Notre Dame University.^^ Most of the Manuals and Floras which have dealt with the genus have for- tunately striven to strike a mean between these two extremes, but frequently with a result which reflects the authors' uncertainty. The specific concept in the present work is largely one of practical application. The watchwords have been qualitative constancy and geographical unity. Above all, the paramount necessity of a workable key has been recognized and supplied to the best of the author's ability. In taxonomic work the key is frequently, as indeed it should always be, the essence of the author's conception of his problem; and by his keys a taxono- mist 's work is most frequently liable to judgment. In this revision, a key to all the species and varieties of the genus Apocynum is offered for the first time. Among the de- ficiencies of the monograph by B^guinot & Belosersky,^^ one of '^ It is unfortunate that a considerable number of Greene's herbarium names were published as nomina nuda in the synonymy of B^guinot & Belosersky's monograph. " B^g. & Bel. I. c. 1913. [Vol. 17 70 ANNALS OF THE MISSOURI BOTANICAL GARDEN the most conspicuous is a complete lack of keys or any other analytical differentia. This lack renders the work not only ex- tremely ambiguous, but of little use to the botanical public. From a glance at the Key to the Species which follows, it will be seen that the genus is roughly divisible into two groups, namely, species with drooping or spreading foliage and corolla at least twice the length of the calyx-lobes; and those with ascending fohage and corolla barely longer than the calyx-lobes. Within the first category there are two subdivisions : species with drooping leaves and corolla at least thrice the length of the calyx-lobes; and species with leaves spreading or rarely {A, Suksdorfii and A. Jonesii) ascending and corolla about twice the length of the calyx-lobes. The second category is divided into a group with leaves evidently petioled, narrowed at the base or the very lowest obtuse and subsessile; and a group with leaves mostly cordate or obtuse and sessile, especially below. Roughly, these four un- named divisions correspond with Bdguinot & Belosersky's sections "Androsaemifolii," "Medii," "Cannabini," and "Hypericifolii," with which those authors sought to subdivide the genus. A study of the plants in the field, however, shows that there are no sharp lines of cleavage in the genus, and that the division of the genus into named sections is both unnecessary and artificial. It is found upon close study, even in the herbarium alone, that the species are so confluent as to make sectional lines extremely obscure. The cause of this specific intergradation, it is believed, is due to mass mutation through hybridism. Evidently the flowers of all species of Apocynum are self- sterile, for otherwise they would yield theoretically 100 per cent fruit, since the anthers press so closely about the stigma that self-pollination is unavoidable. As a matter of fact, however, relatively few flowers are actually fertilized, say 10 per cent at most, which bespeaks the presence of another agency than the mere dropping of the pollen upon the stigma of the same flower. In the summer of 1928, bagging experiments were conducted in experimental plots of the Missouri Botanical Garden and in the field with evidence supporting the assumption of self-sterility. In the succeeding season efforts were made to self -pollinate A. cannahinum, with negative results. ) 1930] WOODSON — STUDIES IN APOCYNACEAE. I 71 It appears probable that through the glandulosity of the clavuncle, cross-poUination is effected by insects freeing the stigmatic surface of the flower in a violent attempt to reach the nectar secretions, thus allowing the subsequent cross-pollination by the same, or another visitor. The violence and frequency of insect visits to the flowers of Apocynum is attested by the nu- merous flies which are found trapped in them. Lotsy,^^ the greatest exponent of the theory of the origin of species by hybridization and mass mutation, summarizes his observations as follows: "It was, to my way of thinking, the coming into the world of sexual reproduclionj not response to stimuH, which made evolution possible. I may also express this so: that the difference is a double one, in as much as the theory of hybridization not only substitutes crossing for variabihty as the cause of evolution, but also considers those groups of individuals which systematists call species as mere remains of formerly very diverse swarms arisen from a cross, and not as the progeny of a single individual or two individuals, which, except in sex, were alike." Without attempting to insert here a criticism of Lotsy's the- ories, especially with regard to the logical consequences of sub- stituting in toto variations by hybridization for the infinitely slower accumulation of Darwinian variations, let it suffice to say that with regard to the genus Apocynum^ the theories of Lotsy can be very favorably applied. Reasons for assuming that the species of Apocynum are self- sterile and depend largely upon cross-pollination for sexual re- production have already been given from a mechanical viewpoint. The obvious result of extensive hybridization, mass mutation, is found on every side by those who are willing to study the plant in the field. Clearly, in the case of Apocynum^ the plants intergrade to such an extent that the taxonomist may well be led to agree with Lotsy that species arise by swarms and the subsequent falling out of the intermediates to form more or less isolated enti- ties. Conservative botanists must allow great variability if they acknowledge a small number of species; others can recognize in the place of specific variability a large number of distinct species, as did Greene. *° Lotsy, J. P. Evolution considered in the light of hybridization, p. 59. Christ- church, N. Z. 1925. [Vol. 17 72 ANNALS OF THE MISSOURI BOTANICAL GARDEN Field observations have been a frequent source of information throughout the course of these studies; and in the case of the suspected hybridism of the genus they have apparently shed much light. Upon several occasions, notably in Indiana and Colorado, the hypothetical hybrids and the suspected parents have been found growing intermingled or in adjacent areas. In the summer of 1928, in company with Dr. J. A. Nieuwland of Notre Dame University, and upon another occasion with Dr. Edgar Anderson of the Missouri Botanical Garden, the writer visited a field at Notre Dame in which A. androsaemifolium, A. cannahinum, A. hypericifolium, and A. medium were abundant. Early in September, when the field was last visited, there was abundant fruiting material of the first three species; none, how- ever, was apparent of A. medium. During the sunmiers of 1927- 29; field studies were made in the state of Colorado. At several points varieties of A. medium were discovered in areas adjoining one or both of the suspected parent species. Sterility of the hyx)othetical hybrid was always conspicuous.^^ Near Bayfield, Colorado (see also p. 48), thousands of plants of A, medium var. lividum were discovered infesting a field of about fifty acres in extent. Throughout the entire field (pi. 8) not a single indication of fructification was found upon examina- tion. Near Lyons, Colorado, the same variety was discovered in clones of several hundreds of plants each. No fructification was visible, although plants of A. androsaeinijolium var. incanum, just across the road, and A. hypericifolium var. salignum, nearby, supposed to be the parents of the hybrid, provided abundant fruiting specimens. At the base of Eagle Cliffs, in Moraine Park, Colorado, there exist intermingled clones of A. androsaemifolium var. incanum and A. medium var. florihundum. This particular location was visited in 1927 and again in 1929; and in neither year was fruiting material observed upon the A . medium, although f oUicles •^ It is not meant to convey the impression that necessarily all of the supposedly hybrid species and varieties are absohitely sterile, however. The common eastern, or typical, variety of A. medium frequently produces folhcles, and for that reason, as well as for its abundance and wide distribution, might be regarded as of earlier origin than the more restricted and apparently sterile western varieties of the same Bpecies. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 73 were borne on A. androsaemifolium. This location is especially interesting because of the fact that in 1929 one small plant was found in flower which was strikingly intermediate between the two species. This plant is described as A. androsaemifoliu7n var. intermedium in the taxonomic section which follows. The foliage of this plant is similar to that of androsaemifolium^ but the flow^ers are strikingly intermediate, since, although the corolla-tube is cylindrical and only about twice the length of the calyx-lobes as in A, medium var. floribundum^ the lobes are recurved as in A. androsaemifolium var. incanum. Statistically, the external measurements of the flowers of the three varieties growing at Eagle Cliffs are as follows; the figures are in millimeters: A, androsaemi- folium var. in- canum A. androsaemi- folium var. in- termedium A. medium var. floribundum Length of corolla-tube, including the limb Width of the corolla- 6.5-7.5 4.5-5.5 3.5-4.0 tube at base 2.0-2.5 3.0-3.5 1.5-2.0 \^^idth of the corolla- tube at orifice 5.0-G.5 3.0-3.5 1.75-2.0 Length of the calyx- lobes 1.0-1,25 2.0-3.0 2.0-2.5 Naturally, it is appreciated that in order to firmly substantiate the hypothesis of the speciation of Apocynum through hybridiza- tion, cytological studies should be made of each species, especially with regard to meiosis in the pollen-mother-cells. Unfortunately, however, the cytology of Apocynum presents serious difhculties. The formation of the tetrads occurs at an extremely early stage in the formation of the bud, young buds in which the stamens are mere filaments less than a millimeter in length invariably being found to contain tetrads. When one remembers that the sporangia occupy only about one-eighth of the mature stamen, the difficulty of the problem may be appreciated. The omnipresent latex also offers difficulties in killing and fixing. Since the primary object of this monograph is taxonomic and morphological, there- fore it was decided to defer attempts at a cytological study. 74 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vou 17 However, certain cytological characteristics of hybrids are deducible through morphology. One of the most easily perceived of these is the irregular division of pollen-mother-cells producing polycary and polyspory through the independent organization of lagging univalent chromosomes. Of course hybridization may occur without such chromosomal incompatibilities or other phe- nomena may produce them, but the work of Rosenberg^^ q^ Drosera and other genera, and the work of Jeffrey ^^ and others, have demonstrated clearly that hybrids can usually be detected through observation of the meiotic divisions of the pollen-mother- cells, and therefore by an examination of mature pollen, since hybrids very frequently betray the nature of their origin through the production of sterile and polysporous pollen. Accordingly studies of the pollen of the species of Apocynum were made, both in the fresh state and in microscopic prepara- tions. Since the pollen grains are held permanently in tetrads throughout the genus, polyspory can easily be detected in the forms which exhibit it. A certain amount of sterile pollen was evident in all of the species, and polysporous tetrads were found in others. The results may be summarized by stating that the four classic species of Apocynum {A. yumilum, A. androsaemi- folium, A. hypericifolium, and A. cannabinum) have the smallest percentage of sterile pollen and that the intermediate species and varieties have a greater sterility. These findings are in accord with the taxonomic arrangement of the species in this monograph. With the combination of taxonomic, morphological, and dis- tributional observations, a phylogenetic chart has been prepared which illustrates, from the author's point of view, the speciation of the genus. Such visualizations are doubtless extremely liable to equivocation, but furnish a certain intellectual fascination which at times has led to deductions of value. Text-fig. 11 is a phylogenetic chart illustrating the supposed speciation of the genus Apocynum through the hybridization of four potential ancestors. A few words of explanation and inter- pretation should perhaps be added. The broken lines present the ^ Rosenberg, O. Cytologische und morphologische Studien an Drosera longi- Jolia X rotundifolia. Kgl. Svenska Vet.-Akad. Handl. IV. 43: 3-64. 1909. ^ Jeffrey, E. C, and Hicks, G. C. The reduction division in relation to mutation in plants and animals. Am. Nat. 59: 410-426. 1925. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 75 4> VI ffj d uj "fi ■^ ^ ^" ^ ^ ^^■^ ^ ^ rf ■H g .•J o CM o «M V * ■ >< >4 <0 4) b8 — * 4 > in ^ << ^ o ■ > > o x; 1 Id > J > 1 • 4 ^ ^^M 1 ^ Fig. 11. Phylogenetic chart of the genus Apocynum, Explanation in text. hypothetical rudiments of the genus; the soHd lines, the supposed relationship of the known species and varieties. The length of the vertical solid lines indicates the supposed age of the species 76 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 and varieties; the thickness indicates their relative abundance. It will be noticed that of all the species and varieties diagrammed, only six continue straight to the ground-line. These species and varieties, A. pumilum (typical glabrous variety), A. andro- saemif olium var. incanum, A. cannahinum and its two varieties, and A, hypericifoUum (typical variety), are those whose origin cannot be hypothesized from any other existent forms, and which are apparently coeval. Of the other species and varieties, only three (A. pumilum var. rhomboideum, A. Suksdorfil var. anguslij oliuvi y and A. hypericifoUum var. cordigerum), which are merely foliage varieties, are not suspected to be of hybrid origin. Hybridity is indicated by a connecting horizontal line, between the vertical lines of the two supposed parents, upon which is imposed a vertical line of length and thickness proportional to the relative abundance and suspected age of the species or variety. Four species are considered older than the others, and are diagrammed as such. They fall naturally into three groups, which have been named for the sake of convenience. The first group, designated as the ''pumiloid lineage" contains a single "primeval" species. This group is characterized by corollas thrice or more exceeding the calyx, drooping foliage, mostly al- ternate branching, and above all, by the erect, or nearly erect position of the follicles. The second group, the ''androsaemi- folioid lineage" likewise contains a single primeval species, and differs from the preceding group chiefly in the position of the follicles, which is decidedly pendulous. The third group, or "cannabinoid lineage" contains two primeval species which are characterized by corollas barely exceeding the calyx, erect or ascending foliage, opposite or subopposite branching, and pendu- lous follicles. Species and varieties having follicles which are strikingly intermediate between the erect and the pendulous, corollas about twice exceeding the calyx, and spreading (i. e, neither drooping nor ascending) foliage, are considered as indi- cating a nature intermediate between two of the three groups, and are diagrammed as hybrids upon the chart. Species and varieties characterized by the intermediate foliage and flowers moreover display in addition alternate branching by means of the abortion of an opposite branch-bud, an intermediate character of significance. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 77 In constructing the chart, attention has been paid the geo- graphical distribution of the species and varieties as well as their morphology. Instances of such use of geography are A . pumilum var. rhomhoideum and A. mediu7n (typical variety). A. pumilum var. rhomhoideum is taken as an offshoot of A. pumilum (typical variety) instead of mce versa, because the latter has a range ex- tending from the Pacific Coast to the Rocky Mountains, while the former inhabits only the region from the Pacific Coast to the Sierra Nevadas and Cascades. A. medium (typical variety) is taken as a product of A. hypericifolium and A. androsaemifolium var. incanum rather than of A. cannabinum and A. androsaemi- folium var. incanum, because in Newfoundland and the New England states and adjacent Canada, where A. cannabinum is rare or lacking (see pi. 1), A. medium is coexistent with A. andro- saemifolium var. incanum and A. hypericifolium, which are com- mon species in the above-cited regions. From an examination of the chart it will be perceived that specific lines in the genus Apocynum do not appear very "natural" (i. e., monophyletic). The outstanding example of this feature is the species A. medium, the varieties of which are believed to have arisen in different parts of North America from several different crosses between other species of the genus. These vari- eties, then, although supposedly of no direct relationship to each other, are grouped together as varieties of a single "species" be- cause their taxonomic characters are so similar that, barring the consideration of independent origin by hybridization, they would appear quite like ordinary variations of a "natural" (i. e., mono- phyletic) species. The geographical distribution of many varie- ties of the genus appears to support the assumption of speciation through hybridism. For example, A. medium var. sarniense is known from only three stations : one in Ontario, one in Michigan, and one in British Columbia. A. hypericifolium var. Farwellii is known from one station each in New York, Michigan, Indiana, and Texas. A. medium var. lividum has been reported from several stations in Colorado and one station each in Washington and California. Such examples are of frequent occurrence, and together with the morphological and distributional considerations which have preceded, have lent encouragement to the drawing of preliminary conclusions from the mass of circumstantial evidence. 78 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 Even if the speciation of the genus through hybridization is disputed or disproved, the chart will serve to illustrate the appar- ent relationship and relative abundance of the species and vari- eties plotted. Breeding experiments are now under way to test the veracity of the chart, and it is hoped that within a few years more tangible knowledge than mere external similarities will be forthcoming. An unfortunate anomaly is discovered in charting the typical variety of A. androsaemifoUum. Linnaeus** described A. andro- saemifolium as utrmque glabra. A. de Candolle^^ also regarded the typical condition of the species as glabrous, and described the variety with the under surface of the leaves puberulent as A. androsaemifoUum var. incanum. As a matter of fact, the var. incanum is to all appearances the primitive condition of the species, and by far the more common, having a range covering nearly that of the genus; while the glabrous or "typical" variety is evidently a hybrid, possibly between A. -pumilum (typical variety) and A. androsaemifoUum var. incanum, with a range limited to the region west of the Rocky Mountains. Hybridity is suspected between the species suggested because of the campanulate corolla (characteristic of A. androsaemifoUum var. incanum) and com- plete glabrity (characteristic of the typical variety of ^4. pumilum), but above all because of the erratic follicles (intermediate between the erect follicles of A. pumilum and the drooping follicles of A. androsaemifoUum var. incanum) . In spite of the probability that Linnaeus could have carelessly described the puberulent leaves of the eastern variety of the species as glabrous, since material from the Rocky Mountain region was scarcely well known to him, the fact remains that he so described them, and the example of de CandoUe should be followed in the matter. r Thephylogenetic method in Biology, and particularly in Botany, has numerous friends and foes. Hall and Clements/® continuing the theories of Bessey, have been outstanding in the advocation of phylogenetic speculation- Crow/^ more recently, has criti- »*L. Sp. PL ed. 1. 213. 1753. "A. DC. in DC. Prodr. 8: 439. 1844. Hall, H. M., and Clements, F. E. The phylogenetic method in taxonomy, etc. Cam. Inst. Washington, Publ. 326: 3-31, 1923, "Crow, W. B. Phylogeny and the natural system. Jour. Genetics 17: 85-155. 1927. 86 1930] WOODSON STUDIES IN APOCYNACEAE- I 79 cized the phylogeneticists helpfully. According to him, phylo- genetic Biology has apparently failed in its objective^ and he remarks : "• . . current zoological and botanical literature reflects a loss of interest in phylogenetic research which contrasts strongly with that seen during the immediate post-Darwinian period, but which is obviously not due to the attain- ment of a fixed scheme of classification." Continuing further, he observes: '*This state of affairs appears to be the outcome of at least two different trends of investigation. On the one hand, systematists themselves, finding the results of their detailed studies impossible to frame in a simple phylogenetic series, have frequently condensed their results in such a manner that obvious anatomical facts were overlooked, and fantastic interpretations placed on structure, in order to bring the facts into line with a preconceived simple scheme ... On the other hand, the possibility of formulating a natural system has been chal- lenged by the introduction of Mendelian ideas into heredity. . . It is sufficient here to point out that the criticism by the Mendelian theorists of the older phyletic theorists rests on the assumi)tion that a knowledge of the method of evolution is necessary for the construction of a phylogenetic hypothesis^ They agree that if Darwin's view, that gradual variations were chiefly responsible for evolution, was the last word on the subject, then the reconstruction of phylo- genesis would be possible, but they think they have discovered some new facts regarding the nature of the variations responsible for evolution, and that these facts make phyletic reconstruction impossible." The statements of Crow in criticism of the older phylogenet- icist's methods, it will be noted, are in harmony with the view- point adopted in this study. After a prolonged consideration of the problem of species in representative fields of Biology, Crow concludes : '*(!) That the construction of phylogeny is not an arbitrary matter, but depends on certain facts concerning the natural system. (2) Phylogeny consists of theories and hypotheses which do not differ in their nature from those of other branches of science. A satisfactory theory of the phylogeny of a group must however state the characters on which it is based. (3) ConvergencCj which has hitherto been urged as one of the greatest objec- tions to phylogeny, is a result of phylogeny and cannot be upheld as an objection to the phylogenetic method. (4) Regression, although probably undetected in many cases, is Ukewise not a real obstacle to phylogenetic research, (5) Mendelism and consequent possibilities, even if accepted, do not affect phylogenetic conclusions. (6) The confusion of phylogeny with genealogy and the consequent mis- interpretation of the aims of phylogeny has led to objections to the phylogenetic method which are without foundation . . . Thus in tracing the phylogeny 80 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 of species no account should be taken of the descent of individuals (genealogy) within the species any more than . . . one should take account of purely specific characters in tracing the descent of genera. (7) The comparison of phylogeny with earlier forms of thought cannot in any way be regarded as an objection to it. Similar comparison can be applied to any branch of scientific thought." The foregoing quotation from Dr. Crow is given because in a fair degree it coincides with the spirit which dominates this in- vestigation of the Apocynoideae. However, some dissimilarity, especially in Crow's section (6) of his conclusions should be explained. Contrary to his views, it appears to the author that the importance of individual variation to the specific concept, and the specific variation in turn to the generic concept, especially in the light of evolution by means of hybridization, can scarcely be overestimated. That species arise from individuals, or at any rate from groups of individuals, and genera arise from species, appears to be unavoidable. Thus it appears somewhat illogical to divorce phylogeny from ''genealogy," at least in theory, as indeed it is impossible practically to do so. GEOGRAPHICAL DISTRIBUTION The genus Apocynum is peculiar among the Apocynaceae in being entirely temperate in habitat. Confined to the continent of North America, by far the greatest number of species occurs only within the borders of the United States. All of the species and varieties recognized in this monograph occur in that country, with four species entering the Dominion of Canada, and three reported from Mexico. Three species are also native to New- foundland. The only region of the United States in which the genus is apparently unknown is peninsular Florida. With its center of distribution in the United States, Apocynum finds its northern limit at about the 55th parallel, and its southern limit at about the 30th parallel. As illustrated by the map in pi. 1 the species of the genus represent a case of progressive endemism. It will be seen that A. androsaemifolium and A. cannahinum have by far the greatest distribution of any of the species, extending throughout the entire range of the genus. A. medium, A. hypericifolium, and A. pumi- lum in the order named have the next largest distribution, while 1930] WOODSON — STUDIES IN APOCYNACEAE. I 81 the other species are shown with still more restricted ranges. Of the latter, A. cannahinum var, Greeneanum and A. Jonesii are the most local, each having been reported from a single locality. The species of Apocynum have become as widely distributed ecologically as geographically. Then- range of habitat is from saline and fresh water beaches to deep soil and subalpine situa- tions. In general, however, they thrive best in a good field loam, growing with pasture plants with which they furnish lively com- petition in the struggle for existence. Species growing in such a situation develop robust plants, profusely branched and leaved, and attaining large proportions; while those growing in more extreme conditions are usually reduced in size, and may be spar- ingly branched and leaved. Plate 7 illustrates a typical clone of A. cannahinum var. glaherrimum in a cultivated field near Caron- delet, Missouri. Because of their frequency in cultivated fields, certain species have become common in ballast, and have tremendously increased their distribution in that manner. There are few roadsides or railroad right-of-ways in the United States beside which Apo- cynum is uncommon. In a previous section of this study, reasons were advanced for supposing that hybridization has played a considerable part in the speciation of Apocynum, Support for such a view is afforded in a map of the species and varieties (pi. 1) where it is seen that those species and varieties assumed to be of hybrid origin exactly overlap in distribution that of the supposed parent stock. ECONOMIC USES The economic uses of Apocynum have been numerous and varied. Among primitive peoples it is frequently used as a source of thread and cord, the strong fibres of the plant earning the popular name of "Indian hemp" for A. cannahinum. Pick- erings* records that Janus Verazzanus, a Florentine mariner sent by Francis I to a "new country westward from Madeira/' pre- sumably North America, came upon natives wearing clothing composed of "certe erbe che stavano pendenti i rami degli alberi (Tillandsia usneoides L.)" interwoven with "canape silvestra " Pickering, C. Chronological history of plants, p. 866. 1879. [Vol. 17 82 ANNALS OF THE MISSOURI BOTANICAL GARDEN {Apocynum canndbinum L.)." Fish-netSj bags, and even cloth have also been manufactured from the fibre of the genus by natives of North America. Fibre of Trachomitum and Poacynum have similarly been employed by the inhabitants of south-central Asia.*^ A decoction of Trachomitum is also used as a tannin by Central Asiatic tribes. The active principle contained in the plant juices of Apocynum is apocynin, a drug which is soluble in both water and alcohol. An injection of apocynin has been shown to have much the same effect as that of digitahn, according to Planchon.^° The powder of the root is frequently used in pharmacognosy as a substitute for ipecac. An indirect economic use of Apocynum is found in the western United States, where fields are allowed to become crowded with the plants as a source of nectar for honey-bees. Recently the economic potentialities of the latex and fibre of Apocynum have been demonstrated. As early as 1912 the abundant latex content of ^4. androsaemifolium was declared to be a neglected source of rubber by Fox,^^ who found the rubber content of the latex of that species to be 2.36 per cent. The rubber was found to coagulate spontaneously, and to be yielded in greater quantities from plants grown upon dry sandy soil than from j)lants growing in agriculturally more desirable land. The natural latex from dry-land Apocynum was found to contain 72.29 per cent water, 26.21 per cent solids, and 1.59 per cent ash. In spite of the favorable demonstration of Fox, little further was done with respect to the economic possibilities of Apocynum until during the Great War of 1914-18, when industrial sub- stitutes were in great demand. Accordingly a survey of western North American rubber-producing plants was begun in 1917 as *' [Ringer, J.] Flora 54; 222. 1871. "In ganz Sibirien, am kaspischea Meere, in Turkestan, Taschkund und in den Steppen Suedrussland wird der bast von Apocynntn venetum und A. sibincum, vielfach zu Geweben und Geflechten benuizt. Die auf der nationalen russischen Industrie-Ausstellung in St. Petersburg (1870) vorhandenen Gewebe, aus Suedsibirien eingeschickt, waren schoen glaenzend, fein und weich. Ferner waren auch, Fischernetze, Jaegertaschen und Schuhe aus demaelbcn Stoff vorhanden. Die Faser ist ungemein theilbar, mehr noch als der Flachs, den sie an Weisse und Glanz weit uebertrifft." ••* Planchon, C. Produits des Apocynac^es. p. 213. 1880. " Fox, C. P. Another Ohio grown rubber. Ohio Nat. 12: 469-471. 1912. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 83 a war-emergency project under the State Council of Defense of California by H. M. Hall and F. L. Long/^ and published in 1921 by the Carnegie Institution of Washington. The results of Hall and Long show that of the 225 species and varieties studied, the species of Asclepias and Apocynum indicated the greatest promise industrially. Instead of A. androsaemifoliumy A. cannahinum is viewed by Hall and Long as the species most favorable as a substitute rubber plant, and a rubber yield of somewhat higher than 5 per cent was obtained from this species. Concerning the fibre of Apocynum j especially A. cannahinum ^ there appears to be unanimity of approbation from commercial investigators. From both the experiments of Dodge^^ and of Hall and Long, the fibre of A. cannahinum was shown to be of great strength and fineness; and of great promise, especially in the manufacture of paper. Agriculturally, Apocynum offers great advantage over flax in the production of fibre. Unlike Linum usitatissimum and other commercial fibres of the flaxes, A. cannahinum thrives at its best upon arid land unfit for any other agricultural use. The studies of Fox, as well as those of Hall and Dodge, demonstrate that both the fibrous and the latex content of all species of Apocynum reaches its optimum upon such land. With square miles of such waste land lying idle in western North America, the value of such a crop should be enormous. TAXONOMY APOCYNUM [TOURN.] L." Apocynum [Tourn.] L. Sp. PL ed. 1, 213. 1753; and ed. 311. 1763; Gen. PL ed. 5, 101. 1754; Juss. Gen. 143. 1789; Sibth. FL Graeca 1: 166. 1806 (as to Trachomitum) ; Pursh, *'^ Hall, H. M. & Long, F. L. Rubber-content of North American plants. Cam. Inst. Wash. Publ. 313: 65 pp. S pis, 1921. ^3 Dodge, C. R. Descriptive catalogue of useful fiber plants of the world. U. S. Dept. Agr. Fiber Investig. Rept. 9: 62-64. 1897. *^ Apocynum L. char, emend. Calyx 5-partitus glaber vel pubescens eglandulosus. Corolla cami)anulata vel cylindrica, limbo 5-partito aestivatione sinistrorsum contorta, squamellis 5 separatim in tubo corollae insertis. Stamina 5 tubo corollae affixa inclusa cum corollae sequentis alternatis, filamentis brevibus cylindricis, 84 ANNALS OF THE MISSOURI BOTANICAL GARDEN (Vol. 17 Fl. Am. Sept. ed. 2, 1: 179. 1816; Nutt. Gen. 1: 161. 1818 Roem. & Schult. Syst. 4: 405. 1818; Michx. Fl. Bor.-Am. 1 : 121 1803; Hook. Fl. Bor.-Am. 1: 51. 1840; A. DC. in DC. Prodr. 8 439. 1844; Benth. & Hook. Gen. PI. 2: 716. 1876; Gray, Bot Cal. 1: 473. 1876; Coulter, Man. Bot. Rocky Mts. 237. 1885 Gray, Syn. Fl. N. Am. 2^: 82. 1886; K. Sch. in Engl. & Prantl Nat. Pflanzenfam. 42; 179. 1895; Chapm. Fl. South. U. S. ed 3, 341. 1897; Howell, Fl. N. W. Am. 1: 439. 1901; Britt. Man PL 738. 1901; Rydb. Fl. Colo. 269. 1906; Robinson & Fernald in Gray, New Man. Bot. ed. 7, 662. 1908; B^g. & Bel Rev. Monogr. Gen. Apocynum, in R, Accad. Lincei, Atti, Mem. Sci. Fis.V. 9:1-144. 1913; Britt. & Brown, 111. Fl. ed. 2, 3: 21. 1913; Small, Fl. Southeast. U. S. ed. 2, 936. 1913.; Wooton & Standi. Contr. V. S. Nat. Herb. 19: 505. 1915; Rydb. Fl. Rocky Mts. ed. 2, 668. 1917; Standi. Contr. U. S. Nat. Herb. 22«: 395. 1921; Jepson, Man. Fl. PL Calif. 768. 1925; Tidestrom, Contr. U. S. Nat. Herb. 25:419. 1925. Apocynastrum Heistr. ex Fabric. Enum. Meth. PL 256. 1759. Apocinum Juss. Gen. PL 146. 1789. Cynopaema Lunell, Am. Mid. Nat. 4: 508. 1916. Herbaceous lactescent perennials from somewhat thickened fibrous rhizomes. Stems erect or ascending, cylindrical, striate, glabrous or pubescent; branches ascending, alternate to opposite. Leaves opposite, petiolate or sessile, membranaceous, entire, pen- ninerved, glabrous or pubescent, ovate to linear-lanceolate, bearing 2-6 small, attenuate, pectinate glands at the base of the petioles. Inflorescence a trichasial cyme, glabrous or pubescent, bracteate or ebracteate. Calyx regularly 5-lobed, glabrous or pubescent, the tube adnate to the receptacle, eglandular within. Corolla gamopetalous, campanulate to cylindrical, regularly 5-parted, glabrous, bearing within 5 distinct, sagittate appendages adnate to the base of the tube and opposite the lobes of the limb; antheris sagittatis basi cum auriculis parallelis non verentibus instructis. Ovarium bilobum, lobis singularibus unilocularibus, ovulis in quoque loculo binis superpositis. Nectaria 5 integra supra discum propriam intra tubum calycis immersum. Folliculi teretes recti vel rarius falcati. — Herba perennis e rhizomate horizontali erecta ramosa, ramis ascendentibus plerusque alternis. Folia opposita petiolata vel sessilia et subamplexicaulia membranacea integra penninervia, Cymae pedunculatae terminales plus minusve compositae glabrae vel pubescentes. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 85 limb erect, spreading or reflexed, sinistrorsely contorted in aestivation. Stamens 5, alternate with the appendages and attached to the base of the corolla-tube, convergent about the pistil; anthers of two small unilocular sporangia borne ventrally near the apex of an enlarged sagittate, peltately affixed connective; auricles of the anther connective parallel, with an open sinus, not convergent; filament short, cylindrical; pollen borne in per- sistent tetrads. Carpels 2, unilocular, sessile, united at the apices by an ovate-cylindrical clavuncle bearing the terminal stigma; ovules many, anatropous, borne upon an axile, binate placenta^ Receptacle totally immersed in the calyx-tube, entire; nectaries 5, separate, ovate-cylindrical, situated upon the receptacle in a ring about the carpels. Follicles 2, divaricate, terete, containing many dry, apically comose, terete exalbuminous seeds; embryo straight. Type species : Apocynum cannahinum L. Sp. PI. ed. 1, 213. 1753. KEY TO THE SPECIES a. Leaves drooping or spreading (except 4-6)] corolla at least twice the length of the calyx-lobes. b. Leaves drooping; corolla at least three times the length of the calyx- lobes (except Ih), c. Corolla campanulate, 5-10 mm. long, the orifice of the tube more than twice the width of the base; inflorescence rarely axillary aa well as terminal; folhcles normally pendulous; coma of the seeds 15-17 mm. long. d. Lower surface of leaves glabrous 1, A. androsaemifolium dd. Lower surface of leaves tomentose. . la. A. androsaemifolium var. incanum cc. Corolla cylindrical, 4-6 mm. long, the orifice of the tube about equalling the width of the base; inflorescence usually axillary as well as terminal; follicles normally erect; coma of the seeds 12-15 mm. long. d. Corolla about twice the length of the calyx-lobes; follicles unknown, but supposedly pendulous, .lb. A. androsaemifolium var. intermedium dd. Corolla at least three times the length of the calyx-lobes; foUicles frequent, erect, or nearly so. e. Leaves glabrous 2. A, pumUum ee. Leaves variously pubescent Sa, A. pumUum var. rhornboideum bb. Leaves spreading or ascending; corolla about twice the length of the calyx-lobes. c. Leaves spreading; corolla relatively large, 4-5 mm. long. d. Corolla campanulate, the tube scarcely longer than the spread of the limb; primarily eastern varieties. e. Calyx and leaves variously tomentose; calyx-lobes lanceolate, acuminate, colored. (Vol. 17 86 ANNALS OF THE MISSOURI BOTANICAL GARDEN f. Corolla glabrous without; upper surface of leaves glabrous or glabrate S. A. medium ff. Corolla hirtellous without; upper surface of leaves softly pubescent Sa. A, medium var. sarniense ee. Calyx and leaves glabrous; calyx-lobes oblong to ovate^ abruptly acute, colorless 36. ^4. medium var. Icuconeuron dd. Corolla cylindrical, the tube conspicuously longer than the spread of the limb; western varieties. e. Plant completely glabrous .5c. A. medium var. fiorihundum f. Plant glabrous except the under surface of the leaves; calyx- lobes ciliate-erose Sd. A, medium var. lividum ff. Entire plant pubescent; calyx lobes entire Se. A. medium var. vestitum cc. Leaves ascending; corolla 2-3 mm. long. d. Cal3TC-lobes ovate, about one-third the length of the corolla; corolla urceolate, about as long as broad 4, A. Jonesii dd. Calyx-lobes lanceolate, about one-half the length of the corolla; corolla cylindrical, about twice as long as broad. e. Leaves oblong, slightly spreading 5. A. Suksdorfii ee. Leaves narrowly lanceolate, strictly ascending 5a, A. Suksdorfii var. angustifolium aa. Leaves ascending, corolla barely exceeding the calyx. b. Leaves evidently petiolate, narrowed at the base, or the very lowest obtuse or sessile; bracts of the inflorescence scarious and aristate, not semifoliaceous or laminate; follicles relatively long and usually falcate, 12-20 cm. long, the coma of the seeds 20-30 mm. long. C. Calyx-lobes decidedly shorter than the corolla; corolla 2-3 mm. long. d. Plant pubescent, at least the lower surface of the leaves. e. Plant glabrous, except the lower surface of the leaves . .6. A, cannabinum ee. Inflorescence and both surface of the leaves pubescent 6a. A. cannabinum var. pubcscens dd. Plant entirely glabrous 6b. A. cannabinum var. glaberrimum cc. Calyx-lobes equalling or slightly surpassing the corolla; corolla 3-4 mm. long 6c. A. cannabinum var. Greeneanum bb. Leaves nearly or quite sessile, especially the lowermost, cordate or obtuse at the base; bracts of the inflorescence semifoliaceous or laminate; follicles relatively short and straight, 4-10 cm. long, the coma of the seeds 8-20 mm. long. c. Corolla sphaerico-cyhndrical, about as broad as long; coma of the seeds 8-12 mm. long; eastern and middle-western species. d. Leaves oblong to lanceolate, the bases obtuse to cordate. e. Plant glabrous 7. A. hy peridfolium ee. Plant pubescent 7a. A. hy peridfolium var. Farwellii dd. Leaves ovate to ovate-oblong, the bases deeply amplexicaul. . . 7b. A. hyperidfolium var. cordigerum cc. Corolla tubulo-cylindrical, longer than broad; coma of the seeds 15-20 mm. long; western species 7c. A. hyperidfolium var. salignum 1. Apocynum androsaemifolium L. Sp. PL ed. 2, 311. 1762; Walt. Fl. Carol. 107. 1788; Lam. Encycl. 1: 213. 1783; Michx. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 87 Fl. Bor. Am. 1: 121. 1803; Pursh, FL Am. Sept. 1: 179. 1814; Ell. Sketch Bot. S. C. & Ga. 1: 314. 1821; Torr. Fl. N. & Mid. States, 275. 1824; Raf. Med. Bot. 1: 49. 1828; Darl. Fl. Cestr. 167. 1837; G. Don, Hist. Dichlam. PI. 81. 1838; Strong, Am. Fl. 4: 72. 1850; Wood, Classb. Bot. ed. 29, 456. 1853; Darby, Bot. South. States, ed. 2, 434. 1860; Piper, Contr. U. S. Nat. Herb. 11: 453. 1906. Apocynumfol. androsaemi L. Sp. PI. ed. 1, 213. 1753. Apocynum androsaemifolium L. var. glabrum Macoun, Cat. Can. PL 2:317. 1884. Apocynum amhigens Greene, PI. Baker. 3: 17. 1901; Rydb. Fl. Colo. 269. 1906; Brown, Univ. Mo. Studies 2^: 193. 1911; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 505. 1915; Rydb. Fl. Rocky Mts. ed. 2, 669. 1917; Standi. Contr. U. S. Nat. Herb. 22^: 395. 1921; Tidestrom, Contr. U. S. Nat. Herb. 25: 419. 1925; Garrett, Spring Fl. Wasatch Reg. 120. 1927. Apocynum scopulorum Greene ex Rydb, Fl. Colo. 269. 1906; Brown, Univ. Mo. Studies 2^: 193. 1911; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 505. 1915; Rydb. Fl. Rocky Mts. ed. 2, 668. 1917. Apocynum androsaemifolium L. var. microphyllum Beg. & Bel. Atti R. Accad. Lincei, V. 9: 87. 1913. Apocynum amhigens Greene var. hicolor Beg. & Bel. I. c. 89. 1913, non MacGregor. Apocynum androsaemifolium L. var. typicum Beg. & Bel. I. c. 80. 1913. Apocynum occidentale Rydb. ex Beg. & Bel. I. c. 1913, nomen, in synon. Apocynum androsaemifolium L. var. orbiculatum (Greene) Beg. & Bel. l. c. 81. 1913. Cynopaema androsaemifolium (L.) Lunell, Am. Mid. Nat. 4: 509. 1916. Apocynum macranthum Rydb. Fl. Rocky Mts. ed. 2, 669. 1917. Stems erect or ascending, 2-5 dm. tall, glabrous, freely and rather dichotomously branched, branches ascending, alternate to subalternate; leaves opposite, petiolate, drooping, ovate to oblong- lanceolate, membranaceous, glabrous, entire, mucronate, tapering at both base and apex; corolla campanulate, the base of the tube 88 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vou 17 less than one-half the breadth of the limb, 4-9 mm. long, the limb reflexed, 5-10 mm. broad at the orifice, calyx-lobes 1-3 mm. long, lanceolate to ovate-oblong, slightly flushed or colorless, glabrous; follicles 6-15 cm. long, pendulous, divergent, straight, not falcate, coma of seeds 1.5-2 cm. long, pale tawny-colored, seeds ovoid, 1 mm. long. Distribution: woodlands, northwestern United States and ad- jacent Canada, eastward to the Dakotas, southward to California, Nevada, and Utah. Specimens examined: Canada : British Columbia: bluff above Lardo, alt. 2000 ft., June 15, 1905, Shaw 680 (US, NY, MBG). United States: South Dakota: Sylvan Lake, alt. 7000 ft., July 20, 1892, Rydberg 871 (NY) ; Black Hills, near Ft. Meade, July 27, 1887, Farwood 252 (US) . Nebraska; War Bonnet Canyon, alt. 5000 ft., June, 1890, Williams (US, MBG). Texas: Davis Mts., Jeff Davis Co., upper Limpia Canyon, July 9-12, 1921, Ferris & Duncan 251^2 (NY, US, MBG). Montana: Bozeman, Sept. 5, 1901, W. W. Jones (G); Deer Lodge, 1901, Scheuher 1 (NY); Glacier Park, Aug. 18, 1919, Standley 17846 (NY, US); Nevada Creek, hillsides, July 11, 1883, Canhy 226 (ANSP, US) ; Spanish Basm, Gallatin Co., June 23, 1897, Rydberg & Bessey 4701 (ND); Belton, Oct. 3^, 1919, Standley 18821 (US) ; Grinnell Glacier, Glacier Nat. Park, July 31, 1919, Standley 16845 (US); Bigfork, alt. 3000 ft., Aug. 6, 1908, M. E. Jones 8881 (P) ; Darby, alt. 4000 ft., July 30, 1909, M. E. Jones (P); Yellowstone Park, Aug. 13, 1911, TreakU 159 (P); Rexford, Sept. 16, 1908, M. E. Jones 8880 (P). Wyoming: open woods, Black Hills, July 25, 1910, A. Nelson 9491 (G); Whalen Canyon, July 19, 1894, E. Nelson 535 (US); Jaw-bone Gulch, July 9, 1896, E. Nelson 2027 (P); Sheridan Co., above Big Horn, June 26, 1897, Pammel & Stanton 71 (MBG) ; Jackson's Hole, Lincoln Co., alt. 6700 ft., dry roadside, Aug. 3, 1920, Payson & Payson 2186 (MBG). Colorado: Horsetooth Mt., July 20, 1897, Crandall 1013 1930] WOODSON — STUDIES IN APOCYNACEAE. I 89 (NY, US); Boulder Canyon, July, 1891, Penard 500 (NY); Mor- rison, June 19, 1881, B. H. Smith (ANSP); hills, Larimer Co., June 15, 1895, Patterson 6863 (P) ; Dome Rock in Platte Canyon, Aug. 8, 1878, M. E. Jones 572 (P); hillside near Odessa Lake, Larimer Co., elev. 10000 ft., Aug. 15, 1927, Woodson 1917 (MBG); dry slopes, Gregory Canyon, near Boulder, frequent, alt. 6000 ft., June 17, 1921, H. C. Hanson 276 (MBG); Steamboat Springs, Routt Co., shady hillsides, July 20, 1903, Goodding 1622 (MBG, NY, ANSP, US, F); Black Canyon, alt. 7000 ft., June 20, 1901, Baker 202 (MBG, US, NY, W); clearings in rocky coniferous forests, above Cub Lake, Larimer Co., July 5, 1929, Woodson S: Anderson 29120 (MBG). New Mexico: data lacking, Fendler 690 (MBG, G); Wing- field's Ranch, Sacramento Mts., July 8, 1895, Wooton (US). Idaho : Challis Creek, Custer Co., base of rock slide in canyon, alt. 6000 ft., July 19, 1916, Machride & Payson 3327 (G, P, MBG, US); Priest Lake, Aug., 1901, Pi^er 3786 (G, US); crevice in recent lava flow, Martin, Blaine Co., alt. 6000 ft., July 5, 1916, Machride & Payson 3040 (G, NY, MBG); mountains, exact lo- cality lacking, July, 1892, Mulford (G, MBG) ; Spencer, July 10, 1916, Rust 247 (US); Middle Fork of Weiser River, Washington Co., Aug. 3, 1899, M. E. Jones (P) ; Divide between Mullan and Canyon Creek, Coeur d'Alene Mts., alt. 1020 m., Aug. 7, 1895, Leiherg 1500 (MBG, G, US, NY, F); Priest River Exp. Sta., alt. 2700 ft., July, 1923, Eyling 6088 (MBG) ; Trinity, Elmore Co., Aug. 8, 1910, Machride 549 (MBG, US, W) ; Salmon, Lemhi Co., shaded roadside, alt. 4500 ft., July 3, 1920, Payson & Payson 1886 (MBG, NY) ; hills southeast of Victor, Teton Co., alt. 6500 ft., July 28, 1920, Payson & Payson 2170 (MBG, "NY, G). Utah: City Creek Canyon, July 23, 1884, July 26, 1883, Leonard (G); Provo, moist semi-shaded places, June 16, 1902, Goodding 1127 (G, MBG, US, NY, ANSP, F, P); East Bountiful, Oct. 22, 1909, CUmens (NY) ; Parley's Canyon, Salt Lake Co., July 7, 1909, Garrett 2450 (MBG); Mt. Nebo, July 15, 1905, Rydherg 7761 (NY, US); Big Cottonwood Canyon, below Silver Lake, July 8, 1905, Rydherg 6772 (NY, US); Provo, July 2, 1894, M. E. Jones 5560 (P, US) ; Provo Slate Canyon, gravel, alt. 6000 ft., July, 1894, M. E. Jones 5613 (P, MBG); Ogden, June 28- Aug. 1, 1890, Pammel 5 (MBG). Vol. 17 90 ANNALS OF THE MISSOURI BOTANICAL GARDEN Nevada: Lee Canyon, Charleston Mts., alt. 8000 ft., July 25, 1913, Heller 10982 (MBG, US, G, NY); mountains west of Carson City, alt. 5000 ft., July 13, 1909, Heller 9815 (US). Washington: Three Tree Point, common in open places. May 30, 1910, Bardell & Frye (MBG); Wenatchee, Aug. 6, 1896, Whited 14£ (US); Loon Lake, July 20, 1897, Wurston (NY); Stehekin, Chelan Lake, July 6, 1911, M. E. Jones (P); Pullman, July, 1896, Elmer 6863 (P) ; Pullman, on north hillside-meadows, July 4, 1896, Elmer 138 (P); Pullman, July 25, 1900, Piper 3508 (US, W) . Oregon: Trail Creek Canyon, alt. 1600 ft., June 2, 1897, Sheldon 8208 (MBG, US); Emigrant Creek, Jackson Co., May 19, 1898, Ehner 2216 (US); Corvallis, common along roads, May, 1922, Epling 557 J^ (MBG). California: data lacking. Bridges 160 (NY); Strawberry, July, 1920, Evans (P). There have been several misunderstandings concerning A. androsaemifolium which have had widespread circulation. One in particular concerns the pubescence or glabrity of the foliage. To practically all botanists who have had to do with the genus, the typical form of the species has been that of the collections from the eastern United States and Canada, the leaves of which have a densely tomentulose under surface. The western form with completely glabrous leaves has therefore frequently been relegated to some other designation, or ignored altogether. That the two forms are significant, however, their geographical distri- bution is proof. Linnaeus, in describing the species, spoke clearly of the foliage as ''utrinque glabris," and A. deCandolle separated the glabrous typical form from a variety, var. incanum, a form with leaves ''subtus pubescentibus." As a matter of fact, the possibility that Linnaeus's description was a careless description of the tomentulose form rather than a careful description of the glabrous form appears likely because of the abundance of the former. At any event, nomenclatorially, Linnaeus's A. andro- saemifolium clearly describes the western glabrous variety, while deCandolle's var. incanum is as clearly the correct designation for the common dogbane of the eastern United States and Canada. It is very rare to find the publication of A. androsaemifolium 1930] WOODSON — STUDIES IN APOCYNACEAE. I 91 L. correctly cited. Linnaeus first published the species (Sp. PI. ed. 1, 213. 1753) as "Apocynum fol, Androsaemi." Notwith- standing, the original publication of A. androsaemifolium L. (Sp. PL ed. 2, 311. 1762) is almost invariably given in our floras as well as monographic work^* as "Sp. PL ed. 1, 213. 1753/' perhaps in an effort to save priority and therefore usage for the common appellation. Such an effort, however, is unnecessary, since Lin- naeus's earlier name was never taken up as such, and was there- fore long ago invalidated by the so-called "Fifty-year Rule" of the International Code of Nomenclature. la. Apocynum androsaemifolium L. var. incanum A. DC. in DC. Prodr. 8: 439. 1844; Wood, Classb. Bot. ed. 29, 456. 1853; B^e. & Bel. Atti R. Accad. Lincei V. 9: 84. 1913. Plate 9. Apocynum muscipulum Moench, Meth. 464. 1794. '^Apocynum androsaemifolium L." ace. to Hooker, Fl. Bor.-Am. 1:51. 1840;Torr.FLN.Y. 2:116. 1843; A. Gray, Bot. Cal. 1: 473. 1876;Coulter, Man. Bot. Rocky Mts. 237. 1885; A. Gray, Syn. FL N. Am. 2^: 82. 1886; Coulter, Contr. U. S. Nat. Herb. 22 : 262. 1892; MacMillan, Metaspermae Minn. Valley, 421. 1892; Coville, Contr. U. S. Nat. Herb. 4: 149. 1893; K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4^: 179. 1895; Chapm. Fl. South. U. S. ed. 3, 341, 1897; S. Coulter, Fl. PL Ind. 880. 1899; Rydb. Mem. N. Y. Bot, Card. 1 : 311. 1900; Mohr, Contr. U. S. Nat. Herb. 6: 674. 1901; Howell, Fl. N. W. Am. 1: 439. 1901; Piper, Contr. U. S. Nat. Herb. 11: 453. 1906; Rydb. FL Colo. 269. 1906; Robinson & Fernald in Gray, New Man. Bot. ed. 7, 662. 1908; Small, FL Southeast. U. S. ed. 2, 936. 1913; Britton & Brown, 111. Fl. 3: 21. 1913; Rydb, FL Rocky Mts. ed. 2, 668. 1917; Bergman, Fl. Rocky Mts. 668. 1917; Bergman, Fl. N. D. 229. 1918; Tidestrom, Contr. U. S. Nat. Herb. 25: 420. 1925; Garrett, Spring FL Wasatch Reg. 120. 1927. Apocynum androsaemifolium L. f. pau^iflora Peck, N. Y. State Mus. Rept. 47: 158. 1894. Apocynum incanum (A. DC.) G. S. Miller, Proc. Biol. Soc. Wash. 13:81. 1899, in synon. Apocynum silvaticum Greene, Leafl. Bot. Obs. & Crit. 2: 179. 1912. »'cf. B6g. & Bel. Monogr. Apocynum. Atti R. Accad. Lincei, V. 9: 79. 1913. 92 ANNALS OF THE MISSOURI BOTANICAL GARDEN rvoL. 17 Apocynum androsaemifolium L. var. puherulum B^g. & Bel. Atti R. Accad. Lincei V. 9: 81. 1913. Lower surface of leaves distinctly tomentulose or pubescent^ in all other essential characters similar to the species. Distribution: woodlands, occasionally in waste-lands, all over temperate North America, with a probable center of distribution in the northeastern United States and adjacent Canada. Specimens examined: Newfoundland: open woods near the confluence of the Exploits River and Badger Brook, Aug. 12, 1894, Robinson &, Schrenk 96 (G, MBG, US, F); ledges and talus, north bank of Exploits River below the falls, Grand Falls, July 22, 1911, Fernald & Wiegand 6087 (G); Bishop Falls, July 28, 1911, Fernald & Wiegand 6088 (G); "Sunnyside," Bay of Islands, fields, date lacking, Waghorne (MBG). Canada : Nova Scotia: dry open soil, Middleton, July 21, 1920, Bissell, Pease & Long 22315 (G); dry exposed serpentine ledges in thin upland woods, Middlefield, June 26, 1913, Fernald & Long 102 J^O + (G); near "Sunny Brae," Pomquet, July 11, 1927, Woodson 1713 (MBG); sandy thicket, Kentville, Aug. 22, 1902, Fernald (G); fields near Tarbet Vale, Cape Breton Island, July 23, 1914, Nichols 289 (G); Ridge Road, above Springfield, Aug. 6, 1910, Greene (US, ND); edge of fields, near Windsor, July 21, 1857, Howe (NS). Prince Edward Island: dry clearing, Alberton, July 11, 1912, Fernald & St. John 7950 (G). New Brunswick: Shediac, roadside, July 30, 1925, Knowlton (G); rich shaded soil, Shediac Cape, July 23, 1914, Hubbard (G); near Moncton, July 13, 1927, Woodson 171 J^ (MBG); near McKinnon, July 13, 1927, Woodson 1651 (MBG). Quebec: gravelly beaches and flats, between Bald6 and the Bale des Chaleurs, Bona venture River, Aug. 5-6-8, 1904, Collins, Fernald & Pease 5877 (G) ; Riviere Vaureal : le long du rivage de la riviere, July 28, 1925, Victorin & Germain 21988 (G); near Ouatchouan Falls, Lake St. John, Aug. 29, 1904, Wright 229 (G, US); mossy woods along the trail between Mont St. Louis and Lac Mont Louis, July 30, 1923, Fernald, Dodge & Smith 25988 19301 WOODSON — STUDIES IN APOCYNACEAE. I 93 (G); vicinity of Longueuil, Sept. 19, 1919, Vidorin 9652 (G); Lac Tremblant, Terrebonne Co., Aug. 4, 1922, Churchill (MBG, G, NY, US). Ontario: Fort William, July 31, 1912, Williamson 1692 (ANSP); Lake Cache, June 26, 1900, Macoun (US); cemetery near London, Saunders, Anderson & Woodson (MBG); vicinity of Ottawa, July 11, 1906, Rydherg (NY); St. Thomas, June 25, 1906, Fisher (D) ; near Heyden, 75 mi. north of Mackinac Straits, Aug. 2, 1906, E. B, Williamson (B); island in Lake Timagami, July, 1929, Gilmour (MBG). Manitoba: Lake Winnipeg Valley, 1857, Bourgeau (G, US); Brandon, July 20, 1896, Macoun (US) ; Athabasca, Aug. 12, 1901, Prehle & Preble 175 (US). Saskatchewan: exact locality not cited, 1857-58, Bourgeau (G). British Columbia: Coldstream, alt. 2200 ft., Aug. 3, 1905, Shaw 1064 (G, MBG, US, NY, F); hillside, Carbonate, alt. 2750 ft., July 9, 1904, Snyder 217 (G, US); Bald Mt., Cowichan Lake, June 17, 1907, Rosendahl 1835 (NY, G, MBG); Telegraph Creek, Cassiar district, July 20, 1910, Preble & Mixter 549 (US); Glacier, Illecillewaet Valley, Loop Trail, alt. 3500^000 ft., July 20, 1906, Brown 655 (G); lower Frazer River, 1859, Lyall (G); open ground, Victoria, date lacking, J, R. Anderson (W) ; orchards, Crawford Bay, date lacking, Harrison (W); river-bank, Skeena River, Comox District, Sept. 10, 1910, J. R. Anderson 500 (W). United States: Maine: gravelly thicket, Sangerville, June 29, 1895, Fernald 245 (G, MBG, US, NY) ; gravelly river-beaches. Fort Fairfield, July 16, 1902, Williams, Collins & Fernald (NE); Caribou, July 18, 1902, Williams 1 (G); open woods, Green Mt., Mt. Desert, July 18, 1899, Williams (G); Augusta, July 10, 1886, E. C, Smith (MBG); Roque Bluff, dry sandy fields and roadsides, Aug. 11, 1907, Cushman & Sanford 1621 (NE); Seal Harbor, Wildwood Farm, Sept. 1, 1892, Rand (NE); Standish, sandy clearings, Aug. 28, 1916, Fernald & Long 14398 (NE, ANSP); Winthrop, 1864, Sturievant (MBG) ; Topsham, bank of Androscoggin River, Aug. 26, 1912, Furbish (NE); Bridgton, July"9, 1912, Martin (NE); Swans Island, clearing on Stockbridge Hill, July, 1914, Hill 1 540 (NE) . [Vol. 17 94 ANNALS OF THE MISSOURI BOTANICAL GARDEN New Hampshire: Glen, White Mts,, Sept. 17, 1892, Kennedy (G) ; Jaffrey, July 23, 1896, Day 70 (G, US) ; Randolph, along roadsides, July 27, 1897, Williams (G); Conway Center, Aug. 4, 1875, Lane (G); Gorham, R. R. tracks, Sept. 6, 1918, Pease 17345 (NE); Gilmanton, summit of Shannon Mt., exposed granite, July 22, 1907, Cushman & Sanford 1296 (NE) ; Merrimack, open roadside bordering swamp, Aug. 11, 1917, Batchelder (NE). Vermont: Derbyline, dry knoll, July 19, 1914, Churchill (NE); Pownal, roadside, Aug. 2, 1898, Churchill (G); Arlington, July 3-4, 1903, Eggleston 3291 (G, US); Manchester, roadside, July 14, 1898, Day 19 (G, US); Mt. Mansfield, June 25, 1900, Howell (US); Peacham, July 12, 1885, Blanchard (MBG); Caledonia Co., July 2, 1884, Blanchard (MBG); Fairlee Center, Sept. 3, 1927, E. Anderson 2700 (MBG) ; Vergennes, July 9, 1908, Kennedy (NE); East Wallingford, meadow, July 1, 1897, Kent (NE). Massachusetts: West Tisbury, along road, Sept. 5, 1917, Seymour 1431 (G); Conway, July 12, 1907, Williams (G); Worth- ington, open, slightly sandy roadside, Aug. 14, 1912, Robinson 741 (G) ; West Roxbury, date lacking, Faxon (G) ; Tewksbury, June F 27, 1853, Boott (G) ; Ipswich, borders of thickets, July 21, Sept. 10, 1874, Morong (MBG); Granville, Sept. 19, 1913, Seymour 63 (MBG, G); Holbrook, June 18, 1899, Greenman 616 (MBG); river-bank, Shelburne Falls, Franklin Co., July 1, 1921, Churchill (MBG, NY, US); Oak Island, July 9, 1898, Greenman 3106 (MBG); Nonquit, 1888, Sturtevant (MBG); Alford, Aug., 1898, Milligan (MBG); South Framingham, July 3, 1890, Sturtevant (MBG); Pride's Crossing, dry loamy gravel, July 11, 1913, Hubbard (NE); Framingham, dry woods, July 13, 1911, Fames (NE); Lexington, moist pastures, Aug. 4, 1906, Knowlton (NE). Rhode Island: North Smithfield, damp ditch on roadside, Aug. 4, 1910, Hope 453 (NE) ; Foster, July 7, 1913, Torrey 748 (NE) ; Cumberland, roadside. May 30, 1911, Ware (NE) ; Tiverton, roadside, July 7, 1909, Sanfmd (NE) ; Middletown, July 17, 1898, Simmons (NE). Connecticut: Hartford, July 16, 1900, Driggs (NE); Trum- bull, June 22, 1892, E. H. Eames (NE) ; Southington, woods and roadsides, conmion, June 14-Aug. 20, 1899, Bissell (NE); Water- 1930) WOODSON — STUDIES IN APOCYNACEAE. I 95 bury, July 27, 1912, Blewitt 3674 (G); East Greenwich, 1861, Fitch (US). New York: Sylvan Beach, Oneida Co., July 11, 1905, House 1208 (G, NY); Lisbon, dry roadsides, July 9, 1914, Phelps (G); South Bay, Wellesley Is., June 28-30, 1902, T. R. Robinson & W, R. Maxon 37 (G); Cape Vincent, July, 1905, Keller (ANSP); Fleischmann's, Delaware Co., July 13, 1892, Schrenh (MBG); Granville Tp., Washington Co., Aug. 19, 1921, Drushel 2^34 (MBG); Little Moose Lake, Herkimer Co., July 5, 1906, Van Pelt (ANSP); Goat Island, Niagara River, July 19, 1870, Boice (ANSP); near Oneida, July 8, 1909, Mrs. Maxon 46^1 (ANSP); Ithaca, Cascadilla Creek, June 18, 1877, Trelease (MBG); New London, June 28, 1921, House 8294 (NY); Ithaca, dry woods, June 28, 1916, Munz 599 (P) ; Norwich, Chenango Co., June 24, 1887, Fitch (P). New Jersey: Windham road, July, 1889, R. C. Perry (MBG) ; Netcong, Morris Co., Aug. 19, 1906, Mackenzie 2290 (MBG, NY, US); Lakewood, Ocean Co., sandy woods. May 30, 1908, Mackenzie 3102 (MBG, NY). Pennsylvania: Bushkill, Pike Co., June 30, 1917, Bartram (G, ANSP); Binton, Columbia Co., July 8, 1921, Meredith (ANSP); Tannersville, July 4, 1901, S. Brown (ANSP); Mead- ville, June 30, 1893, C. D. Curtis (P); Columbia, June, 1889, Ehy (MBG); moist places along the Susquehanna trail near Selinsgrove, Aug. 4, 1926, Heller 14232 (MBG, US); bluff along the Susquehanna River above Nescopec, about or near trees, Aug. 3, 1926, Heller I4204 (MBG); Northampton Co., 1880, Rau 66 (MBG); Mt. Alto, 1909, Ulick (MBG). Delaware: Centerville, July ,1865, Commons (MBG, ANSP). Maryland: Oakland, Garrett Co., July 7, 1881, Donnell- Smith (US, G); Oakland, July 12, 1899, Greene (US). Virginia: Stony Man Mt. and vicinity, near Luray, alt. 3700 ft., Aug. 15, 1901, Steele & Steele 238 (MBG, US, NY); Shannon Gap, Walker Mt., Smyth Co., June 21, 1892, Britton, Britton & Vail (NY); Peaks of Otter, Bedford Co., July 26-Sept. 1, 1871, Curtiss (G, US). West Virginia: Bucklin, Upshur Co., June 10, 1895, Pollock (MBG, US). 96 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 North Carolina: open woods and fields, Biltmore, Bun- combe Co., July 16, 1897, Biltmore 80b (MEG, P, G, US, NY); "mountains of North Carolina," exact locality and date lacking, Ashe (US). Ohio: Put-in-Bay, July 10, 1920, Mosely (G); Rocky River, near Cleveland, July 13, 1896, Greenman 1437 (G, MBG); Columbus, date lacking, SuUivant (G); Painesville, June, 1885, Beardslee (US). Michigan: Stevensville, sandy ground near railroad, July 3, 1911, Lansing 3213 (G, F); Burt Lake, roadside, July 27, 1917, Ehlers 559 (G, US); Hamlin Lake, Ludington, July 25, 1910, Chaney 92 (G); Turin, Marquette Co., July 8, 1901, Barlow (G); Iron Mt., June, 1885, Rydberg 130 (NY); Schoolcraft, railroad banks, June 6, 1903, Burgess 62 (F) ; roadside banks 15 miles west of Ann Arbor, Aug. 22, 1926, Heller 14258 (MBG); Lansing, July 7, 1867, Sleeper (MBG); Gt. Manitou Is., Lake Superior, July 26, 1886, Wislizenus 579 (MBG); Houghton Co., Aug., 1892, Eby (MBG); Ypsilanti, June 23, 1892, Farwell (US). Indiana: Gibson, sandy ridges, Aug. 10, 1910, Lansing 2820 (G, US, F) ; Millers, dry copse, July 7, 1908, Lansing 2764 (G, F) ; Whiting, Aug. 29, 1893, Britton (NY); Notre Dame, Aug. 1909, Nieuwland (MBG); Bluff ton, along fence 14 mile east of town, June 24, 1906, Deam 1125 (MBG, D); Michigan City, July 3, 1903, Mell2S (MBG, US); Gary, July 19, 1907, Greenman 1864 (MBG); in wooded lot across railway from Notre Dame LTniversity, June 29, 1928, Nieuwland, Slavin & Woodson 2 (MBG); bank of the St. Mary's River about 2 mi. southeast of Pleasant Mills, Adams Co., June 25, 1925, Deam 41342 (D) ; Fulton Co., along the Lake Erie Railroad, between Rochester and Tiosa, not frequent, June 25, 1911, Deam 8923 (D); high bank of the Salamanie River, 1 mi. east of Montpelier, Blackford Co., June 22, 1925, Deam 41320 (D); dry wooded bank of Cedar Creek, about 3 mi. above Cedarville, Allen Co., June 14, 1914, Deam 14283 (D) ; black oak sand hill along roadside, 5 mi. west of Gifford, Jasper Co., July 14, 1920, Deam 31710 (D); Marshall Co., on the border of Lake Maxinkuckee, just south of Culver, July 2, 1911, Deam 8997 (D); on wooded dune about 2 mi. west of Michigan City, not frequent, June 18, 1911, Deam 8848 (D). 1930] WOODSON — STUDIES IN APOCYNACEAE. I 97 Wisconsin: Brown Co., 1886, Scheutte (G); River Falls, Aug., 1903, Powell (G); Waupaca, 1907, Garesche (MBG); woods near Mirror Lake, Sauk Co., Aug. 23, 1903, Eggert (MBG); low open ground, near lake, Odanah, Ashland Co., Sept. 11, 1925, E. J. Palmer 28651 (MBG) ; Burg road, near Garrett Bay Inn, Ellison Bay, Door Co., July 16, 1918, M. T. Greenman 52 (MBG); St. Croix Falls, Polk Co., July 10, 1899, C. F. Baker (P); Arena, date lacking, Carr (P). Illinois: Urbana, date lacking, Gleason (G); Flossmoor, Aug. 1, 1909, Greenman 2815 (G, MBG, NY); Decatur, 2 miles from Moffit's Bridge, on the Moffit road, June 29, 1895, Gleason SS6 (G); bank of Desplaines river, Maywood, June 24, 1898, A. Chase (MBG, US); steep clay bank near Wady Petra, June 30, 1900, V. H. Chase 647 (MBG, US); Illinois State Park, Starved Rock, LaSalle Co., June-Sept. 1921, Thone 61 (MBG, F); Cook Co., July, 1870, Babcock (MBG, US); Lake Forest, May-Sept., 1897, Jensen (MBG). Minnesota: Clithrall, July, 1897, Camphell (G); Bemidji, Beltrami Co., June, 1902, Brand 591 (F); Taylor's Falls, July 28, 1900, Banks 21 (ND, US) ; Norway Beach, Cass Lake, sandy soil, border of woods, July 9, 1925, Pammel 864 (MBG); Itaska Lake, June 29, 1892, Sandberg (US); Soudan, St. Louis Co., July, 1896, Eby (MBG); Center City, July, 1892, Taylor (W), borders of thickets, Fish Lake, Mahnomen Co., July 21, 1908, Chan- donnet (D). Iowa: Iowa City, 1889, Hitchcock (MBG); Ames, Aug. 26, 1897, Johnson (MBG) ; border of woods, Fayette Co., July, 1894, Fink (US). Missouri: Ironton, June 23, 1897, Savage & Steele 310 (F); Joplin, rocky ground, June 7, 1909, E. J. Palmer, 2129 (G, MBG, US); Dumas, rocky banks, July 6, 1909, Bush 5871 (MBG, G, US) ; Eagle Rock, uncommon, near spring, June 26, 1897, Bush 15 (G, MBG, US); near Allenton, high sandy hill, June 30, 1912, Letterman (MBG); Howell Co., Aug. 12, 1892, Bush (MBG); rocky hills, St. Louis Co., May 15, 1896, Eggert (MBG); "rare and local, a clump at Castle Rock," Turkey Creek, near Joplin, June 7, 1909, E. J. Palmer 2172 (MBG, US); Jerome, June 16, 1914, Kellogg (MBG); open rocky woods, north slope along 9. 17 118 ANNALS OF THE MISSOURI BOTANICAL GARDEN Apoq/num hypericij olium Ait. var. pseudosuksdorfii B^g. & Bel. I c. 118. 1913. "Apocynum Suksdorfii Elm. non Greene" ex B^g. & Bel. I c. 1913, nomen in synon. Stems erect or slightly ascending, 4-6 dm. tall, glabrous throughout, freely and somewhat fastigiately branched, the branches opposite to subopposite; leaves opposite, petiolate, oblong-ovate, 4-8 cm. long, 1.5-2.5 cm. broad, acute at both base and apex, membranaceous, entire, mucronate; corolla cylin- drical, 2-3.5 mm. long, about 1.5-2 mm. broad at the orifice, the base of the tube about equalling the breadth of the orifice, limb erect; calyx-lobes lanceolate to oblong, 1-1.5 mm. long, entire, colorless or slightly tinted; follicles 9-10 cm. long, falcate, pendulous, seed 3.5^ mm. long, coma of seed 2-2.5 cm. long, slightly tawny. Distribution: generally throughout the Pacific Coast region, eastward to New Mexico and southern Colorado. Specim.ens examined : Colorado: New Windsor, Aug. 8, 1898, Osterhout (NY); steep banks of Mancos River opposite mouth of Johnson's Canyon, Southern Ute Indian Reservation, Montezuma Co., June 27, 1929, Woodson & Anderson 29075 (MBG). Utah: Salt Lake City, July 16, 1880, alt. 4300 ft., M. E. J ones 1863 (P). Nevada: King's Canyon, Ormsby Co., alt. 1700-2000 m., Aug. 20, 1902, Baker 1508, in part (MBG, P) ; Hmnboldt Canyon, West Humboldt JNIountains, Humboldt Co., elev. 5100 ft., July 31, 1912, Heller 10612 (MBG, G, NY, F); Wadsworth, July 26, 1919, Tidestrom 10727 (G, US); Palisade, Elko Co., elev. 5066 ft., Aug. 11, 1909, Heller 9940 (G). Arizona: Navajo Reservation, Nitsie C, a few dense patches, July, 1916, Vorhies U (G, MBG, US, NY); Chevlon, w. of Hol- brook, July Canyon, Painted Desert, plentiful, July 25, 1920, Clute87 (G, US); Willow Spring, June 10-20, 1890, E. Palmer 511 (G). Washington: Klickitat Co., high sandy bank of Columbia River, July 22, 1893, Suksdorf 1522 (G, MBG, US, F) ; Wenatchee, date lacking, Whited (W); North Yakima, Aug., 1895, Watt (W). 1930] WOODSON — STUDIES IN APOCYNACEAE. I 119 Oregon: on rocks in North Fork of Umpqua River, between Rock Creek and Steamboat Creek, alt. 900-1200 ft., Elmer (US). California: Waysides, Palomar Mt., San Diego Co., alt. 5600 ft., Aug. 14, 1918, Spencer 995 (G, P). 5a. Apocynum Suksdorfii Greene var. angustifolium (Wooton) Woodson, n. comb. Apocynum angustifolium Wooton, Contr. U. S. Nat. Herb. 16: 159. 1913; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 506. 1915. Apocynum hypericifolium Ait. var. angustifolium (Wooton) B4g. & Bel. Atti R. Accad. Lincei, V. 9: 120. 1913. Apocynum pseudolaurinum B4g. & Bel. var. typicum B^g. & Bel. I. c. 1913, nomen. Leaves lanceolate to linear-lanceolate; in all other essential characters similar to the species. Distribution : New Mexico and southeastern Utah. Specimens examined: New Mexico: Gila River bottoms, near Cliff, Grant Co., June 13, 1903, Metcalfe 132 (MBG, G, US, NY); Mimbres, Grant Co., July 1, 1904, Metcalfe 1070 (G, MBG, P, US); near Pecos, San Miguel Co., alt. 6700 ft., Aug. 18, 1908, Standley SOU (MBG, US). Utah: San Juan River, July 14, 1895, Eastwood 71 (G). 6. Apocynum cannabinum L. Sp. PL ed. 1, 213. 1753; Michx. FL Bor. Am. 1: 121. 1803; Pursh, Fl. Am. Sept. ed. 1, 1: 179. 1814; Hook. Fl. Bor.-Am. 1 : 51. 1840; A. Gray, Bot. Cal. 1 : 473. 1876; Coulter, Man. Bot. Rocky Mts. 237. 1885; A. Gray, Syn. Fl. N. Am. 2': 82. 1886; Chapm. Fl. South. U. S. ed. 3, 341. 1897; Howell, Fl. N. W. Am. 1: 439. 1901; Piper, Contr. U. S. Nat. Herb. 11: 453. 1906; Robinson & Fernald in A. Gray, New Man. Bot. ed. 7, 662. 1908; Small, Fl. Southeast. U. S. ; Britt. & Brown, 111. Fl. ed. 2, 3: 22. 1913; ed. 2, 936. 19K Rydb. Fl. Rocky Tidestrom, Contr. U. S Nat. Herb. 25: 420. 1925; Garrett, Spring Fl. Wasatch Region 120. 1927. Plate Apocynum platyphyllum Greene, Leaf!. Bot. Obs. & Crit. 2: 166. 1912. [Vol. 17 120 ANNALS OF THE MISSOURI BOTANICAL GARDEN Apocynujn cannaUnum L. var. puberulum B^g. & Bel. Atti R. Accad. Lincei, V. 9: 101. 1913. Apocynum cannahimim L. var. incanum B^g. & Bel. L c. 1913. Cynopaema cannahinum (L.) Lunell, Am. Mid. Nat. 4: 509. 1916. Steins erect or ascending, 3-6 dm. tall, glabrous, freely and somewhat fastigiately branched, branches ascending, opposite to sub-opposite; leaves opposite, petiolate, ovate to lanceolate, 4-10 cm. long, 1.5-4 cm. broad, acute to obtuse or rounded at both base and apex, membranaceous, entire, mucronate, glabrous above, tomentulose or pubescent beneath ; inflorescence terminal, the bracts scarious, inconspicuous; corolla sphaerico-cylindrical, about as broad as long, 2-3 mm. long, 2-2.5 mm. broad, the limb erect; calyx-lobes lanceolate to ovate-oblong, 1.5-2 mm. long, entire, glabrous, scarious; follicles relatively long and subfalcate, 12-20 cm. long, pendulous, seeds 4-5 mm. long, coma of seed 2.5-3 cm. long, white. Distribution: generally throughout the eastern half of the United States, infrequent in New England; a common field and roadside weed in the central states. Specimens examined: Vermont: banks of Connecticut River, cove below Bradford, Aug. 18, 1927, Mathias 275 (MBG). Massachusetts: Plymouth, date lacking, Oakes (G); Non- quit, July 14, 1888, Sturtevant (MBG); Chestnut Hill, Easton, Porter (US). Rhode Island: Smithfield, July 23, 1845, Olney (G). Connecticut: Bridgeport, July 8, 1895, Eames (G, US); Southington, sandy fields, frequent, July 10-Aug. 25, 1899, Bissell (NE) ; Stratford, July 25, 1894, Eames (NE) ; Milford, Aug. 9, 1893, Eames (NE); New Haven, 1859, Eaton (US). New York: near Syracuse, 1891, Strauh (US); Valley Head, July 18, 1898, Ruth 501 (US). Pennsylvania: Bradford Hills, Chester Co., June 26, 1910, Bariram 1007 (ANSP, G); vicinity of Millersville, Lancaster Co., July 10, 1890, Smull (NY, P). Delaware: Centerville, 1866, Commons (MBG) ; same locality, June 24, 1879, Commons (US, G); EUendale, July 23, 1878, Canby (US) . 1 930] WOODSON — STUDIES IN APOCYNACEAE. I 121 Maryland: Berwyn, June 5, 1912, Greerie (MBG, US, ND); Deer Park, Aug. 4, 1906, Carter (ANSP) ; Hancock, June 11, 1927, Woodson 1730 (MBG). Virginia: Dam Neck, Princess Anne Co., low ground, meadow, June 27, 1922, Randolph & Randolph 481 (G) ; near Ocean View, Norfolk Co., July 8, 1898, Kearney 1472 (US); Suffolk, Nanse- raond Co., June 8-13, 1893, Heller 970 (US); Lacey Springs, June 12, 1927, Woodson 1655 (MBG); Dumfries, June 13, 1927, Woodson 1679 (MBG); roadside near Charlottesville, June 22, 1927, Woodson 1746 (MBG); Buckroe, sandy meadow near sea- shore. May 18, 1912, Robinson 862 (G); Glen Carlyn, June 25, 1905, House 1060 (NY) ; Norfolk, May 25, 1922, Meredith (ANSP) ; Great Falls, dry field, Sept. 9, 1909, Bartlett 1768 (P, D). North Carolina: moist grounds, [near Biltmore?], July 12, 1897, Biltmore Herb. 79b (MBG, US, NY). South Carolina: Anderson, suburbs, dani]) banks, Aug. 15, 1919, Dams 4984 (MBG); Anderson, dry soil, Aug. 15, 1919, Davis 8426 (MBG). Georgia: along railway, McDufhe Co., vicinity of Thomson, Aug. 5, 1909, Bartlett 1690 (P). Florida: Jefferson Co., June-July, 1898, Hitchcock (MBG). Mississippi: Ackerman, June, 1905, Jensen 7 (MBG); Dlo, edge of pine forest, April 19, 1927, Woodson <& Anderson 1551 (MBG). Louisiana: exact locality and date lacking, Hale (MBG). Ohio: New Westerville, roadsides, June 10, 1927, Woodson 1701 (MBG) ; Pittsfield, Lorain Co., date lacking, Ricksecker (US). West Virginia: near Wheeling, June 10, 1927, Woodson 1676 (MBG); field, six miles north of Martinsburg, June 12, 1927, Woodson 1682 (MBG) ; Harper's Ferry, banks of the Shenandoah, Oct. 21, 1913, Greene (ND). Michigan: edge of swampy ground on Belle Isle, July 16, 1892, Farwell (US) ; Flint, date lacking, Canby 3422 (US) . Indiana: Wells Co., Harrison Tp., on dry hills along Erie right-of-way, June 30, 1905, Deam 11 (MBG, US); Steckland, 1889, Evermann (US); roadside along woods on the east side of Pitcher's Pond, about 5 mi. southwest of Mt. Vernon, Posey Co., June 14, 1918, Deam 25405 (D) ; in woods, north boundary. State [Vol. 17 122 ANNALS OF THE MISSOURI BOTANICAL GARDEN Reservation, Clark Co., June 30, 1910, Beam 6901 (D); bank of the White River, near Blue Bluffs, Morgan Co., July 13, 1907, Deam 2276 (D); roadside bank 1 mi. east of Willow Valley, Martm Co., July 13, 1915, Deam 17217 (D); in alluvial soil in the open woods in Jackson Park, about 13^ mi. west of Centerville, Wayne Co., July 3, 1913, Deam 13562 (D); in a deep black and white oak woods 4 mi. southeast of Corydon, Harrison Co., June 27, 1916, Deam 20521^ (D). Kentucky : Lexington, date lacking, Griswold 1 (US) . Illinois: Carbonate, 1878, French (US); dry prairie, Wabash Co., July 2, 1902, Schenck (US). Minnesota: roadside, near Forest Lake, June 28, 1928, Anderson & Woodson 5040 (MBG). Iowa: exact locality lacking, Aug. 9, 1872, Arthur (MBG, US). Missouri: Allenton, June 30, 1912, Letterman (MBG) ; Jerome, June 6, 1914, Kellogg 470 (MBG) ; Jefferson Barracks, St. Louis Co., June 17, 1890, Hitchcock (MBG); rocky ledges, high hills. Galena, Stone Co., May 28, 1914, E. J. Palmer 5792 (MBG); near summit, Pilot Knob, Iron Co., May 21, 1927, Woodson 2703 (MBG) ; barrens, Wild Horse Creek Valley, St. Louis Co., June 16, 1918, Hoffmann (MBG). Texas: in groups, margins of water-courses, Pierdenales and Three Creeks, June, 1847, Lindheimer 659 (MBG); rocky creek- beds, Lindendale, Kendall Co., May 26, 1916, E. J. Palmer 9893 (MBG); Purdy Creek, Gillespie Co., date lacking, Jermy 196 (MBG). New Mexico: High Rolls, Otero Co., May 31, 1902, Viereck (ANSP). 6a. Apocynum cannabinum L. var. pubescens (Mitchell) A. DC. in DC. Prodr. 8: 440. 1844; B6g. & Bel. R. Accad, Lincei, V. 9:103. 1913. cynum cannabinum Michx. Fl. Bor.-Am. 1: 121. 1803, non L. Apocynum pubescens Mitchell ex R. Br. Mem. Wem. Soc. 1: 63. 1809. "Apocynum pubescens R. Br." ex Ell. Sketch Bot. S. C. & Ga. 1: 315. 1821; Torr. Fl. N. & Mid. States. 276. 1824: G. Don. 1930] WOODSON — STUDIES IN APOCYNACEAE. I 123 Hist. Dichlam. PL 4: 81. 1838; Darby, Bot. South. States, ed. 2. 434. 1860; Britt. & Brown, 111. Fl. 3: 23. 1913; Millsp. Fl. W. Va. 331. 1913; Hitchc. & Standi. Contr. U. S. Nat. Herb. 21: 230. 1919. " Apocynum cannahinum L. var. puhescens (R. Br.) DC." ex A. Gray, Man. Bot. ed. 1, 364. 1848; Robinson & Femald in A. Gray, New Man. Bot. ed. 7, 662. 1908; Stone, PI. South. N. J. 646. 1911 ; Wiegand & Eames, Fl. Cayuga, etc. 646. 1926. Apocynum palustre Greene, Leafl. Bot. Obs. & Crit. 1 : 58. 1904. Apocynum Behbianum Greene, I. c. 2: 168. 1912. Apocynum didyotum Greene, I. c. 169. 1912. Apocynum cinereum Nieuwl. Am. Midi. Nat. 3: 56. 1913. Apocynum tomentellum Nieuwl. I. c. 55. 1913, non Greene. Apocynum tomentulosum Nieuwl. I. c. 166. 1913. Apocynum cannahinum L. var. puhescens (R. Br.) A. DC. forma pennsylvanicum B6g. & Bel. Atti R. Accad. Lincei, V. 9: 103. 1913, nomen, Apocynum cannahinum L. var. palustre (Greene) B^g. & Bel. I. c. 1913. Stems tomentulose or glabrate; leaves tomentulose or pubes- cent upon either surface; inflorescence tomentose; in all other essential characters similar to the species. Distribution: generally throughout the southeastern and cen- tral United States, also in north-central California; a frequent ballast weed. One collection has also been made in extreme southern Ontario. Specimens examined: Canada : Ontario: Leamington, July 20, 1906, Fisher (D). United States: Massachusetts: Centreville, sandy copse, Scudder's Bay, July 5, 1896, Williams (G). Rhode Island : Block Island, dry roadside banks and thickets, northeast of Great Salt Pond, Aug. 20, 1913, Femald, Hunnewell & Long 10241 (NE). Connecticut: Southington, July 15, 1902, Bissell (NE); East Lyme, New London Co., June 26, 1903, Graves (NE); Fairfield, Sept. 26, 1895, Eames (NE) ; Southington, July 17, 1903, Andrews (ND). 124 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 New York: ''western New York," exact locality and date lacking, Torrey (G, NY); moist field near Larch Meadows, Ithaca, Aug. 23, 1916, Munz 601 (P); Junius, marly moor of Lowery's Pond, Aug. 4, 1919, Fames & Wiegand 12761 (G); Ithaca, gravelly roadside, June 28, 1919, Wiegand 12760 (G). New Jersey: along railroad, Lakewood, Ocean Co., May 30, 1908, Mackenzie 3093 (MBG, US); Mine Brook, July, 1891, Perry (MBG); Glassboro, July 4, 1865, Redficld 4871 (MBG). Pennsylvania: McCall's Feriy, July 2-6, 1904, Britton (NY); Brownsville, June 10, 1927, Woodson 1656 (IVIBG); open places in the forest at Penryn, in old red sandstone, Aug. 7, 1920, Heller 14239 {UV>0). Delaware: Wilmington, in woods, July 1, 1923, Tidestrom 11516 (G, US); dry field near Georgetown, Sussex Co., July 5-6, 1908, Britton 48 (NY, MBG). Maryland: Silver Spring, Montgomery Co., June 28, 1899, Miller (US, P); Ocean City, July 12, 1909, Carter (NY); Kensing- ton, June 11, 1905, House 1007 (NY). Dlstrict of Columbia: B. & 0. railroad embankment, July 17, 1893, Bocttcher I4I (G, MBG, US, NY); Woodridge, July, 1913, Greene (MBG, US). Virginia: Suffolk, June, 1893, Heller 464 (US); Chain Bridge, Fairfax Co., July 12, 1899, W. Palmer 7 (G); Falls Church, July 30, 1874, Carter (ANSP) ; near Maurertown, June 12, 1927, Wood- son 1672 (MBG); Arlington Junction, Aug. 3, 1912, Greene (MBG, US); ten miles north of Amelia in moist soil along the highway in a wooded area, July 9, 1926, Heller 13998 (MBG); exact locality lacking, 1865, Glatfelter (MBG). North Carolina: coastal plain at White Lake, in sand, July 16, 1926, Bladen Co., Heller I4IO3 (MBG); sand banks near Beaufort, 1906, Lewis 220 (NY); Mt. Airy, June 19, 1909, Rushy (NY). South Carolina: paper mill, Hartsville, May 22, 1909, Coker (NY) ; Santee Canal, May, year lacking, Ravenel (G) ; Abbeville, July, 1855, Hexamer & Maier (G). Georgia: edge of pine barren, vicinity of Thomson, McDuffie Co., July 26, 1909, Bartlett 1661 (P). Alabama: Clanton, Apr. 21, 1927, Woodson & Anderson 1608 1930] WOODSON — STUDIES IN APOCYNACEAE. I 125 (MBG); Valley Head, July 18, 1898, Rvth 463 (NY); Auburn, Apr. 25, 1898, Eark 16U (NY). Mississippi: Ridgway, Apr. 18, 1927, Woodson & Anderson 1641 (MBG); Jackson, Apr. 18, 1927, Woodson & Anderson 1549 (MBG). Ohio: exact locality and date lacking, Throop Herb. (P); Rocky River, near Cleveland, July 8, 1896, Greenman 1438 (MBG) ; near Cincinnati, June 20, 1888, Lloyd 296 (MBG). Indiana: Bluffton, Sept. 1, 1904, Beam & Gleason (G); Notre Dame, along railroad, Oct. 8, 1913, Nieuwland 11603 (MBG); Notre Dame, near ice house, June 23, 1913, Nieuwland 11316 (MBG); in wooded lot across railway from Notre Dame Uni- versity, June 29, 1928, Nieuwland, Slavin & Woodson 3 (MBG); on rocky exposed hillside between Madison and North Madison, Jefferson Co., May 28, 1911, Beam 8489 (D); in dry channel of creek near Langdon, Jackson Co., June 30, 1913, Beam 13517 (D); along east side of Winona Lake, Kosciusko Co., July 28, 1904, Beam (D) ; on the bank of the Ohio River at Vevay, Switzer- land Co., July 25, 1913, Beam 13806 (D) ; low place near a wagon road in woods on the "knobs," 9 mi. north of Rockport, Spencer Co., June 9, 1918, Beam 25230 (D) ; roadside about 5 mi. west of Danville, Hendricks Co., July 13, 1913, Beam 13673 (D); sandy prairie roadside 1 mi. southwest of Bee Hunter, Greene Co., July 3, 1918, Beam 25647 (D); dry open woods along Little Indian Creek, about 5 mi. northwest of New Albany, June 21, 1916, Beam 20346 (D). Kentucky: along roadsides, Lexington, July 20, 1923, McFar- land 131 (MBG); Iron Hill, Lyon Co., June 18, 1909, Eggleston 4805 (NY). Tennessee: Knoxville, June, 1897, Ruth 3088 (NY); Obion Co., fields, in thin soil, June 28, 1893, Bain 160 (G) ; Franklin Co., woods near Anderson, June 7, 1897, Eggert (MBG); open banks, low ground, near Memphis, Shelby Co., ^May 18, 1920, E. J. Palmer 17528 (MBG). Wisconsin: Racine Co., July, 1882, Hasse (NY) ; Preble, Brown Co., July 31, 1886, Schuette (NY). Illinois: Winnebago Co., Fountaindale, date lacking, Bebb (ND, US); rock road near Fish Lake, July 16, 1898, Norton [Vol. 17 126 ANNALS OF THE MISSOURI BOTANICAL GARDEN (MBG); slough, near Columbia, Sept. 1, 1927, Woodson 1926 (MBG); East St. Louis, June 19, 1879, Eggert (MBG); prairie, near Wady Petra, June 19, 1900, V. H. Chase 631 (MBG, US); Odin, June 12, 1898, Greene (US). Minnesota: cultivated fields north of Faribault, June 26, 1928, Anderson & Woodson 5013 (MBG). Iowa: Polk Co., July, 1847, C. L. Parry (NY). Missouri: Courtney, dry ground, Aug. 2, 1906, Bush 4083 (G, US, MBG); Swope Park, barrens, July 21, 1915, Bush 7053 (G, MBG, US, NY) ; Webb City, prairies, June 23, 1 91 1, E. /. Palmer 34^6 (G, MBG); Mansfield, frequent, open woods, etc., alt. 1480 ft., June 5-12, 1911, Lan^ng 3171 (G, F); Sibley, dry ground, June 27, 1909, Bush 5849 (G, IMBG, NY); Courtney, common everywhere, July 17, 1899, Bush 314 (G, MBG, NY, US); Jackson Co., July 16, 1893, Bush 243 (G, MBG) ; AUenton, July 30, 1901, Letterman (MBG); near Clifif Cave, St. Louis Co., July 29, 1886, Eggert (MBG); Moscow Mills, July 23, 1927, Woodson 1719 (MBG) ; Forest Mill, Jasper Co., June 23, 1909, E, J. Palmer 2336 (MBG); near Webster Groves, railroad embankment, July 25, 1927, Woodson 1729 (MBG); along Mo. Pac. Ry. 1. mi. s. of station, Webb City, June 23, 1911, E. J. Palmer 3426 (MBG); dry open slope beside Ely Street, Hannibal, Marion Co., Aug. 21, 1914, Davis 2947 (MBG); south of Oakwood, Ralls Co., July 16, 1915, Davis 4762 (MBG); Turkey Creek, Jasper Co., July 10, 1897, Trelease 709 (MBG) ; Shut-in, Arcadia, Iron Co., May 8, 1925, Woodson 384 (MBG); JefTerson Barracks, St. Louis Co., June 17, 1890, Hitchcock (MBG); upland prairies near Asbury, Jasper Co., June 23, 1928, E. J. Palmer 34676 (MBG). Arkansas: low open ground, Corning, Clay Co., June 25, 1914, E. J. Palmer 6092 (MBG). North Dakota: Valley City, Aug. 6, 1917, Mabhott 257 (NY). Kansas: Riley Co., open ground, 1895, Norton 324 (G, US, NY, MBG). Oklahoma: Copan, Washington Co., in thickety field, Aug. 15, Stevens 2105 (G); Sapulpa, common, July 30, 1894, Bush (MBG) Co., June 2, 1913, Ruth 387 (NY) 1930) WOODSON — STUDIES IN APOCYNACEAE. I 127 California: Little Chico, June 1897, Bruce 2093 (P); sloughs, Butte Co., May, 1898, Bruce (P); colonies in the marshes on drier spots, Suisun, Solano Co., June 6, 1903, Baker 321^7 (US, ND, MBG, P, G, NY). Apocynum cannabinum L. var. pubescens is probably the most variable variety of the genus, and several foliage variations might be considered worthy of description in the absence of intergra- dations. Especially noteworthy in this instance is a race of obovate-leaved plants common in southern Missouri. 6b. Apocynum cannabinum L. var. glaberrimum A. DC. in DC. Prodr. 8: 439. 1844; Rydb. Mem. N. Y. Bot. Gard, 1 : 311. 1900. Apocynum canadense Shecut, Fl. Carol. 1 : 187. 1806. Apocynum piscatorium Dough ex A. DC. L c. 1844, nomen in synon. Apocynum alburn Greene, Pittonia 3: 230. 1897; G. S. Miller, Proc.BioLSoc. Wash. 13:88. 1899; Millsp. FIW. Va. 331. 1913. Apocynum nemorale G. S. Miller, Proc. Biol. Soc. Wash. 13: 87. 1899; B6g. & Bel. R. Accad. Lincei, V. 9: 110. 1913. Apocynum cannabinum L. var. nemorale (G. S. Miller) Fernald, Rhodora 10: 55. 1908. Apocynum missouriense Greene, Leafl. Bot. Obs. & Crit. 2: 165. 1912; Beg. & Bel. Atti R. Accad. Lincei, V. 9: 109. 1913. Apocynum isophyllum Greene, L c. 166. 1912. Apocynum littorale Greene, L c. 171. 1912. Apocynum arenarium Greene, I. c. 173. 1912; B^g. & Bel. /. c. 109. 1913. Apocynum Bolandri Greene, Z. c. 175. 1912. Apocynum Carolini Nieuwl. Am. Midi. Nat. 3: 53. 1913, Apocynum pseudolaurinum Beg. & Bel. var. latifolium Beg. & Bel. L c. 100, 1913, nomen in synon. Apocynum cannabinum L. var. album (Greene) B^g. & Bel. /. c. 107, 1913. Apocynum cannabinum L. var. isophyllum (Greene) Beg. & Bel. I c. 1913. Apocynum cannabinum L, var. Bolandri (Greene) B^g. & Bel. I c. 108. 1913. [Vol. 17 128 ANNALS OF THE MISSOURI BOTANICAL GARDEN Apocynuni cannabijium L. var, jiorihundum B^g. & Bel. I. c. 1913. Apocynum nemorale G. S. Miller var. glabnim B^g. & Bel. I. c. 109. 1913. Entire plant glabrous throughout; in all other essential char- acters similar to the species. Distribution: probably common in every state in the United States, and sparingly in Canada; a common roadside and field weed. Specimens examined : Canada : Quebec: St. Lambert, pres du pont Victoria, vicinity de Lon- gueuil, Sept., 1919, Victorin 965S (G). Ontario: Point Abino, WellandCo., Aug. 23, 1886, Couille (US). Alberta: Calgary, July 16, 1913, Hoodie 116 (NY). United States: New Hampshire: Alstead, by Cold River, Aug. 2, 1900, Fernald 375 (G) . Vermont: Salisbury, among thickets on the shore of Lake Dunmore, July 14, 1908, Williams (NE); wet banks, Queechee Gulf, July 4, 1910, Britton (NY); Leicester, damp ground, July 26, 1924, Button (MBG, G). Massachusetts: Sharon, June, 1905, Poole 86 (G); Hunting- ton, rocky banks of small run, alt. 350 m., Aug. 17, 1912, Robin- son 563 (G); along brook in meadow, Granville, June 25, 1914, Seymour 200 (MBG) ; stony shore, Stockbridge Brook, Berkshire Co., July 31, 1917, Hoffman (MBG). Connecticut: Wethersfield, Gray (G). New York: Ithaca, Triphammer Falls, gravelly shore, Fall Creek, July 17, 1913, E, J, Palmer 965 (G) ; Savannah, Wayne Co., Crusoe Prairie, July 10, 1918, Wright & Griscom 10620 (G) ; bed of Chenango River, June 25, 1882, Lucy 5389 (MBG); marsh, Ithaca, July, 1878, Trelease (MBG) ; Fall Creek, June 26, 1889, Norris (MBG); Cayuga Flats, near Ithaca, July 15, 1893, Schrenk (MBG); East Greenwich, Long Island, 1867, Fitch (US); Chenango Forks, July, 1897, Maxon (US). New Jersey: open woods. Cranberry Lake, Sussex Co., June 24, 1906, Mackenzie 2125 (MBG, NY, D). 1930) WOODSON — STUDIES IN APOCYNACEAE. I 129 Pennsylvania: Inglesnook, Dauphin Co., July 16, 1912, Wil- liamson (ANSP); rocks above Zimmemian's Ferry, Monroe Co., June 23, 1918, Bartram (G, ANSP) ; Susquehanna, July, 1889, Eby (MBG); Conewango Mts., Sept. 7, 1898, Eisenhower (MBG); gravel, Easton, July 13-Nov. 28, 1895, Porter (US). Mahyiand: Great Falls, open ground along north side of canal, June 21, 1917, Hitchcock 12909 (G, US); on sand-bar, Plum Point, Aug. 5, 1902, Shull 173 (US, G, MBG, NY); on bar north of Swan Creek, July 26, 1902, Shull 109 (MBG, US). District of Columbia: at Chain Bridge, original station [of A. album], by canal, June 23, 1912, Greene (US, MBG, NY); exact locality and date lacking, Canby (US); flats near Chain Bridge, June 20, 1899, Lyon (US, P). Virginia: Quinnunond, Aug. 25, 1899, Pollard & Maxon (US); Blue Ridge Mts. east of Waynesville, June 12, 1927, Woodson 1698 (MBG). Georgia: Wrightsboro, date lacking, Chapman (MBG); pine barrens, 1907, Bartlett 1124 (P); sandy field, MacDuffie Co., Sept. 18, 1908, Bartlett 1464 (P). Florida: Mary Esther, Apr. 28, 1908, Tracy 9449 (MBG, G, NY); old fields. Lake City, July 14, 1893, Quaintance (MBG); Aspalaga, 1898, Chapman (MBG); Apalachicola, date lacking. Chapman (NY). Alabama: Jackson, Apr. 20, 1927, Woodsoii & Anderson 1548 (MBG); along A. G. S. Railroad, Valley Head, July 18, 1898, Ruth 481 (MBG, NY). Louisiana: exact locality and date lacking, Hale (NY, G); borders of fields, near St. Martinsville, June 26, 1893, Langlois (ND). Ohio: Garrettsville, July 25, 1901, Webb (G); Hannony, June 9, 1927, Woodson 1650 (MBG). West Virginia: pebbly shore of the Blackwater River, Hen- dricks, Tucker Co., Sept. 10, 1904, Moore 2097 (G); Berea, dry soil, waste field, Aug. 21, 1922, Randolph & Randolph 1356 (G). Indiana: on border of swamp, Blackford Co., June 25, 1905, Deam (US); Roby, bog, Sept. 1, 1907, Lansing 2679 (G, F); Spencer, moist sandy roadside, July 18, 1915, Deam 17555 (G, MBG, D); wooded bank of Claypole Pond, Knox Co., July 8, [Vol. 17 130 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1915, Deam 17004 (D) ; on the east bank of the north fork of the Muscatatack River, about 1 mi, north of Vernon, growing in rather open and rocky woods, July 9, 1911, Deam 9135 (D); on a rocky ripple in Laugherty Creek back of Versailles, July 23, 1919, Deavi 13766 (D); on a stony bar of White Water River ^ mi. northwest of Metamora. This bar overflows much of the wet season and plants are mostly decumbent, Sept. 12, 1924, Deam 41020 (D); Dubois Co., on the border of the artificial lake about 1 mi. north of Jasper, July 2, 1912, Deam 11566 (D). Kentucky: Lexington, June, 1836, Peter (NY); North Bend, date lacking, Short (NY); Southern Hills, three miles south of Louisville, Aug. 15, 1892, Bergmann (MBG); Bowling Green, June 19, 1897, Pnce (MBG). Tennessee: Robertson Co., woods near Greenbrier, July 13, 1897, ^-^f/er^ (MBG). Wisconsin : near Portage, June 22, 1927, Woodson 1660 (MBG) ; Milwaukee, Aug. 3, 1884, Hasse (NY). Illinois: St. Clair Co., French Village, Sept. 6, 1892, Eggeii (MBG); Illinois State Park, Starved Rock, LaSalle Co., June- Sept., 1921, Thone 33 (MBG); Olney, Richland Co., July 17, 1927, Ridgway 2832 (MBG) ; near Falling Springs, St. Clair Co., Sept. 30, 1928, Greenman 4594 (MBG); near 119th Street, West Pulhnan, Chicago, Aug. 8, 1907, Greenman 1869 (MBG, F). Minnesota: Hennepen Co., Aug. 1889, Sandberg (ND). Iowa: fields and wastes, Decatur Co., Aug. 22, 1904, /. P. An- derson (MBG); exact locality lacking, July 20, 1875, Arthur 23 (MBG). Missouri: Webb City, rocky woods, Aug. 6, 1920, Bush 9054 (NY, US, MBG); CartervUle, July 13, 1920, E, J, Palmer 18311 (NY, MBG) ; Cedar Gap, hillsides, alt. 1675 ft., May 22-June 2, 1911, Lansing 2992 (G, F); Jackson Co., Aug. 8, 1883, Bush (MBG) ; Reeds, Jasper Co., June 28, 1913, E. J. Palmer 4OO6 (MBG); St. Louis, July 3, 1895, Glatfelter 338 (MBG); Creve Coeur Lake, July 4, 1906, Johnson (MBG) ; Silex, June 24, 1912, Davis 1326 (MBG); St. Louis Co., in open woods, rather damp places, June, 1833, Engelmann (MBG); Joplin, July 4, 1902, common in woods, E. J. Palmer 208 (MBG); Sarcoxie, rocky woods, Sept. 18, 1910, E. J. Palmer 3188 (MBG) ; Mine La Motte, 1930] WOODSON — STUDIES IN APOCYNACEAE. I 131 Madison Co., June 22, 1897, Monell (MBG) ; Sheffield, common in fields, July 10, 1899, Bush 169 (MBG); bottoms, Cass Co., June 14, 1865, Broadhead (MBG) ; common in open grounds. Clay Co., Randolph, July 17, 1898, Mackenzie 253 (MBG) ; Allenton, June 8, 1896, Kellogg (MBG); near Gihnore, July 18, 1927, Woodson 1747 (MBG); Frisco R. R. embankment, Webster Groves, July 22, 1927, Woodson 171^5 (MBG); Cliff Cave, St. Louis Co., May 9, 1879, Eggeri (MBG); uncultivated field, near New London, Aug. 24, 1927, Woodson 171^8 (MBG); roadside south of Festus, May 31, 1926, Woodson 681 (MBG). Arkansas: Benton Co., date lacking, Plank (MBG); Benton, Saline Co., dry open ground, June 24, 1915, E. J, Palmer 8133 (MBG). North Dakota: Leeds, Aug. 21, 1907, Lunell (NY). Nebraska: canyons, prairie & ruderal, Mitchell Co., Aug., 1899, Hedgcock (MBG); Republican Valley, alt. 2000 ft.. May 21, 1894, Layhoume 86 (MBG); Republican River, June 27, 1856, Lieut. Bryant's Expedition, H. Engelmann (MBG); Belmont, July 25, 1889, Wehher (MBG); Wahoo, June, 1890, Rydherg (NY). Kansas: Riley Co., wet places, Sept. 14, 1895, Norton 324 (G, MBG, NY). Oklahoma: vicinity of Fort Sill, June 17, 1916, Clemens 11728 (MBG); Sapulpa, July 22, 1894, Bush 359 (MBG); low grassy place, near Cleo, Major Co., June 8, 1913, Stevens 798 (G, MBG). Texas: moist river banks, Mertzon, Irion Co., July 4, 1917, E. J. Palmer 12^20 (MBG); Sherman, wet banks, June 23, 1872, E. Hall 614 (G); locality lacking, 1835, Drummond 232 (G); Fredericksburg, June, 1847, Lindheimer 658 (G, MBG); Camp 36, Canadian, near sandhills, June, 1853, Bigelow 691 (G); San Marcos, 1897, Stanjield (NY); Dallas, low rich lands, 1874, Reverchon 599 (MBG); Coombs Ranch, Aug. 24, date lacking, Reverchon (MBG). Montana: Box Elder Creek, Valley Co., July 14, 1900, Blank- inship (MBG). Wyoming: Hot Springs Bar, 20 mi. south of Jackson, July 19 1901, MerHll & Wilcox 1042 (G). Colorado: Montrose, July 17, 1897, Shear 4808 (NY); Canyon City, 1872, T. S. Brandegee 472 (MBG) ; Evans, 1909. E. L. John- [Vou 17 132 ANNALS OF THE MISSOURI BOTANICAL GARDEN son 295 (MBG) ; Clear Creek Canyon, alt. 7000 ft,, July 18, 1896, Crandall (MBG). New Mexico: Pecos, San Miguel Co., alt. 6700 ft., Aug. 18, 1908, Standky 60U (G, US) ; exact locality lacking, 1847, Fendler 692 (MBG) . Idaho: Priest Lake, Aug. 1901, Piper 3702 (US, Epling Herb.) ; Clarke's Fork Valley, below Parma Mt., alt. 680 m., Aug. 21, 1895, Leih^g 1557 (G, US, MBG); Twin Falls and Shoshone Falls, steep moist slope, alt. 3700 ft., July 26, 1911, A. Nelson & Machride 1361 (G, MBG, NY, F, W); open plains, Boise, alt. 2880 ft., July 13, 1911, Clark 113 (G, MBG, NY, F). Utah: Peterson Canyon, alt. 8000-10000 ft., July 19, 1902, Pammel & Blackwood 4009 (G); Little Falls, July-Oct. 1879, Ward (US); St. George, alt. 2700 ft., Sept. 29, 1894, M. E. Jones 6099 (US, MBG, P); Provo, gravel, alt. 4500 ft., June 25, 1894, M. E. Jones 5490 (MBG, P). Nevada: St. Thomas, May 25, 1919, Tidestrom 9157 (US). Arizona: White Mountains, Aug. 5-15, 1903, Gri£[iths 6397 (US, MBG); Huachuca Mts., alt. 6000 ft., Sept. 3, 1903, M. E. Jo7ies (P); Fort Whipple, river bottoms, Rio Verde, Sept. 9, 1865, Coues tended by attcn- Petioles subtended by at- Petioles unappendaged uate gland-hke append- or ages. tenuate gland-like ap- subtended by a pair of pendages. punctate glands. summarized onsidered by in the tribe Apocynaceae has i^robably progressed from a primitive condition represented by Ajmcynum, through a condition repre- Trachomitum, to the relatively advanced Poacynum. This attitude is taken from a consideration of the inflorescence, the corolline appendages, the nectaries, the receptacle, and the petiolar appendages, as indicated on pages 162-1G3. The geographical distribution of the three sented at condition would Apocynum further substantiate such a viewpoint, since the older territory than the genera Trachomitum aznd Poacynum, which are almost entirely restricted to the Tertiary sea troughs of southern Europe and south-central Asia. ABBREVIATIONS the preceding taxonomic treatment noted herbaria in which specimens have been examined MBG G NY US F ANSP Herbarium of the Missouri Botanical Garden. Gray Herbarium of Harvard University. Herbarium of the New York Botanical Garden. United States National Herbarium. Herbarium of the Field Museum of Natural History. Herbarium of the Academy of Natural Sciences of Philadelphia. (Vol. 17 164 ANNALS OF THE MISSOURI BOTANICAL GARDEN LIST OF EXSICCATAE numbers Unnumbered indicated by a dash. The numbers in parentheses are the species numbers used in this monograph. Aitchison, J. E. T. 115 (la). Ascherson, P (1). Ball, J. (1). Bernhardi Herb. Boagard, G. H. Burkhardt, — . - Burmeister, David, A. (D- - (la). (1). (1). (1). Engelhardt, V. (1). Fiori, A. & B^guinot, A. 1109 (1). Fischer, —. 8^ (1). Gandoger, M. 1661 (1). Grout, Dr. — . & Hemsdall, Rev. (1). Herb. Acad. Petrop. Karelin, — . & Kiriloff, (1). (1); (2). Hohenacker, R. F. — , 397 (1). Kammorer, P. (!)• Kettner, Lagger, (1). (1). Ling, K. 5^4 {\y,9582 (lb). Martens, F. W. — (1). Nervessky, — . — (1). Pease, A. S. 9213 (1). Penzig, 0. 1169 (1). Perry, W. (1). Pichler, — . 188 (1). Prihler, Th. Regel, A. — Schrenk, Timsfcgen, Vaccari, A. Wiedemani (1). (1). (1). (1) (1). (1). IV. A Revision of the Genus Poacynum The genus num was segregated from Apocynum by Bailloni" jn 1888, who based it upon Apocynum pidum Schrenk The only distinctive characters which were recorded by the alternate leaves and pelviform corolla, hence the has been widely misunderstood Baillonio2 remained ed of the validity of his genus Poacynum in subsequent works, as was also Schumann, ^o' who organized the Apocynaceae for Engler & Prantl's ^Die natiir- hchen Pflanzenfamilien.' However, Danguy,"" and B^guinot and Beloserskyi"^ are of a different opinion, and consider Poacynum as representing merely an element of the Linnaean genus Apocynum. Danguy, while considering Poacynum as congeneric with Apocy- num, recognized that the former genus differs from the latter by "iRaill. Bull, Soc. Linn. Paris 1: 757. 1888. "^ Ibid. Hist. PI. 10 : 208. 1891. »<»K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4»: 179. 1895. •'^P. Danguy in Lecomte, Not. Syst. 2: 135. 1911. lo* B4g. & Bel. Atti R. Accad. Lincei, V. 9: 77. 1913. 19301 WOODSON — STUDIES IN APOCYNACEAE. I 165 such important characters as the alternate leaves, the pelviform corolla, and an annular "disc," thus representing a most unusual element in the latter genus. B^guinot and Belosersky, however, reduced Poacynum to a section, Picti, of Apocynum, with the following diagnosis i^''^ " . . sparse foliage, gradually attenuate to a point, coriaceous, of a glaucous greenish color, with nerves more or less visible, . . pelviform corolla, and . . pendulous flowers." Several morphological reasons (cf. pp. 162-163) have already been given in a previous section of this study in an attempt to iustifv the retention of Poacynum as a genus of the subfamily ynoideae knowledge of the genus Poacynum, which extremely rare in continental herbaria, and which was entirely absent from American collections until the present studies were undertaken, the author is indebted to M. P. Danguy of the Herbarium of the Museum of Paris for fragments of the type specimen of P. grandifiorum Danguy, to Mr. Alfred Rehder of the Arnold Arboretum of Harvard University for a fragment of the type specimen of Apocynum Hendersonii and a photograph of Poacynum pictum (Schrenk) Baill. which he obtained from the Jardin Principal Botanique de Petrograd through the kindness of Prof. Boris Fedtschenko, and to Dr. Carl C. Epling for a fragment of an authentic specimen of Apocynum Hendersonii. Poacynum Baill. Bull. Soc. Linn. Paris 1: 757. 1888; Hist. PI. 11: 208. 1891; K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 41:179. 1895. Apocynum § Picti B^g. & Bel. Atti R. Accad. Lincei, V. 9: 77. 1913. Herbaceous lactescent perennials from a somewhat thickened fibrous rhizome. Stems erect or ascending, terete, striate, gla- brous; branches ascending, alternate to subalternate ; leaves i«B^g. & Bel. Atti R. Accad. Lincei, V. 9: 77. 1913. "Compreude [Stirps A. Picti] due specie confinate nell' Asia centrale: e, cioe, A. pictum della Songaria, ed A. Hendersonii, pure di questo paese, non che del Turkestan cinese e del Kan-sou mongolico. Ambedue convengono per alcuni caratteri di stlrpe, quali le foglie sparse, insensibilmente attenuate in picciuolo, eoriacee, di un colore verde-glauco, con nervi poco o punto visibili, per la corolla piu grande, pelviforrae e per i fiori penduli." 166 ANNALS OF THE MISSOURI BOTANICAL GAKDEN [Vol. 17 alternate, petiolate, oblong to lanceolate, membranaceous, re- motely denticulate along the margin, and frequently also along the midrib of the lower surface, mucronate, or without a mucro, penninerved; petioles subtended by two punctate glands or eglandular. Inflorescence a terminal monochasium, bearing paired pedicels. Calyx regularly 5-parted, hispidulous, the tube adnate to the rece])tacle, eglandular within. Corolla pelviform, regularly 5-parted, conspicuously granulo-papillate without, bearing within 5 elongate-apiculate appendages coalesced into a shallow ring adnate to the base of the tube, the apices opposite the lobes of the corolla; limb widely spreading, dextrorsely con- volute in the bud. Stamens 5, alternate with the apices of the appendicular ring and attached to the base of the corolla-tube, anthers aj^pendaged, convergent about the pistil ; filaments short, flattened; pollen borne in tetrads, mostly sterile. Carpels 2, unilocular, sessile, united at the apices by an ovate-cylindrical clavuncle bearing the terminal binate apiculate stigma; ovules many, anatropous, borne upon a ventral, lateral placenta. Re- ceptacle somewhat elevated above the orifice of the calyx-tube, entire; nectaries 5, more or less coalesced into a ring about the gynoecium. Follicles terete, pendulous, 15-30 cm. long, divari- cate or somewhat falcate, containing many dry, apically comose, clavate-terete, exalbuminous seeds; embryo straight. Type species: Poacynum pidum (Schrenk) Baill, Bull. Soc. Linn. Paris 1 : 757. 1888. KEY TO THE SPECIES a. Calyx-lobes ovate; leaves linear 1. Poacynum pictum aa. Calyx-lobes lanceolate; leaves lanceolate to oblong-ovate. b. Leaves cuncate-lanccolate, decidedly mucionate; nectaries evident, about 34 the length of the carpels S. Poacynum gramlijlorui/i bb. Leaves oblong-lanceolate to oblong-ovate, neither cuneate nor mucro- nate; nectaries inevident, about }4 the length of the carpels 2. Poacynum Henderaonii 1. Poacynum pictum (Schrenk) Baill. Bull. Soc, Linn. Paris 1 : 757. 1888; Hist. PL 10 : 208. 1891 ; K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4»: 179. 1895. Apocynum pictum Schrenk, Bull. Acad. St. Petersb. 2: 115. 1844; B^g. & Bel, Atti R. Accad. Lincei, V, 9: 77. 1913, 1930J WOODSON — STUDIES IN APOCYNACEAE. I 167 Stems erect or ascending, 5-10 dm. tall, glabrous, freely branched; branches alternate to subalternate, ascending; leaves petiolate or subsessile, linear or very narrowly lanceolate, acute to acuminate at base and apex, membranaceous, glabrous, den- ticulate, mucronate; inflorescence few-flowered, bracts evident, scarious; corolla pelviform, granulo-papillate, about 1-1.5 cm. broad; calyx-lobes ovate, about 3-5 mm. long, hispidulous; folli- cles unknown. Distribution: Songaria. Specimens examined: Songaria: "in deserto songooro-kirghisico ad m. Tamgaly- Tas," Schrenk (L, type, .MBG photograph). 2. Poacynum grandiflorum P. Danguy in Lecomte, Not. Syst. 2:136. 1911. Apocynum grandiflorum P. Danguy in Lecomte, Z. c. 137. 1911. Stems erect or ascending, 5-10 dm. tall, glabrous, freely branched; branches alternate, ascending; leaves petiolate, cuneate- lanceolate, 3-5 cm. long, 1-1.5 cm. broad, remotely denticulate, decidedly mucronate, membranaceous; inflorescence relatively few-flowered; corolla pelviform, 1.5-2 cm. broad, granulo-papil- late; calyx-lobes lanceolate, about 2 nmi. long, hispidulous; nectaries evident, about two-thirds the length of the carpels; follicles terete, pendulous, 1.5-3 cm. long; seeds clavate, 3-4 mm. long, coma 15-20 mm. long. Distribution : Turkestan. Specimens examined: Turkestan: Cha-yar, prov. of Chouchar, May, 1907, Vaillant 266 (MP TYPE, MBG fragment). 3. Poacynum Hendersonii (Hook, f.) Woodson, n. comb. Apocynum Hendersonii Plook. f. in Henderson & Hume, Lahore to Yarkand, 327. 1878, Beg. & Bel. Atti R. Accad. Lincei, V. 9: 78. 1913. Stems erect or ascending, 5-10 dm. tall, glabrous, freely branched; branches alternate, ascending; leaves petiolate, oblong- lanceolate to oblong-ovate, 3-5 cm. long, 1-2 cm. broad, neither cuneate nor mucronate; inflorescence relatively many-flowered; [Vol. 17 168 ANNALS OF THE MISSOURI BOTANICAL GARDEN corolla pelviform, granulo-papillate, 1.5-2 cm. broad; calyx-lobes lanceolate, about 3 mm. long; nectaries inevident, about one- third the length of the carpels; follicles unknown. Distribution: Turkestan and Songaria. Specimens examined: Turkestan: Kashgar, alt. 4000 ft., date lacking, Creswell (BM, MBG, fragment); Yarkand, sandy desert, alt. 4000 ft., May 20, 1874, Henderson & Hume (L, MBG fragment). BM L MBG MP ABBREVIATIONS Herbarium of the British Museum Herbarium of the Jardin Principal de Leningrad Herbarium of the Missouri Botanical Garden Herbarium of the Museum of Paris V. Index to Species New genera, species, varieties, and corabinations are printed in bold face type; ms Page Anechites Neriura 147 .Apocinum 84 Apocynastrum 84 Apocynum 83 abditu7n 107 Acouci 144 ajricanuvi 144 agglovieratwn 144 album 127 altcrnijlorum 144 arnbigens 87 var. bicolor 87 Andrewsii 107 androsaernifolium Forsk., non L 144, 159 atidrosaemifoliurji Holm, non L. 107 androsaemifolium L 86 f. pauciflora 91 subsp. deionsum 105 var. Andrewsii 107 var. glabrurn 87 var. mcanum var. intermedium Page 91 100 var. microphyllurn 87 var. nevadense 102 var. oblongifolium 108 var. arhiculatum 87 var. griseum 111 var. puherulum 92 var. pumilum 101 var. typicum 87 ^* androsaemifolium L." ace. to Hook 91 angustifolium Sesse & Mocino, non Wooton 145 angustifolium Wooton, non Sesse & Mocifio 119 apiculatum 145 arciiatum 105 127 Austinac 101 Bebbianum 123 bicolor 101 Bolandri 127 BreweH 142 arenanum 19301 WOODSON — STUDIES IN APOCYNACEAE. I 169 Page bursiflorum 145 calophyllum 101 canadense 127 canariense 145 rannabinum L. nan Michx 119 var. album 127 var. Bolandri 127 var. estillinum 133 var. floribundum 128 var. glaberrimum 127 var. Greeneanum 132 var. hypericifolimn 133 var. incanum 120 var. isophyllum 127 var. lividum 113 var, nernorale 127 var. oliganthum 117 var. palitstre 123 var. puhcndu7n 120 var. pubescena (Mitchell) A. DC 122 var. ^^pubescens (R. Br.) A. DC." 123 f. pennsylvanicum 123 var. Suksdorfii 117 cannobinum Michx. non L 122 cardiophyllum ' . . 101 CaroUni 127 cercidium 101 cervinum 141 ciliolatum. 115 cinereum Heller, non Nieiiwl ... 105 cinereuvi Nieuwl., non Heller. . . 123 ciirijolium 145 coconinuni 108 cowpressum 145, 159 cordatnm Mill., non Thunb .... 145 cordaiitm Thunb., non Mill .... 145 cordifolium Sesse & Mocifio, non Tenore 145 cordifoluim Tenore, non Sesse & Mocifio 145 cordigeram 141 coiinifoliuyn 145 crassifolium 145 cwnanense 145 cnspidatum 142 densijlorum 142 denticidatum 115 Page dictyotum 123 divergent 107 diversijolhim 105 ellipticum 107 Elmeri 107 eredum 145 ericifoliwn 105 estillinum 133 exijniwn ....". 101 Farivellii 140 f. verticillare 140 var. glaucum 134 f. anomalum 134 f. iernarium 134 fdiforme 145 fimbriatufn 145 jioribunduvi 113 fioristraturn 146 Joetidum. 146 Jol, androsaemi 87 frudu spinoso 146 frntescens Afzel., non L 146 frutescens L., non Afzel 146 glaucum 113 grandifiorum 146, 167 Greeneanum. 133 griseum HI hasiaiwn 146 Hendersonii 146, 167 hypericifolium 133 f. arcnarium 133 var. angustifolium 119 var. cordigerum 141 var. Farwellii 140 var. iidermedium 101 var. intennediuni 134 var. latifolium 142 var. myrianihum 117 var. nevadense 142 var, oblongum 142 var. pscudo.suksdorfii 118 var. salignum 141 var. typicum 134 imbricahim 146 incaman Greene, non G. S. Miller 116 incanum G. S. Miller, non Greene 91 ind'icum 146 insigne 107 isophyllum. 127 170 ANNALS OF THE MISSOURI BOTANICAL GARDEN iVoL. 17 Page ithacrfi.se 133 Jonesii 117 Jxiventas 146 lanceolatiun 146 laurinwn 117 leuconcuroa 112 lineare 146 littorale 127 lividum 115 var. ahdiium 107 var. Elnieri 107 var. floribundwn 113 var. tcxanum 134 var. typicum 115 var. vacillans 113 var. viarum 113 longijolium 142 luridum 101 Macounii 142 macranthum 87 macrophylluvi 107 maculatxim 146 medium Greene 107 var. floribundum 113 var. leuconeuron 112 var. lividum 115 rar. sarniense Ill var. vestitum II6 '^medium G. S, Miller, not Greene" 107 viexicGnum 146 Milleri 107 var. paucifloTum 108 minuium 146 missourieitse 127 inolle 105 fnucronahnn 147 mtcscipuluni gi imjrianthum II7 riemorale 127 var. glahrum 128 neogeum 133 Nerium 147 nervosum 147 nevadense 141 obliquum 147 occidcnialc 87 odoratissiinum 147 oligaiithufn 1 17 Page orixcnse 147 ovalifolium 101 pahialre 123 paniculatufn- Greene, non Lam., 101 pC— X samiense Kvidum, A^S7£k A , medium VB ^ A. Jonesii, m A, Suksdorfii, Sulzsdorfii var. angmiifoUu cannabinum, cannahinuyn var, vuhcscens •oooo A. cannab A, ainuabinum var. Grceneanu\ TAT4T A. cannabinum var. hypericifol A. cannabinum var. FarweUii, nuny -4. cannabinum var. cordigerum A. cannabinum var. saHanum. Ann. Mo. Bot. Gaud., Vol. 17, 1930 Plate 1 WOODSON— APOCYNACEAE [Vol. 17, 1930 ] 176 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 2 Fig. 1. Longitudinal section of calyx-lobe of single-flowered form of Nerium Oleander: x, place of attachment to the receptacle; c', calyx-lobe proper; sq, calycine squamellae; 6, peripheral glandular tissue; a, central "conductive*' tissue; c, occa- sional vascular traces leading to squamellae, X 20. Fig. 2. Longitudinal section of calyx-lobe and supernumerary petal of double- flowered form of Nerium Oleander: x, place of attachment to the receptacle; sp, supernumerary petal; c\ calyx-lobe proper; sq, squamella. X 20. Fig. 3. Diagram of a median longitudinal section of single-flowered form of Nerium Oleander illustrating the vascular anatomy: c\ calyx; c", corolla; or, ovary; dj clavuncle; sg, squamellae. Fig. 4. Diagram of a cross-section of single-flowered form of Nerium Oleander: c', calyx-lobes; c", corolla; ov, ovary; 35, squamellae; si, staminal traces. Fig. 5. Cross-section of a supernumerary petal of double-flowered form of Nerium Oleander. X 20. Ann. Mu. Bot. Gaud., \"ul. 17, lu:Jl) Pj.ate 2 1 ^^^OODSON^APOCYNACEAE [Vol. 17, 1930] 1/8 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 3 ■ Fig. 1. Diagrammatic longitudinal section of flower of Aponjnum androsaemi- foliuin^ indicating the vascular anatomy: a. Corolla; bj calyx; c, ovary; rf, clavuncle; e, stamen;/, nectary. Fig. 2. Diagrammatic cross-section of flower of Apocynum androsaemijolmm through the region of the gynoecium, indicating the vascular anatomy: a, adnate staminal filaments; 5, dorsal corolline traces, c, ventral corollinc traces; d, ovarj'*; Cy nectaries. Fig. 3. Diagrammatic cross-section of flower of Apocynum androsaeyiiifolium through the region of the androeciumj indicating the vascular anatomy: a, ventral corolline traces; 6, dorsal corolline traces; c, anther; d, clavuncle. Fig. 4. Longitudinal section through the place of attachment of the staminal filament to the base of the corolla-tube of Apocynum androsnemifoUum: a^ staminal filament; 6, corolla; c, cal3''x-lobe; d, ovary-wall. X 150. Fig. 5. Pollen grains of Apocynum medium var. leuconeuroHy illustrating sterility and polyspory. Ann. M(j. Bot. Gaud., Voj-. 17, 1030 1'lA'I'E o )y e c b 4 2 3 WOODSON— Ar( )CYNACEAE [Vol. 17, 1930 180 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 4 Fig. L Cross-section of young root of Apocynwn hypcricijolium: a, epidermis; b, cortex; c, endodermis; d, xylcm; e, phloem;/, pericycle. X 150. Fig. 2. Cross-section of gemmiferous root of Apocyiium cannabinum var. glaber- rimum: a, phcUem; 6, phellogea; c, phelloderm; d^ xylera; c, phloem; /, cortex; g, cambium; 7i, latex tube; k, lenticel; /, storage ray. X 150. Fig. 3. Diagrammatic cross-section of node of stem of Apocynum caimahinm/t, immediately below foliar gap. Fig. 4. Diagrammatic longitudinal section of node of stem of Apocynum carina- hiniim. Fig. 5. Embryo of Apocynum hypcridjolium. X 30. Fig. 6. Tracheid from secondary xylem of stem of Apocynu?7i cannabinum.^ illustrating bordered pits. X 1000. Ann. Mo. Bot. Gard., Vol. 17, 1930 Plate 4 5 WOODSON— APOC^'NACEAE [Vol. 17, 1930] 182 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 5 Fig. 1. Sector of a cross-section of a stem of Apocyniun ccmnabinum: a, epid h, hypodermis; c, cortex; d, external phloem; e, cambium; /, internal phloem; g, xylem; h, pith; {, starch-containing cells of pith; j, perlcyclic fibers associated with latex tubes; m, latex tubes in pith. X 150. Fig. 2. Sector of a cross-section of a stem of Apocynum androsaemifolium: a, dead ei)idermal cells; &, periderm; c, cortex; dj external phloem; e, cambium;/, internftl phloem; g^ xylem; h, pith; j, pericyclic fibers associated with latex tubes; 771, internal latex tubes. X 150. Fig. 3. Sector of a cross-section of a fully mature rhizome of Apocynum cannahi- nrim: a^ dead epidermal cells; 6, periderm; r, cortex; d, external phloem; e, cambium; /, internal phloem; g, xylem; h, pith; j, pericyclic fibers associated with latex tubes; At, vascular raj'. X 150. Fig. 4. Habit of the subterranean axes of Apocynuin cannabinum: a, rlxizome; 6, cataphylls; c, young root-shoot enveloped in cataphylls; d, young gemmiferous root l)earing a young root-shoot; ff, fully mature gemmiferous root bearing several rhizomes; /, strictly absorptive root. X %. Fig. 5. Starch grains from cortical cells of Apocynum cannabinwn. Ann. Mo. Bot, Card., Vol. 17, 19?0 Plate 5 C 1 2 WOODSON— APOCYNACEAE COCKAYNE, BOSTON 184 ANNALS or THE MISSOURI BOTANICAL GARDEN [Vol. 17. 1930] Explanation of Plate PLATE c Fig. 1. ndrosaemi folium var. incanum through the region of the corollar appendages: a, dorsal vascular trace; &, ventral trace; c, latex tubes. X 150. Fig. 2. Longitudinal section of a nectary of Apocijmim andromcmijolium var. incanum^ indicating the vascular anatomy; a, dorsal vascular trace of nectary; 5, sector of corolla-tube; c, sector of ovary wall; d, latex tubes. X 150. Fig. 3. Longitudinal section of the clavunole of Apocijnum androsaemijoliuw , indicating the vascular anatomy: a~h, continuation of the dorsal vascular trace of either carpel of the gynoecium; c, latex tubes; J, glandular cells of the stigmatic surface. X 150. Fig. 4. Cross-section of a leaf of Apocyyium cannabinum var. glabcrrimiim: a, upper epidermis; b, lower epidermis; r, palisade parenchyma; d, spongy parenchyma. X 100. Fig. 5. Cross-section of a leaf of Apocyiium androaaemijolium var. incanum: a, upper epidermis; b^ lower epidermis; r, palisade parenchyma; J, spongy i>arencliyma, X 150. Ann. Mo. But. (!aiu>., Vol. 17, lUoO 1*LATE t) 6 h 4r d -b 5 \\( )ODS( )X— A P( )CYX ACEAE [Vol. 17. 1930] 186 ANNALS OF THE MISSOUIil BOTNAICAL GARDEN Explanation of Plate PLATE 7 A typical clone of Apocynum cannabinum, near Carondclct, Missouri Ann. Mo. Bot. GARn., Vol. 17. 19?0 Plate 7 n o n > w O y: H O 2: o o O O IVoL .17, 1930] 188 ANNALS OF THE MISSOUKI BOTANICAL GARDEN Explanation of Plate PLATE 8 Ajpocynuin medium var. lividam, infesting cultivated field, near Bayfield, Colorado Ann. Mo. Bot. Card.. Vol. 17, 1930 Plate 8 WOODSON— APOCYNACEAE COCKAYNE. BOSTON [Vol. 17. 1930] 190 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 9 Habit of Apocynuvi androsaemifoUum L. var. inamum A. DC. X enlarged flower, X 2. del. J. A. Stoyermark Ann. Mo. Bot. Gard.. Vol. 17, 19?0 Plate 9 WOOD SON — APOC YN ACE AE COCKAVNE, BOSTON % . [Vol 17. 1930] 192 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 10 Habit of Apocynum puiyiUum (A. Gray) Greene var. rhomhoideum (Greene) B6g. & Bel X }4] enlarged flower, X 2. del. J. A. Steyermark. Ann. Mo. Box. Card., Vol. 17, 1930 Plate 10 WOODSON— APOCYNACEAE COCKAYNE, BOSTON 194 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17. 1930] Explanation or Plate PLATE 11 Habit of Apocynum medium Greene, X 3^; enlarged flower, X 2. del. J. A. Steyermark Ann. Mo. Box Card., Vol. 17. 1930 Plate 11 rv^)- WOODSON— APOCYNACEAE COCKAYNE, BOSTON IVOL. 17, 1930 J 196 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 12 medium Greene var. florihundum H enlarged flower, X 3» del. J, A. Steyermark. Ann. Mo. Box. Card., Vol. 17. 1930 Plate 12 WOODSON — APOCYNACEAE COCKAYNE, BOSTON 198 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17, 1930] Explanation of Plate PLATE 13 Habit of Apocynum medium Greene var. vestilum (Greene) Woodson, X M; enlarged flower, X 3. del. J. A. Steyermark. Ann. Mo. Bot. Gard,. Vol. 17, 1950 Plate 1? I WOODSON — APOCYNACEAE COCKAYNE. BOSTON [Vol. 17, 1930) 200 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 14 Habit of Apocymim Jonesii Woodson, X l^] enlarged flower, X 4. del. J. A. Stovermark. y Ann. Mo. Box. Card., Vol. 17, 1950 Plate 14 WOODSON —APOCYNACEAE COCKAYNE, BOSTON [Vol. 17. 1930] 202 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 15 Habit of Apocynum Suksdorfii Greene, X Hi enlarged flower, X 4. del. J. A. Steyermark. Ann. Mo. Bot. Card., Vol. 17, 19?0 Plate 15 WOODSON— APOCYNACEAE COCKAYNE. BOSTON [Vol. 17, 19301 204 ANNALS OF THE MISSOURI BOTANICAL GARDEN - Explanation of Plate PLATE 16 Habit of Apocynum cannabinum L. X M; enlarged flower, X 4. del. J- A. Steyermark Ann. Mo. Box. Card.. Vol. 17, 1930 Plate 16 woo DSON — APOCYNACEAE COCKAYNE, BOSTOl* [Vol. 17, 1930] 206 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 17 Habit of Apocynum cannahinmn L. var. Grteneanum (B^g. & Bel.) Woodson, X }/2\ enlarged flower, X 3. del. J. A. Steyermark. Ann. Mo. Box. Gard., Vol. 17, 1950 Plate 17 WOODSON— APOCYNACEAE COCKAYNE. BOSTON [Vol. 17. 19301 208 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 18 Habit of Apocynum hypericifolium Ait. X Ml enlarged flower, X 3. del. J. A. Steyermark Ann. Mo. Box. Card.. Vol. 17, 1930 Plate 18 f WOODoON - APOCYNACEAE COCKAYNK, BOSTON [Vol. 17, 1930] 210 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 19 Habit of Apocynum hyperidfolium Ait. var. salignum (Greene) Woodson, X H; enlarged flower, X 3. del. J. A. Steyermark. Ann. Mo. Hot. Card.. Vol. 17. 19i0 Plate 19 ^ •^ 'r WOOr SON — APOC YN ACEAE COCKAYNE. BOSTON [Vol. 17. 1930] 212 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 20 Comparative morphology of Apocynum, Trachomitnm ^ and Poacynum: 1^ inflores- cencej £, exterior of flower; 5, interior of corolla, indicating the corollar appendages; 4» rear of stamen, illustrating the auriculate anther; 5, gynoecium, illustrating the clavuncle, carpels, and nectaries. Ann. Mo. Bot. Gaiux, Vol. 17, 1930 Plate 20 1 2 4 Jxpocyn vim 5 irsicnomituiii X o *,cy n u^txi. \\UOD«<>N— ArOCVNACEAE Aimals (»f the Missouri Botanical Garden Vol. 17 SEPTEMBER-NOVEMBER, 1930 Nos. 3-4 STUDIES IN THE UMBELLIFERAE. IIP A Monograph of Cymopterus Including a Critical Study OF RsiiATED Genera MILDRED E. MATHIAS Formerly Research Ass'stant, Missouri BoU >f Washi Table of Contents I Introduction 214 11 216 History of the genus Cymopterus General morphology Comparative morphology of the genera with special reference to the genus Cymopterus • • Evaluation of characters used in classification ■ 223 Summary Geographical distribution Generic and specific relationships Economic uses Ill Key to the genera Review of related genera 250 253 375 Taxonomy of the genus Cymopterus List of exsiccatae Abbreviations • AfiO Index to species 1 An investigation carried out at nhe Missouri Botanical Garden in the Graduate Washingt mitted as a thesis in partial fulfillment of the requirements for the degree of doctor ^f r^l.1■^^or^r.Tlv I'n fVi^ TTPTirv Rhfl-W Srhool of Botanv of Washington University. Issued December 27, 1930. Ann. Mo. Bot, Qabd., Vol. 17, 1930. (213) [Vol. 17 214 ANNALS OF THE MISSOURI BOTANICAL GARDEN I Introduction The problem of generic delimitation in the Umbelliferae of western North America has caused great confusion in recent years. This confusion has been particularly evident in the group of genera centering around the Cogswellia and Cymopterus types. Some workers are inclined to treat these groups as composed of many comparatively small genera; others have on the contrary considered them as two "polymorphous" genera composed of a number of distinct sections or subgenera. The present study was undertaken in an endeavor to give a comprehensive mono- graphic treatment of the genus Cijmopterus. As the study developed the necessity arose for defining and delimiting the various generic elements which have been confused with Cymop- terus. Thus the present paper is a monographic treatment of that genus as well as a survey of the Cymopterus complex. There are some fifteen genera involved, among them P hellopterus , Pseudocymopterus, and Aulospermum. Several of these genera are founded on segregates from Cymopterus while others are so closely related as to cause confusion. In the first part of this paper Cijmopterus is discussed in detail and the relationships of the other genera to the Cymopterus type stressed. In the re- mainder of the paper the various genera are delunited and given a taxonomic treatment. At this time the writer wishes to express her appreciation and gratitude to the people who have made this study possible. Thanks are due Dr. George T. Moore, Director of the Missouri Botanical Garden, for the use of the excellent library and her- barium facilities which this institution afford. Sincerest apprecia- tion is due Dr. J. M. Greenman, Curator of the Missouri Botanical Garden, under whose guidance this study has been carried on, for the aid and advice which were so willingly given at all tunes. Special thanks are due those in charge of the herbaria of the following institutions for their courtesy in extending the privilege of examining material deposited there and for the loan of material necessary for this study: the Gray Herbarium of Harvard Uni- versity, the New York Botanical Garden, the Brooklyn Botanic 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 215 Garden, the New York State; Museum, the Academy of Natural Sciences of Philadelphia, th(3 United States National Museum, Carnegie Museum, Pittsburgh, the Field Museum of Natural History, the Herbarium of the University of Chicago deposited in the Field Museum of Natural History, the Edward L. Greene Herbarium at Notre Dame University, Iowa Agricultural College, the University of Colorado, Pomona College, the California Academy of Sciences, and the Dudley Herbarium of Leland Stanford, Jr. University. For the loan of material in connection with this study the writer is indebted to those in charge of the herbaria of the Uni- versity of Wyoming, the University of California, the University of Oregon, the State College of Washington, the University of Minnesota, Nevada Agricultural Experiment Station, Brigham Young University and the National Museum of Canada. The writer has appreciated the opportunity of examining specimens in the private herbaria of I'rof. M. E. Jones, Claremont, Cali- fornia, and Mr. George J. Goodman, Missouri Botanical Garden. Appreciation is due Dr. J. H. Barnhart of the New York Botanical Garden, Dr. John Briquet of the Botanical Conser- vatory of Geneva, and Dr. i^.lfred Rehder and Miss E. M. Tucker of the Arnold Arboretum for their assistance in bibliographical matters concerning various genera; to Dr. A. B. Rendle of the British Museum and Dr. A. W. Hill and Mr. T. A. Sprague of the Kew Herbarium, for their kindly assistance in locating the type specimen of Cymopterus and for the photograph of it which was magnanimously supplied; and to Dr. L. Diels of the Botanical Garden and Museum of Berlin-Dahlem for a fragment of the type specimen of Musinewn Ehrenhergii. Sincere appreciation is due Dr. D. H. Linder of the Missouri Botanical Garden for his valuable assistance in the preparation of microphotographs, and Dr. R. V. LaGarde and Mr. Paul Kohl of the same institution for the preparation of photographs of specimens used in this paper. The author is indebted to many others for their assistance in the numerous details of this study. 216 [Vol. 17 GARDEN II History of the Genus Cymopterus The genus Cymopterus was first described by Rafinesque^ in his 'Prodrome des nouveaux Genres de Plantes' published in 1819. The genus was based on Selinum acaule of Pursh' and on Thapsia glomerata of Nuttall.^ These two species were correctly treated as conspecific by Rafinesque; because of priority the first named must be taken as the historical type of the genus. Selinum acaule was based on plants collected by John Bradbury "on the alluvion of the Missouri, from the river Naduet to the Mahas"* in "upper Louisiana" [from a few miles north of the present site of St. Joseph, Missouri, to Sioux City, Iowa]. Thapsia glomerata was based on a collection of Thomas Nuttall "On the open plains of the Missouri, commencing 40 miles below the confluence of White river "^ [southern South Dakota]. These plants were collected in 1811 on a journey which Nuttall and Bradbury made from St. Louis, Missouri, to the Mandan Villages on the upper Missouri River, now near Bismarck, North Dakota. The first review of the genus following its original pubhcation was by de Candolle" in 1830 in the Trodromus.* Only one species was recognized, namely, Cymopterus glomeratus. Nuttall^ in 1834 described another species, Cymopterus glaucus. A comprehensive treatment of the genus was made by Torrey and Gray' in 1840. They recognized and described eight species, namely, Cymopterus glomeratus, C. montanus, C glaucus, C. campestris, C. ter ehinthinus , C. foeniculaceus , C. alhiflorus, and C. thapsoides, which were placed in four sections — Eucymopterus of Torrey and Gray, Phellopterus, Leptocnemia, and Pteryxia, of Nuttall. The genus Cymopterus has been surveyed in the various floras of the western region of North America where it is of common » Rafinesque, Jour. Phys. 89: 100. 1819. »Pursh, FI. Am. Sept. 2: 732. 1814. 'Nuttall, Gen. 1: 184. 1818. * Bradbury, Cat. 1819 [reprinted in Thwaite, 'Nuttall, Gen. 1: 185. 1818. « De Candolle, Prodr. 4: 203. 1830. ^ Nuttall, Jour. Acad. Phila. 7: 28. 1834. « Torrey and Gray, Fl. N. Am. 1: 623. 1840. West. Trav. 5: 318. 19041. 1930) MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 217 11 occurrence. Between 184Ci and 1888, when the 'Revision of North American Umbellife:.-ae* by Coulter and Rose^ appeared, fourteen species and two varieties were pubUshed under the generic name Cymopterus. Bentham^° recognized the genus and attributed to it some ten species. The new genus Phellopterus was described by him and based on Cymopterus littoralis. Watson in 1887 described the gems Podistera, basing it on Cymopterus nevadensis . In the 'Revision' of Coulter and Rose^^ only one species was added to the genus but a number of previously published species were transferred to other i?enera, as only thirteen species were recognized in the genus Cymopterus. The new genus Coloptera contained three species which were considered closely related to Cymopterus; and the ncjw genus P seudocymopterus also con- tained three species, two of which had previously been described under Cymopterus. In Drude's^' survey of the Umbelliferae for 'Die naturlichen Pflanzenfamilien' these three genera were recognized and treated according to the Coulter and Rose 'Revision.' The next treatment of the genus Cymopterus was by Coulter and Rose^* in their 'Monograph' in 1900. They recognized only seven species; the genus Coloptera was reduced to synonymy under Cymopterus; howev<;r, P seudocymopterus was retained as a generic element. Specieis which had originally been described under Cymopterus were to be found in eight other genera, namely, Podistera, Oreoxis, Rhysopterus, Phellopterus, Aulospermum, Pteryxia, P seudocymopterus , and Glehnia. Jones*"* in 1908 surveyed the genus Cymopterus, and reduced the genera Rhysopterus, Phellopterus, Aulospermum, Pteryxia, Oreoxis, and P seudocymopterus to sectional rank under Cymopterus, recognizing forty-four species in the genus. In 1913 Rydbergi" desc]*ibed the genus Coriophyllus, formerly » Coulter and Rose, Rev. N. An. Umbell. 1888. "Bentham, in Bentham and Hooker, Gen. PI. 1: 911. 1867. Watson 1887. " Coulter and Rose, Rev. N. Am. Umbell. 1888. "Drude, in Engler and Prantl, Nat. Pflanzenf. S^: 221. 1898. " Coulter and Rose, Contr. U. M. Nat. Herb. 7: 179. 1900. " Jones, Contr. West. Bot. 12: lfr-29. 1908. "Rydberg, Bull. Torr. Bot. Club 40: 69. 1913. 218 rVou 17 GARDEN treated as a section of Cymopterus by Jones. He also described the genera Pseudopteryxia^^ and Pseudoreoxis,^' both based on species of Cymopterus. Thus recent workers have been doubtful as to the course to pursue in treating these genera. There are two methods of procedure in use — that of Jones, who treats all the genera men- tioned as mere variants of one large generic concept, and that of Coulter and Rose, and of Rydberg, who recognize the different elements as being generically distinct. Both methods of pro- cedure have been followed in recent years by authors of local manuals. General Morphology COMPARATIVE MORPHOLOGY OF THE GENERA WITH SPECIAL REFERENCE TO THE GENUS CYMOPTERUS Habit and Habitat. — The various genera included in this survey have for the most part a xerophytic or semixerophytic habit. They occur practically without exception in the dry, sandy, or alkaline regions of western North America and usually in more elevated habitats. They are all herbaceous perennials, glabrous or pubescent, frequently caespitose, caulescent or acaulescent. The genus Cymopterus occurs in arid or semi-arid regions from central Saskatchewan to northern Mexico and from western Minnesota and central Nebraska to eastern California in eleva- tions from approximately one thousand to eleven thousand feet. The plants are glabrous except for an occasional pubescence of the fruit and the puberulent foliage of one species, C. cinerarius; and are strictly speaking acaulescent although certain species become subcaulescent through the development of a pseudo- scape. Roots. — The roots are those typical of most Umbelliferae. With few exceptions they are fleshy and tuberous and much elongated, seldom branching, and of the perennial, or at least biennial, type. In certain species a bulbous growth is developed; but in no case are the roots fibrous or filiform. The genus Cymopterus never develops a bulbous root, but has a fleshy, sub- fusiform, usually much elongated tap-root, rarely branching. " Rydberg, BuU. Torr. Bot. Club 40: 71. 1913. "/bid., 73. 1913. 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 219 Stems. — Certain genera are typically caulescent with usually simple un branched grooved stems. The stems are mostly glabrous, but one example of a pubescent condition is the hirtel- lous development at the base of the umbel in certain species of the genus Pseudocymopierus. Probably the predominating condition is the development of only subterranean stems, giving the plant an acaulescent appearance. In certain genera a so- called "pseudoscape" is dev^eloped which is comparable to the stem and morphologically equivalent to it (pi. 25, fig. 1). It is produced by the subterranean stem becoming elongated and bearing at its apex a cluster of leaves and peduncles — a situation typified by the genus Aulosj'ermum. This pseudoscape, however, may never be raised above; the surface of the ground, so the plants retain their acaulescont appearance. The genus Cymop- terus is essentially acaulesc(;nt, some species growing practically prostrate with only leaves and peduncle above the surface of the ground. The development of a pseudoscape always char- acterizes certain species. The peduncle is of varying length — shorter than or exceeding tlie leaf cluster — and always glabrous. Leaves. — The leaves are alternate, and usually develop a dilated petiole which forms a sheath at the base. They may be ternate, palmate or pinnate^ and generally become much divided. There is a variation from coriaceous and subcoriaceous leaves with cartilaginous margins to thin, submembranous leaves with entire undifferentiated margins. The ultimate leaf divisions are in many species definitely linear; in other species they attain a width of about three centimeters. For the most part the leaves are glabrous but exceptions occur in several species, the pronounced being in the case of Glehnia leiocarpa, which always has a dense tomentum developed on the under surface of the leaf. The leaves of the genus Cymopterus in those species developing a pseudoscape appear whorled at its apex. The petioles are sheathing and surround thci pseudoscape or peduncle base. The leaves are bipinnate to quadripinnate, appearing as ternate in certain species due to the confluence of the pinnae. They vary from coriaceous and subcoriaceous with cartilaginous mar- gins to thin, submembranous ones with undifferentiated margins. most [Vol. 17 220 ANNALS OF THE MISSOURI BOTANICAL GARDEN The ultimate leaf divisions in some species are linear and remote, in others more or less confluent. Certain species, such as C. Newherryi, have mostly broader ultimate leaf segments. The leaf margins may be lobed or entire and the apex of the ultimate segment may be apiculate. Inflorescence. — The inflorescence is in all cases a compound umbel, and the number of umbellets varies. The umbellets may be sessile or borne on long rays, resulting in the former case in a dense globose inflorescence and in the latter in a widely spreading type. The involucre is mostly absent but may occur in the form of a low sheath, one or two small, inconspicuous, linear, foliaceous bracts or even develop more or less conspicuous, partially united, scarious bracts which characterize the genus P hellopterus . The involucel varies from small, inconspicuous, thin bracts to the conspicuous, subulate bracts in species of Pseudocymopterus; to the prominent, dimidiate, foliaceous bracts of some species of Cymopterus; and to the large, scarious, some- times many-nerved, bracts of Phellopterus, The inflorescence is usually glabrous, but in some species the rays and pedicels may be scabrous-pubescent and the involucre and involucel also may become pubescent. The inflorescence of the genus Cymopterus varies greatly. In certain species, such as C. descriicola and C. glohosus, the primary and secondary rays have become aborted and fused, forming a discoid inflorescence. In other species the umbellets may be borne on long rays, giving a spreading inflorescence type. All intermediate stages occur, resulting in a variation from the discoid rayless inflorescence through a globose type with very short rays to a widely spreading umbel. The involucre may be entirely absent, or present as a low, inconspicuous sheath, occasionally toothed and sometimes definitely bracteoid. The involucel is characteristic for the genus, being conspicuous in most species and composed of dimidiate, scarious, whitish, purp- lish or green, typically foliaceous bracts. In such species as C. glohosus and C. deseriicola , in which the inflorescence is discoid, the involucel bracts are inconspicuous, scarious, and paleaceous Flower. — The flowers are small, epigynous, with five sepals which are often inconspicuous, five petals and stamens, and a bicarpellary pistil although three carpels may rarely occur. 1930] MATHIAS — STUDIEP. IN THE T7MBELLIFERAE. Ill 221 Calyx. — The calyx lobes are usually inconspicuous and not persistent, although exceptions occur in such genera as Musineon and Pseudocymopterus, with conspicuous, persistent, subulate calyx lobes. The genus Cymopterus has mostly non-persistent calyx lobes but in some species the teeth are conspicuous at the apex of the mature fruit. Corolla. — The corolla is that typical for the family. The petals are concave with iaflexed tips, and vary from white or yellow to purple. Stamens. — There are five alternate introrse stamens which are similar in all the genera. Pistil. — The pistil is bicarpellary and surmounted by a disk which in a number of species has been confused with the stylo- podium. This disk may become quite fleshy and prominent. With the exception of the ^;enus Podistera no stylopodium occurs in the group. The style may become leflexed and is usually persistent. However, in the genus Cymopterus the style rarely persists. Fruit. — The fruit may be sessile or pedicellate. It is bicar- pellary (rarely a tricarpelliry condition occurs), the two carpels being in contact on their commissural surfaces (pi. 21, fig. 3b, fig. 4h). In most of the genera the commissural surface is broad and the carpels are in cbse contact until maturity; in other genera the commissure is quite narrow, the two carpels appearing distinct, as in Harhouria. The carpophore (pi. 21, fig. la), a portion of the receptacle prolonged between the carpels as a central axis, may be entire or two-parted, and may or may not be persistent. The fruit may be compressed laterally, at right angles to the commissural plane, or dorsally, parallel to the commissural plane, or this (Compression may be practically absent. The carpel surface is usually marked by five longitudinal ribs or wings. The two nearest the commissure are known as the lateral wings (pi. 21, fig. 2a, fig. 4e); those on the back of the carpel have been referred to as the dorsal wings (pi. 21, fig. 2b, fig. 4a). Technically the wing on the center back of the carpel is the dorsal wing and the ones between it and the laterals are the "intermediate" wings. In this paper the intermediate and dorsal have been merged under the latter term. When only [Vol. 17 222 ANNALS or THE MISSOURI BOTANICAL GARDEN one dorsal wing is present it may be inferred that it is normally centrally placed on the back of the carpel; in some fruits the so-called ''intermediate" wings develop and the "dorsal" wing remains as a rib; in others three or four wings may be developed on the dorsal surface. Thus it has seemed best to include all the wings, excepting the laterals, under the general term of dorsal wings. There may also be developed secondary ribs. The condition in the group thus varies from a carpel with inconspicu- ous ribs to one with both dorsal and lateral wings conspicuously developed. The wings and ribs vary greatly in size and shape. The ribs may appear as inconspicuous protuberances on the carpel surface or may become quite conspicuous as in Aletes. The wings in cross-section may be short or extended linear pro- jections; may be constricted at the base, then abruptly broadened and acuminate at the apex; or they may be inflated at the base and abruptly constricted to a linear projection toward the apex (pi. 22, fig. 11). The texture of the wings varies from a thin, scarious one through a spongy tissue to a definite corky develop- ment. A subcartilaginous epidermal structure may even develop. Oil tubes are present in all the genera; their size, number and position, however, is variable even within the genus. They may be solitary in the intervals as in species of Oreoxis and Aleies; or there may be as many as eight or more oil tubes in each interval; on the commissural surface the oil-tube number varies from two to twenty or more. In certain species there is an accessory oil tube in each rib. A number of species have oil tubes in the ribs or wings — at the wing-base, in the inflated portion of the wing, or in the apex. The oil tubes are usually arranged in definite areas surrounding the seed but at times they may be scattered throughout the pericarp. Strengthening cells may or may not be developed at the wing bases. This strengthening tissue may be composed of vascular elements or of only a small-celled parenchyma. As a rule it merges gradually into the other tissue. The fruit is usually glabrous but a pubescent condition occurs in certain species. A thick tomentum may cover the carpel, the apex only may be pubescent, usually with multicellular hairs, or the commissural surface may be puberulent to scaly- 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 223 tomentose. In other genera the fruit is merely tuberculate — the degree varying from a slight tuberculation between the ribs to a condition covering the entire fruit. The genus Cymovterus presents a number of interesting fruit norm characters. The fruit is j^edicellate or sessile, and compressed dorsally. The outline of the fruit varies from ovate or ovate-oblong to narrowly cuneate or oblong with an abrupt constriction below the middle. This latter outline is largely due to the compression witliin the umbel, a varying condition in each species. Lateral wini;s are always developed; these may or may not be constricted at the base in cross-section; in certain species they are definitely inflated at the base. The dorsal wings vary greatly. In C. acaulis and C. Fendleri one to three dorsal wings occur; both the dorsals and laterals are constricted at the base. Cymopierus Newherryi shows a variation from a carpel with only lateral wings to one with dorsals fully developed. In those forms with only lateral wings developed the sections show evidence of abortion cf the dorsals, the prevalent condition throughout this species. Cymopterus deserticola presents a condition with lateral wings fully developed and the dorsals remaining as slight projections on the carpel surface. Cymopterus Coulteri, C. glohosus, and C. zinerarius have carpels with the wings rarely constricted at the base but usually tapering gradually from base to apex. The oil-tube number is also variable in the genus; there may be one to many oil tubes in the intervals and two to many on the commissure. Certain species may have an accessory oil tube developed in each winj?. Strengthening cells may be formed at the wing base but in some species they are always absent. The fruit may be entirely glabrous or puberulent to scaly- tomentose on the com mist ural surfaces and pubescent at the tip with multicellular hairs as is the case in C. Tnegacephalus and C. deserticola. EVALUATION OF CHAR^lCTERS USED IN CLASSIFICATION The Umbelli ferae as a family has always caused much per- exity. The absence for the most part of floral characters useful [Vol. 17 224 GARDEN in diagnosis has created its reputation as a difficult group. The North American genera of the family have received various treatments, largely due to the small amount of material available for study. The treatment of Torrey and Gray/^ the first comprehensive survey of the North American genera, was based on a relatively small amount of material. As a consequence their genera at times were exceedingly heterogeneous as to morphological rela- tionships. Gray^s opinion is expressed in a letter which he wrote to Sir George Bentham in 1867: — "As to Umbelliferae, I wish you joy of the job, and do hope you will reduce the genera twenty percent at least. I never could take the least satisfaction in them. I never could collate our Umbelliferae with European genera, and I have no clear conception of more than half a dozen of our genera . . ."^° One of the difficulties encountered in the earlier treatments of the North American genera was this attempted identification with European forms when many times the characters in common were merely superficial ones of appearance. An opposing view is given by an earlier worker : — " it is, . . . in Umbelliferae, . . . indispensible that the genera should be confined within the most exact limits ; and it is far better that this should be effected by the creation of many new genera, than that it should not be done at alL^^i This illustrates the conflicting views concerning the treatment of the North American Umbelliferae. The question has been as to what constituted a genus — is the family made up of a number of small or even monotypic genera or is it composed of relatively few large polymorphous groups? One of the purposes of this study has been to ascertain what characters can be considered as constant in the delimitation of both genera and species. The stress has formerly been laid on detailed fruit characters, at times to the elimination of all other possibilities. As a con- sequence heterogeneous groups have been merged because of such microscopical characters as oil-tube number. The statement of T ■ " Torrey and Gray, Fl. N. Am. 1: 623. 1840. » Gray, J. L. ed., Letters of Asa Gray 2: 553. 1893. "Lindley, in Edwards' Bot. Reg. 18: t. i486. 1832. 19301 MATHIAS — STUDIES JN THE UMBELLIFERAE. Ill 225 Coulter and Rose in their 'Revision' illustrates this point of view: ''The order must still be considered a difficult one to the ordinary- student, inasmuch as it demands careful sections of the fruit and an examination of rather minute structures. "^^ The classificatory value cf various characters has also been briefly discussed by them: "We have therefore divided our Umbelliferae into two series', based upon the development of secondary ribs or not, . . . Our second series, which is by far the larger, we have broken into groups based upon the flattening of the fruit, which seems to be a very reliable character, and one which best groups together related genera. Characters which are used in further subdivisions, enumerated in the order of their general importance, are as follows: flattening of the carpel, nature of the stylopodium, winging of the fruit, character of the seed-face, number of oil-tubes, and general habit. Thus genera are reached, and in a way that seems to express relation- ship as well as can be done in any lineal arrangement."23 It may be seen from the above that practically all the generic characters are based on th(j fruiting condition, and many of these are apparent only after detailed examination and sectioning. In the course of this study material has been obtained from the larger herbaria of the (Country as well as from the local herbaria in the regions where the genera are indigenous. A detailed morphological study has been made of every species to determine the range of variation of all its characters. The characters previously used in the classification of the groups studied have been tested w'.th the unusually large amount of material available. In the casie of the genus Cymopterus a detailed study of every specimen has been made especially in connection with fruit characters. Free-hand or microtome sections were made of all fruiting material ia order to show the variation within the species, individual umbel, and between the two mature carpels of one fruit. The herbarium material fcr free-hand sections was soaked in distilled water or a two per C(Jht solution of potassium hydroxide from twelve to fortv-eight hours, boiled one-half to five minutes « Coulter and Rose, Rev. N. Am. irmbell. 3. 1888. « Ibid., -15-16. 1888. 226 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 in water or sectioned without boiling, sectioned in pith, dehy- drated in absolute alcohol, and mounted in glycerine. At times material was boiled immediately in water without the long preliminary soaking. There was no apparent difference in the sections due to the various treatments given the fruits. The fruits used for microtome sections were soaked in distilled water over night, boiled one-half to five minutes in a two per cent solution of potassium hydroxide, dehydrated, and embedded in paraffin as usual. Serial sections were cut at a thickness of six microns and stained with a one per cent aqueous solution of gentian violet and a one-half per cent clove-oil solution of ery- throsin. The method was adapted from that used by Mrs. Haber-^ in her studies on Euphorbia, with changes made to suit the material. These sections proved very satisfactory for a study of the fruit — shape and composition of the wings; origin, number, and variation of oil tubes; and vascular structure. Since the fruit characters have been considered of prime importance in the family they shall be discussed first. The general outline of the carpel has been found to be a verj^ constant character within a species or group of species. The outline is only rarely affected by the compression within the umbel or by the drying of the specimen. When the outline is due to the compression within the umbel it is as a rule constant for the species— as, for example, the narrowly cuneate or abruptly constricted fruits of Cymopterus glohosus. The compression of the fruit is a much more variable character. As a rule this compression remains within certain limits in each genus, and laterally compressed fruit may readily be distinguished from that dorsally compressed. But there are a great many intermediate stages — some fruits practically non-compressed, others so inter- mediate that the interpretation depends on the individual point of view. For those groups with definite laterally or dorsally compressed fruits this character has been found very useful gen- erically. For the remaining groups the character is very doubt- ful and has not been found of any value. The winging of the fruit is a character of great importance « Haber, JuUa M. Ann. Bot. 39: 661. 1925. 19301 MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 227 but one frequently over-sti'essed. An excellent example is the case of the species, Cymopterus Newhemyi. A superficial examina- tion of a small amount of material would probably lead one to the conclusion that only lateral wings are developed, and rarely one dorsal wing appears as a ridge on the carpel surface. Yet an examination of a series of fruits has shown the variation to range from a carpel with only laterals to one with both laterals and dorsals. An extreme example of this is illustrated by fig. 3 of plate 25, which shows a cross-section of the two carpels of one fruit. An examination of microtome sections has shown that when the dorsal wings have not developed the vascular tissue and oil tubes indicate thei]' normal position and the fact that they have been aborted (pi, 25, fig. 2). In this group of genera lateral wings are present in all the winged fruits. Certain genera have only rib development. Dorsal wings may be absent entirely, absent through abortion as illustrated (pi. 25, fig. 3), present as one to three inconspicuous ridges on the dorsal surface of the carpel, or there may be one to four dorsals as conspicuous as the laterals and of the same foi-m. This variation is of value both specifically and generically if a large series of specimens has been critically examined. Certan genera are characterized by the constant presence of both lateral and dorsal wings, others by the presence of laterals only. The genus Cymopterus, however, presents both conditions. ]3ut in this genus if the dorsals are ]letely absent vascular evidence indicates their point of com ongm. And the variation within an umbel may cover the range of variation for the genus. This character when used in com- bination with others has proved of specific value in this genus, — for example, C. deserticola (])1. 23, fig. 3) has a carpel with only dorsal ridges developed; its closest relative, C. megacephalus (pi. 23, fig. 4), develops a prominent wing on the dorsal surface of the carpel. The dorsal wings as a nJe are not greatly convoluted, but Cymopterus corrugatus is chaiacterized by the constant corrugated appearance of the dorsal wiags — a character which will readily separate this species in the Iruiting stage from all others. The shape of the wing in cross-section is a very constant character and usually one of great value in diagnosis: as both a [Vol. 17 228 BOTANICAL GARDEN generic and specific character it has been found most useful. Every genus as a rule possesses a predominating wing outline. The condition in the genus Cymopterus is usually that with a constricted wing-base. Wing types for the various genera and species are illustrated in plates 22-25. The range of variation within the group is shown (pi. 22) and also within the species figs. 1-14) and It may be seen that the wing outline is extremely constant within the species The texture of the wings can also be used to some extent in diagnosis, although degrees of difference in texture are difficult to express objectively. There is a generic and specific variation in cellular structure resulting in scarious to subcartilaginous wmgs. length of the wing in cross-section as compared with carpel breadth is a character frequently of value in specific analysis. The number of oil tubes has been used as a very important character in separating the family into groups. The key given by Coulter and Rose^^ in their 'Monograph' has as the third main subdivision of the genera the presence of a solitary oil NUMBER OF OIL TUBES IN INTERVAL Fig. 1. Graph showing the range of variation in oil-tube number in the intervals of mericarps of Cymoptems Fendleri. tube as contrasted with several oil tubes in the intervals. Too much stress has been placed on the value of oil-tube number as a diagnostic character. It is a valuable character to a very limited degree. Again it is most essential to know as nearly as possible the range of variation within a species and within a single umbel. Also the plane in which the section is cut is of » Coulter and Rose, Contr. U. S. Nat. Herb. 7: 21. 1900. 19301 MATHIAS — STUDIES III THE UMBELLIFERAE. Ill 229 great importance. Serial se(jtions in paraffin of a number of carpels show a wide range of variation from base to apex in the same carpel. If all sections were made in the median plane a basis of comparison might be reached, but here again the personal equation interferes. Figure 1 shows graphically the range of variation within th3 species Cymopterus Fendleri. It is recognized that generally the presence of only one oil tube in the interval is fairly constant and dependable; yet exceptions occur as illustrated in pi. 22, fig. 7, and pi. 23, fig. 2. The num- ber of oil tubes on the coirmissural surface is more limited in variation but not of any gres.t value specifically. The size of the oil tubes, as a rule, depends on the number present. When the oil tubes; are solitary in the intervals they are much larger than in those cases with numerous oil tubes in the intervals. The shape vai'ies also, appearing in cross-section from round to elongated-elliptical and lenticular. The ] of the oil tubes presents ms In the majority of genera studied the oil tubes are arranged in a definite row around the seed, as exemplified by Cymopterus acaulis (pi. 24, fig. 21). The genus Neoparrya (pi. 22, fig. 2), on the other hand, has oil tubes scattered throughout the pericarp. In addition to these there may be accessory oil tubes located singly in the wings. They are mud smaller than the others, appearing at the base, apex, or in the iiiflated portion of the wing, usually in connection with strengther ing tissue. The vascular tissue in the carpels is at times apparently absent, at others conspicuous — a cor dition almost specific in its occur- rence. The present study h£.s shown the occasional presence of one to three strands near th'3 wing base and vascular elements in the carpophore, but the vascular study has not been sufficiently complete to warrant any coni elusions concerning its phylogenetic significance. The persistence of the carpophore has been found to be usually of generic constancy and is an easily observable character in any fruiting specimen. There are a number of inflorescence characters which have not been used to any great extent beyond occasional specific delimita- tions. The length of the primary and secondary rays is a specific [Vol. 17 230 ANNALS OF THE MISSOURI BOTANICAL GARDEN character — varying for each species within more or less definite limits. The complete abortion of the rays in certain species of Cymopierus resulting in a discoid inflorescence is a constant character for an entire group of species. This discoid develop- ment is apparently due to a shortening and fusion of the primary and secondary rays and is correlated with other specific differ- ences, indicating that it is not a chance fasciation occurring in any species. Cymopterus acaulis is characterized by relatively short rays resulting in a globose inflorescence, while in C. Fendleri and C. Newherryi the rays are as a rule much longer, resulting in a more or less spreading type of inflorescence. The involucre and involucel (pi. 21, fig. lc,e) are characters of great value specifically and generically — being apparent in either fruiting or flowering condition. Rafinesque in his 'Scadio- graphy'made an interesting comment concerning these characters: ''and the essential form of the common and partial involucres no longer neglected, since they are as important as those of the perianthe of compound flowers, being both bracts assuming a floral importance, when the flowers are so much alike. "It is the same for the bracteal glumes of Grasses that give generic characters, and wherever the bracts offer more diversities than the uniform flowers. "^^ The presence or absence of the involucre is a relatively uniform character for the species. At times the type of involucral bracts is a reliable character but in some species the range of variation is too great in this respect. The texture of the bracts is in some cases constant for an entire genus, as, for example, Phellopterus, which is always characterized by a scarious involucre. The involucel is a more prevalent structure and fairly constant as to form and texture for the genus. The dentation and venation of the involucel bracts are valuable characters specifically. The flowers in this group are too uniform to present any characters of classificatory value. Their color has been used to some extent but the variation may be from white to purple within a species or on the same plant. In the young carpel a number of the fruit characters are visible — wing number and » Rafinesque, C. S. Scadiography or 100 G. of Ombelliferous plants etc. The Good Book 1 : 60, 1840. 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 231 position and occasionally oil-tube number can be determined from sections of very young carpels. The peduncle presents an interesting character in its length. Some species have very sho]-t peduncles, the inflorescence being PEDUNCLE LENGTH Fig. 2. Correlation table illustrati ng the relation between peduncle length and leaf Fendlen. The measurements are from one hundred typical individuals of each species. XX C. Fendleri oo C. acaulis practically sessile in the axils of the leaves; others have peduncles much longer than the leaves. The comparative relation of peduncle and leaf l<}ngth character for species, and a study of this ratio has been made in some detail for two closely related species— Cymopterus acaulis and C. Fendlen. The results of tliis study are set forth in the above 232 (Vol. 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN ency ion table (fig. 2). It may be seen that in spite of the I^ing each species shows a different developmental tend- The peduncle becomes lengthened as the plant matures PEDUNCLE LENGTH Fig. 3. ems -•"--• C acaidis C. Fendleri 00 -co C. Newberry O, .0 ticola nahip expressed by the species peduncle and leaf. The limits ly of all available material. C megacephalus -• C. globostts )0 (7. cinerarivs X C Covlteri corrugatus but a more or less constant ratio is preserved between the peduncle length and leaf length. Figure the various species of Cymopt lengths of peduncle and leaf. Qstrates the relationship c based on the comparativ 1Q30] MATHIAS — STUDIES IN THE UMBELLIFERAE, III 233 The division of the leaves is constant within the species ternate or twice- tern ate, ar.d simply pinnate to quadrip innate. The frequent confluence of the ultimate segments makes it difficult at times to determine the degree of division. Some assistance is obtained by tracing the vascular system through the leaf. In this study the length and breadth of the ultimate leaf segments (pi. 21, fig. 1^,0 have been measured, and the character has been used in a definite delimitation of the species. The ratio of ultimate segment length to breadth is a constant specifically. The leaf-length /leaf-breadth ratio has also been studied. Here again it is relatively constant for the species. The general outline of the leaves also adds a useful character. The leaf texture is specific bat, as in the case of wing texture, the difficulty arises in expressing the differences objectively. The pseudoscape develop-ment is quite constant throughout some genera and characterizes many species. This is a character easily seen in herbarium material and present in the vegetative condition. It was first made use of and defined by Rydberg'^ and has proved of great value. Pubescence is relatively i^are in this group but when present is an excellent diagnostic character. Examples of its occurrence are the hirtellous condition of the foliage of Cymoplerus mierarius^ the multicellular pubescence on the tips of the fruit of C. deserti' cola and C. megacephaluSj the scaly tomentum of the com- missural surface of the cai'pel of C. deserticola, the hirtellous pubescence at the base of the umbel in P seudocymoylerus mon- tanus, and the tomentose lower leaf surface of Glehnia leiocarpa. Phellopterus monianus exhibits a peculiar foliage condition somewhat resembling pubeucence. The epidermal layer of the leaves has become very irregular, giving the leaf surface an uneven and roughened appearance; moreover, the leaf margins and veins develop a subcar tilaginous denticulation. The genus Harhouria exhibits a definitB tuberculation of the fruit. The roots are mostly tuberous. Occasionally a bulbous condition occurs. The general habit is of some value, particularly in a superficial determination of the genus — the variation ranging from prostrate, solitary, a(;aulescent plants through prostrate- 2» Rydberg, Fl. Rocky Mts. 607. 1917. 234 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 caespitose to a tall, erect, caulescent condition — a specific char- acter of value. Summary 1. The comparative morphology of the group of genera studied is discussed, stressing the gross morphology of the genus Cymop- terus. 2. The generic and specific characters used by previous workers in a classification of the family are reviewed. 3. There is a detailed treatment of the following morphological characters used in classification, and an evaluation of these characters is made, based on the results of this study: habit of plant; type of root; pseudoscape development; leaf outline, texture, division and margin, outline and length/ breadth ratio of the ultimate leaf segment; the correlation of peduncle length to leaf length; type, texture, venation, and dentation of the involucre and involucel; length of rays; flower color; presence or absence of stylopodium; persistence of carpophore and calyx; outline of carpel, compression of the mericarp, winging of the fruit, convolution and texture of wings, shape and dimensions of wing or rib in cross-section, number and position of oil tubes, vascular and strengthening tissue in the carpels. Geographical Distribution The genera included in this study occur indigenously throughout western North America from western Missouri to the Pacific Coast and from southern Canada to northern Mexico. One exception, Glehnia littoralis, is limited in distribution to eastern Asia, occurring along the coast from southern Siberia to Formosa and in Japan. The presence of the genus Glehnia on the western coast of North America and the eastern coast of Asia presents an interesting situation paralleled by many other genera. A more detailed discussion of the genus has been given in a previous paper." The genus Cogswellia has the widest area of distribution — the limits of its range coinciding with those for the entire group. The genus Cymopterus has only a slightly more limited dis- tribution and presents many interesting problems. Cymopterus "Mathias, Ann. Mo. Bot. Gard. 15: 91-108. 1928. 1930J MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 235 acaulis (fig. 4), undoubtedly one of the oldest species in the genus, extends from Minnesota to eastern Oregon and from central Saskatchewan to southern Colorado. The most Fig. 4. Map of western North Anerica showing the geographical distribution of the species of Cymopierus, C, acaxilis C.Fendleri C. Nev^erryi C. globosua O © C. CouUeri C. corTugatvs ♦ 4- X C. megacephalua C. deserticola. O O C. cinerarixis eastern station recorded is i'rom a collection of Lapham in the "south bend of the Red River of Minnesota" [latitude 46 longitude 96°] on the western boundary of that state. The most western station is in the Malheur Valley in eastern Oregon. The species also occurs in i:he Snake River plains of southern Idaho. It has doubtless passed across the mountains at some 236 [Vol. 17 GARDEN break in the region of Yellowstone National Park. Since this species is probably of pre-glacial origin its distribution may best be explained as a northern, eastern and western expansion from the unglaciated areas of Colorado where it now occurs — a situa- tion not improbable. Its nearest relative, C. Fendleri, occurs from west-central Colorado and eastern Utah to northern Mexico and from central New Mexico to west-central Arizona. Cymop- terus Newhcrryi has a more limited range in east-central and southwest Utah. Cymopterus glohosus has a somewhat compar- able distribution in the state of Nevada as shown in the accom- panying map (fig. 4). The other species of the genus are chiefly local — C. deserticola is known only from the Mohave Desert of California; its closest relative, C. megacephalus , occurs in the Little Colorado River Valley of Arizona. These two species are probable branches of one ancestral type common in the inter- vening area before the invasion of troughs during the early periods of the Cenozoic which separated large tracts of southern California and Arizona as islands. Cymopterus cinerarius is the only species found at higher altitudes, occurring at elevations from eight thousand to eleven thousand feet in the Mono National Forest region of eastern California and adjacent Nevada. At present its range is very limited but will doubtless be extended as a result of further explorations in that vicinity. Cymopterus corrugatus is known only from the Carson and Humboldt Lakes region in western Nevada, and C. Coulteri only from the valleys of the Sevier River and its tributaries in Utah. The distribution of the other genera introduces a number of cases of local or at least very limited ranges. Rhysopterus, a monotypic genus as now defined, is known only from the Malheur Valley and adjacent regions in eastern Oregon (fig. 5). Neoparrya has been reported only from the type locality in northern New Mexico near Servilleta, Taos County (fig. 5). Other monotypic genera are Podisiera and Harhouria. The former, an alpine representative of the group, is known only from the region of Mt. Dana in California and a collection from the San Bernardino Mountains (fig. 5). Harhouria is commonly distributed in the lower mountains from southeastern Wyoming to northern New Mexico (fig. 7). 1930] MATHIAS — studied; IN THE UMBELLIFERAE. Ill 237 Oreoxis is primarily an alpine and subalpine member of the group, represented most widely by 0. alpina which occurs in the higher mountains throughout Colorado and extends westward to Fig. 5. Vmerica showing the geographical distribution Musineon, Neoparryj, Rhysopterns, and Podistera Mueineon Ho )k€ri O o M. tenuijolium &0 M. vaginalum A M. Hnsare. O Neoparrya lithophila ^O Khysopterus plurijug XX Podistera nevadensis . Oreoxis humilis. the Henry Mountains, Utah (fig. 8). other hand, is known only from the vicinity of n the Peak, Colorado, a region from which 0. alpina has never been reported. Oreoxis Bakeri occurs pririarily in southwestern Colorado and adjacent Utah and northe]"n New Mexico. The only species of the genus found at lowei* elevations is 0. MacDougali until recently known only from the type collection in the Grand Canyon of the Colorado, I Subsequent collections have rvoL. 17 238 GARDEN now given it a somewhat discontinuous range in the natural- bridge region of southeastern Utah and in Mesa Verde National Park, Colorado. In the latter locality it forms large colonies along the upper ledges of Spruce Canyon. Further collections in northeastern Arizona and adjacent regions, particularly on the mesas and in the canyons, will doubtless show this species to be common in that region. The genus Aletes, represented by two species, A. acaulis and A. huviilis, is best known by the former which extends from southern Wyoming in the lower mountains through central New Mexico to western Texas (fig. 7). Aletes humilis is at present known only from the region of Dale Creek, Larimer County, in northern Colorado. The distribution of the genus Musineon parallels that of Cymopierus acaulis. Musineon divaricatum extends from the Missouri Kiver in South Dakota to northwestern Montana, while its variety Hookeri is widespread from the western Dakotas to western Montana, with one collection from eastern Nevada, and from central Saskatchewan to central Colorado (fig. 5). The other species of the genus are more local in distribution; M. tenuifoliuvi is common in southeastern Wyoming and adjacent regions; M, vagwatum, is known from the Bridger Mountains of southern Montana and the Big Horn Mountains of northern Wyoming; and M, lineare has been collected only in the type locality in Logan Canyon, Cache County, Utah. The remaining genera of the group are among the best known and most widely distributed of western North American Umbelli- ferae. The genus Phellopterus is common from the plains of central Kansas and Nebraska west to eastern California (fig. 6). Pliellopterus montanus and P. macrorhizus are the more eastern species of the genus ; the former extends from central Kansas and Nebraska to western Wyoming and south to northern New Mexico; P. macrorhizus is limited to Texas, the only species of the group common in that state. Phellopterus multiner vaius is the most southern representative of the genus, extending from western Texas to southeastern California; P. purpurascens has been reported from about the fortieth parallel in western Nevada and central Utah south to the Little Colorado River in Arizona; 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 239 P. hulbosus occupies a more: or less narrow north and south belt west of the Rocky Mountains from southwestern Wyoming through Utah and Colorado to east-central Arizona but in the south extends east to Texa^ and then north to central Colorado, Fig. 6. Map of western North America, showing the geographical distribution of the species of Phellopterus. • — P. montaniis A A P. macrorhizua ooo P.m%UHnervcUu8 x>'Y); rolling plains between Sheridan and Buffalo, 3500-5000 ft. alt., 15 June-15 July 1900, Tweedy 3374 (NY, R, WSC); grassy fields, Hanna Creek (upper), 15 June 1909, Willits 100 (R\ Colorado: Boulder, 1896, Andrews (P 83208); dry hills, Denver. Denver Co., 1600 m. alt., 25 May 1921, Clokey 4227 [Vol. 17 264 ANNALS OF THE MISSOURI BOTANICAL GARDEN (M, MU, NY, PA, R); foothills, Larimer Co., 6000 ft. alt., 11 May 1895, Cowen (NY); bluffs north of La Porte, 5500 ft. alt., 2 May 1896, Crandall (M); hills north of La Porte, 7 May 1898, Crandall 1438 (MU, PA); bluffs, north of La Porte, 2 May 1896, Crandall 1439 (R, WSC) ; American Plains Flora, Lat. 41°, 1862, Hall & Harbour 214 (M, PA in part); Adam's Ranch, Grover, 1 May 1926, E. L. Johnston 190 (R); bluffs n. w. of Grover, 24 May 1926, E. L. Johnston 206 (R); 1907, E. L. Johnston 458 (R); granite slopes. Pike's Peak region, Machride 2679 (M); Mt. Golden, 29 May 1872, Parry (M); near Boulder, 1700 m. alt., 30 May 1905, Ramaley 1015 (R); frequent, on mesas, Boulder, 7 May 1914, Ramaley 9789 (R) ; plains near Denver, 1500 m. alt., 8 May 1900, Rydherg & Vreeland 5839 (NY); 1869, Thomas (G). Idaho: Beaver Canon, 27 June 1895, Rydherg (NY). Nevada: Cherry Creek, 6000 ft. alt., 19 June 1906, M. E. Jones (NY, P 82864); Muncy, 6000 ft. alt., 25 June 1906, M. E. Jones (NY, P 82865); Streptoc Valley, 13 July 1891, M. E. Jones (P 82862). This variety grades into the species but by a combination of characters may be separated. Although described by Coulter and Rose as usually having soUtary oil tubes in the intervals, that condition is comparatively rare. The more usual condition, probably mistaken by them for a solitary oil tube, is the presence in each interval of one large oil tube accompanied by two com- paratively small inconspicuous ones. 2. M. vaginatum Rydb. Mem. N. Y. Bot. Card. 1: 288. 1900; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 350. 1909; Rydb. Fl. Rocky Mts. 615. 1917, and ed. 2. 615. 1922. "Musenion vaginatum Rydb." ace. to Heller, Cat. N. Am. PI. ed. 2. 149. 1900. "Musenium vaginatum Rydb." ace. to Ind. Kew. Suppl. 2: 121. 1904. Plants caulescent, 0.4-3 dm. high; stems purplish- tinged, glabrous except for a slight scaberulence at the base of the umbel; leaves broadly ovate to oblong-ovate in general outline, excluding the petiole, 1-10.5 cm. long, 1.5-4 cm. broad, twice to thrice ternate, ultimate segments distinct, linear, mucronate, 2-11 ram. 1G301 MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 265 long, about 1 mm. broad, raruly scaberulent on the upper surface, petioles 1-4 cm. long, inflated at the base into a conspicuous, purplish, scarious-margined .sheath; peduncles usually exceeding the leaves, umbels several-rayed, rays 0.4-1.7 cm. long, involucre absent or rarely present as a low sheath, involucel bracts linear- lanceolate, shorter than the white or yellowish flowers; fruit ovate-oblong, 3-4 mm. long, 1-2 mm. broad, somewhat glabrous to densely scaberulent, oil tubes mostly three in the intervals, 4-5 on the commissure, strengthening cells absent, seed face plane. Type specimen: Rydberg ■& Bessey 4G26, Bridger Mountains, Montana, 15 June 1897 (type in the Herbarium of the New York Botanical Garden, cotypes in the Edward L. Greene Herbarium of Notre Dame University and the Rocky Mountain Herbarium of the University of Wyoming). Distribution: known only from the Bridger Mountains, Montana, and the Big Hon Mountains of northern Wyoming. Specimens examined: Montana: limestone shingle, Mt. Bridger, near Bozeman, 8000 ft. alt., 5 July 1905, Blankinship 218 (M, PA, E); Bridger Mts., near the Pass, 8000 ft. alt., 28 July 1896, Flodman 695 (M, NY); Bridger Mountains, 7000 ft. alt., 14 June 1897, Ryd- berg & Bessey 1^625 (NY) ; Bridger Mountains, 7000 ft. alt., 15 June 1897, Rydberg & Besi-ey 4626 (NY type, N, R). Wyoming : headwaters of Tongue River, Big Horn Mountains, July 1898, Tweedy 58 (NY) , rolling plains between Sheridan and Buffalo, 3500-5000 ft. alt., 15 June-15 July 1899, Tweedy 2424 (WSC) ; rocky hills, Sheridai Co., 4000 ft. alt., July 1899, Tweedy $424 (NY). 3. M. lineare (Rydb.) Mathias, new comb. Aletes (?) tenuifolia Coult. & Rose, Contr. U. S. Nat. Herb. 7: 108. 1900. Daucophyllum lineare Rydb. Bull. Torr. Bot. Club 40: 69. 1913; Tidestrom, Contr. U. S. Nat. Herb. 25: 393. 1925 [Fl. Utah & Nev.]. D. linearis Rydb. Fl. Rocky Mts. 616. 1917, and ed. 2. 616. 1922. "Aletes (?) tenuif alius C!oult. & Rose" ace. to Wolff, Pflan- zenreich90: 141. 1927. 266 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 Plants subcaulescent, subcaespitose, 0.8-2.5 dm. high; stems glabrous, slender, lax, usually longer than the leaves; leaves narrowly oblong in general outline, excluding the petiole, 2-5 cm. long, about 2 cm. broad, simply pinnate, ultimate' segments (pinnae) distinct, linear, acute, 1-3.5 cm. long, about 1-2 mm. broad; petioles 3-14 cm. long; umbels several-rayed, rays 2-5 mm. long; involucel conspicuous, subdimidiate, bracts distinct, linear-lanceolate, acute, greatly exceeding the yellow flowers; fruit ovate-oblong, 3-4 mm. long, 1-2 mm. broad, minutely scaberulent; oil tubes 2-4 in the intervals, 6-8 on the commissure. Type specimen: Rydberg, near Logan, Utah, 9 Aug. 1895 (type in the Herbarium of the New York Botanical Garden, cotypes in the United States National Herbarium and the Rocky Mountain Herbarium of the University of Wyoming). Distribution: known only from the vicinity of Logan, Utah. Specimens examined: Utah: Logan, Aug. 1895, Rydberg (NY); Logan, 9 Aug. 1895, Rydberg (NY type, R 29913, US 391721, 1010452); rocks, Logan, 9 Aug. 1895, Shear 3162 (NY, US); cliffs, Logan Canyon, Logan, Cache Co., 24 April 1911, C. P. Smith 2337 (NY, R); Cache, 28 April 1911, C. P. Smith 2338 (NY). 4. M. tenuifolium (Nutt.) Coult. & Rose. PI, 22, fig. 3; pL 26, fig. 2. Musenium tenuifolium Nutt. in Torr. & Gray, Fl. N. Am. 1: 642. 1840 ; Walp. Rep. Bot, Syst. 2 : 427. 1843 ; Wats. Bibl. Ind. 1: 426. 1878; Coult. Man. Bot. Rocky Mt. 115. 1885; Coult. & Ruse, Rev. N. Am. Umbell. 111. 1888; Webber, Cat. Fl. Nebr. 124. 1890; Nels. Fl. Wyo. 115. 1896. Adorium tenuifolium (Nutt.) Kuntze, Rev. Gen. PI. 1: 264. 1891; Rydb. Bot. Surv. Nebr. 3: 37. 1894; Britt. Mem. Torr. Bot. Club 5: 242. 1894; Rydb. Contr. U. S. Nat. Herb. 3: 501. 1896 [Fl. Black Hills]. "Musineon tenuifolium (Nutt.) Torr. & Gray" ace. to Coult. & Rose, Bot. Gaz. 20: 259. 1895. ''Musineon tenuifoliuin Nutt." ace. to Britt. & Brown, 111. Fl. 2: 527. 1897, and ed. 2. 2: 644. 1913; Britt. Man. 677. 1901, and ed. 2. 677. 1905; Coult. & Nels. Man. Bot. Cent. 19301 MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 267 Rocky Mts. 350. 1909; Nels. Spring Fl, Intermt. States, 114. 1912; Petersen, Fl. Nebr. 161. 1912. ^'Musenion tenuifolium Nutt." ace. to Heller, Cat. N. Am. PL 97. 1898, and ed. 2. 149. L900. "Musineon tenuifolium Nutt. in Torr. & Gray" ace. to Coult. & Rose, Contr. U. S. Nat. Herb. 7: 78. 1900. Daucophyllurn tenuifolium. (Nutt.) Rydb. Bull. Torr. Bot. Club 40: 69. 1913; Rydb. Fl. Rocky Mts. 616. 1917, and ed. 2. 616. 1922. Plants acaulescent, erect, liubcaespitose, 0.6-3 dm. high; leaves narrowly oblong in general outline, excluding the petiole, 1.5-10 cm. long, 0.5-2.5 cm. broai, simply pinnate to tripinnatisect, ultimate segments distinct, linear, acute, 2-30 mm. long, 0.5-1 mm. broad, petioles 1-12 (3m. long; mature peduncles always exceeding the leaves, hirtellous at the base of the umbel; umbel 8-30-rayed, rays of each umbel subequal, 2-20 mm. long, fre- quently scabrous; involucre mostly absent (when present varying from one inconspicuous, short, linear-lanceolate bract to two prominent, linear-lancBolate bracts equalling the rays in length); involucel of several inconspicuous, linear, acute, bracts exceeding the white or yellow flowers ; fruit ovate t green rowly oblong, slightly constricted at the apex, 2-5 mm. long, 1-2 mm. broad, granular-scabrous, oil tubes mostly 3 in the intervals, 2-4 on the commissure, seed face plane. Type specimen: Nuttall, "Rocky Mountains," probably on the upper Platte River in Wyoming (probable type in the Herbarium of the New York Botanical Garden, cotypes in the Gray Herbarium of Harvard University, and the Herbarium of the Royal Botanic Gardens, Kew). Distribution : western South Dakota and Nebraska and adjacent w Wyoming and Colorado. Specimens examined: South Dakota: Elk Mti?., 5000 ft. alt., 6 July 1906, Cary 34 (US) ; Black Hills, near Ft. Meade, 5 May 1887, Forwood U6 (US) ; Fort Meade, 11 June 1887, Forwood I46 (M); sand hills and on Running Water, 12 Aug. 1S53-4, Hayden (M); Running Water, 14 Aug. 1853-4, Hayden (M); hills surrounding lake, Sylvan Lake, Black Hills, 9 Aug, 1926, Hayward 318 (NY); exposed 268 [Vol. 17 GARDEN hillsides, Deadwood, 19 June 1927, Hay ward 1161 (R); red foothills, Buffalo Gap, 27 June 1927, Hayward 15J^2 (R); Black Hawk, on shale hillsides, 10 May 1924, Lee (US 1244505); Spring Creek, Rockerville Road, 20 April 1925, Lee (R 106147); rocky ridges, Tilford, Elk Creek, Piedmont, 16 May 1924, Mcintosh 65 (R); head of City Creek, Deadwood, 5300 ft. alt., 10 July 1909, Murdoch 3538 (US); rim of Spearfish Canyon near Savoy, 5500 ft. alt., 21 June 1910, Murdoch 4128 (F, NY); rock crevices, Rapid Canyon, Pennington Co., 27 June 1914, Over 1642 (US); hillsides, Mayo, Custer Co., 15 June 1914, Over 1792 (US); high limestone ridges, Box Elder Creek, Black Hills, June 1887, W. S. Rushy (NY); Custer, 5500 ft. alt., 4 June 1892, Rydberg 718 (US); Sylvan Lake, 7000 ft. alt., 8 June 1892, Rydbcrg 718 (US); Hot Springs, 3500 ft. alt., 13 June 1892, Rydhcrg 718 (NY, US); Custer, 5500 ft. alt., 1 Aug. 1892, Ryd- berg 718 (NY); hills west of Rapid City, 1 Aug. 1908, Visher (F 244470); hillside, Indian Creek, Fall River Co., 1 Aug. 1911, Visher 2671 (F) ; on dry knolls in the Black Hills, Rockerville, June 1909, White (M); Hot Springs, 27 Aug. 1898, C. S. William' son (PA). Nebraska: Cheyenne Co., Aug. 1901, H. P. Baker (M); Pine Ridge, Dawes Co., 4600 ft. alt., 29 June 1889, C. E. Bessey (US); Dawes Co., July 1889, Bessey & Webber (CH 3G1700); on rocky habitats, Sioux Co., June- July 1927, Kramer 19 (M) ; hills. Banner Co., Aug. 1890, Rydberg 74 (NY); Court House Rock, Cheyenne Co., 4 July 1891, Rydberg 127 (US); hills south of Pumpkin Seed Valley, Scott's Bluff Co., 6 July 1891, Rydberg 127 (US); hills in Pleasant Valley, Scott's Bluff Co., 28 July 1891, Rydberg 127 (F, NY); high rocky ridges, Belmont, Dawes Co., July 1889, Webber (NY); Pine Ridge, 20 July 1889, Webber (M) ; high rocky prairies. Hot Creek Basin, Sioux Co., 2 Aug. 1889, Webber (US); War Bonnet Canon, 5000 ft. alt., 23 June 1890, T. A. Williams (US 750360); War Bonnet Canon, 23 June 1891, T. A. Williams (M); War Bonnet Canon, 5200 ft. alt., 23 June 1890, T. A. Williams 308 (US). Wyoming: Douglas, 1 June 1891, Bates (B); top of a flat desert hill, T. B. Ranch, Carbon Co., 20 June 1901, Goodding 47 (F, M, NY, R, US) ; barren ground, top of Como Bluff, Albany 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 269 Co., 26 May 1929, Goodman 349 (GO); Sheep Creek, 1 June Hatcher (C) 1923, Kemp 16 (NY) Lusk, 28 June 1896, Knoulton 108 (NY, US); along hillsides, Beaver Creek, Harper Ranch, near Newcastle, 24 Aug. 1929, Mathias 790 (M); Chimney Rock, 8 Aug. 1896, A. Nelson (R 7221); Laramie Hills, May-June 18— A. Nelson 72 (CH, US); Laramie Hills, 7 June 1894, A. Nelson 176 (B, M, MU, NY, PA, US, WSC); Horse Creek, Tune 1893, A. Nelson I404 (R); in rich soil in ravines, Pine Bluffs, Laramie Co., 14 May 1897, A. Nelson 2878 (US); stony slopes, Laramie Hills, Albany Co., 10 June 1897, A. Nelson 2878 (C, M, MU, NY, R, US 43302 in part); in dry gravelly .soil, Laramie Hills, Albany Co., 16 June 1897, A. Nelson 316t' (US); Laramie Hills, Albany Co., Aug. 1897, A. Nelson 3168a (US); rocky ledges and ravines, Laramie Hills, 7 June 1894. A. Nelson 3941 (176) (R); Laramie Hills, 25 June 1898, A. NeUon 4356 (F, R 12351, 12737 in part); Platte Hills, 11 July 1894, A. Nelson 4819 {391) (R); stony ridges, Laramie Hills, Albany Co., July 1903, A. Nelson 8977 (C, M, MU, NY, R); stony ridges, Orin Jc, May 1907, A. Nelson 9132 (R); Laramie Lills, Laramie, 4 June 1900, E. Nelson 216 (NY); Rocky Mts., Nuitall (NY type, G, K, M photo- graph); Rocky Mts., Nuitall (NY, PA); bad lands, Niobrara Co., 6 Sept. 1924, Over 16375 (R); rock crevices, Freezeout Hills, north of Medicine Bow, 21 June 1926, Payson & Payson 4808 (R); dry hills. Pine Bluffs, Laramie Co., 28 July 1889, Rodin (MU 164447) ; in a ditch east of Laramie, 7 June 1902, Sellon 29 (R); on the hills east of Laramie, 31 May 1902, Sellon 64 (R); Rocky Mountains, 1869, T nomas (US 215914). Colorado: Pawnee Butt(!s, Weld Co., 17 June 1919. Osterhout (R) EXCLUDED SPECIES Musineon alpinum Coult. & Rose, Bot. Gaz. 20: 260. 1895 = Tauschia alpina (Coult. & Rose) Mathias, new comb. "Musenium alpinum Coult. & Rose" ace. to Ind. Kew. Suppl 1: 284. 1906 = Tauschia alpina (Coult. & Rose) Mathias. Musineum Ehrenhergii Wc Iff, in Fedde, Repert. Nov. Sp. 8 : 524 1910 = Tauschia EhrenbergiL (Wolff) Mathias, new comb. Musineon alpinum Coult & Rose and Musineum Ehrenherah [Vol. 17 270 ANNALS Wolff are not congeners of the Musineon of Rafinesque. Their exact position is somewhat doubtful but the material at hand, including the type specimen of M. alpinum and a fragrant of the type of M. Ehrenbergii, would warrant their transfer to the genus Tauschia as defined in the broader sense by Macbride.*^ The genera included in the Tauschia complex have been so poorly defined and delimited that it seems best to treat them as members of a polymorphic genus until a detailed monographic study of the group can be made and more material obtained of some of the questionable species. Musenium GreeneiGrsiy, Proc. Am. Acad. 8 : 387. 1872 = Aletes acaulis (Torr.) Coult. & Rose, Rev. N. Am. Umbell. 126. 1888. Ill HARBOURIA Coult. & Rose Harbouria Coult. & Rose, Rev. N. Am. Umbell. 26, 125. 1888; Engler & Prantl, Nat. Pflanzenf. S^: 187. 1898; Coult- & Rose, Contr. U. S. Nat. Herb. 7: 92. 1900; Rydb. Fl. Cob 252. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 350- 1909; Daniels, Fl. Boulder, Col. 182. 1911; Clem. & Clein. Rocky Mt. Fl. 229. 1914; Rydb. Fl. Rocky Mts. 617. 1917, and ed. 2. 617. 1922; Wolff, Pflanzenreich 90: 71. 1927. Erect, branching, subcaespitose perennials with long slender tap-roots. Stems grooved, glabrous or pubescent. Leaves petiolate, thin, tri-quadri-pinnatisect; ultimate leaf divisions linear, distinct, acute, mucronulate; petioles slightly sheathing at the base. Inflorescence spreading, more or less scaberulent; peduncles terminal (usually in pairs), much exceeding the leaves; involucre usually absent; involucel bracts few, inconspicuous, lanceolate; calyx teeth inconspicuous; flowers yellow; stylopodium lacking. Carpophore entire. Fruit ovate, granular-roughened, flattened laterally, conspicuously constricted at the commissure, — w carpels appearing nearly distinct, subterete in cross-section with 4-6 distinct, subequal, corky ribs; oil tubes large, mostly solitary in the intervals, 1-3 on the commissural side ; strengthen- ing cells present or absent. Type species: Harbouria trachypleura (Gray) Coult. & Rose, Rev. N. Am. Umbell. 125. 1888. "Macbride, Contr. Gray Herb. N. S. 56: 28. 1918. 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 271 1. H. trachypleura (Gray) Coult. & Rose, Rev. N. Am. Umbell. 125. 1888; Nels. Yl Wyo. 115. 1896; Heller, Cat. N. Am. PI. 97. 1898, and ed. 2. 149. 1900; Coult. & Rose, Contr. U. S. Nat. Herb. 7: 93. 1900; Rydb. Fl. Col. 252. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 350. 1909; Daniels, Fl. Boulder, Col. 182. 1911 ; Rydb. Fl. Rocky Mts. 617. 1917, and ed. 2. 617. 1922; WoliF, Pflanzenreich 90: 71. 1927. PI. 27, fig. 2. Thaspiu7n trachypleurum Gray, Proc. Acad. Phila. 1863: 63. 1864; Port. & Coult. Syn. Fl. Col. 52. 1874. Cicuta (?) trachypleura (Gray) Wats. Bibl. Ind. 1: 417. 1878; Coult. Man. Bot. Rocky Mt. 116. 1885. '^Thaspium montanum, \ar. tenuiflorum. Gray, Am. Jour. Sci. 2. 33. 408." ace. to Wats. Bibl. Ind. 1: 417. 1878, err. typ. Thaspium montanum var. tenuifolium Gray, Am. Jour, Sci. II, 33: 408. 1862, as to Parry no. 159, not T. montanum var. tenuifolium Gray, Smiths. (]Jontr. 5: 65. 1853 [PI. Wright. 2: 65. 1853], namely WHght r,o. 1107. Plants caulescent, 0.8-5.5 dm. high; stems glabrous except for hirtellous pubescence at the base of the umbel; leaves mostly basal, ovate-oblong in general outline, excluding the petiole, 2.5-20 cm. long, 2-12 cm. broad, tri-quadri-pinnatisect, ultimate segments 2-30 mm. long, 0.5-1 mm. broad, petioles 0.8-20 cm. long, cauline leaves similar to the basal leaves; umbels 8-30- rayed, rays of each umbel Bubequal, 5-35 mm. long; involucre occasionally present in the form of one or two inconspicuous lanceolate bracts; involucel bracts linear, acute, usually shorter than the pedicels; fruit 3-6 mm. long, 3-4 mm. broad, ribs mostly obtuse in cross-section, oil tubes mostly solitary in the intervals, 1-3 on the commi;5sural side. Type specimen: Hall & Harbour 215 "On the mountains, at middle and lower elevatiiDns," Colorado, 1862 (type in the Gray Herbarium of Harvarcl University, cotypes in the herbaria of the Missouri Botanical ( warden, the United States National Museum, the Academy of Natural Sciences, Philadelphia, the Field Museum of Natural History, and the Brooklyn Botanic Garden) . Distribution: southern Wvomine to central New Mexico. (Vol. 17 272 ANNALS Specimens examined: Wyoming: Laramie Mts., east of Laramie, 18 June 1909, Cary 301 (US); aspen groves, Sheep Mountain, Albany Co., 2 Sept. 1903, Goodding 2100 (C, M, NY, R, US); along brook, near Chimney Rock, Laramie Plateau, 29 July 1929, Greenman & Greenman 6062 (M); Laramie Hills, May-June 1893, A. Nelson J^ (F, US); Pole Creek, hills near Table Mt., 2 June 1894, A, Nelson 160 (MU, PA); Table Mt., 2 June 1894, A. Nelson 160 (C, G, M, MU, NY, US); Laramie Hills, June 1893, A. Nelson 1399 (R); Laramie Hills, 13 June 1896, A. Nelson 19^8 (MU); Wood's Creek, 11 Aug. 1896, A. Nelson 2099 (R); hillsides. Medicine Bow Mts., Albany Co., 10 Aug. 1897, A. Nelson 3365 (R, US); Pole Creek, near Table Mt., 2 June 1894, A. Nelson {160) (R); Laramie Hills, 25 June 1898, A. Nelson 4356 (C, R 12737 in part); dry stony hillsides, Sand Creek, Albany Co., 1 June 1900, A. Nelson 7011 (C, G, M, MU, NY, R, US); dry, stony bottoms, Centennial, Albany Co., 27 July 1902, A. Nelson 8687 (F, G, M, NY, R, US); head of Pole Creek, 3 July 1901, E. Nelson 603 (G, NY); La Barge, Uinta Co., 17 July 1894, E. Stevenson 31 (US); Little Laramie River, Sheep Mts., 23 July 1897, P. A. Williams (NY). Colorado: foothills, w. Ft. Collins, 5500 ft. alt., 24 May 1890, C. F. Baker (F, M, N 16866, NY) ; foothills, 6 mi. w. Ft. Collins, 5500 ft. alt., 24 May 1896, C. F. Baker 1 (US); hills, Larimer Co., 5500 ft. alt., 12 ]\Iay 1895, C. F. Baker {Patterson 3988) (M) ; Log Caiion and Rist Canon, 74 mi, north of Denver on U. P. RR., 6000 7000 ft. alt., 31 May 1896, Baker & Holzingcr 4 (US); near Golden, 1887, Berthoud (CH 361696); Morrison, Oct. 1897, Bethel (US 329880); Morrison, June 1895, Bethel 36-^12 (US); Gore Mts., 9500 ft. alt., Aug. 1895, Bethel 36-13 (US); dry soil, Morrison, Jefferson Co., 1960 m. alt., 3 June 1921, Bethel & Clokey 4222 (F, M, MU, PA, R, US) ; Grand Canon of the Arkansas, 1873, T. S. Brandegee (M, PA); mts., Aug. 1871, Canhy (G); dry hills, Boulder, Boulder Co., 1635 m. alt., 19 June 1921, Clokey 4224 (R) ; Long's Peak Inn, Estes Park, 9000 ft. alt., 18 May 1908, Cooper 14 (R); Estes Park, 7500 ft. alt., 13 July 1904, Cooper 45 (R) ; Long's Peak Inn, Estes Park, 9000 ft. alt., 10 June 1908, Cooper 59 (R); Estes Park, 9000 ft. alt., 17 Aug. 1930] MATHIAS — STUDIES tN THE UMBELLIFERAE. Ill 273 1906, Cooper 156 (R); Estes Park, 7500 ft. alt., 13 July 1904, Cooper 261 (R); St. Vrain Canon, 26 May 1873, Coulter (PA, US); foothills, Larimer Co., 6000 ft. alt., 21 May 1892, 25 June 1895, Cowen (NY); Boswoith's Ranch, Stove Prairie, 20 Aug. 1895, Cowen (NY); foothills, Larimer Co., 6000 ft. alt., 9 May 1896, Cowen 27 (NY); Howe's Gulch, 1 May 1890, Cowen U02 (US); Bosworth's Ranch, Stove Prairie, 7500 ft. alt., 20 Aug. 1895, Cowen 1405 (NY); foothills, Ft. Collins, 6000 ft. alt., 9 May 1896, Cowen U07 (G, MU, NY, R) ; Rist Canon, 30 May 1891, Crandall (US 216531); foothills, 5500 ft. alt., 13 June 1896, Crandall (M) ; Pennock's, 13 June 1896, Crandall (NY) ; mountains, Larimer Co., 7500 ft. alt., 13 June 1896, Crandall (R 8401); foot- hills, 5500 ft. alt., 12 May 1893, CrandalU{V^) ; hillsides, foothills, 6500 ft. alt., 27 May 1893, Crandall 188 (US); Gregory Canon, 6400 ft. alt., 16 July 1906, Daniels 157 (M); St. Vrain Creek, 9 June 1906, Dodds 1890 (R); Central City, Oct. 1892, Eastwood (F, US); Genesee Mt., 27 May 1916, Eastwood 5^37 (US); rocky mountain tops, near Spring Creek, 8500-9000 ft. alt., 31 July 1874, G. Engelmann (M); rocky slopes at Idaho and Fall River, 26 Aug. 1874, G. Engelmann (M); lower mts., Greene 580 (G); Rocky Mountain Flora, Lat. 39°-4r, 1862, Hall & Harbour 215 (G TYPE, B, F, M, Vk, US); common, dry slopes. Ward, 9500 ft. alt., 25 June 1921, H. C. Hanson C 260 (M); infrequent among pines, upper mesas, near Boulder, about 5800 ft. alt,, 27 May 1921, H. C. Hanson C 263 (M); Estes Park, 22 June 1913, EL. Johnston 751 (]vl, NY); Estes Park, 22 June 1913, E. L. Johnston 895 (US); foothills near Golden, 6500 ft. alt., 20 June 1878, M. E. Jones 278 (C, NY); Penn Gulch, Boulder Co., 7700 ft. alt., July 1885, Letterman (F 348463); Penn's Gulch, 20 July 1885, Letterr.ian (G); Penn's Gulch, near Sunset, 30 July 1886, Letterman (M 776554, NY, PA); Howe's Gulch, 13 June 1899, Marshall (NY); Howe's Gulch, 13 June 1899, Mar- shall I4O8 (F, R, US); Masonville Road, west of Ft. Collins, 20 June 1929, Mathias 362 {M); Fern Lake Trail, Estes Park, 20 June 1929, Mathias 436 (M); Aspenglen Camp, Fall River Road, near Estes Park, 22 June 1929, Mathias 44^ (M); marshy field. Middle St. Vrain Canon, near Estes Park, 24 June 1929, Mathias 478 (M); Golden City, Au?. 1871, Meehan (PA); Estes Park, [Vol. 17 274 GARDEN 20-30 Aug. 1898, Moyer (MU 163846); mts., 1894, Osterhout (PA); near Estes Park, June 1894, Osterhout (MU 163845); foothills, Larimer Co., July 1895, Osterhout (G, PA, US 231800); from the head-waters of Clear Creek, and the alpine ridges lying east of "Middle Park," 1861, Parry 159 (G); Bear Creek Canon, 24 June 1929, Phelps (CAS 164795); Rocky IMts., near Idaho Springs, 2 Aug. 1872, Portei' (PA) ; near Boulder, 20 June 1900, Ramaley A. 84 (R) ; Nederland, 2500 m. alt., 1 July 1905, Ramaley 1136 (R); Smartweed Lake, near Rollinsville, 9 Aug. 1911, Ramaley 0020 (R); Boulder, 28 July 1913, Ramaley 9706 (R); near Ward, Boulder Co., about 9500 ft. alt., 18 July 1907, Rama- ley, Dodds & Robbins 3190 (R); Arkansas Canon, 22 July 1872, Redfield 470 (M, PA); Boulder, 2 Sept. 1895, Rydberg (NY); Clear Creek Caiion, 5 June 1893, Schneck (M); dry hills in Platte Canon, 4 July 1885, B. H. Smith (PA); Rocky Mountains, 1869, Thomas (G, US); Golden, 22 Sept. 1887, S, M. Tracy (CH 375310); mountains between Sunshine and Ward, Boulder Co., 8000-9500 ft. alt., Aug. 1902, Tweedy 4990 (NY, R); plains and foothills near Boulder, Boulder Co., 6000 ft. alt., July 1902, Tweedy 4991 (NY, R) ; mountains between Sunshine and Ward, Boulder Co., 8000-9500 ft. alt., Aug. 1902, Tweedy 4992 (NY, R); Empire, Clear Creek Co., 8500 ft. alt., 15-25 July 1903, Tweedy 5600 (NY, R); Rocky Mountains, Lat. 40°-4r, 1868, Vasey 225 (G); Boulder, Boulder Co., 25 May 1912, Vestal 370 (M). New Mexico: vicinity of Las Vegas, San Miguel Co., June 1920, Aned 157 (NY); vicinity of Las Vegas, San Miguel Co., Oct. 1919, Aned 279 (US); Las Vegas, Romersville, 2000 m. July 1926, Ars^ne & Benedict 15456 (US); Oscuro Mts July 1898, F. S. Earle 112 (M, NY); elevated rocky region E of M (G, M); Las Vegas June 1903, G. B. Grant 544 (F) ; brushy hillside, vicinity of Ute Park, Colfax Co., 2200-2900 m. alt., 20 Aug. 1916, Standley 13390 (US); oak thicket, vicinity of Ute Park, Colfax Co., 2200-2900 m. alt., 24 Aug. 1916, Standley 13687 (US). Utah: Gilbert's Meadow, Uinta Mts., 10,000 ft. alt., 28 July 1873, Porter (PA). 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 275 IV RHYSOPTEKUS Coult. & Rose Rhysopterus Coult. & Rose, Contr. U. S. Nat. Herb. 7: 185. 1900, in part; Tidestrom. Ccntr. U. S. Nat. Herb. 25: 398. 1925 [Fl. Utah & Nev.], as to nair.e only. Low, herbaceous, subprostrate, acaulescent (appearing caules- cent with the development of a pseudoscape), glabrous perennials with long, simple tap-root;?. Leaves petiolate, subcoriaceous, ternate, then subpinnate; leaf segments 1.5-3 cm. long, 2.5-4 cm. broad; petioles slightlj' sheathing, mostly scarious- winged throughout their length. Inflorescence spreading, peduncles shorter than to slightly exceeding the leaves; involucre absent; involucel conspicuous, dimidiate, foliaceous; calyx teeth conspicu- ous ; flowers white ; stylopodi jm lacking. Carpophore obsolete in the mature fruit. Fruit ovate, boat-shaped, flattened laterally with 7 conspicuous, equal, obtuse, corky ribs; oil tubes solitary in the intervals, 2 on the commissure, solitary in the apex of each rib; seed face concave. Type species: Rhysopterus plurijugus Coult. & Rose, Contr. U. S, Nat. Herb. 7: 186. 1900. 1. R. plurijugus Coult. & Rose, Contr. U. S. Nat. tierb. 7: 186. 1900. PL 22, fig. 5; pi. 27, fig. 1. Cymopterus plurijugus (Coult. & Rose) Jones, Contr. West. Bot. 12: 25. 1908. Plants, including the pseudoscape, 10-15 cm. high; pseudoscape 4-13 cm. long; leaves ovate- oblong in general outline, excluding the petiole, 1-4 cm. long, 1.5-5 cm. broad, ternate or biternate, then pinnatisect with the se|;ments confluent, giving the appear- ance of a lobed leaflet, ultiir.ate segments (lobes) 1-4 mm. long, about as broad, mucronulabe, petiole 0.5-3 cm. long; umbels several-rayed, rays stout, 5-14 mm. long, subequal in each umbel, reflexed in the fruiting stage, involucel bracts ovate-oblong, acute, mucronulate, frequently scarious-margined, exceeding the pedi- cels; fruit ovate to orbicular, 3-4 mm. long, 2-3 mm. broad, sometimes purplish-tinged, lateral and dorsal ribs sunilar, in the young fruit appearing wing-like; oil tubes solitary in the intervals, 2 on the commissure, solitar\ in the apex of each rib. (Vol. 17 276 ANNALS Type specimen: Leiherg SS40, ''black serpentine dykes around a peak rising from Malheur Valley about 12 km. west of con- fluence of Cottonwood Creek and Malheur/' near Harper Ranch, Malheur Valley, Oregon, 1000 m. alt., 10 June 189G (type in the United States National Herbarium, cotypes in the herbaria of the California Academy of Science, the Field Museum of Natural History, State College of Washington, and the University of California and the Gray Herbarium of Harvard University) . Distribution: Malheur Valley and adjacent regions, Oregon. Specimens examined: Oregon: Indian Creek, Malheur Co., June 189G, Leiherg (US 258913); loose soil on black volcanic dykes, Malheur Valley, near Harper Ranch, 1000 m. alt., 10 June 1896, Leiherg 22/^0 (US TYPE, CAL, CAS, F, G, WSC); dry sandy ground. Narrows, Harney Co., 4 July 1912, Peck 6W (F). EXCLUDED SPECIES R. corrvgatus (Jones) Coult. & Rose, Contr. U. S. Nat. Herb. 7: 187, 1900 = Cymopterus conugatus Jones, Am. Nat. 17: 973. 1883. R. Jonesii Coult. & Rose, Contr. U. S. Nat. Herb. 7: 186. 1900 = Cymopterus Coulteri (Jones) Mathias, new comb. V NEOPARRYA Mathias Neoparrya Mathias, Ann. Mo. Bot. Gard. 16: 393, pi 33. 1929 PL 22, fig The type species is Neoparrya lithophila Mathias, Ann. Mo. Bot. Gard. 16: 393, pi 33. 1929. The genus is known only from the type locality in northern New Mexico, It differs from Cymopterus in its spreading in- florescence with reflexed rays and in the development of small ribs and scattered oil tubes. VI ALETES Coult. & Rose Aletes Coult. & Rose, Rev. N. Am. Umbell. 27, 125. 1888; Engler & Prantl, Nat. Pflanzenf. S^: 193. 1898; Coult. & Rose, 1930] MATHIAS — STUDIES ]N THE UMBELLIFERAE. Ill 277 Contr. U. S. Nat. Herb. 7: l(i6. 1900; Rydb. Fl. Col. 252. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 351. 1909; Daniels, Fl. Boulder, Col. 182. lOll; Nels. Spring Fl. Intermt. States, 115. 1912; Clem. & Clem. Rocky Mt.Fl. 226. 1914; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 479. 1915 [Fl. New Mex.]; Rydb. Fl. Rocky Mts. 616. 1917, and ed. 2. 616. 1922; Wolff, Pflanzenreich 90: 140. 1927. Herbaceous, caespitose, acaulescent, glabrous or pubescent perennials, with long, more Dr less slender roots. Leaves petio- late, thin, simply to tri-pinnjitisect; ultimate leaf divisions 0.5-10 mm. long, 0.5-5 mm. wide; petioles slightly sheathing at the base. Inflorescence spreading, ped mcles shorter than or exceeding the leaves; involucre mostly absent; involucel bracts inconspicuous, linear to triangular; calyx teeth conspicuous; flowers yellow; stylopodium lacking. Carpophore persistent, 2-parted. Fruit ovate-oblong to oblong, subterete in cross-section, sometimes flattened laterally; ribs pnssent or obsolete, corky; oil tubes large, solitary in the intervals, 2 on the commissural side, some- times solitary in the apex of the ribs in cross-section, strengthen- ing cells inconspicuous; seed face plane or concave. Type species: Aletes acaulis (Torr.) Coult. & Rose, Rev. N. Am. Umbell. 126. 1888. KEY TO SPECIES A. Peduncles exceeding the leaves 9-27 cm. long; rays of the umbel 5-30 mm . long; ribs of the frut well-developed 1. A. acaulis AA. Peduncles shorter than the leaves, 1-4 cm. long; rays of the umbel 20-45 mm. long; ribs of the fruit obsolete S- A. humilis 1. A. acaulis (Torr.) Coult. & Rose, Rev. N. Am. Umbell. 126' 1888; Heller, Cat. N. Am. 1.^1. 95. 1898, and ed. 2. 150. 1900; Coult. & Rose, Contr. U. S. Nat. Herb. 7: 106. 1900; Rydb. Fl. Col. 252. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 352. 1909; Daniels, Fl. Boulder, Col. 183. 1911; Nels. Spring Fl. Intermt. States, 115. 1912; Clem. & Clem. Rocky Mt. Fl. 226. 1914; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 479. 1915 [Fl. New Mex.]; Rydb. Fl. Rocky Mts. 616. 1917, and ed. 2.616. 1922; Wolff, Pflanzeireich 90: 141. 1927. PI. 22, fig. 4; pi. 28, fig. 1. 278 [Vol. 17 GARDEN Deweya (?) acaulis Torr. Pacif. R. R. Kept. 4: 94. 1856; Bot. Whipple's Exp. 38 [94]. 1857. Oreosciadium acaule (Torr.) Gray, Proc. Am. Acad. 7: 343. 1867; Wats. Bibl. Ind. 1: 427. 1878. Seseli Hallii Gray, Proc. Am. Acad. 8: 288. 1870; Port. & Coult. Syn. Fl. Col. 52. 1874. Musenium Greenei Gray, Proc. Am. Acad. 8 : 387. 1872. Carum (?) Hallii (Gray) Wats. Bibl. Ind. 1 : 416. 1878; Coult. Man. Bot. Rocky Mt. 115. 1885. Zizia Hallii (Gray) Coult. & Rose, Bot. Gaz. 12: 137. 1887, in foot note. Aletes ohovata Rydb. Bull. Torr. Bot. Club 31: 573. 1904; Fl. Col. 252. 1906; Coult. &. Nels. Man. Bot. Cent. Rocky Mts. 352. 1909; Daniels, Fl. Boulder, Col. 182. 1911; Rydb. Fl. Rocky Mts. 616. 1917, and ed. 2. 616. 1922. Plants 0.5-3.5 dm. high; leaves oblong in general outline, excluding the petiole, 2-10 cm. long, 1-4 cm. broad, bi-tri- pinnatisect, ultimate segments linear to ovate, acute, confluent, usually appearing as teeth of obovate to cuneate leaflets of a simply pinnate leaf, sometimes puberulent on the veins and ciliate, 0.5-10 mm. long, 0.5-3 mm. broad, petiole 1.5-15 peduncles exceeding the leaves, 9-27 cm. long, occasionally pu- berulent at the base of the umbel, rays of the umbel 5-30 mm. long, stiff, spreading to reflexed, sometimes puberulent, involucre mostly absent, involucel bracts about equalling the flowers, membranous, narrowly triangular, subacuminate, sometimes ciliate; fruit oblong, 4-7 mm. long, about 2 mm. broad, lateral ribs short and obtuse in cross-section, corky, dorsal ribs 2-3, similar to the laterals, oil tubes solitary in the intervals, 2 on the commissure, accessory oil tubes solitary in the apex of each rib. Type specimen: Bigelow, "in crevices of rocks near Santa Anto- nita, New Mexico," Oct. 1853 (type in the New York Botanical Garden Herbarium, cotype in the Gray Herbarium of Harvard University) . Distribution: mountains of Colorado, south through New Mexico to western Texas. Specimens examined: Texas: rocks, Chisos Mts., Aug. 1883, Havard (US 156546); cm 19301 MATHIAS — STUDIES 1.)^ THE UMBELLIFERAE. Ill 279 in clefts and crevices of por]:hyritic cliffs, near Mt. Livermore, Davis Mountains, Jeff Davi^ Co., 2200 m. alt., 11 June 1926, E, J. Palmer 30773 (M, PA); Star Canyon, 11 April 1917, Young 68 (US). Colorado: Morrison, Oct. 1897, Bethel (US 329878); mts., Morrison, 6000 ft. alt., Maj 1895, Bethel 36-11 (US); dry soil, Mt. Morrison, Jefferson Co., 1960 m. alt., 3 June 1921, Bethel & Clokey 1^228 (F, M, NY, PA, R, US); dry hills, mts., Aug. 1871, Canly (G); Carhton (CH 352079); Golden, 7000 ft. alt., 30 April 1892, Crandall (NY, US 21f.825); gulch south of Boulder, 6000 ft. alt., 23 May 1896, Crandcll (M, NY, R); Clear Creek Caiion, above Golden, 6500 ft. alt., 30 April 1892, Crandall £54 (G); gulch south of Boulder, 23 May 1896, Crandall 1374 (NY, R); foot-hills, Boulder, 6000 ft. alt., 25 June 1906, Daniels 145 (M); near Morrison, 1891, Eastwood (US 44697) ; 1872, Greene (G, NY) ; near Morrison, 27 July 1889, Greene (N 21130, NY); shade, lower mts.. Golden City, <^:c., 1871, Greene 582 (G); E. Hall (CH 366096, US 44711, 7703^ 7) ; low mountains, Rocky Mountain Flora, Lat. 39°-41°, 1862, Hall & Harbour 221 (G, M) ; hillsides. Sunshine, 7500 ft. alt., 16 July 1920, H. C. Hanson C257 (M); in the aspen zone near Golden City, Clear Creek Canon, 5655 ft. alt., 20 July 1899, Holm (F 214725); Clear Creek Canon, near Golden City, 5650 ft. alt., 23 July 1899, Holm 95 (M); Boulder Co., Aug. 1892, M. Holzinger (US 44710); Golden, Lookout Mountain, 4 June 1916, E. L. Johnston 872h (NY, US); Ute Pass, Colorado Springs, 18 Vlay 1878, M. E. Jones 84, in part (NY); South Table Mt., Golden, 17 June 1896, Knowlton 76 (NY, US) ; Penn's Gulch, near Sunset, 30 July 1886, Letterman 35 (M) ; 188-, Mohr 27, in part (US) ; Lower Boulder Caiion, Boulder Co., 25 June 1901, Osterhout 2433 (NY, PA, R); Canyon of Thompson River, Larimer Co., 15 Aug. 1905, Osterhout 3097 (R); Boulder, 1908, Pace 271 (INI); Gold Hill, 1875, Patterson (G); Castle Rock, 6000 ft. alt., 1 July-9 Sept. 1885, Patterson 37 (CH, G, M, US) ; Gold Hill, Boulder Co., Aug. 1882, Patterson & Beaty (G, M, NY, US 1384121); Gregory Cafion, near Boulder, 8 June 1901, Ramaley 719 (R) ; Mont Alto, near Boulder, 2300 m. alt., 16 Aug. 1901, Ramaley 832 (R); Boulder Falls, 2200 m. alt., 2 July 1905, Ramaley 1217 (R); Sugar Loaf Mt., 14 July 1906, 280 ANNALS OF THE MISSOUEI BOTANICAL GARDEN rVoL. 17 Ramaley & Rohbins 1781 (R); ridge, Colorado Springs, 21 July 1909, Rohhins 6U7 (R); Glacier Lake, 23 Aug. 1912, H. H. Rushy (NY); Boulder, 2 Sept. 1895, Rydberg (NY); rocks, Boulder, Sept., 3 Oct. 1895, Shear 4738 (NY, US); Golden, 1887, S. M. Tracy (CH 375311); mountains between Sunshine and Ward, Boulder Co., 8000-9500 ft. alt., Aug. 1902, Tweedy 4986, 4987 (NY, R); Rocky Mountains, Lat. 40°-41°, 1868, Vasey 220 (M); Bear Creek, 12 miles west of Denver, 1868, Vasey 221 (G). New Mexico: crevices of rocks, San Antonita, 9 Oct. 1853, Bigelow (NY type, G); crevices of rocks, ridge between Las Huertas Canyon and Lagunita, Sandia Mts., 10 March, Ellis 396 (US); Socorro, 10 Aug. 1895, Plank (NY); Socorro, May 1881, Vasey (CH 375807, G, US 44676, 156545); Sandia Mount- ains, Sandoval Co., 4 Aug. 1910, Wooton (US 737550). This species has been referred to six genera, as may be seen from the synonymy, and stood under three of them at the same time. Torrey^'^ in 1856 doubtfully referred a plant collected by Dr. J. M. Bigelow '^In crevices of rocks near Santa Antonita, New Mexico," in October 1853, to the genus Deweya, describing it as a new species. This plant, now in the New York Botanical Garden Herbarium, must be taken as the historical type of the genus. Gray" in 1867 transferred the species to Oreosciadium. In 1870 a plant appeared in the Hall and Harbour collection and was described by Dr. Gray^^ ^s Seseli Hallii. In 1872 Gray'-^ described a plant, sent to him by Rev. E. L. Greene, as Musenium Greenei. These three plants have since proved to be conspecific. Thus the same species was known under three generic names at the same time, Oreosciadium, Seseli, and Musenium. Watson^^ in his 'Bibliographical Index' recognized the fact that Seseli Hallii and Musenium Greenei were conspecific and referred both doubtfully to Carum Hallii. He maintained Oreosciadium acaule as a distinct species. In 1887 Watson" recognized the identity of Oreosciadium acaule with his Carum Hallii, and for the first time '"Torrey, Pacif. R. R. Rept. 4: 94. 1856 "Gray, Proc. Am. Acad. 7: 343. 1867. " Gray, Proc. Am. Acad. 8 : 288. 1870. "/&tU387. 1872. "Watson, Bibl. Ind. 1: 416. 1878. " Watson, Proc. Am. Acad. 22: 475. 1887. 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 281 all the plants involved wen? regarded as conspecific. Coulter and Rose^® in 1887 transferred the species to Zizia. In 1888 they" recognized the species as constituting a distinct and un- described genus and so proposed the name Aletes. 2. A. humilis Coult. & Rose, Contr. U. S. Nat. Herb. 7: 107. 1900; Rydb. FL Col. 253. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 352. 1909; Rydb. Fl. Rocky Mts. 616. 1917, and ed. 2. 616. 1922; Wolff, Pflanzenreich 90: 141. 1927. PL 25, fig. 38; pi. 28, fig. 2. Plants 0.2-1 dm. high; leaves oblong in general outline, ex- cluding the petiole, 1.5-4 cm. long, 1-3.5 cm. broad, simply or bi-pinnatisect, ultimate segments linear to ovate-oblong, acute, confluent in the bipinnatisect leaves, appearing as lobes of a leaflet, sometimes puberulenb on the ribs and ciliate, 1-10 mm. long, 0.5-5 mm. broad, petiole 1-6 cm. long; peduncles shorter than the leaves, 1-4 cm. long, sometimes puberulent at the base of the umbel, rays of the umbel 20-45 mm. long, slender, lax, spreading, mostly glabrous, involucre absent, involucel bracts about equalling the flowers, linear, acute, sometimes ciliate; fruit ovate-oblong, 3-4 mm. long, about 2 mm. broad, ribs much re- duced, in cross-section inconspicuous on the carpel surface, oil tubes solitary in the intervals, 2 on the commissure, accessory oil tubes absent. Type specimen: Osterhout 6, Dale Creek, Larimer Co., Colorado, 19 July 1899 (type in the United States National Herbarium, cotype in the New York Botanical Garden Herbarium). Distribution: known only from the region of Dale Creek, Larimer Co., Colorado. Specimens examined: Colorado: stony cliffs, Colo.-Wyo. line, April 1902, Goodding 8850 (R) ; Dale Creek, Larimer Co., 19 July 1899, Osterhout (US 361576); Dale Creek, Larimer Co., 30 May 1901, Osterhout (PA 516511, US 735613); Dale Creek, Larimer Co., 19 July 1899, Osterhout 6 (US type, NY); on cliffs, Dale Creek, Larimer Co., 20 July 1898, Osterhout 22 QsY) ; Dale Creek, Larimer Co., 19-20 *" Coulter and Rose, Bot. Gaz. 12: 137, 1887, in footnote. " Coulter and Rose, Rev. N. Am. Umbell. 126. 1888. 282 ANNALS or THE fVoL. 17 July 1899, Osierlwut 2002 (G, R); Dale Creek, Larimer Co., 24 May 1912, Osterhout 4678 (NY, R). EXCLUDED SPECIES A. (?) Davidsoni Coult. & Rose, Contr. U. S. Nat. Herb. 7: 107. 1900 = Pseudocymopterus Davidsoni (Coult. & Rose) Mathias, new comb. A. (?) MacDougali Coult. & Rose, Contr. U. S. Nat. Herb. 7: 107. 1900 = Oreoxis MacDougali (Coult. & Rose) Rydb. Bull. Torr. Bot. Club 40: 68. 1913. A. (?) tenuifolia Coult. & Rose, Contr. U. S. Nat. Herb. 7: 108. 1900 = Musineon lineare (Rydb.) Mathias, Ann. Mo. Bot. Card. 17: 265. 1930. VII OREOXIS Raf. The genus Oreoxis was described by Rafinesque-^' in 1830 and based on a plant collected by Dr. Edwin P. James in the ''Rocky Mts." in the summer of 1820. Torrey*^ in the account of the James collections recognized and described this plant as a new "Um- bellifera" but failed to eive it a eeneric name. Rafinesque the name Oreoxis to ''le genre Anonymus, n. 179, de Torrey." The number, 179, does not refer to the James plant; but it is the serial number in Torrey's list of the James collection and refers to Heradeum Sphondylium in the list. The descrijition of the new " Umbellif era " follows in an observation and no number is given to it. The type specimen in the New York Botanical Garden Herbarium was examined in connection with this study. It is doubtfully referred to '' Musenium divaricatu7n ," and the herba- rium label has a Latin description of the plant, which, however, does not pertain to M. divaricatuvi but to Oreoxis humilis. The generic name Oreoxis was first mentioned in a letter sent « by Rafinesque to de Candolle; this letter was published in Ser- 4 "Rafinesque, in Scringe, Bull. Bot. 1: 217. 1830. "Torrey, in "Some Account of a Collection of Plants made during a journey to and from the Rocky Mountains in the summer of 1820, by Edwin P. James, M.D, Assistant Surgeon U. S. Army .... Read December 11, 1S26," Ann. Lye. Nat. Hist. N. Y. 2: 207. 1828, gave an excellent description of an umbelliferous plant in an observation following no- 179, Heradeum Sphondylium, but unfortunately failed to suggest a generic name. 1930] MATHIAS — STUDIES IN THE IJMBELLIFERAE. Ill 283 inge's 'Bulletin Botanique' in Geneva in 1830. Rafinesque also mentioned the genera Ador'um, Lomatium, and Cymopterus, and stated: "Ces quatre genres ;5ont 6tablis dans sa Flora Mandanen- sis." The Tlorula MandanSf. Williamson (PA). 3. O. MacDougali (Coult. & Rose) Rydb. Bull. Torr. Bot. Club 40:68. 1913; Fl. Rocky Mts. 617. 1917, and ed. 2. 617. 1922; Tidestrom, Contr. U. S. Nat. Herb. 25: 394. 1925 [Fl. Utah & Nev.]; Wolff, Pflanzenreich 90: 142. 1927. PL 30, fig, 1. Aletes {?) MacDougali Coult. & Rose, Contr. U. S. Nat. Herb. 7: 107. 1900. Plants 8-20 cm. high, glabrous; leaves oblong in general outline, excluding the petiole, 0.5-S cm. long, 0.5-3 cm. broad, mostly simply pinnate, occasionaLy bipinnatisect, ultimate segments distinct, linear, acute (in bipinnatisect leaves somewhat con- fluent, appearing as cuneste to obovate, lobed leaflets), 2-30 mm. long, 1-4 mm. broad, petioles 1-11 cm. long; peduncles exceeding the leaves, 5-15 cm. long, umbels several-rayed, rays 3-17 mm. long, involucre al^sent, involucel bracts several, linear, about equalling the yellow flowers; fruit oblong, 4-8 mm. long, 1-5 mm. broad, wings in crDss-section broadly linear, rounded or acute, oil tubes solitary in the intervals, 2 on the conmiissure. Type specimen: MacDouc;al 192, "on Berry's trail to the Grand Canyon of the Colorado, Arizona," 7000 ft. alt., 28 June 1898 (type in the United States National Herbarium, cotypes in the herbaria of the Brooklyn Botanic Garden, Field Museum of Natural History, and the New York Botanical Garden and the Gray Herbarium of Harvard University, photograph of thetype in the Missouri Botanical Garden Herbarium). Distribution: canyons and mesas of southeastern Utah, adja- cent Arizona and Colorado. -I * a [Vol.17 290 ANNALS OF THE MISSOURI BOTANICAL GARDEN Specimens examined : Colorado: between top of clifif and Needle's Eye on way to Square Tower House, Mesa Verde National Park, 8 July 1929, Mathias 637 (M) ; along cliffs below camp ground, Spruce Canyon, Mesa Verde National Park, 8 July 1929, Mathias 650 (M); on cliff on trail to pictographs on west-facing slope near mouth of Spruce Canon, Mesa Verde National Park, 6800 ft. alt., 16 June 1925, Schmoll 1749 (R). Utah: Courthouse Wash, 16 June 1913, M. E. Jones (P 83107) ; Armstrong and White Canyons, near the Natural Bridges, 1600-1800 m. alt., 4-6 Aug. 1911, Rydberg & Garrett 9456 (NY); Armstrong and White Canyons, near the Natural Bridges, 1600-1800 m. alt., 4-6 Aug. 1911, Rijdberg & Garrett 9458 (NY, US). Arizona: Berry's Trail, Grand Canyon, 7000 ft. alt., 28 June 1898, MacDougal 192 (US type, B, F, G, M photograph, NY). This species, originally referred questionably to Aletes Coulter and Rose, is undoubtedly better placed in Oreoxis because of the fruit structure. Mature well-developed fruit is apparently rarely produced, a condition occurring in certain other species of the family. However, the mature fruit when produced is that typical for the genus Oreoxis in the corky development of the wings, oil- tube number, position, etc. The type specimen, MacDougal no. 192, shows a slight ten- dency toward a caulescent condition in certain cases, but this tendency is apparently a rare development and does not appear in the other material. Mathias no. 660 is a much attenuated form of the species with conspicuously elongated leaf segments but intermediates occur in the same colony. Further collections from the mesas and canyons of northeastern Arizona and adjacent Utah, Colorado, and New Mexico will doubtless extend the range of this species and give it a more con- tinuous distribution than that now known. 4. O. Bakeri Coult. & Rose, Contr. U. S. Nat. Herb. 7: 144. 1900; Rydb. Fl. Col. 252. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 354. 1909; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 480. 1915 [Fl. New Mex.], excluding no. 4318; Rydb. 1930] MATHIAS — STUDIES tN THE UMBELLIFEKAE. Ill 291 Fl. Rocky Mts. 617. 1917, and ed. 2. 617. 1922; Tidestrom Contr. U. S. Nat. Herb. 25: 394. 1925 [Fl. Utah & Nev.]. PI. 29, fig. 2 Cymopterus Bakeri (Coult. & Rose) Jones, Contr. West. Bot 12: 28. 1908. Plants 1-12.5 cm. high, j;labrous except for a slight puberu- lence at the base of the um])el and on the rays; leaves narrowl} oblong in general outline, excluding the petiole; 0.5-5 cm. long 0.3-2 cm. broad, mostly bip innatisect, ultimate segments linear sometimes mucronulaf e, more or less distinct, 1- nim. broad, petioles 0.5-5 cm. long; peduncles mm ceeding the leaves, 1.5-11 cm. long, umbels several-rayed, rays 3-5 mm. long, involucre nostly absent, involucel bracts con- spicuous, about equalling th3 yellow or whitish flowers, obovate, mostly 3-toothed (varying fi'om 2 to 5) at the apex, greenish or purplish- tinged; fruit ovate-oblong, 3-4 mm. long, 2-3.5 mm. broad, usually pur] lish-tinj;ed, wings in cross-section mostly linear, rounded to subacute at the apex, oil tubes 3-4 in the intervals, 5-10 on the commissure. Type specimen: C. F. Baker 12, frequent on high, bald summits, mts. near Pagosa Peak, Colorado, 23 Aug. 1899 (type in the United States National Herbarium). Distribution: high mountains of southern Colorado and adja- cent New Mexico and Utah. Specimens examined: Colorado: frequent on high, bald summits, mts. near Pagosa Peak, 23 Aug. 1899, C. F. Baker 12 (US type) ; near Pagosa Peak, 12,000 ft. alt., Aug. 1899, C. F. Baker 505 (G, M, N, NY, R); Mount Ouray, 12,000 ft. alt , 20 Aug. 1901, C. F. Baker 856 (G, M, MU, N, NY, US); Mt. Hayden, 13,000 ft. alt., 14 July 1898, Baker, Earle & Tracy 577 (NY, R, US) ; alpine summits of Sangre de Xto Range, Aug. 1874, T. S. Brandegee 955 (M) ; among rocks, near Trout Lake, San MigU'sl Co., 12,200 ft. alt., 21 Aug. 1924, Payson & Payson 4206 (G, B,) ; West Spanish Peak, 3000-3800 m. alt., 9 July 1900, Rydherg (fe Vreeland 5809, 5810 (NY); West Spanish Peak, 2800-3000 m. alt., 6 July 1900, Rydherg & Vreeland 5811 (NY); Cumberland Basin, La Plata Mountains, 12,000 ft. alt., July 1896, Tweedy 93 (US); La Plata Mts., 12,000 ft. alt., 15 July 1896, Tweedy 493 (US). [Vol. 17 292 ANNALS OF THJ: MISSOURI BOTANICAL GARDEN New Mexico: Lake Peak, vicinity of Santa Fe, 3700 m. alt., 12 Aug. 1926, Arsene & Benedict 16135 (US) ; top of Pecos Baldy, 12,000 ft. alt., 31 July 1903, Bailey 631 (US); top of Las Vegas Range, above Sapello Creek, 11,000 ft. alt., T. D. A. Cockerell (US G60065) ; top of Las Vegas Range, 11,000 ft. alt., end of June 1901, T. D. A. Cockerell 25 (R); above timber line, Baldy Peak, Colfax Co., about 3600 m. alt., 4 Sept. 1916, Standley 14341 (US). Utah: exposed alpine meadow, Mt. Tomasaki, La Sal Moun- tains, 12,500 ft. alt., 14 June 1914, E. B. Payson 414 (G, M, R); La Sal Mountains, 3300-3600 m. alt., 7 July 1911, Rydberg & Garrett 8694, 8697, 8698 (NY); La Sal Mountains, 3300-3600 m. alt., 7 July 1911, Rydherg & Garrett 8695 (G, M, MU, NY); West Mt. Peal, La Sal Mountains, 3300 m. alt., 10 July 1911, Rydherg & Garrett 8790 (NY) ; Gold Basin, La Sal Mountains, 3000-3300 m. alt., 11 July 1911, R^jdherg & Garrett 8840 (NY); La Sal Mountains, near Mount Peal, 3300-3700 m. alt., 17 July 1911, Rydberg & Garrett 9016 (NY); La Sal Mountains, near Mount Peal, 3300-3700 m. alt., 17 July 1911, Rydherg & Garrett 9017 (NY, R, US). VIII COGSWELLIA Spreng. Cogswellia Spreng. in Linn. Syst. Veg., ed. Roem. & Schult., 6: XLViii. 1820; Britt. & Brown, 111. Fl. 2: 631. 1913; Piper & Beattie, Fl. N. W. Coast, 260. 1915; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 484. 1915 [Fl. New Mex.]; Rydb. Fl. Rocky Mts. 624. 1917, and ed. 2. 624. 1922. Peucedanum of American authors, non [Tourn.] L. ; Nutt. Gen. 1: 181. 1818; DC. Prodr. 4: 176. 1830, as to Am. spp.; Torr. & Gray, Fl. N. Am. 1: 625. 1840; Walp. Rep. Bot. Syst. 2: 409. 1843, as to Am. spp.; Benth. & Hook. Gen. PI. 1: 918. 1867, as to Am. spp. in part; Wats. Bibl. Ind. 1 : 427. 1878; Coult. Man. Bot. Rocky Mt. 119. 1885; Coult. & Rose, Rev. N. Am. Umbell. 20, 53. 1888; Coult. Contr. U. S. Nat. Herb. 2: 142. 1891; Howell, Fl. N. W. Am. 1 : 251. 1898; Drude in Engler & Prantl, Nat. Pflanzenf. 3^: 234. 1898, as to Lomatium Raf. Jour. Phys. 89: 101 Am R Linn. Soc. Bot. 10 : 199. 1810 ; Coult. & Rose. Contr. U 1930] MATHIAS — STUDIES "N THE UMBELLIFEKAE. Ill 293 S. Nat. Herb. 7:204. 1900; Piper, Contr.U. S.Nat. Herb. 11:419. 1906 [Fl. Wash.]; Gray, Maa. ed. 7, 619. 1908; Jeps. Man. Fl. PI. Calif. 719. 1925. ''Cogswellia Raf." ace. to Coult. & Nels. Man. Bot. Cent. Rocky Mts. 362. 1909; Tidestrom, Contr. U. S. Nat. Herb. 25: 400. 1925 [Fl. Utah & NevJ. The type species of the genus is Cogswellia villosa (Raf.) Spreng. in Linn. Syst. Veg., ed. Roein. & Schult., 6: xlviii. 1820, which equals C. foeniculacea (Nutt,) Coult. & Rose, Contr. U. S. Nat. Herb. 12:449. 1909. Synonyms : Ferula foeniculacea Nutt. Gen. 1: 183. 1818. Lomatium villosum Raf. Jour. Phys. 89: 101. 1819. Pastinaca foeniculacea (Nitt.) Spreng. in Linn. Syst. Veg., ed. Roem. & Schult., 6: 587. 1820. Cogswellia villosa (Raf.) Schult. in Linn. Syst. Veg., ed. Roem. & Schult., 6 : 588. 1820. , Lomatium puhescens Raf. in Seringe, Bull. Bot. 1: 216. 1830; Raf. Atlantic Jour., extra ol' number 6: 40. 1833. Peucedanum foeniculaceur i Nutt. in Torr. & Gray, Fl. N. Am. 1:627. 1840. Lomatium foeniculaceum (Nutt.) Coult. & Rose, Contr. U. S. Nat. Herb. 7: 222. 1900. The historical type on which Cogswellia was based is a plant collected by Bradbury on the Missouri and named Lomatium mllosum by Rafinesque. Bradbury and Nuttall collected together in the upper Missouri in the region which is now Nebraska and South Dakota. Rafinesque described some of Bradbury's plants, among them Ferula foeniculacea. These two have since been regarded as conspecific, and a study of the historical material at hand substantiates this view. The Nuttall plant which was col- lected "On the high plaimi of the Missouri, commencing about the confluence of the river Jauke" [James River, South Dakota] is said by Coulter and Rose not to be "in Herb. Philad. Acad., and possibly lost." However, herbarium specimens from the Philadelphia Academy of Natural Sciences which were labeled "Cymopterus glomeratus Raf. {Thapsia Nutt.)," collected in Louisiana by Bradbury and Nuttall, and supposedly a part of the (Vol. 17 294 ANNALS OF THE MISSOURI BOTANICAL GARDEN tyj-e material of Cymopterus, were kindly loaned for this study, and upon critical examination proved not to be Cymopterus but correspond well with the original characterization of Lomatium and probably represent the missing type of that genus. A careful examination of a large amount of widely distributed material of European and American plants previously referred to Peucedanum substantiates conclusions reached by various stu- dents of the Umbelliferae that the American plants differ from the true Peucedanum of Europe in several important morphologi- cal characters, particularly in the absence of a stylopodium, in the usually single umbels terminating simple elongated peduncles, and in the number and disposition of the oil tubes in the fruit. The name Cogswellia was proposed by Sprengel in 1820 for the Lomatium of Rafinesque, as that name was antedated by a Lomatia of Robert Brown. Recent workers have used both Cogswellia and Lomatiuvi for these species which are congeneric with the Peucedanum of American authors, but not of Linnaeus. The name Lomatium was reinstated in 1918 by Macbride" who justified its use by quoting Article 57 of the International Rules of Botanical Nomenclature, which reads in part: "When the difference between two names, especially two generic names, lies in the termination, these names are to be regarded as distinct even though differing by one letter only." However, it is felt that this is not a sufficiently valid reason for retaining both Lomatium and Lomatia and that such retention would be a source of permanent confusion. Article 51, part 4, covers this case by saying that every one should refuse to admit a name "when it becomes a permanent source of confusion and error." The genus Cogswellia is, in the Grayian sense of the word, morphi and of wide distribution in the western part of North America, extending fron Saskatchewan and British Columbia to Mexico, and from western Missouri to California. The most important diagnostic character in distinguishing the genus from other genera under consideration is in the fruit which is ovate-oblong, varying in length, glabrous or pubescent, flattened dorsally, and distinctly winged on the dorsal wings only. The •1 Macbride, Contr. Gray Herb. N. S. 53: 15. 1918. 1930) MATHIAS — STUDIES tN THE UMBELLIFERAE. Ill 295 inflorescence is spreading and the peduncles usually much exceed the leaves. Representative species o\' the genus are Cogswellia ambigua (Nutt.) Jones, C. dasycar^a (Torr. & Gray) Jones, C. Grayi Coult. & Rose, C macrocarpa (Nutt.) Jones, C. nudicaulis (Pursh) Jones, C. tnternaii (Pursh) Jones, and C. utriculata (Nutt.) Jones. IX PSEUDOCYMOPTERUS Coult. & Rose Pseudocymopterus Coult. & Rose, Rev. N. Am. Umbell. 20, 74. 1888; Howell, Fl. N. W. Am 1 : 258. 1898; Engler & Prantl, Nat. Pflanzenf. 3*: 222. 1898; Coult. & Rose, Contr. U. S. Nat. Herb. 7: 187. 1900; Rydb. Fl. Col. 256. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 360. 1909; Nels. Spring Fl. Intermt. States, 119. 1912; Clem. & Clem. Rocky Mt. Fl. 235. 1914; Wooton & Standi. Ccntr. U. S. Nat. Herb. 19: 482. 1915 [Fl. New Mex.]; Rydb. Fl. Rocky Mts. 622. 1917, and ed. 2. 622. 1922; Tidestrom, Co:itr. U. S. Nat. Herb. 25: 399. 1925 [Fl. Utah & Nev.]. Pseudopteryxia Rydb. Bull. Torr. Bot. Club 40: 71. 1913; Fl. Rogky Mts. 623. 1917, and ed. 2. 623. 1922. Pseudoreoxis Rydb. Bull. Torr. Bot. Club 40: 73. 1913; Fl. Rocky Mts. 621. 1917, and ed. 2. 621. 1922. Herbaceous, acaulescent or caulescent, caespitose, glabrous or pubescent perennials from long slender tap-roots. Leaves petio- late, thin, simply to tri-pinnatisect; ultimate leaf divisions 0.1-11 cm. long, narrow; petioles somewhat sheathing. Inflorescence subglobose to spreading, peduncles exceeding the leaves; in- volucre mostly absent, involucel present, mostly dimidiate; flowers white, purple or yoUow; calyx teeth conspicuous; stylo- podium lacking. Fruit narrowly oblong to ovate-oblong, flat- tened dorsally or subtereto; lateral wings present; dorsal wings similar to the laterals or absent through abortion; wings mostly thin, sublinear in cross-section; oil tubes 1-8 in the intervals, 2-8 on the commissure, sometimes present in the wing base; strength- ening cells absent or present; seed face slightly concave. Type species: Pseudocymopterus montanus (Gray) Coult. & Rose. Rev. N. Am. Umboll. 74. 1888. 296 (Vol. 17 GARDEN KEY TO SPECIES A. Plants caulescent. B. Stems few-leaved; fruit ovate to ovate-oblong, 3-7 mm. long, 2-4 mm. broad :f. p. montanus BB. Stems mostly many-leaved; fruit oblong, 3-4 mm. long, 1-2 mm. broad f. P. Davidsoni AA. Plants acaulescent. C. Leaves glabrous except for an occasional pubescence on the veins and nodes. D. Ultimate leaf segments mostly appearing as lobes of a subcuneate leaflet; rays of the umbel conspicuously unequal; dorsal wings of the fruit aborted 5. P. anisatus DD. Ultimate leaf segments distinct, mostly Hnear; rays of the umbel subequal; dorsal wings of the fruit usually developed. F. Involucel bracts prominent, mostly exceeding the yellow flowers; calyx teeth conspicuous. 4. P. Hendersoni FF. Involucel bracts not prominent; calyx teeth not conspicuous. G. Umbels few-rayed, mature rays less than 5 ram. long. 5. P. huynholdtensis GG. Umbels several-rayed, mature rays over 5 mm. long. ./. P. montanut CO. Leaves rough-puberulent. E. Leaves bi-tri-pinnatisect, ultimate segments crowded; rays of the umbel 1-17 mm. long 6. P. hipinnatu% EE. Leaves simply or bi-pinnatisect, ultimate segments not crowded; rays of the umbel less than 5 mm. long 7. P. nivalis !• P. mcntanus (Gray) Coult. & Rose, Rev. N. Am. UmbelL 74. 1888; Heller, Cat. N. Am, PL 98. 1898, and ed.*2. 151. 1900; Coult. & Rose, Contr. U. S. Nat. Herb. 7: 188. 1900; Rydb. Fl. Col. 256. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 361. 1909; Clem. & Clem. Rocky Mt. Fl. 235. 1914; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 483. 1915 [Fl. New Mex.l; Rydb. Fl. Rocky Mts. 622. 1917, and ed. 2. 622. 1922; Tidestrom, Contr. U. S. Nat. Herb. 25: 399. 1925 [Fl. Utah & Nev.]. PL 25, figs. 4-13; pis. 32-33. Thaspium (?) montanum Gray, Mem. Am. Acad. N. S. 4: 57. 1849 [PL FendL]; Torr. Bot, Whipple^s Exp, 38. 1857. T. (?) montanum var. {?) tenuifolium Gray, Smiths. Contr. 5: 65. 1853 [PL Wright. 2: 65. 1853]. Ligusticum montanum (Gray) Benth. in Benth. & Hook. Gen. PL 1: 912, 914. 1867; Port. & Coult. Syn. FL Col. 51. 1874; Rothr. Rept. U. S. Geogr. Surv. Wheeler 6: 134. 1878 [PL Wheeler.]; Wats. Bibl. Ind. 1: 426. 1878; Coult. Man. Bot. Rocky Mt. 117. 1885. 19S0] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 297 Peucedanum Lemmoni Coult. & Rose, Bot. Gaz. 14: 277. 1889; Heller, Cat. N. Am. PI. 98. 1898, and ed. 2. 152. 1900. Ligusticum monianum var. tenuijolium (Gray) Wats. Blbl. Ind. 1 : 426. 1878. Pseudocymopierus montanus var. tenuijolius (Gray) Coult. & Rose, Rev. N. Am. Umbell. 75. 1888; Contr. U. S. Nat. Herb. 7: 188. 1900. P. montanus var. purpurevs Coult. & Rose, Rev. N. Am. Umbell. 75. 1888; Heller, Cat. N. Am. PI. 98. 1898, and ed. 2. 151. 1900; Coult & Rose, Contr, U. S. Nat. Herb. 7: 189. 1900. "P. montanus var. tenuijolius (Wats.) Coult. & Rose" ace. to Heller, Cat. N. Am. PI. ^'8. 1898, and ed. 2. 151. 1900. Lomaiium Lemmoni Coult. & Rose, Contr. U. S. Nat. Herb. 7: 231. 1900. Pseudocymopierus sylvaticus Nels. Bull. Torr. Bot. Club 28: 224. 1901; Rydb. Fl. Co!.. 257. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 361. 1909; Daniels, Fi. Boulder, Col. 185. 1911; Rydb. Fl. Rocky Mts. 623. 1917, and ed. 2. 623. 1922. P. montanus var. multifidus Rydb. Bull. Torr. Bot. Club 31: 574. 1904. P. multifidus Rydb. Bull. Torr. Bot, Club 33: 147. 1906; Fl. Col. 257. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 361. 1909; Daniels, m. Boulder, Col. 185. 1911; Wooton & Standi. Contr. U. S. Nat. Herb. 19: 483. 1915 [Fl. New Mex.]; Rydb. Fl. Rocky Mts. 623. 1917, and ed. 2. 623. 1922. P, purpureus (Coult. & I^ose) Rydb. Bull. Torr. Bot. Club 33 : 147. 1906; Fl. Col. 257. 1906; Coult. & Nels. Man. Bot. Cent. Rocky Mts. 361. 1909; ^V^ooton & Standi. Contr. U. S. Nat. Herb. 19: 482. 1915 [Fl. ^ ew Mex.]; Rydb. Fl. Rocky Mts. 623. 1917, and ed. 2. 623. 192^1; Tidestrom, Contr. U. S. Nat. Herb. 25: 399. 1925 [Fl. Utah root crowned with persistent leaf bases, 1-3.5 dm. high, glabrous; leaves narrowly oblong in general outline, excluding the petiole, 3.5-14 cm. long, 1.5-3.5 cm. broad, bipinnatisect, ultimate segments rigid, acute, mostly confluent, appearing as lobes of a subcuneate leaflet, 1-G mm. long, 1-1.5 mm. broad, petiole 3 11 cm. long; peduncles usually exceeding the leaves, umbels 6~9-rayed, rays of each umbel conspicuously unequal, 0.5-6 cm. long, involucre absent, involucel bracts conspicuous, much exceeding the yellow flowers, 0.3-1.5 cm. long, linear-lanceolate, entire; fruit narrowly oblong, 4-6 mm. long, 2-3 nmi. broad, calyx teeth persistent, conspicuous, lateral wings present, linear or sublinear in cross- section, acute at the apex, dorsal wings abortive, appearing as minute projections on the dorsal surface, oil tubes 1-3 in the intervals, 2-4 on the commissure, strengthening cells absent or present. Type specimen : Hall & Harbour 222 "drj' hills in the middle mountains," Rocky Mountains, Colorado, Lat. 39°-41°, 18G2 (type in the Gray Herbarium of Harvard University, cotypes in the herbaria of the Field Museum of Natural History, the Missouri Botanical Garden, the Academy of Natural Sciences, Philadelphia, and the United States National Museum). Distribution: Colorado, in the region of El Paso County. Specimens examined: Colorado : north of Cheyenne Canon, neighborhood of Pike's Peak, 6 July 1895, E. A. Bessey (NY) ; Minnehaha, 8600 ft. alt., 31 July 1895, E. A. Bessey (NY); Greenhorn Mts., 1874, T. S. Brandcgee 954 (M); mts., Aug. 1871, Canhy (G); dry ground. Bear Creek, Brookvale, Clear Creek Co., 13 June 1918, Churchill (G, M 900325); golf Imk, Manitou, 1900, F. Clements (NY); Ruxton, 1896, F. Clernents 211 (NY); Manitou, 19 Aug. 1900, F, C. Clements (US 370000); Minnehaha, 2600 m. alt., 16 June- 22 July 1901, Clements & Clements 94 (G, M, NY, US); Chiann Mt., July 1892, Eastwood 6136 (NY) ; Pike's Peak, July, Aug. 1901, Fossler (M); Lake Ranch, French (G); on gravel slide, Engelmann Canon, Pike's Peak, 8500 ft. alt., 8 July 1921, H. M. 10301 MATHIAS — STUDIES ::N THE UMBELLIFERAE. Ill 319 Hall 11400 (US); Rocky Mi;. Flora, Lat.'39°-41°, 1862, Hall & Harbour 222 (G type, F, M, PA, US) ; Cascade Canon, Aug. 1900, Harver & Harper (M ^i69893); Ute Pass, Colorado Springs, 18 May 1878, M. E. Jones H in part (NY); Ute Pass, Colorado Springs, 29 May 1878, M. E. Jones 119 (NY); Peter's Dome, Petrified Stump, El Paso Cd., 3 Aug. 1884, Leiterman 220 (M, US); slides, Pike's Peak, 14 Aug. 1913, Macbride 2674 (M); near Golden City, Aug. 1871, Meihan (PA); South Cheyenne Canyon, Colorado Springs, 1800-2000 m. alt., 10 May 1900, Rydherg & Vreeland 5815 (NY, US); Rocky Mts., 1869, Scovell (US 44727); below Half Way House, Pilce's Peak, 3 July 1896, Shear 3709 (NY, US); gravelly slopes. North Cheyenne Canon, El Paso Co., 6 Aug. 1892, Sheldon 105 (US); gravelly slopes. North Cheyenne Canon, El Paso Co., 8500 ft. alt., 6 Aug. im2, Sheldon 428 (US); gravelly slopes, North Cheyenne Canon, El Paso Co., 8500 ft. alt., 6 Aug. 1892, Sheldon 5839 (NY). In the enumeration of the Parry collection of 1861 in the Rocky Mountains Gray listed Parry no. 157 as ''Cymopterus terebinthinus , Torr. & Gray, var. C. foeniculaceus Nutt."^" How- ever, this was a case of mistaken identity which he corrected in 1864 in the list of the Hall and Harbour collections, when 7io. 222 was described as a new upecies, Cymopterus f anisalus, with the following comment: "called 'C. terebinthinus, vdi. foenicula- ceus' in Parry's 1861 collection (No. 157) ; but it can hardly be either of Nuttall's species under those names, on account of the very long and subulate leaflets of the involucel as well as calyx-teeth, yet apparently related to them; the foliage, &c., very similar. Mature fruit not collected; some of the present collection pretty well forme5e [250J (272); 4356 (269. 272); -^574, 4576 389, 394); 7/ (372); 40 (287); 67 (390); 204 (342); 168 (374); 277 (358); 180 (394); £76, 459 (312); 500 (311). Palmer, E. J. 13289 (367); 50775 (279); 50755 (299); 31160 (392); 5i;874 (302). Pammel, L. H. and R. E. Blackwood, 3827 (324). (346); 4584 (336); 4604, 4625 (387); P^^^h, S. B. 974^ (380); 19235 (361). 4645 (336); 4709 (369); 4710 (346); Parish, S. B. and 4505 (322); 4819 [391] (269); 6975 (361). W (387); 7011 (272); 7022 (365); 7065 Parker, C. S. 306 (334). (336); 7095 (259); 7263 (263); 7562 Parry, C. C. — (264, 286, 332, 336, 369, (322); 7377 (263); 7667 (299); 8687 (272); 8845 (365, S87); 5546 (388); 8977 (269); 9004 (369); 5252 (269); 374, 388, 392) ; 31 (365) ; 65 (332) ; 80 (359); 55 (394); 84 (358); 227, 227 I4«] (365); 122 (323); 255 (388); 167 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 405 (323); 158 (286); 159 (274); 160 (366, Ramaley, F., G. S. Dodds and W. W. 389); 161 (302); SS4 (260). Robbins, 3190 (274). Patterson, H. N. — (279, 286, 302, 323); Ramaley, F. and W. W. Robbins, 1781 37 (279); 38, 39 (286); 40 (323); 4^ 42 (302) ; 5390 (286) . (280); 2489 (287); 6755 (302). Redfield, J. H. 470 (274); 2567 (302). Patterson, H. N. and F. Beaty, 287). (279, Reverchon, J. [Curtis 1031], S60, 610, 2008, 200Sa, 3564 (367). Payson, E. B. 29, 223, 244 (369); S55 Robbins, W. W. 6447 (280). (357); 274 (391); 414 (292); 1579 Rodin, J. E. (269). (289) . Rose, J. N. 48, 533 (340) . Payson, E. B. and G. M. Armstrong, Rose, J. N., P. C. Standley and P. G. 3207 (388); 3270 (346); 3337 (323); Russell, 12203 (367); 12320 (374). 3359 (346); 3452 (336); 3557 (323); Rothrock, J. T. — , 253, 755 (311). 3563 (336); 3781 (346). Ruby, C. (365). Payson, E. B. and L. B. 1726 (347); 1866 Rusby, H. H. (327); 1910 (263); 2073 (336); 2132 (346); 2281 (323); S5eS (336); 2579 7 (280, 306, 311, 374); , 147H, 148, I48y2 147 (374); 14 (306); 65/, 6SS (311); e5.5H (392). (369); 2585 (336); ^6^56 (346); ^757 Rusby, W. S. (268). (336); 3040 (346); 3827 (357); 5567 Rust, H. J. 537 (327). (391); 3941 (324); ,^05/ (307); 4038 Ruth, A. 2 (367). (287); 4109 (307); ^^06 (291); 45^0 Rydberg, P. A. (264, 266, 274, 280, (346); 4662 (336); 4757 (263); 4S08 (269) Pearson, E. 239 (336). 287, 323, 324); 74, 127 (268); i50 (364); 718 (268); 7/5 (259); 726 (364); 7^7 (386); 4650 (327); 6005 (34S). Rydberg, P. A. and E. A. Bessey, 4615 (263); 4625, 4626 (265); 4628, 4629, 4630 (327). Peebles, R. H., G. J. Harrison and T. Rydberg, P. A. and E. C. Carlton, 6370, Pease, F. N. (302). Peck, M. E. eS13 (332); 6448 (276); 10064 (334). H. Kearney, 2512, 4484 (311). — (289). Penard, E. — Phelps, K. E. Pilsbry, H. A. Piper, C. V. - 2980 (331). Plank, E. N. - Pope, J. — - (274). - (305). (333); 2340, 2795 (334); (280). (370). Popenoe, E. A. (302). Porter, D. and H. Fulton, 2840 (311). Porter, T. C. Pringle, C. G. Purpus, C. A. (349); 97 (357); 157 (302); 196 (349); (274, 336). (311); 1250 (312). - (324); 46 (369); 83 6389, 6598, 7085 (324); 7089 (308); 7096 (324); 7114, 7127, 7164, 7166, 7167 (308); 7207 (324); 7254 (308); 7557 (324); 7340, 7389, 7426, 7437, 7439, 7440, 7671, (308). Rydberg, P. A. and A. O. Garrett, (308); 8694, 8695 (292); 8696 (308); 8697, 8698 (292); 8708 (308); 8723, 8724 (324); 8788, 8789 (308); 8790 (292); 8833 (308); 884O (292); 8892, 8927, 8928 (308); 9015 (324); 9016, 9017 (292); 9070, 9234 (308); 9456, 9458 (290); 9536, 9715 (308); 9756 (287); 5760, 9761 (324). i97 (357); 46/, 544 (302); 55-5 (391); Rydberg, P. A. and F. K. Vreeland, 5288, 5393 (360); 5792 (342); 5866 (372); 56)70 (311). Ramaley, F. A 84 (274); 719, 832 (279) /OW (389); 1015 (264); /7S6 (274) /^/7 (279); 1235 (302); 50^0 (274) 9665 (302); 9706 (274); 9755 (264) 9829 (260). 5697, 5790, 6791, 5792, 5793, 5794, 5795, 5796, 5797, 5799, 5800, 5801, 5802 (303); 5809, 5810, 5811 (291); 5812 (287); 5813, 5814 (323); 5815 (319); 55/6, 55/7, 5818, 5819, 5820, 5821 (389); 6822, 6823, 5824, 5825 (369); 5559 (264). 406 [Vol. 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN St. John, H. 35 18, 8051 (334). Sonne, C. F. (339, 381). St. John, II., W. D. Courtney and C. S. Soth, Mrs. M. E. 7, 8, 14, 15 (303), 221 Parker, 4943 (331). (333). St. John, H., G. Cowan and F. A. Spreadborough, W. [4966, 4967, 85015] Warren, 3301 (334). (387). St. John, H., W. J. Hardy and F. A. Standley, P. C. — , 4027, 4319, 4586, Warren, 3072 (331); Sill (334). St. John, H., G. N. Jones, C. English and L. A. Mullen, 8111 (331). St. John, H., F. L. Pickett, J. A. Carey and F. A. Warren, 6308 (331). 4811, 6177, 6626, 7546 (306); 13390, 13687 (274); I434I (292); 14345 (306). Standley, P. C. and H. C. Bollman, 11058, 11111 (306); 11165 (323). Starz, E (263, 327). St. John, Mrs. O. (366, 370). Sandberg, J. H. and J. B. Leiberg, — ■, 230 (331). Schedin, L. M. and N. T. 292 (389); 296 (287). Scheuber, E. W. — — (327). Schmoll, H. M. 470, 473a (389); 1093 Stearns, E. 350 (306). Stevens Expedition, — ■ (259). Stevens, G. W. 232 (386); 300 (365); 41s, 3032 [Coll. R. L. Clifton]. 3052, 3184 [Coll. R. L. Clifton] (386). Stevenson, E. S. 12 (336); 31 (272); SS (336); 40 (36iO. (303); 1226 (357); 1235 (303); 1690 Stokes, S. G. — (348, 352, 358, 372, 394). (357); 1749 (290). Stretch, R. H. (381); 162 (382). Schneck, J. (274, 365). Studhalter, R. A. 1116 (369). Schneider, E. C. 58 (303). Scovell, J. T. (319). Sturgis, W. C. Suckley, G. — (306, 389). Scribner, F. L. 66c (327). Sears, F. C. (347). Sellon, 0.1.29,64 (269). Sharpies, S. P. 163 (331). Shear, C. L. 3162 (266); 3291 (303) 3359, 34O8, S425 (263); 3652 (303) 3709 (319); 3851 (287); 4^93 (303) 4382 (369); 4^02 (303); 4527 (323) 4648 (287); 4738 (280); 5021 (263) 6202 (303). — (386). Suksdorf, W. N. — , 129, 317, 1201 (331); 11480 (332). Taylor, Dr. 22 (387). Taylor, W. P. 4O8 (332). Thomas, C. (264, 269, 274). Thompson, Mrs. E. P. — (358, 374, 393); 11 (358). Thompson, J. W. 4^78 (332). Thompson, W. Thurber, G. — - (374, 394). (373, 392); 143 (373). Shear, C. L. and E. A. Bessey, 4193 Tidcstrom, I. 206, 249, 367 (309); 1016 (303). Sheldon, C. S. 105 (319); 124 (303); 311 (289); 428, 5839 (319). Shockley, W. H. — (372, 382); 208 (372, 382). Shreve, E. B. - Siler, A. H. — (311). — (287). Smiley, F. J. 97, 706 (339). (348); 1811 (309); 2052 (348); 2127, 2132, 2148 (369); 2167 (358); 2190 (391); 2196 (369); 2231 (303); 2454, 2915 (309); 9416 (358); 9882, IOI24 (342). Torrcy, J. — , 167 (339). Toumey, J. W. — (374); 190, 192 (311). Townsend, C. H. T. and C. M. Barber, 29 (312). Smith, B. H. - (274, 303, 389); 44 Tracy, J. P. 75^8 (339). (303). Smith, C. P. — , 1519, 1523, 1626, 1564, Tracy, S. M. (274, 280); 201 (306); 235 236 258 (311)' 621 (324). 1577 (348); 1681, SS25 (336); 2330 Tracy', S. M. and EvLns, (7^/(324). (348); 2337, 2338 (266); 2345 (348). Smith, E. C. (303, 389). Trapliagen, F. W. (263). Smith, R. A. 58 (365); 59 (263). Treakle, Mrs. Trelease, W. - (374). (287, 303). 1930) MATHIAS — STUDIES IN THE UMBELLIFEKAE. Ill 407 * Tweedy, F. (327, 387); 1 (327); 2 Wetherill (304). (387); 68 (265); 93 (291); SO4 (303); Wheeler SIS (335); 309 (323); 488, 49S (303); 49s (291); 853 (263); 241S4 (265); S425 936 (369). White (268). (263); 3S69 (323); S570 (285); 3374 Whited, K. ;?57 (334). (263); 4505 (38^); 45i;? (303); 4513 Wilcox, E. V. /5, 134, (387). (SSe); 457>^ (323); 49S6, 4987 (280); Wilcox, T. E. — (386); 481 (312). 4989 Williams, P. A. (272). 5599 (260); 5600 (274); 550/ (304); Williams, R. S. — 13 (387); 16 (259); 5503 (2i7). j;e (327). Underwood, L. M. and A. D. Selby, 313, Willi 358 (304). . Vasey, G. R. Will (268, 391); 505(268). (268, 289, 304, 389). (280, 304, 306); 220, Willi ts, V. 100 (263). 221 (280); 221 (287); 225 (274); f^5, Wilson, N. C. — (374); 188 (372). 226a (304); ;2g5 (352); 309 (334). Wislizenus Vestal, A. G. 353 (363, 389); S70 (274). Wolf (304). (287); 7n (389); 716, 716-18, Visher, S. S. — , ;?57; (268). Vorhies, C. T. 110 (311). Voth, H. R. — , 85 (370); 85 (392). Waldron, L. R. 2316 (262). Walker. E. P. 242 (309) ; ^55 (287) 719, 720 [667], 724 [127], 724 (304); 25, 731 [160] (287). / Wolf (324). Ward S30, 494, 31 453 [330] (309). Ware, R. A. 2681c (339). Warren, E. R. 25, IO84 (304). Waring, M. G. IS (392). Watson, S. — (364); 158 (327); 448 (328); 449, 449a (382); 4^56 (372); 719, 720, 724 (304); 723 [Coll. Loew] (306); 7^5 [Coll. Loew] (312); 852 (287). (311, 312, 392); 25 (372); ^^. c. N. 4, ^5a, 214 (351). Woolson, G. C. 96 (367). Wooton (280, 306, 307, 312); 360, 41 6 iZ07); 420 (312). Wooton, E. 0. and P. C. (307). 450 (352); 46 1 (348); 432 (342); 4^5 Wright (392); 233 (367); ii07 (324). Webber, Wentz. (307). (268). (364). Young, M.S. 68 (279). Zuck, M. (312, 392). ABBREVIATIONS The following abbre\ have been used in citations to indicate the different herbaria from which material has been obtained for study: A = Herbarium of the New York State Museum, Albany. B = Herbarium of the Brooklyn Botanic Garden. = Herbarium of Brigham Young University, Provo, Utah. = Herbarium of the Carnegie Museum, Pittsburgh. = Herbarium of the University of California. = National Herbarium of Canada. = Herbarium of the California Academy of Sciences. = Herbarium of the University of Chicago, deposited in the Field Museum of Natural History. BYU C CAL CAN CAS CH 408 [Vol. 17 GARDEN D F G GO lAC J K M MU N NEV NY P PA PAR R UC US wsc Dudley Ilorbarium of Lcland Stanford, Jr. University. Herbarium of the Field Museum of Natural History, Gray Ileibarium of Harvard University. Herbarium of G. J. Goodman, Missouri Botanical Garden. Herbarium of the Iowa Agricultural College. Herbarium of M. E. Jones, Clareniont, California. Herbarium of the Royal Botanic Gardens, Kew. Herbarium of the Missouri Botanical Garden. Herbarium of the University of Minnesota. Edward L. Greene Herbarium of Notre Dame University. Herbarium of the Nevada Agricultural Experiment Station. Herbarium of the New York Botanical Garden. Herbarium of the University of Oregon. Herbarium of Pomona College. Herbarium of the Academy of Natural Sciences, Philadelphia. Parry Herbarium deposited in the Iowa Agricultural College. Rocky Moimtain Herbarium of the University of Wyoming. Herbarium of the University of Colorado. United States National Herbarium. Herbarium of the State College of Washington. « Index to Species New species, combinations, and species emended are printed in bold face type; Bynonj'ma in italics; and previously published names in ordinary type. Page Adphylla Forst. in part 376 Adorium Raf 255, 256, 283 A. crassifolium Raf 257 A. divaricatum (Pursh) Raf 257 Page 214, 217, 219, 239, 244, 246, 251, 252, 253, 342 m 241, 344, 352, 378, 394, 395 A. divancatxm (Pursh) Rydb 258 A. angustatum Osterh 345 A . Hookeri (Torr. & Gray) Rydb.. . 261 A. angustum Osterh 345 A. lucidum (Nutt.) Kuntze 258 A. basalticum (Jones) Tidcstrom. . 344,354,895,39;) A. Belheli Osterh 357 A. tenuijolium (Nutt.) Kuntze 266 Aletes Coult. & Rose 222, 238, 245, 246, 250, 276, 281, 290 A. cineranum (Gray) Coult. & Rose A. acaulis (Torr.) Coult. & Rose. . 352, 361, 377 238,245,270,277,320 A. duchesnense (Jones) Tidestrom A. (?) Davidsoni Coult. & Rose. 282, 316 241,344,355,395 A. humilis Coult. & Rose. .238, 277, 281 A. glaucum (Nutt.) Coult. & Rose A. (?) MacDougaliCowli. &. Rose.282, 289 240, 343, 350, 375, 396 A. ohovata Rydb 278 A. ibapense (Jones) Coult. & Rose A. (?) tenuifolia Coult. & Rose.. 265, 282 A. (?) tenuifoliua Coult. & Rose 265 A. Jonesii Coult. & Rose 240, 343, 349, 396 Ammineae 244 241, 344, 358, 396 Anonymus Torr 282 A. longipes (Wats.) Coult. & Rose Apioideae 244 Aulospermum Coult. & Rose 240, 246, 343, 345, 396 A, minimum Mathias. . . .241, 344, 353 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 409 Page A. panamintense Coult. & Rose. . . 241, 344, 359, 397 A. panamintense var. acutifolium oblong Page 353, 378, 394 C. ahoriginwn var. ovalu Jones. 353, 394 Coult. & Rose 241, 344, 361 C. aboriginum var. subtematus Jones A. planosum Osterh 240, 343, 348 A. purpureum (Wats.) Coult. & Rose 241,344,356,397 353, 394 C. acaulis (Pursh) Raf 223, 229, 230, 231, 235, 238, 247, 249, 376, 377, 384 A. Rosei Jones 241, 344, 355, 398 C. acaulis (Pursh.) Rydb 384 A. Watsoni Coult. & Rose 240,343,351,396,398 C. albiflorus (Nutt.) Torr. & Gray 216, 339, 394 Carum (?) Hallii (Gray) Wats. .278, 280 C. alpinus Gray 284, 287, 394 Cicuta (?) trachjpleura (Gray) Wats. 271 C. ? anisatus Gray 317, 319, 320, 321, Cogswellia Raf 293 Cogswellia Spreng 214, 394, 395 Rose) 234, 245, 246, 248, 250, 292, 293, 294 C. basalticus Jones 395, 393 C. ambigua (Nutt.) Jones 295 C bipinnatus Wats .325, 395 C. dasycarpa (Torr. & Gray) Jones 295 C. bulbosus Nels 368, 370, 395 C. foeniculacea (Nutt.) Coult. & C calcarcus Jones 334, 395 Rose 293 C. calif amicus (Coult. & Rose) Jones C. Grayi Coult. & Rose 295 337, 395 C lapidosa (Jones) Rydb 345 C. campesiris (Nutt.) Torr. & Gray C. Lemmoni (Coult. & Rose) Jones 298 216, 363, 395 C. macrocarpa (Nutt.) Jones. . . . .295 C. cinerariu3 Gray emend. Mathias C. nudicaulis (Pursh) Jones 295 218, 223, 233, 236, 361, 377, 378 C. triternata (Pursh) Jones 295 C. cinerarius Gray. . .352, 377, 378, 395 C. utriculata (Nutt.) Jones 295 C. corrugatus Jones C. villosa (Raf.) Schult 293 227, 236, 247, 276, 377, 383 C villosa (Raf.) Spreng 293 C. corrugatus var. CouUeri Jones. . . 382 Coloptera Coult. & Rose 217, 376 C. corrugatus var. scopulicola Jones. 382 C. Jonesil Coult. & Rose 393 C. Coulteri (Jones) Mathias C. Newberryi (Wats.) Coult. & Rose 393 223, 236, 247, 276, 375, 377, 382 C. Parryi Coult. & Rose 384 C. decipiens Jones 390 § Coriophyllum O. E. Schulz 343 C. deserticola Brandg Coriophyllum Setchell & Gardn. . . 342 220, 223, 227, 233, 236, 377, 379 § Coriophyllum Jones 342 C. duchesnensis Jones 356, 395 Coriophyllus (Jones) Rydb 217, 342 C. Elrodi Jones 333, 395 C, basaliicus (Jones) Rydb 399 C. Fendleri Gray 223, 229, C Jonesil (Coult. & Rose) Rydb.. . 353 C. purpureus (Wats.) Rydb 357 C Rosei (Jones) Rydb 355 Cynwptertbus Buckl 362 C. Tnacrorhizus Buckl 363, 366, 396 Cymopterus Raf 214, 215, 216, 217, 218, 219, 220, 221, 223, 225, 227, 228, 230, 232, 234, 241, 244, 245, 246, 247, 248, 250, 251, 252, 253, 283, 294, 320, 375, 383 C aboriginum Jones 352, 394 230, 231, 236, 247, 248, 249, 377, 390 Fendleri var. Newberryi (Wats.) Jones 393 foeniculaceus (Nutt.) Torr. & Gray 216, 332, 395 glaher •av^u^ Wats 220, 223, 226, 236, 247, 377, 380 C. glomeratus (Nutt.) DC. 216, 249, 384 C. glomeratuA (Nutt.) lUf 293, 384 Vol. 17 410 ANNALS OF THE MISSOURI BOTANICAL GARDEN Page C glorneratus var. Leibt rgii (Coult. & Rose) Jones 385 C. glomerahis var. Parryi (Coult. & Rose) Jones 385 C. humboldlensis Jones 325, 396 C. ibapensis Jones 349, 396 C, Joncsii Coult. & Rose. . .35 ', 382, 396 C lapidosus Jones. 345, 396 C. lapidosus va,r. dcserli Jones. .345, 396 C, Lcibergii Coult. & Rose 3^'5 C ligusticoides Jones 297, 396 C. ligusticoides var. iennijolius (Gray) Jones 298, 396 C. (?) litioralis Gray 217, 396 C. longipes Wats 345, 396 C lucidiis Ostcrh 385 C{ymopleribus) macrorhiziis Buckl. 363, 366, 396 C. megacephalus Jones 223, 227, 233, 236, 377, 370 C. monfanus Nutt 363, 3S0, 397 C. monfanus (Nutt.) Torr. & Gray 216, 249, 363, 396 C. montanus var. globosus Wats. . . . 370, 380, 397 C. montanus var. pedunadalus Jones 366, 397 C vionlanus var. purpurascens Gray 370, 397 C. ? nevadensif< Gray 217, 253, 397 C. Ncwberryi (Wats.) Jones. .220, 223, 227, 230, 236, 245, 247, 248, 377, 392 C. Ncubcrryi var. alaiu^ Jones 393 C. Ncwberryi var. Jona^ii (Coult. & Rose) Jones 393 C nivalis Wats 327, 397 C. owenensis Jones 397 C. panaminiensis Coult. & Rose. 359, 397 C Parryi (Coult. & Rose) Jouea. . . 384 C. petraeus Jones 341, 397 C plurijugus (Coult. & Rose) Jones 275, 397 C purpurascens Jones 373, 397 C. purpurascens (Gray) Jones 370, 373, 397 C purpureas Wats 356, 397 C Rosei Jones 355, 398 C. terebinlhinus Dougl 330, 398 Page C. terebinlhinus (Hook.) Torr. & Gray 216,329,398 C. terebinthinus var. albijlorus (Nutt.) Jones 339,398 C. terebinthinus var. califormciis (Coult. & Rose) Jeps 337, 398 C. terebinthinus var. foeniculaceus Gray 321,398 ^*C. terebinthinus Torr. & Gray, var. C focniculaceus Nutt," 319 C. thaploides Torr. & Gray 332, 398 C thapsoides (Nutt.) Torr. & Gray 216, 332, 398 C. idahe7isis Jones 370, 398 C utahensis var. Eastwoodae Jones 36S, 398 C, utahensis var. monocephalus Jones 370, 398 C Waisoni (Coult. & Rose) Jones 351, 398 Cynomaraihrum Macbridci Nels. . . . 325 Daucophyllum (Nutt.) Rydb 256 D. lineare Rydb 265 D. linearis Rydb 265 D. tcnuifolium (Nutt.) Rydb 267 Dewcya Torr. & Gray 280 D. (?) acaulis Torr 278 § Eucyniopterus Torr. & Gray. .216, 375 Euphorbia L 226 Ferula foeniculacea Nutt 293 F. Newberryi Wats 393 F. f Palmella Hook 384 Glehuia Schmidt 217,234,244,246,253,375 G. leiocarpa Mathias 219, 233, 246, 375, 396 G. littoralis Schmidt. .234, 249, 375, 395 Harbouria Coult. & Rose. 221,233,236,245,250,270 H. trachypleura (Gray) Coult. & Rose 270,271 Heracleura sphondylium L 282 Laserpilium terebinthinum Dougl. . . 829 § Leptocnemia Nutt 216, 361 Ligusficum montanum (Gray) Benth. 296 L. uiontanwn var. tenuifoliuyn (Gray) Wats 297 Lomatia R. Br 292, 294 1930] ^ ^ ■'■m ^ . - MATHIAS — STUDIES IN THE UMBELLIFERAE III 411 h B ^ ^ '^ ■^ - * - ^ ■, ^ q Page Page Lomalium Raf. . ..:. . . . , . .283, 292, 294 M. peduncidatum Nels. .:.•.:: .258, 260 ^_ r^ L. foenicidaceum (Nutt.) Coult- & ilf. tenuifolium Nutt. . ■ ^ » « 266 1 ■ > ■^ ^. Rose. ^ . . p r 4 « * 293 M. tenuifolium Nutt. in Torr. & ^ -^ ■ r I f ■ ■ ^ * "^ r b * V - '\ L. lajndosum (Jones) Garrett- .. . . • 345 Gray ^ ■ . ^ '• L ^ > _ r ^, ^ *>^ -. H 267 'n L.Xem7woniCoult.& Rose........ 297 M. tenuifolium (Nutt) Coult. & ■> X* L. pt/5escen5 Raf . . . ... . . . .... . . . - 293 Rose.. . .,,,.... . :.238, 245, 257, 266 L, i;i"Ko5WW Raf..., ;.;... .... 293 ilf. ienm/oKwm (Nutt.) Torr. & Gray 266 Marathrum Humboldt & Bonpland " "i ^ . M. trachyspermum Nutt. . . . ... ... 261 255,256 M-vaginatumRydb...... 238, 257,264 Marathrum Raf .255, 256 Musineum Raf , ... . . . .... 256 M. divaricatum (Pursh) Raf. 257 M. Ehrenbergii Wolff. .... .215, 269, 270 ilfiisenwrnRaf,. ... .... .. .... 256 Neoparrya Mathiaa. ........... .. » ^■. M. divaricatum (Pursh) Coult. & Rose 229,236,245,248,250,276 258 N. lithophaa Mathias 276 M. divaricatum (Pursh) Nutt. /. ... 258 Oreosciadium Wedd. ...... .... . . . 280 M. HookeH Nutt. .... ... ........ 261 O. acavJe (Torr.) Gray.... . . . . .278, 280 M. HookeH Torr. & Gray. , ... . , . .261 Oreoxia Raf.. 217, 222, M* tenuifolium Nutt.. ....... .... . 267 237, 244, 246, 251, 252, 282, 283, 290 M. trachyspermum Nutt. . . . . . .'. . . 261 O, alpina (Gray) Coult. & Rose. . JIf. pofl'inaiwm Rydb... ........... 264 Musenium Nutt. . . : .... . .255, 256, 280 0. Bakeri Coult. & Rose 237, 284, 394 M. alpinum Coult. & Rose, . . . .... 269 237, 246, 284, 290, 395 M. angustifolium Nutt .... . ..,..: . 261 O. hunmia Raf. .... . . 237, 282, 284, 287 M, divaricaium Nutt. . - . . • ■ r h ^ - .-K- ' ^ J '" r- W. ' - -^ , 'r '■* ■■^ ^ 257, 261, 282 ^. '". M. divaricatum var, Hookeri (Nutt.) r- T r- O. MacDougali (Coult. & Rose) ; Rydb.. . . .237, 246, 282, 283, 284, 289 Pastinaca :M foemculacea j j;^ (Nutt.) XL L L T w^ -- ^. ..-^-^ /, > T - - ■ > -.^. <-*-,- 293 ^ . "^ Torr. & Gray.. 260 ^:: -■ ; b^ ^. . . . - - . . . . ... .- • M. Greenei Gray: :;:;:..: 270, 278, 280 :; Peucedaneae. rl^rf; f^^^ J - ^ ^ -J--':-:-^ ^ ^ ■* rS ■ * _ - :r -^ ^ - ■% M. HookeH Nutt.; .t. . :rC;. ;.;::. 260 iPewcedantm Am;Authr;vr:fv;292, 294 r- . S ' -^ -. l_¥'J^ f-^ J ■'*'■" '] ^f|B iPhvL ti^^j i'l^* ' m - ' — '*-| -I ^fc^ ^'bJi-i > ikf. pedunculatum Nels^Vv^:; V: :^^; . 258 i PeucedanumlTourn.] L.. v:^f.v.292, 294 ■x'" _ .-L ■ '^■ ■■'■ *^ J ■' ^' -^ ^--■^' 4 « » * « .A-" -Y*^ * Kh*. 266 V ^ P.foeniculaceum Nutt ; J h . vi^ 1 V . ; v 293 *_ --: "■■-» ,^ ■ r-" .^.^■ ^ - r ^ + h. 1 X L H I fc > , 'I- ' -^ ilf. tenuifolium Nutt. ... . M. trachyspermum Nutt^v Vv::\^> ;' 260 ^^ P. foptdoww Jones.;, .v^;;.: ;:\;:: 345 Af. vaginatum Rydb;; . >?V?: .^i vV. 264 P. Lemmoni Coult. & Rose •^.^.^ 297, 314 Musineon Raf . V. . . v;; : , >; r: .^;^v;^^^^^^ 221, 238, 245, 246, 250, 255, 256, 269 KPhdlopteris montana Nutt..T;T; . f: 363 Phellopterm Benth/.^/ If I :v ;:^217, 361 s , - _> ■ M. alpinum Coult. & Rose ..... 269, 270 ^ > ^ * ■^ , -^ K^ ■V. ^■ ^ ;..' *. ■^w 1* -V , -r\ M. an^usf^Kum Nutt. .;:;■:. v.. 261 ^^^hdlo'fderus Nutt. ^ .v. ..;■■:•. .216, 362 M. divaHcaium (Pursh) Coult. &:J \T -■ ^ fc^ V p* ' ^ "^r . a- 220, 230, 238, 244, 246, 253, 361 h ^ h H ■^-.-^ .K ,h, ".h M* divaricatum (Pursh) Raf.f ; ivV\ -■> r ^ fc ^ ^ 1 ■k y'^ ..'.". " '"^^ ^ -^- ^ ■ ^ - \ h r J i'- N ' ■_ --■ \ ; *^^^■■V^ 238, 257, 260 ■^, "■ " - ^:-' ■ ■-,- M. divaricatum var. Hookeri Jr- 3 «'« *._■ #f ' m '" -^.:.. '-V ^ .^- *^ _ I 1 (Nutt.) Mathiaa. .... . . 238, 257, 260 P. camporum Rydb^ P. ffZoMciM Nutt.;?i^.ST; .¥■!;; 350, 374 k. HookeH Nutt. . /. . . . :: . f : . : , : 261 1 ll*- Z^^*^ (Coult. & Itese) RydK. 1 ■ > -.<- 375, 382 ^ ^ ■ M. HookeH (Nutt.) Torr. & Gray.■;:•^■261^.;:-£-■^•s:iimre::^.^^^>l^S;i•^- ;: P. KftoroKs (Gray) Benth.;;vt. 249, 375 M. Hookeri (Torr. & Gray) Nutt.^::^ 261 ' -m M. lineare (Eydb.) Mathias.. . ; . . . -■■i . .' -:/ r ' iBb ^'*. ^ _>_ J - *J^ V^^!^^ P^\/^ 238, 257, 265, 282 ;: P. macrocarpu$ Osterh. . - J. /'^ v-^"<-^ .* - 363 ■ ^ ^ -i^ C * J q J h ^ i - - ■ J. -^ . ^" h 1 ■l J- ■-.■ ■> 1 ^ . I - J A _ T t K - ^ '' H ^ _ - ^ ^ ^ r L P 1 L '- ' ■ ^ ^ L V ^ - / s - X " y ^ I J h < - ^ -, ■- r -H T - 'i \ - f ^ r h ¥ r_ Hr 4 h . ^ 412 ^ - {Vol. 17 ANNALS OF THE MISSOURI BOTANICAL GARDEN - \ ^ 4 ^ y ^ ' ^ * ■^ \ 1 ■ J \ _ T ^ ' ^ ^ - ^ H __ - ' - ^ _ ^ ^ ^ - \ \ k ^ ^, Jm > ^ 1 S ^ ^ ^ rf k I 1- - T , L ^ ^ '^ ^■^ - V » 1 . ^ 1 fc. F 1 ■- >F Page ^ =4 \ * * H - 4 ■ h ■^ ^ - "r Page P. macrorhizus (Buckl.) Coult. & \r Rose. . . . . . . . . .238, 362, 366, 396, 397 P. montanus Nutt r^ p. movianus v&T. tenuifolius (Wats.) Coult. & Rose. ... . :. .: .... .... 297 P. 7n«Z«^dMS Rydb. ........... 297, 315 233, 238, 246, 249, 363, 395, 396, 397 P. nivalis (Wats.) Mathias. . . ... ^ 4 P. multinervatiis Coult. & Rose-. . , ^ H . k \ >T 242,296,327,397 ^ ^ . 238, 362, 372, 397' P. purpurea (Cotdt. & Rose) Rydb. 297 P. purpurascens (Gray) Coult. & P. sylvaticus Nels. . : . . .... 297, 314, 315 1 ^ 1 > Rose......... ........ .....368,371 P. tenuifolius (Gray) Rydb.; . ; .:. 297 P. purpurascens (Gray) Coult. & . Rose, emend. Mathias. ........ P. Tidestromii Coult. & Rose.. .298, 315 P. versicolor Rydb. . ......... .298, 315 238, 362, 370, 397, 398 Pseudopteryxia Rydb... .. .218, 295, 320 J 1 P. purpurascens var. Easiwoodae (Jones) Coult. & Rose. 368 P. uiahensis (Jones) Woot. & Standi. ¥ 4 Podistera Wats 368, 371 P. alelifolia Rydb. . ...... .242, 246, 318 P. anisata (Gray) Rydb. ..... .317, 321 P. Hendersoni (Coult. & Rose) ............:.. 322 "■ -. ^ -r- j ^ Rydb - T- + ^ 217, 221, 236, 244, 245, 248, 250, 253 P. longiloba Hydh . . ...... . . . . .321, 322 , r - Pseudorepxis Rydb. . ... . . .M . .218, 295 *■ - * L _ \ ^ ■> h ^ ^ ^ 1 P.aZfeeTwfs Jeps. ................ . 254 P. 6^p^nnai^^5 (Wats.) Rydb/. . . . .. 325 P.nevadensis(Gray)Wats..:..253, 397 P.nivalis (Wats.) Rydb. ......... 327 n H Pseudocymopterus Coult. & Rose IPteryxia Nutt ; . .216, 328 * -r t ■_ 214, 217, 219, 220, 221, 242, 244, 246, Pteryxia Nutt. . . . . .... . . . . . ... . * 250, 251, 252, 253, 295, 319 217, 241, 244, 246, 251, 252, 328 , J-"^ - f I _ P. aletifolius Rydb.,... . ... .317, 320, 321 P. albiflora Nutt. . . ... ....... . . . . 339 ■^ P, anisatu^ (Gray) Coult. & Rose. ■-b. -- \ * P. calcarea (Jones) Coult, & Rose. . 334 ■- - ? ^ - *- ' * '1 ■ ? - \ _-*i^ .. ' H n : 242, 249, 317, 320, 321- P. califomica Coult. & Rose.. .... .337 ■ ^ \ * y X b J ^ P. anisatus (Gray) Coult. & Rose, emend. Mathias. ........: .... . . .. L_ * ^ > ^ '-.■.*, \ . P. JS^Zrodt (Jones) Rydb ..:...;,.; 333 P. foeniculacea Nutt. ........... i . 332 V - 1 ■*>* 4-- , * V H * , ^ 242, 246, 296, 317, 320, 321, 394 P. /oenicuZacea (Torn i& Gray) Nutt. 332 P. petraea (Jones) Coult. & Rose, . , 241,329,341,397 P. terebinthacea Nutt. . , . , ... . . , , . 330 H ^ P, antsafzw var, longUohua (Rydb,) : ;] Tidestf:;.. ■.........-:....:.. 322 ' - -- . ... p. bipinnatus (Wats.) Coult. & Rose . ■ K ^ '■ ^ Jr -1 h h . ^ r T *^^ 7 r X > -" T ' 4 h -. *J H > ^ F - ^ 242, 246, 296, 325, 395 P. terebinthina (Hook.) Coult. «& ^ h - ^ \ - ^ -r p. Davidsoni (Coult. & Rose) ^"^ t Rose. ....... . .241, 329, 340, 341, 398 .^ ■ ^ *■ Mathias, . . , . .,244, 246, 282, 296, 316 P. terebinthina var. albiflora ^ _ r m P.filidnus Wooton & Standi . . . 316 P. Hendersoni Coult. & Rose . . 321 r ?• Hendersoni Coult. & Rose, 4 n E ^ (Nutt.) Mathias . I 241,329,339,394,398 emend. Mathias 4, P. terebinthina var. calcarea (Jones) Mathias.. .241, 329, 334, 340, 341, 395 ■ >, i M- ^ - s ^ 242, 249, 296, 320, 321, 395,398 P. terebinthina var. califomica 4. r- _ T r ^ ■ ^ - *" * k i > " ^ p. humboldtensis (Jones) Mathias. . » ^ (Coult. & Rose) Mathias. . . . . . ; \ '- / Kj "-H ^ . . h n ^ ' ^ H ^ t 242, 246, 296, 325, 396 . > 1 . ^ ^ *. 241, 329, 337, 340, 341, 395, 398 ^ * P. montanus (Gray) Coult. & Rose 233, ?• terebinthina var. foeniculacea * > ^ ^ V - r^ . 1. 242, 246, 295, 296. 313, 314, 315, 396 ♦ , P. montanus var, maUifidus Rydb . . J h - J ^ ' ^ ' H \ 1 r (Nutt.) Mathias..:... ...:..... 241, 246, 329, 332, 340, 341, 395, 398 . • • • • • « Ot5i6 r » b^ >x 297, 316 P. thapsoides Nutt. . . . . , . . T , X p. marUanus var. purpureus Coult. & Rose...T,. ..........;,^. 297,314 P. vtontanua vax. tenuifolitia (Gray) -■ h r - ■ CouIt.&Rose......^... .;..... 297 Rhysopterus Coult. & Rose. ...... H L - y .. \ *- J h '- F r , » r ■_ > L 'fi ^ '* i ^ * 217, 236, 244, 245, 250, 275 - ^ ^ -v R. corrugaius (Jones) Coult. & Rose r- ^ ■ r *- T r f- ^ -f ■ v ^ ^ ^ + X H. * ^ -t -1 * *- h ^ ' .^ -t >* fc- ^ >_^^ - ^ *■ * ^ n1 ■- • ■, V ^ > L * "1 - ^ ■• k \ f h ■", - ■^1" > ■ > . ■_ + _.^ ^ ^1 ^ '.J ' H ^ - '^ 1 1 ^ 276, 383 L I rt .- r - - « > J '. V ^L ^^ * 1 r - X L V ' ■ ^ > ^J ^ - J ^ ^ ■ t J -S ^ -H ^ . ^ .V - ■, 4 ' - L - - V* + y -■ _- ^ * > ■J ,\. J - _ 1 " T A ^_ k * . ^ y -. - _ \ ^^ L" ^ ' ^ ^ J - ■ - ■■ s I ■^ r. ^ i ^ ■ , ^ d . - ^ £ r X _ -» ,y. ^ V I ^ "■ \' - -r -^^ > ^ T H .^- ^^ , K ^ - X-L -t J ^ ^ ^ - ^_'- ,> - .- -. K -^ b>' 1. .V - . '. ^- ^ r L '^ ^> -' 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 413 Page Page R. Jonesii Coult. & Rose 276, 382 Tauschia Schlecht 270 R. plurijugus Coult. & Rose 275, 397 T. alpina (Coult. & Rose) Mathias 269 § Scopulicola Jones 342 T. Ehrenbergii (Wolff) Mathias. . . 269 Selinum acaule Cav 384 T. Parishii (Coult. & Rose) Macbr. 397 S. acaule Pursh 216, 384 Thapsia Nutt 293 S. acaule Turcz. ex Bess 384 T. glomeraia Nutt 216, 384 S. terebinthinum Hook 329 Seseli L 280 Thaspium (?) montanum Gray 296 T. montanum var. tenuiflorum Gray 271 S. divaricalum Pursh 255, 257, 260 T. montanum var. tenuifolium Gray S. Hallii Gray 278, 280 271, 296, 313 S. lucidum Nutt 255, 260 T. trachypleurum Gray 271 S. terebinthinum Hook 330 Zizia Koch 281 Smyrnieae 244 Z. Hallii (Gray) Coult. & Rose. ... 278 [Vol. 17, 1930] 414 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 21 Composite diagram illustrating the morphological characters used in a classification of the various genera. Fig, 1, Diagram of a complete plant; a, carpophore; 6, pedicel or secondary ray; c, involucel; d, primary ray; e, involucral bract; /, leaf -blade length; h, loaf-blade width; p, pseudoscape; r, root; s, ultimate-segment length; /, ultimate-segment breadth; ?/, peduncle. Fig. 2, Dorsal surface of a typical mericarp; a, lateral wing; 6, dorsal wing. Fig, 3, Ventral or commissural surface of a typical mericarp; a, persistent calyic teeth; b, commissural surface showing longitudinal striations marking the position of oil tubes and the carpophore; c, fruit length; d, fruit breadth. Fig. 4. Cross-section in the median plane of a typical mericarp showing the position of oil tubes around the seed, etc.; a, dorsal wing; b, accessory oil tube at the wing-base; c, interval between the wings; d, seed; e, lateral wing; /, strengthening tissue at the wing-base — coUenchyma or vascular elements; h, commissural surface. Ann. Mo. \Ur\\ (Iakd., \'(>l. 17, liKJO Pj.ATH 21 4 b y 3 MATH IAS— STUD IKS L\ LMBELLIFERAE [Vol. 17, 1930] 416 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 22 Microphotographs of cross-sections in the median plane of the fruit, illustrating the various genera. X 12. Fig. 1. Podistcra nevadensis (Gray) Wats. Collected at Mt. Dana, California, July 1902, Hall & Bahcock 3607 (Missouri Botanical Garden Herbarium). Fig. 2. Neoparrya lithophila Mathias. Collected on rocks, Huerfano Mts., New Mexico, 1867, Parry SS (Missouri Botanical Garden Herbarium), isotype. Fig. 3. Miisineon tcnidfolium (Nutt.) Coult & Rose. Collected in Laramie Hills, Albany Co., Wyoming, July 1903, A, Nelson 8977 (Missouri Botanical Garden Herbarium). Fig. 4. Aides acauUs (Torr.) Coult. & Rose. Collected at Castle Rock, Colorado, 1 July-9 Sept. 1885, Patterson 37 (Missouri Botanical Garden Plerbarium). Fig. 5. Rhysopleriui plurijugtis Coult. & Rose. Collected in Malheur Valley, near Harper Ranch, Oregon, 10 June 1896, Leiberg 22Ifi (U. S. National Herbarium), type. Fig. 6. Oreoxis alpina (Gray) Coult. & Rose. Collected in Colorado Territory, 1872, Parry (Missouri Botanical Garden Herbarium). Fig. 7. Phdlopterns monfamis Nutt. Collected at Fort Collins, Colorado, 20 May 1896, C. F. Baker (Missouri Botanical Garden Herbarium). Fig. 8. Phelloptertis macrorhizus (Buckl.) Coult. & Rose. Collected at Dallas, Texas, 10 April 1900, Reverchon 200Sa (Missouri Botanical Garden Herbarium). Fig. 9. Aulospertmim longipes (Wats.) Coult. & Rose. Collected at Sandy, Utah, 24 May 1895, M. E. Jones (Missouri Botanical Garden Herbarium). Fig. 10. Pseudocymopterus bijdnnatus (Wats.) Coult. & Rose. Collected in Bridger Mts., and Trail Creek, Park Co., Montana, 29 June^2 July 1899, Blankinship S25 (Missouri Botani<'al Garden Herbarium). Fig. 11. Aulosperviam Jonesii Coult. & Rose. Collected at Frisco, Utah, 22 June 1880, Af. E. Jones 1S08 (Missouri Botanical Garden Herbarium), cotype. Ann. Mo. Dot, G.\i{U., Vol. 17, 1930 Plate 22 MATI11.\S STUDIES IN I'MBELLIFERAE [Vol. 17. 19301 418 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 23 Microphotographs of cross-sections in the median j^lane of the fruit of Cymopterus. X 12. Fig. L C. cinerarias Gray, emend. Mathias. Collected in the Mono National Forest, Nevada, 19 Aug. 1912, Holton & Maule (U. S. National Herbarium 583073). Fig. 2, C globosus Wats. Collected at Carson City, Nevada, 29 May 1897, M, E, Jones (Missouri Botanical Garden Herbarium). Fig. 3. C. deseriicola Brandg. Collected at Kramer, California, May 1913, K. Brandegee (Herbarium of the University of Cahfornia 173143), type. Fig. 4. C. megacephalns Jones. Collected at Little Colorado, N. Arizona, 10 June 1890, M. E. Jones (U. S. National Herbarium 47080), type. Fig. 5. C corrugaivs Jones. Collected at Humboldt Lake, Nevada, 17 June 1882, Af. E. Jones 3886 (U. S. National Herbarium). Fig. 6. C. Covlteri (Jones) Mathias. Collected at Sevier Bridge, Utah, 26 April »10, M. E. Janes (He Fig. 7. C. Fendlen lissouri Botanical G Jones Fig. 8. C. Newberryi (Wats.) Jones. Collected at St. George, Utah, 1877, E. Palmer 180 (Gray Herbarium). \\N. Mo. Birr. Hahd., Vol. 17, I'M) I'l.ATio '2:\ M ATI HAS-^STU DIES IX UMBELLTFERAE [Vol, 17. 19301 420 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation op Plate PLATE 24 Cross-sectiona in the median plane of the mature fruit of Cymoplerus. X 6. Figs. 1-14. C. Fendleri Gray, showing the variation within the species. Fig. L Collected at Grand Junction, Colorado, 22 May 1895, M, E. Jones (Her- barium of Pomona College 82247). Fig. 2. Collected at DeBeque, Mesa Co., Colorado, 19 May 1911, Osterhout US7 (Herbarium of Pomona College). Fig. 3. Collected at Green River, Utah, 23 May 1914, M. E. Jones (Herbarium of Pomona College 8:W96). Fig. 4. Collected at San Rafael Swell, Utah, 19 May 1914, M. E, Jones (Herbarium of Pomona College S3099). Fig. 5. Collected at San Rafael Swell, Utah, 8 May 1914, M. E. Jones (Hcrl>ariura of Pomona College 8309S). Fig. 6. Collected at San Rafael Swell, Utah, 28 May 1914, M, E. Jones (Herbarium of Pomona College 83100). Fig. 7. Collected at McElino Creek, Utah, Eastwood (Herbarium of Pomona College 82241). Fig. 8. Collected at La Bajada, New Mexico, 15 May 192G, Benedict 2/fi2 (U. S. National Herbarium). Fig. 9. Collected in Sandia Mts., near Madera, New Mexico, Ellis 263 (U. S. National Herbarium). Fig. 10. Collected at La Bajada, New Mexico, 15 May 1926, Benedict 2^02 (U. S. National Herbarium). Fig. 11. Collected in Sandia Mts., near Madera, New Mexico, Ellis 263 (U. S. National Herbarium). Fig. 12. Collected at Grand Falls, Cascade of the Little Colorado, Arizona, 13 May 1901, Ward (U. S. National Herbarium 410080). Pig. 13. Collected at Holbrook, Arizona, 18 June 1901, Ward (U. S. National W rbarium 410081). Fig. 14. Collected at X Ranch, IS miles north of Holbrook, Arizona, May 1901, ■. Hough 104 (U. S. National Herbarium). Fig. 15. C. cinerarius Gray, emend. Mathias. Collected in the Mono National Forest, Nevada, 19 Aug. 1912, Holton tfc Maule (U. S. National Herbarium 583073). Fig. 16. C. globosvs Wats. Collected at Reno, Nevada, 12 May 1896, Ilillman (Herbarium of Pomona College 82070). Fig- 17. C. megacephulus Jones. Collected at Little Colorado, N. Arizona, 10 June 1890, M. E. Jones (U. S. National Herbarium 47080), type. Fig. 18. C. deserticola Brandg. Collected at Kramer, California, May 1913, K. Brandegee (Herbarium of the University of California 173143), type. Fig. 19. C, Coulteri (Jones) Mathias. Collected at Sevier Bridge, Utah, 26 April 1910, M. E. Jones (Herbarium of Pomona College 82077). Fig. 20. C corrngatxis Jones. Collected at Humboldt Lake, Nevada, 17 June 1882, M. E. Jones SSS6 (U. S. National Herbarium). Fig. 21. C. acaulis (Pursh) Raf. Collected in the **R. Mts.," Nutlall (Herbarium of the Academy of Natural Sciences, Philadelphia). Figs. 22-23. C. Ncwberryi (Wats.) Jones. Collected at Moab, Utah, May 1892, Eastwood 3 (U. S. National Herbarium). ,\.\N. Mo. But. Uaud., \'ol. 17, VJ'M Platk 24 X 2 5 M.VrHIAS— STl'DlEri IN UMBELLIFKKAI 422 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vol. 17 Explanation or Plate PLATE 25 long Fig. 1. Cn & Rose. X 4 Fig. 2. Cross-section of a portion of the mcricarp of Cyvioplerus Newberryi (Wats.) Jones, showing the oil tubes and strengthening tissue at the base of an aborted wing. X 45. Fig. 3. Cross-section of a fruit of Cymopterus Newberryi (Wats.) Jones, showing 'seudo the variation in wing structure in the two mericarps. X 8. Figs. 4-13. Cross-sections in the median plane of the mature mericarp of I cymopterm nwntanu^ (Gray) Coult. & Rose, indicating the variation within the species. X 6. Fig. 4. Collected at Tie City, Albany Co., Wyoming, 20 July 1900, A. Nelson 7667 (Herbarium of Pomona College). Fig. 5. Collected at Coppermine Creek, New Mexico, 1851, Wright 1107 (Gray Herbarium). Wooion 1853-4 Fig. 6. Collected at Continental Divide, New Mexicc National Herbarium 737584). Fig. 7. Collected at Sandia Mts., New Mexico, Oct. barium). Fig. 8. Collected at Panguitch Lake, Utah, 7 Sept. 1894, M. E. Jones G015x (U. S. National Herbarium), Fig. 9. Collected at Bromide Pass, Utah, 27 July 1894, M. E. Jones 56051 (U. S. National Herbarium). Fig. 10. Collected near Pagosa Peak, Colorado, Aug. 1899, C. F, Baker SOS (Missouri Botanical Garden Herbarium). Fig. 11. Collected at San Francisco Mts., Arizona, 26 Aug. 1889, Knowlton 120 (U. S. National Herbarium). Fig. 12. Collected at Williams, Arizona, 8-25 July 1903, Griffiths 4930 (U. S. National Herbarium). Fig. 13. Collected near Flagstaff, Arizona, 23 Aug. 1922, //. C. Hanson A215 (Missouri Botanical Garden Herbarium). Fig. 14, Cross-section in the median plane of the mature mericarp of Pseudocy- mopierus Davidsoni (Coult. & Rose) Mathias. X 6. Collected at Clifton, Arizona, Pseudo 1900, Davidson 161a (Missouri Botanical Garden Herbarium), Figs. 15-16. Cross-sections in the median plane of the mature mericarp oi cymoplcrus anisaius (Gray) Coult. & Rose, emend. Mathias. X 6. Fig. 15. Collected in the Rocky Mountains, Colorado, 1862, Hall & Harbour 22S (Missouri Botanical Garden Herbarium), cotype. Fig. 16. Collected at Minnehaha, Colorado, 16 June~22 July 1901, Clements & Clements 94 (Missouri Botanical Garden Herbarium). Figs. 17-18. Cross-section in the median plane of the mature mericarp of Pseudo- cymopterits Hendersoni Coult. & Rose, emend. Mathias. X 6. Fig. 17. Collected at Teton Mts., (U. S. National Herbarium). Fig. 18. Collected at Teton Mts., (Missouri Botanical Garden Herbarium). Fig, 19. Cross-section in the median Wyoming I'V Wyoming Pseudocy 1930] MATHIAS — STUDIES IN THE UMBELLIFERAE. Ill 423 Explanation of Plate PLATE 25 (Continued) mopterns humboMtensis (Jones) Mathias. X 10. Collected in the East Humboldt Jones CO type. 20-21. Cross-sections in the median plane of the mature mericarp of Pteryxia terebinthina (Hook.) Coult. & Rose. X 6. Fig. 20. Collected at Umatilla, Oregon, 1 June 1905, M. E, Jones (Missouri Botanical Garden Herbarium). Fig. 21. Collected near Boardman, Morrow Co., Oregon, 14 June 1928, J. W. Thompson 4778 (Missouri Botanical Garden Herbarium). Figs. 22-23. Cross-sections in the median plane of the mature mericarp of Pteryxia na X 6. Fig. 22. Collected in the Sierra Nevada Mts., California, 1875, /. G. Lemmon (U. S. National Herbarium 44683). Fig. 23. Collected at Little Summit near Butte Meadows, Butte Co., California, 22 June 1914, A. A. Heller (Missouri Botanical Garden Herbarium 748222). Figs. 24-26. Cross-sections in the median plane of the mature mericarp of Pteryxia terebinthina VSLT. foeniculacea (Nutt.) Mathias. X 6. Fig. 24. Collected in the Blue Mts., Washington, 4 July 1892, Lake & Hull 535 (Gray Herbarium). Fig. 25. Collected at Wallowa Mountains, Oregon, 23 Aug. 1898, Cusick 20S5 (Herbarium of the University of Minnesota). Fig. 26. Collected at Wallowa Mountains, Oregon, 23 Aug. 1898, Cusick 20S5 (U. S. National Herbarium). Fig. 27. Cross-section in the median plane of the mature mericarp of Pteryxia terebinthina var. albijlora (Nutt.) Mathias. X 10. Collected at Red Lodge, Montana, 26 July 1893, Rose 48 (U. S. National Herbarium). Figs. 28-29. Cross-sections in the median plane of the mature mericarp of Pteryxia terebinthina var. calcarexi (Jones) Mathias. X 6. Fig. 28. Collected in Bridger Mts., Montana, 26 May, 17 July 1905, Blankinship 226 (Missouri Botanical Garden Herbarium). Fig. 29. Wyoming 1901 Macbride Figs. 30-31. Cross-sections in the median plane of the mature fruit of Pteryxia petraea (Jones) Coult. & Rose. X 6. Fig. 30. Collected at Martin, Blaine Co., Idaho, 5 July 1916, 3053 (Missouri Botanical Garden Herbarium). Fig. 31. Collected at Alvord Desert, Oregon, 30 June 1896, Leiberg 2426 (Gray Herbarium). Fig. 32. Cross-section in the median plane of the mature mericarp of Aulospermum vanamintense var. acutifolium Coult. & Rose. X 6. Collected at Newberry's on California Herbarium 335930), cotype. Fig. 33. Cross-section in the median plane of the mature mericarp of Aulospermum panamintense Coult. & Rose. X 6. Collected at Argus Mts., California, June 1897, Purpus 6393 (Missouri Botanical Garden Herbarium). Fig. 34. Cross-section in the median plane of the mature mericarp of Aulospermum [Vol. 17, 1930] 424 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 25 (Continued) duchcsnense (Jones) Tidestrom. X 6. Collected at Myton, Utah, 20 May 1908, M. E. Jones (Rocky Mountain Herbarium 113304), cotype. Fig. 35. Cross-section in the median plane of the mature mericarp of Aulospermum minimum Mathias. X 6. Collected at Cedar Breaks, Utah, 17 July 1930, Goodman & Hitchcock lf)01 (Missouri Botanical Garden Herbarium). Fig. 36. Cross-section in the median plane of the mature mericarp of Aulospermum liosei Jones. X 6. Collected at Richfield, Utah, 18 June 1898, M. E. Jones 30 (U. S. National Herbarium), type. Fig. 37. Cross-section in the median plane of the mature mericarp of Aulospermum abonginum (Jones) Mathias. X 6. Collected at Bishop, Owen's Valley, California, 15 May 1897, M. E. Jones (Herbarium of Pomona College 82033). Fig. 38. Cross-section in the median plane of the mature mericarp of Aletes humilis Coult. & Rose. X 6. Collected at Dale Creek, Larimer Co., Colorado, 19 July 1899, Osterhout (U. S. National Herbarium 361576). / \nn". Mo. Bot. (Iaud., \\)L. 17, VXli) I'l.ATK 25 1 *i ■^y K irA' ' / -A-^ ^ ^ Av z ■*' "■v-^ *#■ >-^^- 5 H 5 £ 10 11 12 i3 MATIIIAS— STUDIES IN UMBELLIFERAE [Vol. 17, 1930] 426 ANNALS OF THE MISSOURI BOTANICAL GARDEN Exi>LANATION OF PlATE PLATE 26 Fig. 1. Musineon divaricaium (Pursh) Raf. From authentic material, Forwood 144, in the Missouri Botanical Garden Herbarium. Fig. 2, Musineoii Icnuijoliavi (Nutt.) Coult. & Rose. From typical material, A, Nelson S977, in the Missouri Botanical Garden Herbarium. :\\\. Mo. BoT. (!aui)., Vol. 17, 19o0 rj.ATi: LM) r ■\ s -^1 •-. i- 4 ? \ *. s s Ivj si t'j ^ -« IC s: « Vj^. -fh m *;. ^, i— - (Vol. 17, 1930] 428 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 27 Fig. L Rhysopterm plurijugus Coult. & Rose. From the type specimen, Leiherg 2240, in the United States National Herbarium. Fig. 2. Harhouria trachypleura (Gray) Coult. & Rose. From a cotype specimen, Hall and Harbour 21fi, in the Missouri Botanical Garden Herbarium. Ann. Mo. Hot. Gaud., Vol. 17, WVM) P*LATb: 27 S > C/2 y. W to ^ ■- «r V ^- \ ^ . . ' X4^ ^^J H.^r.' 1^ ^-y^'-. ^^-'^ -™4^^ [Vol. 17, 1930] 430 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1 Explanation of Plate PLATE 28 Fig. 1. Aletes acanlis (Torr.) Coult. k Rose. From typical material, Patterson 37, in the Missouri Botanical Garden Herbarium. Fig. 2. Aleies humilis Coult. & Rose. From a cotype specimen, Osterhout 6, in the Herbarium of the New York Botanical Garden. Ann. Mo. Bot. Gard., Vol. 17, 1930 Pla'I'I': 2.S t> > > d t?1 ^. ^ M W I A ^ "^ r v^J t -f* it *- I J J ■IT V^ -1^:. r-vv - -> y V. }k f •H \ -r^^ J * r- - 1^ S _^ r> ■?: r' 4 V ff^" ^ ^ ■:5/^ ?i^ V- J^ Ipin; .v^ ^ I r^ [Vol. 17, 1930] 432 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 29 Fig. 1. Oreoxis alpina (Gray) Coult. & Rose. From the type specimen, Parry 158, and Fig. 2. Oreoxis Bakm Coult. & Rose. From the type specimen, Baker 12, in the United States National Herbarium. \nv. Mo. Rot. Cakd., Vol. 17, \'.)?>Q VhATv: 29 > > ^- > ^^ : 4f. t ^ ^. I 9 ^ ^P > :3e s se "^ 1 Q .K n c o t> r -^vM ^ U b ■-^»- (» > ^J ■^ 1 E fi ^ r <;f^ w_L *J / ? r \ y ' ■ i / r ■■ - .---'-^' % -< vi V « \ _ft_- %- 1% \ * \ 6 ? ■to / / / y---^- ^ M 10 Jj- r r- - ' , 4 A ^ '*> * i J > \ ; y .> V ^ _./ f 5 P n V ^ 1 ^ ^ t ■ V ^ ^ ^ ^ K t y .>1 *« h /' '- ' * y > i r > \ i ■ rf ^ 'iT.. "^ («■ \ n ^ *- rf^ (-- •■ 1 *=* 1 ■* V i^ * « i , ,' \ bv J { a > ^ :* ( x^ , ! ■ >v f ■^ t* ■* *il. ^ -^ ■ .-A^ Nl i > 1 e ^ ■ ^ f J £ 1 i > 1 ] r i ^ V i ii. J > "l ^ ^ 1 * a # ^ tl. J ^^s V t.''' -.^ . ^ i «*> [Vol. 17, 1930) 434 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 30 Fig. 1. Oreoxis MacDougali (Coult. & Rose) Rydb. From a cotype specimen, MacDougal WS, in the Gray Herbarium of Harvard University, Fig. 2. Phelloptervs invltinervatus Coult. k Rose. From typical material, showing the scarious involucre and involucel, Eastwood 8188, in the Herbarium of the Cali- fornia Academv of Sciences. An\. Mo. l^oT. (Iahix, \'ol. 17, WVAi) Platk 30 ► . ffi w CO 7: ^ cr ^ H > H ro V :3 % Y ^ ti. j^ _;. i M 1 ^ ^ 4 s ^ ft^ V [Vol. 17, 1930) 436 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 31 Fig. 1. Phelloplerus bulbosus (Nels.) Coult. & Rose. From the type specimen, A. Nelson 1^709 ^ in the Rocky Mountain Herbarium of the University of Wyoming. Fig. 2. Phellopterits bulbosus (Nels.) Coult. & Rose. From a specimen illustrating the elongated peduncle form, Eastwood, in the United States National Herbarium (type of Cymoplerus ulahensis var. Easlwoodae Jones). These two specimens illustrate the habital variations within the species. Ann-. Mo. lioT. {\ah\k, Vol. 17, l\W ri.ATi; :\\ > r. 1 o re ■k L. I t' 5 _^' ! ^ I in > 4 -^ Jt f -..* V w< y r ^ _ J ■^ 14 ^ ^ J _ t* ^ ^ 4 ■ ^ 1 ^i'. ^ 1 /^ J 1 ^v 4 L r J « „^ V r r K k. J K lO >. > ^ - > 9 3 ^ 3 4 ■rv^ ^ . ^ **. ■f^^r- Hh '.^ [Vol, 17, 19301 438 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 32 Fig. L Pseiulocymnpterus monianiis (Gray) Coult. & Rose, From the type speci- men, Feruller S76, in the Gray Herbarium of Harvard University. Fig. 2. PseudocywojHerus montamis (Gray) Coult. & Rose, From representative material, Wright 1107, in the Gray Herbarium of Harvard University (type of Thaspinm montanum var. tenuifolium Gray). These two specimens illustrate the two extremes in foliar variation which occur in the species. All intermediate forms exist, two of which are illustrated by the following plate. \n\. Mo. Bot. ClAiti)., You 17, lO^iO Platk 32 > 1 > X c/^ ^v ■\ A ^ * 4 M * *■■■;: 4 •^-^ n >. \'- £ Si a L. -J to [Vol. 17, 1930] 440 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 33 Fig. 1. Psendonjmoptcrus montanns (Gray) Coult. & Rose. From typical material, A. Nelson 7667 ^ in tlic Rocky Mountain Herbarium of the University of Wyoming (type of Pseudoajmopteriis sijJvaticiis Nels.). Fig. 2. Pscudocymoptcrus inontanus (Gray) Coult. & Rose. From typical material, J. G. Lemmon 392, in the L"^nited States National Herbarium (type of Peuccdanum Lenimoni Coult. & Rose). These two specimens and those illustrated by the preceding plate serve to indicate the great foliar variation occurring in tliis species. These two specimens are repre- sentative of those elements in which there is a tendency toward the complete or partial abortion of dorsal wings. Anx. Mo. ]\n\ . (i.MU)., \'<)i>. 17, [[j'M) J'lati-: lili > X £ a V^ L ' k. k -J > t I* h > f n # ^ £ I T %AL^A f f rr jr- r r 1 t i ■6 I I i I - -t - 1 > ^ < 4 ••^ .**■ \ -■i. -r. V ^' ■X\ -4N^ «v^-^ .'F"rfW^ I f-' < t ■T>, ■^. W. ^^^^S^ >-_ ■-v l>_ X X -Wh -x^-- -t_ ^ ^ #. '^ o-}- \ rf # .;^^*' 4 V -I' ''S [Vol. 17, 1930] 442 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 34 Fig. L Paeadocytnopterus Davidsoni (Coult. & Rose) Mathias. From the type specimen, Davidson 16 1 a. In the Ignited States National Herbarium. Fig. 2. Pseudocyniopterus bipinnaivs (Wats.) Coult. & Rose. From the type specimen, Ilayden 14, Canhy I4S, and Watson 158 , in the Gray Herbarium of Harvard University. Aw. iMo. Hot. Gakd., Vol. 17, 1!):50 Plati-: 3! > 1 y- ^ r ' r ; 4 •r ■ *■ r I r '/>» ^ r '■^? f ^ i } V J* t V ^ ^1 V r* ^ V h ^ ^ ^^ L ^ ^ ■> L-J ND \l ^ S >/- ; i-: ^ « I 3 i 1- * V -^ ::u ^' > \ ^ # y I r# 5 h [Vol. 17, 1930] 444: ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 35 Fig. 1. Pseudocymo'pterus animtus (Gray) Coult. & Rose, emend. Mathias. From the type specimen, Hall ami Harbmir 222y in the Gray Herbarium of Harvard Uni- versity, Psevdocymopteriis Hendersoni Coult. & Rose, emend. Mathias. Prom typical material, Parry 157, in the Gray Herbarium of Harvard University. Fig. 2. Pseudocymopterus anisatus (Gray) Coult. & Rose, emend, Mathias. From typical material, Clemefils and Clements 94f in the Herbarium of the New York Botanical Garden (type of Pseudopteryxia aletifolia Rydb.). A.w. I\Io. BoT. (iARi)., Vor.. 17, IIKJO Pr.ATE .'55 > 'J. I i ^ f- *,. ^ . 4 P. ^ fr t J - *;^:i i:-^ ?? v^ ■-^ ^*t^ j^A^ r » » U3 d I ^ ^ H w s f > ^ J A ft ^ ^ m ffttr jfcAS*^ ^uu^ [Vol. 17. 19301 446 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 36 Fig. 1. Pseudocymoptcrus humboldtensis (Jones) Mathias. From a cotype speci- men, Af . E. Jones, in the Herbarium of the New York Botanical Garden, Fig. 2. Psewlocyrnopterus nivalis (Wats.) Mathias. From a cotype specimen, Watson LIS, in the Gray Herbarium of Harvard Universitv. \\iN. Mo. lioT. (!aicd., Vor.. 17, lil-'it) Pl.ATR 'Mj > en y: C 7. tr M ! r '■%< I 1 * ^T %V '- ■; i - i 1 ■F J => IVoL. 17, 1930] 448 ANNATES OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 37 Fig. 1. Pteryxia terebinthina (Hook.) Coult. & Rose. From typical material, Af . E. JoneSj in the Missouri Botanical Garden Herbarium. Fig. 2. Pteryxia ierebinthiiia var. californica (Coult. & Rose) Mathias. From the type specimen, H. E. Brown, in the United States National Herbarium. Ann. Mo. Bot. CIakd., Voi.. 17, lOoO I'liATi: .'J7 > K > a to t #IP* *■ V .#--?■ ^^ =i ■^' i ■ —^ J. ■-rV V * ^ » [Vol. 17, 1930] 450 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 38 Fig. 1. Pteryxia terebinthina var. calcarea (Jones) Mathias. From a cotype specimen, M, E. JoneSj in the United States National Herbarium. Fig. 2, Pteryxia petraea (Jones) Coult. & Rose. From a cotype specimen, A/. E. Jones, in the Herbarium of the New York Botanical Garden. \N>r. Mo. BoT. Gaud., Vol, 17, 19o0 l^LATE 3S > H u3 W Is! << r ? ^ ■^6 ^ 1 . ^ V < r r* ^ ^ 4. i ^ 4 # ff f -< £ 1- T i; \ ^ d F n r "^ > ^ * -*- f .^^ IfH fi V 3 A * ■■^f ? 4r -r 3C ^ i I 4r ^ i^r V X ^ > ^, v'^-^. ^ ■K-^*fl» r i A *mf ^ I 7- t^ M t^ I n h* ,^ T^l ^ ^ « ** /< - ^ - r \' ^ ■^ ■V \ ^ K * ? j-p n V s< - V, fr- \ S ^ f -H i ■ 3 -V \ 9 ^ ^ K^ s ± ^ ^ V » a ^^1?** [Vol. 17, 1930) 454 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 40 Fig. 1. Aulospernwm ibapense (Jones) Coult. & Rose, From the type specimen, Af. E, JoneSf in the United States National Herbarium. Fig. 2. Aiilosjicrvium Watsoni Coult, & Rose. From a cotype specimen, Watson 450t in the Gray Herbarium of Harvard University. AulospermiLin glaucum (Nutt.) Coult. & Rose. From the type (?) specimen, Nuttallj and Canhy 147, in the Gray Herbarium of Harvard University. \.\\. Mo. \U)T. CiAHi)., Yin.. 17, HKiO Platk 40 X ^ X Cd PI r ■ r pi t w » r IJ J_. i ■* Z -A r r ■ t % 1^ '^ \ P I ■'-? •! ^^^ 4 ■1 ii ^ f *^. -S£* ■■^ ^^ ■V vW.. ■^: .^^ / ■^^^■+^ >^>K. > * ■> r- <• - w ^^ N-^ t^ *^ ff \ «^ \ i ■^ i ' i I [Vol. 17, 1930] 456 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 41 Fig. 1. Aidospermvm planosum Osterh. From a cotype specimen, Osterhout 2572 j in the United States National Herbarium. Fig. 2. Aulospermutn minimum Mathias. From the type specimen, Mathias 723, in the Missouri Botanical Garden Herbarium. Ann. ]\I(). Hot. Gaud., Vol. 17, IDiiO rLATE 11 > rj2 O H w H ic r- >■ r *^ i V V 'A^-, \ t J, .,^ *fln». «,,*-—*- 1^ ..^-^ ^t. 7 ■o. ■> c% % \ BHB^^^^ »% :..ru V'' ;s. H [Vol. 17, 1930] 458 ANNALS OF THE MISSOUM BOTANICAL GARDEN Explanation of Plate PLATE 42 Fig. 1. Aulospcrmum duchcsnense (Jones) Tidestrom. From a cotype specimen, M, E. JoneSy in the Herbarium of the California Academy of Sciences. Fig. 2. Aulospermiim Rosci Jones. From the type specimen, M. E. Jones 30^ in the United States National Herbarium. Aw. Mo. 15()T. CiAKi)., \'()r.. 17. lUoO Platk 12 t> s^ > I ^4 ^ J A I I ■ * ^ ■-■" . J7 c!- .s^* • ^-it 4 r * * i t t 1 -A -J tc •» -vX 1 I 2 y £ > > > r i' i i ■t u > S,i ■\ - -^ I L^C^ ^^ k^iF-H Jx L L^ ^^. -'>my^-^-- vT^. ,-^" --Vrn-^ [Vol. 17, 1930J 460 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 43 M Fig. 1. Aulosjiervnim aborigimim (Jones) Mathias. From the type specimen, Fig. 2, Aulospcrmxim ahonginum (Jones) Mathias. M svbtematus horigi Ann. Mo. Bot. (iAKi).j \'(>l. 17, liYM) Platk 43 t> a > A r-. m -J ( \ ^ V o ■ t- r= 5 j: ? > I » f •5 4 ( li i . * n. jt ^ p^. / y.y. V .■* X iC 5 «-^ \' #t- ^U '^x^*W / .^ ■^ ♦ ^ «r« f"-i I- I -X "I F ^ - f r' ■ * V m ft- ^t ^^.. ^ ■*fc r^ 'V v r L [Vol. 17, 1930] 462 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 44 Fig. 1. Aulospervium Jonesii Coult. & Rose. From a cotype specimen, Af. E. Jones ISOSf in the Missouri Botanical Garden Herbarium. Fig. 2, Aulosper^mim panamintense Coult. & Rose. From the type specimen, Coville and Funston 50S, in the United States National Herbarium. Anx. Mo. T^ot. Gauu.j Vol. 17, 10;J() Platk II > "1 > cz: "^ ^ o — » D3 ?0 L^j tc ^' V ■, t. ly^* ■ 4 J' [Vol. 17, 1930] 4G4 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate FLATE 45 Fig. 1. Cymopterus cinerarius Gray, emend. Mathias. From the type specimen, Brewer 1899, in the Gray Herbarium of Harvard University. Aidospermum ahoHginmn (Jones) Mathias. From typical material, Brewer 1825^ in the Gray Herbarium of Harvard University. Fig. 2. Cymopterus dncrarius Gray, emend. Mathias. From typical material, Tlollon and Maule. in the United States N>itionjd TTnrl»n.riiirn. Ann. Mo. Ht/r. CiAui)., Vol. 17, VJ'M Plate 45 X rr: — * r K ISD [Vol. 17, 1930] 466 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 46 Fig. 1. Cymoptems megacephalus Jones. From the type spooimen, Af. E. Jones, in the United States National Herbarium. Fig. 2. Cymopterua deserticola Brandg. From the type specimen, K. Brandegee, in the Herbarium of the University of CaUfomia. \nx. Mo. Bot. Ciaui)., \'()r.. 17, VSM l^LATE 40 > c/^ 5 V l^ .^ ^ i 4 ^ ^ t '■> \ <. ;.'' 3 i H 4v « ^ I r 3 £ ff ^ «■ r !-■■ t t 4 ** ^i -A a to r r^. -I 3f 4 Vm [Vol. 17, 19301 468 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 47 Fig. I. Cyjnopterus glohosns Wats. From typical material, M. E. Jones, in the Herbarium of Pomona College. Fig. 2. Cymopterits Newberryi (Wats.) Jones. From the type specimen, Newberry, Parry 83, and Mrs. E. I\ Thompson, in the Gray Herbarium of Harvard University. Aw. Mo. I^oT. CIaiu)., Vol. 17, IWM) Platk 47 t^ > > C/3 ^ ci ^ M W LC [Vol. 17, 1930] 470 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 48 Fig. L Cymoplerus Coulteri (Jones) Mathias. From a cotype specimen, M. E Jones 1691 J in the Gray Herbarium of Harvard University. Fig. 2. Cymoptcrus Coulteri (Jones) Mathias. From typical fruiting material, M E, Jones, in the Herbarium of Pomona College. Ann. Mo. Hot. (Iahd., \'()L. 17, liKJi) Plati-: 4S •/: u/: ^ k.-j K K t ?r si it ■r K « *. ii ti I; ? /• ■^ ^ r f-^^i^-^-h^MMA^^^ ■ iHM |i-M.^L 472 [Vol. 17, 1930] GARDEN Explanation of Plate PLATE 49 Fig. 1. Cymopterus corrugaius Jones, From the type Bpedmen, M. E. Jones 3886, in the United States National Herbarium. Fig. 2. Cymopterua corrugatus Jones. Fr in the Missouri Botanii^al Garden Herbarium. Engelmann Ann. Mo. Boi. (Iakd., \' ?1 ^ — ^ 2 K r^ w I ^-4 ? > l> i ^ -t. ? ^ V --> f * ' p ^ So 1 1 I I L^ T *P- [Vol. 17, 1930] 4/4 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 50 Fig. 1, Cyrnopteriui acaulis (Pursh) Raf. From a photograph of the type speci- men, Bradhvnjj in the Kew Herbarium. Fig. 2. Cymopteriis acaulis (Pursh) Raf. From authentic material, Nutiallj in the Herbarium of the Academy of Natural Sciences, Philadelphia. \\\. Mo. Hot. (Iaud., Vol. 17, I'JoO Pi. ATI-: ."iO > rj: "1 J d 1^ u fi ^ 3 5 /:■ '^y^ *;-. I rsi S.M + ^ vT 3-" K ^ ! ^ \ L X V y W / 5l ,r-^ ,.^'* *i '^^-' ,^ .^ ■-■- >r-rA ^j Y ■^■-VW-^Bir(rt>---> ^ ■v^i' ^ y ■ T ■l i I. t t *!- --* ^ + ■ ?v .f- V -.: -t f S k L -V ^ ^ ^ 1 J^. 'J [Vol. 17, 19301 476 ANNALS OF THE MISSOURI BOTANICAL GARDEN Explanation of Plate PLATE 51 J Fig. 1. Cymopterus Fendleri Gray. From the type specimen, Fendler 274, and Thurher, in the Gray Herbarium of Harvard University. Cymopterus acaulis (Pursh) Raf. From typical material, Parry, in the Gray Herbarium of Harvard University. Fig. 2. Cymopterus Fendleri Gray. From typical material, Kammerer ^7, in the Missouri Botanical Garden Herbarium. These two specimens, Fendler 276 and Kammerer ^7, illustrate the extremes of variation in the species C, Fendleri, \n\. Mo. BoT. (iAKI)., \'(1L. 17, VXW PlaTK :")1 H a vV r \ k \ >. *-\ \ ■A o Si ^^ 'r ■4' \ it & •>■ .< ■■•s> ^^ ^ I c >E 6^ 4 •4 *■ 4 I r.~ ?: ?• I V z'* ** * « -f * "r i I L -HV%iV^ ^ WflJ^KiM-l-^ 4 -'^ .^-*^ I • , N< ^ \ ' / K »T^ k. rf ^ > rc i ■^ A. if r -v >. r 2 O Z m B o ^.^__ f+wii^^i^^^i.^ i > _V "V^ ^-^ B^^J Annals of the Missouri Botanical Garden A Quartf^^'ly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Graduate Labora- tory of the Henry Shaw School of Botany of Washington Uui- versity in aflfiliation with the Missouri Botanical Garden. Information The Annale of the Missouri Botanical Garden appaara four times (iuring the calendar yattff : February, April, September, and November. Four nnmberft constitute a vohime. Subscription Price - — 13.00 per volume Single Numbens — — LOO each Contents of previous issues of the Annals of tho Mis^tun ootanical ( ..irden ^vv, iisitid in t.ue Aerriculiural [ndex, puimshcd hv the H. W. Wilson On. STAFF OF THE MISSOURI BOTANICAL GARDEN Director^ George T. Moore. Assistant to iha Directory Kathbrinb H. Leigh, Hermann von Schrknk, Patho Jesse M. Gheenman, Curator of the Herbarium Edgar Anderson^ GenetidHt. Ernest S. Reynolds, Physlologlfit. David H. Liwoer, Mycoiotfst. A "TON HOGSTAD, Jr., Ph arma cognosds t . Roland V. La Garde, rch Mildred E. Mathias, R W'iJV^C^ rcb Assistant. Robert E. Woouson, Jk., Kosearch Abc^isitant. Nell C. Horner, librarian and Editor of PuDue«»onfi BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN President^ George C. Hitchcock Vice-President^ Samuel C. Davis. Second V ice-President ^ Daniel K. Catlin. L. Ray Carter. Thomas S. Maffitt George T. Moore. ALiiiuttr T. Pbrki" Pmilip C. Scanlai^ FuED G. Zbiuig. EX'OFFICIO MEMBERS- George R. Throop, ObanoiUor of Washing Frederick F. Johnson, Btabop of the !> Victor Louis Satterthwait Missouri. President of The Academy ence of St. Loui^. ArTHUH a. liLUMEYER, t of the Board of Education of St. Louis. Daniel Breck, Secretary.