ANNALS OF THE - MISSOURI BOTANICAL GARDEN FU сако оный un Pm Annals of the Missouri Botanical Garden Volume XXIX 1942 With Thirty-four Plates and Sixty-nine Figures Published quarterly at Fulton, Missouri, by the e of Trustees of the Missouri Botanical Garden, St. Lo Entered as second- age matter at the post- gps at Fulton, Missouri, r the Act of March 3, 19. 59261. Annals of the Missouri Botanical Garden A Quarterly Journal eontaining Seientifie Contributions from the Missouri Botanieal Garden and the Henry Shaw Sehool of Botany of Washington University in affiliation with the Missouri Botanieal Garden. Information The ANNALS OF THE MISSOURI BOTANICAL GARDEN appears four times dur- ing the ealendar year: February, April, September, and November. Four numbers constitute a volume, Subscription vida dat rn per volume Single Numbers.......... 1.50 eaeh* Contents of previous issues of the ANNALS OF THE MISSOURI пени е ач listed in the p pee н Index, published by the H. W. Wilso Company, * Except No. 3 of Vol. 22, price $5.00, and No. 2 of Vol. 25, price $2.50, M TINTA УМ N Ч LIBRARY <) / П < p | А 3 4 د‎ “< Lica, SCH ا0‎ P Se STAFF OF THE MISSOURI BOTANICAL GARDEN irector GEORGE T. MOORE HERMANN VON SCHRENK, ROBERT E. WOODSON, JR. Pathologist Assistant Curator of the Herbarium JESSE M. GREENMAN, HENRY N. ANDREWS, Curator of the Herbarium Paleobotanist CARROLL W. DODGE, ROBERT W. ScH Mycologist Research Ape ass EDGAR ANDERSON, NELL C. HORNER, Geneticist Librarian and Editor of Publications BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN President GEORGE C. HITCHCOCK Vice-President DANIEL K. CATLIN Second Vice-President EUGENE PETTUS L. RAY CARTER RICHARD J. LOCKWOOD DUDLEY FRENCH GEORGE T. MOORE JOHN S. LEHMANN A. WESSEL SHAPLEIGH ETHAN A. H. SHEPLEY EX-OFFICIO MEMBERS GEORGE R. THRO WILLIAM DEE BECKER, Chancellor И Mayor of the City of University St. Louis WILLIAM SCARLETT WASHBURNE D. SHIPTON ir of the Dico of President of the Academy of Missouri Science of St. Louis RUDOLPH HOFMEISTER President of the Board of Education of St. Louis GERALD ULRICI, Secretary TABLE OF CONTENTS Contributions to Our Knowledge of American Carboniferous Floras: I. Scleropteris, gen. nov., Mesoxylon and ось о MEME apes такава Henry N. Andrews И ое cuis ck eat ee ee e ae eas РСР Henry №. Andrews апа Eloise Pannell New Gesneriaceae from Panama.......... C. V. Morton Contributions to Our Knowledge of American Carboniferous Floras. III. Stipitopteris...... L. Wayne Lenz * + 9 ж е э э ө ө ө ө ө ө ө 9 ө ө ө ө ө э + ө 9 5 э «ө ө $9 c$ 9 а ө ө « I. Their Recognition and $ 9 c c5 9 c9 € 9» c$ 9 оо ооо ооо Raees of Zea Mays: Classificiation Dolomitie Glades of East-central Missouri...... Ralph O. Erickson, Louis G. Brenner and Joseph Wraight The Environmental Variables of the Missouri Botanieal Garden Wildflower Reservation at Gray Summi ouis G. Brenner, Jr. Helminthosporium ыр of Citronella and Lemon Grass in Guatemal Carroll William Dodge Commentary on the North American Genera of Commelinaceae.............. obert E. Woodson, Jr. A Revision of the Genus Bumelia in the United IDE TUR E Robert B. Clark Monograph of Malvaviseus.......... Robert W. Schery Contributions to Our Knowledge of American arboniferous Floras: IV. A New Species of Lepidodendron....... Eloi * 9 $9 c9? 9 ооо A m Arauearian Wood from Western Wyo- TOREM Henry N. Andrews and Eloise Pannell ORA of Schoenocaulon........ Robert R. Brinker Contributions toward a Flora of Panama. VI. Collections chiefly by H. von Wedel in Bocas del Toro...Robert E. Woodson, Jr. and Robert W. Schery General Index to Volume XXIX........ Ее а РАСЕ 1- 18 19- 34 35- 58 59- 68 69- 88 89-101 103-135 137-140 141-154 155-182 183-244 245-274 215—282 283—286 287-314 317-378 381-385 Annals of the Missouri Botanical Garden Vol. 29 FEBRUARY, 1942 No. 1 CONTRIBUTIONS TO OUR KNOWLEDGE OF AMERICAN CARBONIFEROUS FLORAS! I. ScLEROPTERIS, GEN. NOV., MESOXYLON AND AMYELON HENRY N. ANDREWS Instructor, Henry Shaw School of Botany of Washington University The occurrence in America of the petrifactions known as coal- balls has been known for many years and in itself needs no review. It is, however, still rather generally accepted that American coal- balls are by no means the equal of the renowned English specimens that have yielded such a wealth of information concerning the flora of Carboniferous times. Much of the American material is highly pyritized, and although pyrite does not necessarily render the petri- factions worthless it is not the most desirable preserving mineral. Pyritized material from the English mines is by no means unknown, while many of the American specimens are quite free from pyrite or contain but insignificant quantities and the plants contained in the coal-balls often exist in an exquisitely fine state of preservation. It has been my privilege to study rather carefully most of the great English collections of petrified Carboniferous plants, and from a comparison of those with some hundreds of Illinois speci- mens that have recently come into my laboratory there is no doubt that the better ones are equally as well preserved as the English. The commonly encountered ‘‘open pit" method of mining in our central states is making huge quantities of material available, and much may be expected of the ensuing studies on American paleo- botany during the coming years. It must be remembered, moreover, ЗА study financed in part by a grant from the Penrose Fund of the American Philo- sophical Society. Issued February 18, 1942. v (1) [Vor. 29 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN that the outstanding English collections and the slides dispensed by professional petrologists represent the choice selections of nearly a century of labor, while almost nothing has been done with Ameri- can coal-balls until a very few years ago. It is safe to assert that when our own flora is better known it will not suffer by comparison. Although a considerable number of species have already been reported from the petrifactions of the midwestern states (Darrah, '41), few comprehensive studies have been undertaken. It is planned to present here, and in the following parts of this series, descriptions of the plants composing the coal-ball flora, chiefly of southern Illinois. As it is not possible to present the results of such a study in phylogenetic order without undue delay, they will appear as adequate material of the genera is collected and studied. All of the specimens on which the present paper is based were col- lected by the author and his students from the open-pit Pyramid Mine of the Binkley Coal Co., located about three miles south of Pinckneyville, Perry County, Illinois. I wish to express my appreci- ation for the very generous assistance and cooperation extended to me during many visits by officials and employees of that mine. The petrifactions are abundant, although, as might be expected, they are by no means uniformly distributed through the coal. One may chance upon a ‘‘nest’’ of half a ton or more or wander along the seam for a mile without encountering any. The better coal-balls are usually found near the top of the seam, sometimes a foot or so below the surface, but more often than not in sheet-like aggrega- tions, their tops flush with the surface of the coal. They are easily detected when the latter is exposed and slightly weathered. The mineral content of these more or less unit aggregations seems to be quite constant, and in many cases they are very nearly free of pyrite. Thus it is a rule that when one or two good specimens are picked up, dozens of others are near at hand. The better coal-balls lend themselves readily to the preparation of satisfactory nitrocellulose peels, although it has been found neces- sary in some instances to use them in conjunction with ground sec- tions. Arborescent Lycopod stem remains constitute the dominant floral feature of our collections. Associated with them are Lepidophyl- lum, Stigmaria and Lepidocarpon. A consideration of the stems and leaves will be presented at a later date, while a description of a re- markably well-preserved gametophyte and the microsporangiate 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. I 3 cones of Lepidocarpon constitutes the second contribution of this series (see p. 19). Stems and roots of Sphenophyllum are well rep- resented and fructifications probably referable to Scolecopteris oc- cur in many specimens. Heterangium is also present. Stems and roots of Mesoxylon are common, and these, together with a new genus of terid ferns, will be considered in the following pages. Scleropteris illinoiensis Andrews, gen. et sp. nov. General Anatomy and Procedure.— Our knowledge of this pteridophyte is derived from a small por- tion of a branched stem which, although fragmentary, is excellently preserved. Since but one specimen has been encountered in some hundreds of coal-balls, it seems desirable to describe it without further delay. As will be pointed out in the following pages, Scleropteris resembles in some respects Botrychioxylon and Zygopteris but certain highly distinctive features of the cortex, as well as the tracheidal pitting, set it apart from these two genera (or one, as Sahni (’32) has shown may be the case). The specimen was about 6 em. long and branched twice. Prior to the first division, an equal dichotomy, the stem measured 1.3 cm. in diameter. Shortly after the dichotomy of the stele the stem became somewhat flattened, measuring 1 x 2 сш., and the two resultant branches each measured 1 cm. in diameter. These figures are neces- sarily approximate due to some slight crushing and distortion. The most distinctive feature of the stem lies in the sclerotic nests scattered through the broad cortex. The stele consists of a mixed primary body surrounded by more or less radially arranged tra- cheids presumably of secondary origin. Adventitious roots are numerous. The specimen was cut into three slices perpendicular to the long axis of the stem. The central section was used for a series of longi- tudinal peel preparations, the other two for transverse sections. Unfortunately, the initial saw cut passed directly through the point of dichotomy of the stele. Preparations made by the nitrocellulose peel technique proved very satisfactory. In fact, in this particular case they were superior both for study and photographic purposes to ground sections. It was found that by photographing the peels unmounted, rough side down, against a black background, excessive reflection was reduced and much finer detail was obtained than when the peels were mounted in balsam. [Vor. 29 4 ANNALS OF THE MISSOURI BOTANICAL GARDEN The Stele.— Figure 2 of pl. 1 represents a transverse section of the entire stem prior to dichotomy, while fig. 1 shows it shortly above the division of the stele. It may be noted that the dichotomy is equal, both branches (pl. 1, fig. 1, sti, 3) being approximately the same size. The stele itself, taken from a point about 1 mm. above the section represented in fig. 2, is shown more highly enlarged in fig. 3. At this point, which is immediately prior to dichotomy, it is oval-shaped but is otherwise nearly circular in cross-section (pl. 3, fig. 11). There the stele measures 2 mm. x 3.5 mm., while after division the resultant branches are approximately 1.3 mm. in diameter. The central part of the stele is mixed, being composed of small tracheids scattered through the parenchyma cells. The latter may be isodiametric but for the most part are vertically elongated, be- ing 2-4 times as long as broad. It is not possible to distinguish definitely protoxylem elements in transverse section. In median longitudinal sections, however, small scalariform tracheids may be observed near the periphery of this primary mass (pl. 2, fig. 7, pa). It would thus seem that the primary wood was mostly centripetal, as is probably the case in Botrychioxylon paradoxum, but this point is not without an element of doubt in both species. Around the mixed primary xylem there is a zone of larger tra- cheids, for the most part radially arranged and presumably sec- ondary in origin. These are uniformly pitted on both radial and tangential walls. Some of the small tracheids located near the inner margin are scalariform but all of the larger ones possess multi- seriate bordered pits although they may be slightly elongated hori- zontally (pl. 2, fig. 7). With reference to the pitting of B. para- dorum, Scott ('12) writes: “Те large tracheids of the secondary wood have sealariform markings, the bars being often short, so that there are two or three series on the same wall." Гр. 376]. His fig. 21, pl. 41, bears out this description. The pits are much less closely compaeted in Scleropteris, and although there are transitions be- tween scalariform and nearly circular bordered pits the latter are distinctly predominant. Since no tangential sections of B. paradoxum were prepared, the presence of wood rays was left in question. Scott writes: ‘It is doubtful whether true medullary rays are present; if so they must be very few in number. Transverse sections afford no decisive proof of their existence.’’ [p. 375]. It is likewise difficult to discern rays 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. I 5 with certainty in the cross-sections of Scleropteris but serial longi- tudinal sections reveal their presence. A tangential section from near the outer margin of the secondary wood (pl. 3, fig. 12) shows uniseriate rays of great height. The secondary wood is not extensive, varying from 5 to 9 cells wide. The cells are, for the most part, radially arranged and occa- sional evidence of cambial activity at the periphery leaves little doubt as to its true secondary origin.’ Although the tissues immediately outside the xylem are not as well preserved, in certain sections there may be observed a narrow phloem, a pericycle two cells wide and what appears to be an endo- dermis the cells of which are filled with a very dark substance. The endodermis is, however, not readily distinguishable in most sections, as might be expected where an appreciable amount of secondary growth has taken place. The Cortex.— The most prominent feature of the stem lies in the abundant sclerotic nests scattered through the cortex (figs. 1, 5, 6 of pls. 1-2). Although rather irregularly shaped they appear more or less 180- diametrie in eross-section and slightly elongated longitudinally (figs. 5,6). The sclereids composing the nests are very thick-walled, in some the lumen being almost lacking. Like the nests as a whole, they are slightly elongated longit The outer parenchymatous cortical бай» contain an abundance of branching fungal filaments. Although they may compose the mycelium of a decay organism, their aggregation within the cells suggests a mycorrhizal relationship.? 1 It is evident from recent pru studies that the distinetion between primary and seeondary vascular tissues is not a arp a one as was onee supposed, and it is possible that these ovd arranged касын have had their origin from a multiple (‘‘primi- tive") eam layer rather than a single row of meristematie cells. а Ж inea and the origin of roots: The presence of mycelium in the tissues of fossil plants is not uncommon, and it occurred to me some few years ago that there might be a deeper lying significance to this association than most authors realized. The idea, and I do not claim absolute priority although I know of no published formulation of it, seems worthy of some thought. Briefly stated, it seems possible that mycorrhizae or a mycorrhizal-like association were a deciding factor in the establishment of the higher land plants and that root hairs and rhizoids are the result of later specialization in the higher vascular plants and the lower vascular plants and bryophytes respectively. It is supposed that mycorrhizal associations existed prior to the evolution of the normal absorbing organs of these plants and that as the aquatic ancestors migrated landward e mycorrhizae made possible this new habitat; and as these newly arrived land plants [Vor. 29 6 ANNALS OF THE MISSOURI BOTANICAL GARDEN The cells of the outermost cortical layers and epidermis are ar- ranged in peculiar mounds which seem to represent the basal por- tions of emergences (pl. 1, fig. 4). Whatever was the nature of such emergences is not known. Most of them terminate in a flat top as shown in fig. 4. That they do represent the basal part of a super- fieial appendage seems certain as their remnants are present in some of the sections and none show evidence of having possessed vascular tissue. Petioles are lacking in the single available specimen. It was thought at first that the small branch shown at the right of fig. 5 might be a petiole but the serial longitudinal peels reveal no positive evidence of a leaf gap at its junction with the main stem stele. The specimen is, then, either a fragment of a plant bearing rather widely separated nodes or it is one in which there is no distinct segregation of the shoot system into stems and leaves as they typically occur in more recent groups. Roots.— Adventitious roots are numerous throughout the specimen. Three may be seen in various stages of departure from the stele in pl. 1, fig. 3. The passage of the root stele through the cortex is irregular but in most cases it appears to be upward at an angle of about 30° with the stem stele. It is not possible to determine for certain the number of protoxylems. There seems to have been some secondary xylem, as well as cork cambial activity, and the cortex is compara- tively broad. Diagnosis.— A dichotomously branching stem 1-1.3 cm. in diameter; stele consisting of a central mixed primary region, probably centripetal in development, surrounded by a zone of radially arranged tracheids became better adapted to their new habitat they developed rhizoids or root hairs and roots which gradually replaced the fungus. There are a number of points that lend sup- port to such a hypothesis. Roots are lacking in the earliest land plants but rhizoids are present and in certain cases fungal associations that may be mycorrhizae; roots are not strongly Il in the higher vascular eryptogams as a whole; in certain seed plants it is well known that germination of the seeds and early growth of the young seedling is difficult or impossible without myeorrhizae The possibility is, of course, clearly енория that the rhizoid-like structures present in the algae may have given rise direetly, in all instances, to the eorresponding structures in early land plants. This наи of the part that mycorrhizae may have played is pre- sented for what it may be wo 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. I T bearing multiseriate bordered pits on radial and tangential walls; prominent sclerotie nests scattered through cortex. Affinities of Scleropteris.— As far as stelar structure is concerned, the affinities of this genus appear to lie with the zygopterid ferns, with which it agrees in the division of the stele into a central mixed pith and a peripheral re- gion of comparatively large tracheids. In both Amkyropteris and Zygopteris a weak development of secondary wood has been re- ported. Compared to these two, the tracheids of Scleropteris seem to be much more regularly arranged and more distinctly secondary. Of the described species of Ankyropteris the closest comparison is with A. corrugata (Holden, ’30) but in that species the tracheids are typically scalariform, quite in contrast to the multiseriate and nearly circular pits of Scleropteris. The closely compacted leaf bases of Zygopteris and the branching of the stem in Scleropteris present contrasting characters which sharply separate these two genera. The cortical sclerotic nests of Scleropteris constitute the most distinctive feature of the fossil and set it apart from any other fern referred to the Coenopteridineae (classification of Hirmer, ’27). These structures are of interest not only for their diagnostic value but because they suggest these possible affinities with the pterido- sperms. First, one of the most prominent features of that group is the dictyoxylon cortex so strikingly displayed in Lyginopteris. It seems evident that this tissue must have had its origin in a less highly organized arrangement of parenchymatous and sclerotic or fibrous cells, and the scattered, slightly vertically elongate nests of Scleropteris may represent a step in that direction. Second, the radial arrangement of the tracheids indicating weak cambial ac- tivity points in a gymnospermous direction. Mesoxylon.— Judging from the frequency of occurrence of Cordaitean stems and roots in the Pyramid Mine petrifactions, this group was well represented in the Carboniferous Lycopod forests of southern Illi- nois. The stem remains described below are referable to Mesoxylon, and although in themselves they present only such features as are already known in the five English species, it has been possible to demonstrate that roots of the Amyelon type, long considered to be of Cordaitean affinities on the basis of association, actually occur in organic connection with Mesoxylon. [Vor. 29 8 ANNALS OF THE MISSOURI BOTANICAL GARDEN The stems are fairly well preserved and the roots exquisitely so. It has not been possible, however, to work out satisfactorily the structure and course of the leaf traces or leaf bases, and as these have been used chiefly in delimiting species it seems advisable to as- sign a new specific name to our specimen. The most significant point in the present discussion lies not in the recording of another species of Mesoxylon based on trivial or even negative characters but in the presentation of conclusive proof that Amyelon is the root of a Cordaitean stem. | The stems attain a diameter of little more than 2 em. including extra-stelar tissues. The pith in the larger specimens averages about 14 mm. and is typically chambered. The peripheral region consists of large parenchymatous cells about 105 и in diameter by 60 u high which are arranged in rather regular vertical rows. Surrounding the pith are prominent masses of primary wood. These are sufficiently well preserved to state with reasonable cer- tainty that the protoxylem is mesarch (pl. 3, fig. 10). The latter seems to be the only distinctive generic character of M esoxylon seg- regating it from Cordaites, and in view of the otherwise close sim- ilarity between the two the validity of its use as a generic character must be looked upon with some doubt. In a previous paper (240) I have dealt with the variability in the primary body of the pterido- sperms and other groups and have pointed out that the position of the protoxylem must be used with reservation аз a taxonomic char- acter. The tracheids of the secondary wood range in cross-section from 15 x 21 и to 21 x 30 и, the tangential dimensions being somewhat greater in most cells. As is usual in Cordaitean stems, there is a broad pitting transition zone bordering the pith, the innermost tra- cheids being spirally thickened. These are followed through a radial distance of about 10 cells by scalariform, retieulate, and pitted cells. The latter are not well preserved, and it is not possible to de- termine the number of rows of pits per cell. The rays are uniseriate and low, ranging from 1 to 3 cells high (pl. 4, fig. 17) with an ocea- sional ray 4 or 5 cells high. Roots.— Well-preserved roots belonging to the form genus Amyelon and associated with Cordaitean stem remains have long been known from the Carboniferous of Europe and have more recently been re- ported in coal-balls from Iowa (Wilson and J ohnston, '40). The 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. I 9 specimens in our collections from the Pyramid Mine consist of minute rootlets bearing well-preserved hairs up to larger roots 3 mm. in diameter. There is some question as to the identity of the smallest rootlets, but their association with the older typical specimens of Amyelon strongly suggests that they belong to that genus. In the youngest roots (text-fig. 1a) the diarch primary xylem consisting of only 8 or 10 cells, may be observed in the center. All of the roots, young or old, studied thus far are diarch, and the de- velopment of the metaxylem is, of course, entirely centripetal. The phloem is not preserved but immediately outside the position it ap- Text-fig. 1. Rootlets, probably of Mesozylon. In a there may be ob- served the large epidermal cells with root hairs, two layers of cortical cells, the dark endodermis and the diarch primary xylem. In b a few secondary xylem cells are present. WCB3IF.T5. x 84. parently occupied is an endodermis the cells of which are filled with a dense brown substance (text-fig. 1, fig. 8). Beyond this is a cortex of larger, thin-walled cells slightly elongated longitudinally. The cells of the epidermis are somewhat larger and thicker-walled. One of the most interesting features of these rootlets is the beauti- ful preservation of the root hairs. These may be seen in pl. 3, fig. 8 and at higher magnification in fig. 9. They are clearly typical root hairs, being outgrowths of the epidermis and composed of but a single cell. Most of the young roots show a trace of secondary xylem starting to form (text-fig. 1b). In all the older roots the secondary wood is strongly developed (pl. 4, fig. 16). This is followed by a narrow band of phloem and a [Vor. 29 10 ANNALS OF THE MISSOURI BOTANICAL GARDEN cortex of rather large cells many of which appear to be resinous. The development of periderm started early and all exeept the smallest roots possess some. The origin of this cork cambial activity was not as deep seated as described by Osborne (709) for the Eng- lish speeimens, since an appreciable thickness of cortex remains in the older roots. As far as this feature is concerned, the roots de- seribed here seem to be intermediate between the English ones and those described by Renault (779) from France. The question of the natural affinities of the roots may now be considered. In 1879 Renault noted their association with Cor- daitean stems and wrote: ** Au milieu des mémes fragments silicifies que renferment les rameaux et les feuilles de Cordaitales on trouve souvent des débris de racines, dont la strueture offre une analogie suffisamment grande avee celle des tiges de ees plantes, pour qu'on puisse les regarder comme ayant appartenu à ees dernieres." Гр. 294]. The apparent although unproven significance of this association seems to have been accepted by most paleobotanists, although nearly half a century later Scott (723) wrote with reference to Amyelon that, ‘‘they agree so well in histological structure with the stem, that there is no reason to doubt the correctness of Renault's conclusions, that they belonged to the same plants, though I am not aware that roots showing structure have yet been found in actual connection with the stem.’’ Гр. 286]. One of our Mesoxylon specimens has been found with Amyelon roots in actual organic connection, and figs. 13-14 of pl. 4 represent three successive ground sections through part of the stem. In the upper right of fig. 13 a noticeable bulge appears in the periphery ; in the next section (fig. 14) the departing adventitious root is very prominent while in the next (fig. 15) it is almost clear of the stem. Fortunately the root was preserved in such a position that immedi- ately after departing from the stem it lay parallel to the latter. Thus, fig. 15 shows the outer part of the root in transverse section and in the next section (fig. 16) the root appears quite separate from the stem and in perfect transverse section. At the left of fig. 15 an- other root may be seen departing from the stem. The American and European specimens of Amyelon are all so nearly alike that it can now be certainly asserted that they belong to Cordaitean stems. Аз is usual with roots, their anatomy is more stereotyped than the other vegetative organs, and there seem to be no known dependable characters that make possible the delimitation 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. I 11 of distinct species of Amyelon. The chief variations in the roots are size, number of protoxylem poles and relative origin of the periderm with reference to the cortex, and none of these (with the possible exception of the last) is sufficient for specific segregation. Mesoxylon Nauertianum Andrews, sp. nov. Diagnosis.— Stems 2 em. or more in diameter; pith large and chambered; peripheral cells arranged in vertical rows; protoxylem mesarch ; secondary xylem tracheids averaging 18 x 25 и in cross-section, pit- ting transition broad, rays uniseriate and mostly 1-3 cells high; roots in organic connection with the stem of the Amyelon type. The species is named in recognition of Mr. N. H. Nauert whose willing assistance in the field and ability as a technician has greatly facilitated my coal-ball studies. Literature cited.— Andrews, H. N. (1940). On the stelar anatomy of the үү with partieular ref- erence to the secondary wood. Ann. Mo. Bot. Gard. 27: 51-118. Darrah, W. C. (1941). Studies of American coal balls. Am. E Sei. 238: 33—53. Hirmer, M. (1927). Handbuch der Palüobotanik. Berlin. Holden, H. S. (1930). On the при and affinities of Ankyropteris corrugata. Phil. Trans. Roy. Soc. London, B 218: 79-114 СЕ zi вв да E ME lateral roots of Amyelon radicans, Will, and their mycor- E 23: ети en (1870). а comparée de quelques tiges de la flore Carbonifére. Paris. Sahni, B. (1932). On the strueture of Zygopteris primaria (Cotta) and on the relations between the genera Zygopteris, Etapteris and Botrychioxylon. Phil. Trans. Roy. Soe. London, B222: 29-45. Scott, D. Н. (1912). On Botrychioxylon Fulco Sp. nov., à biet au fern with sec- ondary wood. Trans. Linn. Soe. London (B E II, 7: 373—389. , (1923). Studios i in Fossil ms d Wilson, L. R., and A. W. Johnston. (1940) re new species 1. Мер: rd from the Penn- ма ban, of Iowa. Bull. Torr. Bot. Club 67: 117-1 [Vor. 29, 1942] 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 1 Scleropteris illinoiensis Transverse section through the stem showing dichotomy of the stele, st. Transverse section through the stem; sc, sclerotic nests; st, stele. WCB90C.T15, Fig. 3. Transverse section through the stele; mp, mixed pith; x, secondary xylem; r, de- parting root. WCB90C.T2, x 20. Fig. 4. Longitudinal section through the outer cortex showing bases of emergences. WCB90B.822, x 17. PLATE 1 1942 Вот. GARD., Vor. 29, ANN. Мо ve ALES i MT VLA) ФА "ае СА - d Us j L] “ee "CP d ۳ „© YN‏ کے جا oF et‏ our‏ LORAS. I 4 NIFEROUS I 'SCAMERICAN CARBO EW DR AN 14 [Vor. 29, 1942] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 2 Scleropteris illinoiensis Fig. 5. Longitudinal section showing branching of stem: st, stele; sc, sclerotic nests. 90B.835, x 4. section through cortex showing an emergence and sclerotic nests. Fig. 6. Longitudinal WCB90B.827, х 8. Fig. 7. Radial longitudinal section through stele: pz, protoxylem (1). WCB90B.822, x 150. PLATE 2 1942 , ANN. Mo. Bor. Garb., Vor. 29 - Жт e, селт C, 7 s^ peated jov - я -- б К+ A SN Tenis. EPA jar КОБО; SEA с, ЖИ САПИИ A ت‎ °» - i ЖҮ. ТТЕ (Жү ыз, + ante mj rt mr > T © ыныр А ek. کک ر‎ А; ©; 4 it V ( у ‚ л, з. Pri d Йй < Sj Aur s Ip. "ow 4 ма) LORAS. I ROUS F 4 ы 4 ARBONII WS—AMERICAN C RE ANDI [Vor. 29, 1942] 16 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 3 Fig. 8. Rootlet, probably belonging to Mesozylon, — root hairs. WCB31F.T4, x 93. Fig. 9. Portion of another rootlet shown at a hi igh xylem. WCB53B-B. 88, stem; r, roots. WCB90C. T9, x Fig. 12. Scleropteris linoiensis. Tangential وا‎ through the secondary xylem showing tracheids and rays. WCB90B. S32, x 70. броен ыы Tangential section through the central portion of the 20. PLATE 3 1942 ANN. Mo. Bor. Garb., Vor. 29 rt M a 25 pias ward M га I —AMERICAN CARBONIFEROUS FLORAS. WS ANDRE [Vor. 29, 1942] 18 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 4 Mesoxylon Nauertianum Figs. 13-15. Stages in the vi hia of a root from the stem: fig. 13, WCB92A.1; fig. 14, WCB92A.2; fig. 15, WCB92A.3. All figures x 5 Fig. 16. Same soot as бен! in ME figures after departure from the stem; д», sec- ondary xylem; ph, phloem; с, cortex; pd, periderm. WCB924A.4, x 20. Fig. 17. Tangential section through the secondary wood of the stem. x 110. PLATE 4 ANN. Mo. Bor. Garb., Vor. 29, 1942 S. I RA EROUS FLO 4 RICAN CARBONII 4 ZWS—AME RI AND CONTRIBUTIONS TO OUR KNOWLEDGE OF AMERICAN CARBONIFEROUS FLORAS! II. LEPIDOCARPON HENRY N. ANDREWS Instructor, Henry Shaw School of Botany of Washington University AND ELOISE PANNELL Assistant in the Henry Shaw School of Botany of Washington University In recent investigations of the American Pennsylvanian coal- balls, no less than three new species of Lepidocarpon have been de- scribed in as many years (Darrah, '41; Hoskins and Cross, '41; Schopf, 38). However, much still remains to be known about this interesting and important fossil, particularly its relationships to the various other Lycopod organs that are frequently found as- sociated with it. It is perhaps especially significant that we are able to present at this time descriptions of both a well-preserved gameto- phyte bearing archegonia and the microsporangiate strobilus of a Lepidocarpon from southern Illinois. Gametophytes have been re- ported for the genus but never undoubted archegonia, so far as we are aware, nor microspore-bearing organs. Materials and Procedure.— All of the material described below was obtained from the Pyra- mid Mine of the Binkley Coal Company, Perry County, Illinois. Our collections contain two species of Lycopod seeds, an Illinio- carpon, probably I. Cadyi Schopf, and the Lepidocarpon described here. Numerous specimens of the latter have been found but thus far we have but one in which the gametophyte is preserved. It was found in a eoal-ball numbered WCB56 in our collections, and all ref- erences in this paper pertain to that specimen. Other parts of this particular seed are not as well preserved as certain of the specimens lacking a gametophyte, particularly WCB55. These two serve as the basis of the following description, although our knowledge of the species has been supplemented by numerous other specimens. The original saw cut through WCB56 was apparently near one 1A study financed in part by,a grant from the Penrose Fund of the American Philo- sophical Society. Issued February 18, 1942. (19) [Vor. 29 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN end of the gametophyte. A ground section was prepared from one half and a series of nearly 200 peels, extending through a radial dis- tance of about one cm., from the other half. In WCB55 a series of 170 peels, as well as some ground sections, was made in order to follow the struetural changes that take place between the proximal and distal ends. Our knowledge of Ше “ае”? strobili is based on abundant micro- spores which occur within the seeds and two partially complete cones bearing mature microspores in situ. Series of transverse and longitudinal sections were prepared through these strobili. Inasmuch as the seeds that we refer herewith to Lepid pon are typical for the genus, it does not seem necessary to describe in detail all of the better-known features. Inclusive generic descriptions may be found in Scott's (201) monumental treatise as well as the more recent work of Hoskms and Cross (241) and Schopf ('41). Our descriptions will stress, rather, those points that are previously un- reported or poorly understood. The Seed.— Seven preparations have been selected from the series made from specimen W CB59 to illustrate in tangential section the struc- ture of a typical seed. These are shown in figs. 1-4, 7, 8 and 13 of pls. 5-7. Figure 1 shows the proximal end of the seed near its attach- ment to the axis; figs. 2 and 3 are from points approximately 14 and 1 mm. farther out, the sporangium appearing in the former; fig. 7 is from the central region; fig. 4 is from a point midway between fig. 7 and the distal end, while figs. 8 and 13 are close to the outside. It may be noted that the seed as a whole is comparatively narrow close to the cone axis and gradually broadens, the base becoming extremely massive toward the distal end. An interesting feature ap- pears in fig. 8 where two loops appear on either side and immedi- ately below the vascular bundle; in fig. 13 these have united, cut- ting off what is, in fact, the lower (dorsal) lobe or ‘‘heel’’ of the sporophyll. The cellular structure of the sporangium wall is well preserved and consists of an outer layer of comparatively thick-walled, brick- shaped cells enclosing a zone of more delicate elongate ones (text- fig. 1c). As will be noted on a later page, the outer mechanical cells are quite different from the corresponding tissue found in the micro- sporangium (text-fig. 1d). At either side of the base of the sporan- gium there is a supporting tissue of thick-walled cells following up 194 ANDREWS & PANNELL—AMERICAN CARBONIFEROUS FLORAS. II 21 the sporangial wall and adjoining portion of the sporophyll for a short distance (text-fig. 2). Although the vascular bundle is well preserved, no part of it can unhesitatingly be called protoxylem. As in Г. Lomasi, transfusion tissue is present at least toward the distal end (pl. 7, fig. 14). Parichnos strands, or the cavities left after the decay of the delicate Cl 7 D ALS KIA R Кусс уы LO TEIN BOT Qu 5 б A <> eS NE И САН ИСААК Igi ө, I DAS 3075 A 25% p IIT a e S| ee Ji p Lu VRAC суай Sr жа] Мао Yeu b d Text-fig. 1. a, Lepidocarpon magnificum, part of the gametophyte bear- the gametophyte versity, x 185; c, a portion of the megasporangium wall of L x 250; d, a portion of the microsporangium wall of L. magnificum, x 250. tissue composing them, are present throughout the length of the sporangial region although they are much more strongly developed in the distal end (figs. 4, 8, 13, of pls. 5-7). The Gametophyte.— A distinctive and rather surprising feature of the gametophyte deseribed here is its small size compared with the seed as a whole [Vor. 29 22 ANNALS OF THE MISSOURI BOTANICAL GARDEN (pl.6,fig. 11). Although both its distal and proximal ends are poorly preserved, the more or less cylindrical form and orientation in life, as shown in fig. 12, remain constant through the series of 200 peels. Its diminutive size, which presents a contrast to other deseribed Lycopod gametophytes, will be discussed in a later paragraph. The gametophyte is composed of two distinct cellular regions as shown in fig. 12 and text-fig. 2: a central core of large isodiametric cells, averaging about 66 и, enclosed by a zone of more or less radi- ally aligned cells, an organization quite similar to that found in the gametophyte of certain living species of Lycopodium. The most striking feature of the specimen lies in the presence, on the upper side, of structures which we believe to be archegonia (pl. 6, fig. 12, ar, and text-fig. 2). Although the neck and venter walls do not ap- pear to have been as generally massive in the fossil, they bear a strong resemblance to the female organs of Lycopodium. Camera- lucida drawings of the archegonia of both genera are presented for comparison in text-fig. 1a and b. The fine preservation of the tissue immediately adjacent to the archegonia seems to preclude the possi- bility that they are artifacts resulting from partial decay of the periphery of the gametophyte. Scott described the rather well-preserved gametophytes of Lepid pon Готаля and L. Wildianum, but they differ in at least two respects from our Illinois specimen. The gametophyte of the English species fills or nearly fills the entire sporangium, while in ours it occupies a much smaller portion and the cellular differentia- tion into a central core and an outer peripheral region is much more pronounced in our specimens. A few other fossil Lycopods have been reported with gameto- phytes, most noteworthy of which are Lepidostrobus V eltheimianus (Scott, '20; Gordon, '10) and Lepidodendron esnostense (Renault, 1896). These species are not as well preserved nor do they present points of close similarity with the one described here. Of the Ameri- ean Lycopod seeds described in recent years (Krick, 32; Reed, 736, 741; Schopf, 38; Darrah, 41; Hoskins and Cross, 241) in only one is the gametophytie tissue preserved. With reference to his Lepidocarpon glabrum, Darrah notes that ‘‘The gametophyte is ex- tensive, nearly filling the whole cavity’’ (p. 99), but his description and illustrations are not sufficiently detailed to allow comparison. The small size of our gametophyte is a feature deserving further comment. Since it is the only one that we have found, it is impossible 1942] ANDREWS & PANNELL—AMERICAN CARBONIFEROUS FLORAS. II 23 ‘to state for certain whether or not it is typical. If so, then it would seem that we are dealing with a more primitive seed than previously described species, such as L. Loma, and опе in which possibly more than one megaspore was functional. WCB55I.i4, WCB55II.03, WCB56B.12, ‘WOBSGE, 39. Itis not entirely improbable that certain of those fossil Lycopods which attained the seed level produced more than one embryo in each seed. Professor A. J. Eames has kindly shown the senior author single gametophytes of living species of Lycopodium col- [Vor. 29 24 ANNALS OF THE MISSOURI BOTANICAL GARDEN lected near Ithaca, N. Y., bearing as many as seven sporelings in ап“ advanced stage of development. Judging from what we know of the structure of mature speci- mens of Lepid pon in general, and that is now quite extensive, they were not seeds constructed to withstand long periods of ad- verse conditions. The survivors must have been those that were de- posited in moist places favorable for growth. A hypothetical Lepidocarpon combining the multiple sporeling character of Lyco- podium with the potentiality of possessing more than one functional megaspore may reasonably be postulated. Deposited, as most of them probably were, under conditions favorable to quick germina- tion such seeds would have been exceptionally effieient, no matter how primitive from a morphological standpoint. That such an or- gan would encounter difficulties under more xerophytie conditions is obvious and probably accounts in large part for the decline of the arborescent forms. The Microsporangiate Cones.— In the earlier phases of this study microspores were observed in considerable numbers within the gametophyte-bearing seed, and further search revealed their presence in other specimens. At least four different species of spores have been found in the seeds, two of which are most likely fern spores (figs. 5, 6, of pl. 5), while other larger ones characterized by a prominent equatorial wing are ref- erable to Endosporites (Wilson and Coe, '40). 'These are briefly mentioned for whatever ecological value they may present. The fourth and by far the most abundant (pl. 6, figs. 9, 10) are believed to be the mierospores of the Lepidocarpon seeds for reasons that will be set forth below. The presence of microspores within the seed is of considerable significance in the establishment of the true morphological status of the genus. With reference to the species that he described, Scott wrote ('01, p. 322) : ** Unfortunately we are at present ч to decide whether мода in Lepido- carpon, took place on the parent plant or not. Mierospores hav been observed within the integument of the megasporangium, nor is there мес Анн of the nature of a pollen-ehamber to receive them. We must therefore take into account the possi- bility that the seed-like organs of Lepidocarpon may have served merely for the protection of the megaspores and prothallus; that they were shed first and received the microspores afterwards. On this supposition, it is possible that all the specimens observed, including those with prothalli, were still unfertilized or even unpollinated. If this were so, the organs in question were certainly not true seeds, though it is pos- sible that the evolution of true seeds may have started in the same way.” 1942] ANDREWS & PANNELL—AMERICAN CARBONIFEROUS FLORAS. II 25 The abundance of mierospores in our seeds renders highly im- probable their presence as the result of chance entry after shedding. It seems certain that they have been deposited through a normal pollination process. We may pass now to a consideration of the microsporangiate cones themselves. Two partially complete strobili have been obtained from the Pyramid Mine, numbered WCB166 and WCB172 in our collections. The former measures approximately 16 cm. long and in life was about 5 em. in diameter, although only one longitudinal half is pre- served. Toward one end the woody axis is complete and quite well preserved, and although most of the sporangia have dehisced a few still retain their spores. In the other specimen (WCB172) the cone is 15 em. long by 5 cm. in diameter and runs through the entire length of the coal-ball. It was apparently somewhat longer in life. The xylem of the cone axis measures 1.4 mm. in diameter and is siphonostelie, enclosing a central pith eavity 0.6 mm. in diameter. The pith, as well as the tissue immediately outside the xylem, is not preserved. About 34 protoxylem points are distributed about the periphery of the woody axis in an exarch position. The structure and arrangement of the sporophylls midway be- tween their proximal and distal ends are shown in fig. 20 of pl. 8. Like the seeds, they become massive toward the outside where parichnos strands are well developed and a ligule is present (pl. 7, fig. 16, pl. 8, fig. 18). The organization of the sporophylls is uniform throughout the length of the cone and likewise the size and form of the contained spores. In fig. 15 a part of one of the cones is illus- trated in which certain sporangia have retained their spores. A por- tion of one of these sporangia is shown at a higher magnification in fig. 19. Two particularly significant features may be noted here: first, the columnar outer cells which are elongated at right angles to the surface (averaging 15 x 45 u) are quite in contrast to the hori- zontally elongated corresponding cells in the megasporophylls (cf. text-fig. 1c and d) ; second, the persistence of the mature spores in tetrads. A few of the latter are shown at a higher magnification in pl. 7, fig. 17. They are uniform in size and shape and average 26 y. If, now, these spores and those present in the seeds (figs. 9, 10) be compared (the two are illustrated at the same magnification), the very close similarity is apparent. However, not only are they strictly comparable in size and form but, as noted above, they fre- quently occur in tetrads in the seeds as well as in the microspor- [Vor. 29 26 ANNALS OF THE MISSOURI BOTANICAL GARDEN angia. The abundance of these microspores in the seeds, combined with the fact that individually they are indistinguishable from those borne in the mierosporangiate cones, leaves no doubt that the latter and the Lepidocarpon seeds belong to the same species. In view of the numerous previously unrecorded features char- acterizing our specimens it seems advisable to introduce a new specific name. Lepid pon magnificum Andrews and Pannell, sp. nov. Diagnosis.— Seeds detached, 19-20 mm. long (parallel to sporophyll pedicel and not including free portion of lamina), approximately 6 mm. broad by 11 mm. high in the central portion; one functional mega- spore occupying about one quarter of the sporangial cavity, col- umnar cells of sporangium wall horizontally elongated, gameto- phyte well developed and composed of a central mass of isodiametric cells surrounded by a broad zone of smaller, more or less radially arranged cells; apparent archegonia borne in the upper portion ; microspores abundant within the sporangium; mierosporangiate cones at least 16 ст. long by 5 em. in diameter, columnar cells of sporangial wall vertically elongated, microspores 26 и in diameter, tending to remain in tetrads when mature. Locality and horizon: Pyramid Mine of the Binkley Coal Company, 3 miles south of Pinckneyville, Illinois ; coal #6, Upper Carboniferous. Discussion.— The microsporangiate cones that we have assigned to Lepido- carpon magnificum agree closely with Lepidostrobus Coulteri Jong- mans as described by Coulter and Land (’11, '21) and Mathews (40). It is highly probable that the latter are specifically сотра- rable with those described here, but although we have no hesitancy in referring our own specimens to L. magnificum it is perhaps best to retain Lepidostrobus Coulteri (for Coulter and Land’s speci- mens) until a monographie work of the group is prepared. In his account of the English species, Scott (’01) mentions a micro- sporangiate cone as questionably referable to L. Готаал but due to the uncertainty involved he did not include a description of it. The supposed relationship of this cone to Lepidocarpon was based оп “The presence of rudimentary integuments, together with agreement in details of structure and the evidence of associa- tion. . . .’’ Гр. 314]. Further discussion of that species can be of little avail until additional colleetions of it are at hand. 1942] ANDREWS & PANNELL—AMERICAN CARBONIFEROUS FLORAS. II 27 It is clear that in the two microsporangiate cones described there is no indication of an integument enclosing the sporangia. They are mature and strictly microsporous and, in themselves, typical Lepidostrobus cones. However, the occurrence of the microspores in the seeds, as noted in detail above, affords sufficient evidence in our opinion for their reference to L. magnificum. The only possible doubtful feature of our taxonomic treatment of the microsporangiate cones and seeds is the contrasting structure of the outer mechanical cells of their respective sporangia. It must be remembered, however, that this tissue is a highly specialized de- hiscence mechanism allowing dissemination of the microspores, and likewise megaspores in forms such as Selaginella and Lepidostrobus where a true seed habit has not been reached. With the evolution of the permanent retention of a single functional megaspore the need for such a tissue diminishes, although the sporangium must of course be opened sufficiently to admit the microspores. Inasmuch as the function of the microsporangium apparently remained more nearly static during this seed development, it is logical that its struc- ture should remain the same. For the same reason the absence of an ‘‘integument’’ from the microsporangium in no way invalidates the apparent relationship to the seed. It becomes increasingly clear that the Lepidocarpaceae is a di- verse group and one whose taxonomy will undergo frequent revi- sions in the future. Acknowledgment.— Weare indebted to Professor A. J. Hames for his helpful criticism and for the opportunity to study his fine collection of Lycopodium gametophytes. Literature cited.— Z ie ae Жа, and W. J. G. Land. (1911). An American Lepidostrobus. Bot. Gaz. 51: —— — — , ——————, (1921). A homosporous American Lepidostrobus. Ibid. 72: 106—108. Darrah, W. C. (1941). The fossil flora of Iowa eoal-balls. IV. Lepidocarpon. Bot. Mus. Leaf. Harv. Univ. 9: 85-100. B W. T. (1910). Notes on the prothallus of Lepidodendron Veltheimianum. Ann. . 94 821-822. Hoskins J. H., and A. T. Cross. VA $ consideration of the structure РА и fasi on a new strobilus from Midl. Nat. 25: 52 Sco Krick, Ha V. (1932). не oí аа pistes found in ш ч A Harris- burg, Illinois. Bot. d 93: 151-172. т G. B. (1940). w Lepidostrobi from Central United States. Ibid. 102: 26—49. Reed, F. D. (1936). Жане ы sporangia from the Upper Carboniferous of Illinois. Ibid. РА 307—316. [Vor. 29, 1942] 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN — — —— , (1941). Coal flora studies: ——Á Ibid. 102: 663—683. m— в. аА Étude des gites mineraux a France. Bassin Houiller et Permien et d'Épinae. Flore Fossile. Paris sr. Ж, М. (1938). Two new Lycopod seeds rn the Illinois Pennsylvanian. Ill. State Geol. Surv. Cir. 28: 139—146. 1941). Notes on the Lepidoearpaceae. Am. Midl. Nat. 25: 548-563. Scott, D. 'H. (1901). On the structure and affinities of fossil plants from the Paleozoic rocks. IV. The seed-like fructifications of Lepidocarpon, a genus of Lycopodiaceous cones ees the Carboniferous a. Phil. Trans. Roy. Soc. Lond. B194: 291-333. , (1920). Studies in Fossil Botany. I. Wilson, L. R., and E. A. Coe. ( bro Deseriptions of some unassigned plant mierofossils from the Des Moines Series of Iowa. Am. Midl. Nat. 23: 182—186. EXPLANATION OF PLATE PLATE 5 Figs. 1-4. Lepidocarpon magnificum. Part of a series of peel preparations starting near the point of attachment to the cone axis and progressing toward the distal end: fig. 1, WCB551.02; fig. 2, WCB55I. - 26. 3, WCB551.013; fig. 4, WCB55ILi2. АП figures x 11 except fig. 4, which is Figs. 5, 6. Foreign spores, probably А Боса found within the seed of Lepidocarpon: fig. 5, WCB56B.12; fig. 6, WCB56B.41. Both figures x 700. PLATE 5 ANN. Mo. Вот. Garp., Vor. 29, 1942 eum Bn : AS: Н LOR Е I x “EROU RICAN CARBONII LL—AME ^ 4 " PANNE SWS & DRI AN [Vor. 29, 1942] 30 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 6 Lepidocarpon magnificwm Figs. 7, 8. Part of the series shown in figs. 1-4, explanation in text: fig. 7, WCB551I.016; fig. 8, WCB55IILO01. Both figures х 10. Figs. 9, 10. Microspores found within the seed: fig. 9 shows characteristic retention of the tetrad n s ipe WCB56B.9; fig. 10, a single spore showing triradiate commissure, WCB56B.23. Both figures x 700. Figs. 11, 12. т of the specimen bearing a gametophyte: fig. 11, from WCB56B.1, x 12; fig. 12, era ET ا‎ ar, archegonia; mw, megaspore wall; from WCB56B. 18, x ANN. Mo. Bor. GARD., VoL. 29, 1942 PLATE 6 iode VE f “ wry E PA | ed х x ha 4 ANDREWS & PANNELL—AMERICAN CARBONIFEROUS FLORAS. II [Vor. 29, 1942] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 7 Lepidocarpon magnificum Fig. 13. Tangential section near the distal end of the sporophyll shown in figs. 1-4, 7 and 8; p, parichnos; i, integument; sp, sporangium wall; h, heel of sporophyll; v, vascular bundle. From WCB55II1.09, x 10. Fig. 14 Enlarged bundle region of the seed, p v, vascular bundle, te, transfusion h part of a microsporangiate cone. Most of the 166B2.83, x 3.1. a microsporophyll Agr near the ter end; sporangium; 1, ligule; nt sporophyll. WCB172C.816, x Fig. 17. A few microspores sh WCB166B1.82, x 700. sp, owing the characteristic леи of the tetrad. , ANN. Mo. Bor. Garp., Vor. 29, 1942 PLATE 7 ANDREWS & PANNELL—AMERICAN CARBONIFEROUS FLORAS. II [Vor. 29, 1942] 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 8 Lepidocarpon magnificum Fig. 18. Tangential section of part of the microsporangiate cone near the distal ex- tremity of the sporophylls. WCB172C. 3.5 Fig. 19. Part of a sporangium containing “iin from the specimen illustrated in fig. 15. WCB166B1.82, х 80. Fig. 20. Tangential section through a mierosporangiate cone, taken about midway be- tween the cone axis and distal end of the sporophylls, showing the dehisced sporangia. WCBI172C2.82, x 3 PLATE 8 ANN. Мо. Bor. Garb., Vor. 29, 1942 AMERICAN CARBONIFEROUS FLORAS. II ANDREWS & PANNELL— NEW GESNERIACEAE FROM PANAMA! C. V. MORTON Assistant Curator, United States National Herbarium, Washington, D. C. The many botanical collections that have been made recently in Panama, chiefly under the auspices of the Missouri Botanieal Gar- den, have added greatly to our knowledge of the family Gesneri- aceae. The plants of this family are exclusively tropical and often epiphytie. Until the last few years fewer than 20 species have been known from Panama. This number has been much more than doubled. There are known now over 30 species of Columnea alone, more than are recorded from Costa Riea, formerly considered one of the richest regons for Gesneriaceae. Almost all the novelties be- long to the subfamily Columneoideae, which reaches its greatest development in Colombia. The greatest number of new species have been discovered by Mr. Paul H. Allen, an exceedingly discerning collector, with an unusual faculty for picking out interesting and rare plants. Most of his re- cent collections have come from a locality cited as ‘‘hills north of El Valle de Anton, Province of Coclé, altitude 1000 meters.’’ This locality is deseribed by Mr. Allen in a letter dated Feb. 2, 1941, as follows: ** About a week ago I had the opportunity of spending four days in El Valle, in Coclé Province, and I am astonished at the tremendous wealth of the place. This time I was able to get up into the high plateau regions to the north, and never in all my collecting experience have I seen such a fantastic region or such fascinating plants. After leaving the crater proper, there is a line of buffer hills to the north, rising in three rounded domes topped with an elfin forest and over which a steady cascade of fog and misty rain pours. To the north of these hills is a maze of valleys, with a most curious set of plateaus, some not over an acre or two and others probably forty or more acres. The tops are almost perfectly flat, with deep rich black soil, supporting a growth of giant trees which literally drip with epiphytes of all descriptions. There are almost always practically sheer drops on nearly all sides of these places, some- times going down for several hundred feet. This last time there I came to such a precipice and could look down six or seven hundred feet below to where the tumbled hills and е valleys ran off to the north. БҮ? the mists would elear fora and far o e Atlantic Ocean could be . . I think I ean guarantee that you m get some interesting plants." Mr. Allen was quite right. On this trip, as on several later ones, he found many new and interesting plants, especially in the Ges- ! Published by permission of the Seeretary of the Smithsonian Institution. Issued February 18, 1942. (35) [Vor. 29 36 ANNALS OF THE MISSOURI BOTANICAL GARDEN neriaceae. The affinity of the flora of this region is decidedly with the mountain flora of Costa Riea. In addition to the plants collected by Mr. Allen, I have had the advantage of studying collections made by Dr. Robert E. Woodson, Jr. and Mr. R. W. Schery. These include several novelties, here de- scribed. A small, but very interesting collection, made by Mr. and Mrs. Terry in the mountains of the Province of Darien, was kindly made available by the Field Museum of Natural History. The present paper lists only new species and new records. A com- plete account of the Gesneriaceae of Panama will be published at a later date. ALLOPLECTUS CORIACEUS (Oerst.) Hanst. Linnaea 34: 363. 1865- 66.—вослз DEL токо: Crieamola, region of Almirante, Jan.-March, 1928, Cooper 207. Distributed as Drymonia sp. Previously known from Costa Rica. ALLOPLECTUS ICHTHYODERMA var. PALLIDUS Morton, Field Mus. Publ. Bot. 18: 1145. 1938.—cutrigui: Casita Alta to Cerro Copete, alt. 2300-3300 m., July 10, 1940, Woodson в Schery 359. Previously known from Costa Rica. ALLOPLECTUS panamensis Morton, sp. nov. Herba lignosa, 1 m. alta; caules non ramosi, crassi, usque ad 12 mm. diam., apice ca. 5 mm. diam., tetragoni, lateraliter compressi, dense flavo-subhirsuti ; folia per paria aequalia, longe petiolata, petiolo 5-11 em. longo, hirsuto; lamina foliorum ovata 21-23 em. longa, 12.5-15 em. lata, breviter acuminata, basi rotundata, aequalis, chartacea, serrulata, supra viridis, pilosa, subtus pallidior, non rubro-maeulata, ubique dense pilosa, venis lateralibus ca. 8-jugis, supra subobscuris; in- florescentia axillaris, densiflora, bracteis rubris, ovatis, са. 1.5 em. longis, acuminatis, denticulatis, pilosis, pedicellis 2-3 em. longis, dense hirsutis; calyx ruber, lobis subaequalibus, oblongis, ca. 2.5 em. longis, 1 em. latis, acutis, herbaceis, glanduloso-serratis (denti- bus ca. 11 utroque latere), externe dense flavo-hirsutis, intus stri- #0815 ; corolla coccinea, 4.5-5 em. longa, in calyce obliqua, basi cal- carata (4 mm.), tubo ca. 4 mm. diam. supra basim, sursum gradatim ampliato, non ventricoso, externe dense hirsuto, intus glabro, fauce 1 cm. lato, obliquo, limbo са. 1.5 em. lato, obliquo, lobis recurvis, sub- aequalibus, suborbiculatis, ca. 4 mm. longis, subintegris, intus glabris; filamenta basi vix in tubum connata, im corolla alte ad- nata, glabra; antherae inclusae, cohaerentes, transverse oblongae, 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 37 ca. 2 mm. longae, 3 mm. latae; ovarium hirsutum; stylus glaber; stigma bilobum; disci glandula postiea solum evoluta, magna, crassa, glabra, emarginata.—cocrÉ: hills north of El Valle de Anton, alt. 1000 m., July 14, 1940, Paul H. Allen 2189 (U. S. Nat. Herb., no. 1,791,876-7, турЕ); same locality June 23, 1940, Allen 2166. On Cerro Campana in the Province of Panama a related species has been collected (Allen 2412), which at first I considered the same. However, I am sure now that it is different, but the material lacks corollas and is insufficient for a description. Alloplectus panamensis belongs to the section Glossoloma, but is not closely related to the other North American species, A. tetra- gonus and A. simulatus, which have appressed pubescence on stems and leaves. They have, also, cuneate-based leaves, those of A. pan- amensis being much broader and rounded-truncate at base. ArnLoPLEcTUS simulatus Morton, sp. nov. Herba 1-2 m. alta; eaules non ramosi, tetragoni, apice ca. 3 mm. diam., appresso- puberuli ; folia per paria aequalia, longe petiolata, petiolo 9.0-6 cm. longo, puberulo; lamina foliorum late elliptica, 14-17 cm. longa, 6—7.5 em. lata, aeuminata, basi late cuneata, paullo obliqua, sub- herbacea, serrulata, utrinque viridis et sparse puberula, pilis flavidis, flaccidis, paucicellularibus, venis lateralibus ca. 9-jugis; inflorescentia axillaris, pauciflora, bracteis deciduis, pedicellis 2-3 em. longis, hirsutis; calyx rubro-aurantiacus, lobis subaequalibus, ovatis, 2.5-2.7 em. longis, acutis, argute serratis, externe hirsutis, praecipue in linea mediali; corolla externe rubro-aurantiaca, intus lutea, ca. 5 em. longa, in calyee obliqua, tubo faucem versus ampli- ato, densissime hirsuto, limbo valde obliquo, ca. 1.7 em. lato, lobis subaequalibus, recurvis, intus glabris.—cnrimiQví: vicinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July 18, 1940, R. Е. Woodson ё В. Ў. Schery 53? (U.S. Nat. Herb., no. 1, 808, 045, TYPE). Closely related to A. tetragonus of Costa Rica, but differing in pubescence. In A. tetragonus the pubescence of the stems and lower leaf surface is closely appressed, but in A. simulatus the hairs are not appressed, but tortuous, very short and few-celled. In the Gesneriaceae, as a whole, the character of the hairs is very constant within the species, so this character alone would be sufficient to separate these two species. There are other differences also. The calyx of A. simulatus is larger and broader, the corolla more densely hirsute and the limb broader. [Vor. 29 38 ANNALS OF THE MISSOURI BOTANICAL GARDEN Bester Allenii Morton, sp. nov. Frutex 1-1.5 m. altus; caules non ramosi, teretes, graciles, са. 3 mm. diam., apicem versus 1 mm. diam., dense et breviter puberuli ; folia per paria aequalia, petiolata, petiolo 0.8-4 em. longo, gracili, puberulo ; lamina foliorum oblonga, 7-13 em. longa, 3-4.7 cm. lata, acuminata, basi late euneata, paullo obliqua, membranacea, irregulariter serrata, supra atroviridis, gla- bra, subtus pallidior, praecipue in venis strigosa, venis lateralibus 10- vel 11-jugis subtus prominulis; inflorescentia axillaris, umbel- lata, 3- vel 4-flora, pedunculo communi ca. 2.5 em. longo, parce pilosulo, pedicellis ca. 12 mm. longis; calyx viridis, lobis paullo in- aequalibus, lobo postico ovato, reflexo, ca. 1.4 em. longo, 7 mm. lato, lateralibus ovato-lanceolatis, obliquis, ca. 15 mm. longis et 6 mm. latis, anticis lanceolatis, ca. 15 mm. longis et 5 mm. latis, omnibus longe et acriter acuminatis, argute serratis, externe hirsutis, intus glanduloso-pilosulis; corolla flava, 3.5-4 ст. longa, in calyce hori- zontalis, basi postice calcarata (7 mm.), tubo basi са. 5 mm. diam., sursum valde ventricoso, faucem versus non contracto, 1.8-2 em. lato, externe glanduloso-pilosulo, intus glabro, limbo magno, termi- nali, bilabiato, lobis latis, rotundatis, intus glabris; filamenta gla- bra; antherae inclusae, cohaerentes, ca. 1.5 mm. longae et 2 mm. latae, loculis confluentibus; ovarium pilosulum; stylus pilosulus; stigma bilobum; discus annularis, antice evanescens, postice mag- nus, crassus, puberulus.—cocré: vicinity of La Mesa, north of El Valle de Anton, alt. 1000 m., May 12, 1941, Paul H. Allen 2371 (U. S. Nat. Herb., no. 1,793,964, TYPE). This species belongs to section Neobesleria subsect. Wend- landianae. From Besleria W endlandiana Hanst., of Costa Rica and Panama, it differs in the strongly ventricose, very broad corolla. From B. columbiana Morton it differs in the shorter peduncle and longer pedicels, the much larger calyx lobes, and the much larger and broader corolla. BESLERIA ALLENI var. paucivenia Morton, var. nov. A var. typica venis lateralibus paucioribus (7- vel 8-jugis), inflorescentia uni- flora, pedunculo communi brevi (1-1.5 em. longo), pedicellis longi- oribus (1.5-2 em. longis) differt.—cocrf: hills north of El Valle de Anton, alt. 1000 m., July 14, 1941, Paul H. Allen 2188 (U. S. Nat. Herb., no. 1,806,127, түрк). BESLERIA BARBENSIS var. hirsuta Morton, var. nov. Herba 1 m. alta; caules non ramosi, ca. 3 mm. diam., apicem versus ca. 2 mm. 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 39 diam., hirsuti; folia per paria aequalia, petiolata, petiolo 2-3 em. longo, gracili, hirsuto; lamina foliorum elliptica, 8-12 cm. longa, 4 5.5 em. lata, breviter acuminata, basi cuneata, aequalis, herbacea, inconspicue serrulata, supra viridis, subtus pallidior, utrinque hirsuta, venis lateralibus 11- vel 12-jugis, subtus prominulis; in- florescentia axillaris, uniflora, pedunculo communi nullo, bracteis nullis, pedicello 3-3.5 em. longo, gracili, hirsuto ; calycis lobi virides, fere liberi, paullo inaequales, lobo postico ovato, ca. 12 mm. longo et 7 mm. basi lato, alteris oblongis, ca. 13 mm. longis, 5 mm. latis, omni- bus acutis, argute serratis (dentibus subulatis, apice glandulosis), externe hirsutis, intus glabris; corolla aurantiaca, ca. 2 cm. longa, in calyce erecta, basi vix saccata, tubo supra basim са. 4.5 mm. lato, sursum paullo ventricoso et 7 mm. lato, externe apice hirsuto, intus lineam horizontalem sparse pilosam gerente, limbo satis magno, ca. 12 mm. lato, terminali, bilabiato, lobis orbiculatis, pat- entibus, ca. 5 mm. longis et 7 mm. latis, rotundatis, subintegris, ex- terne pilosis, intus glabris; filamenta glabra; antherae cohaerentes, inclusae, quadratae, ca. 1.5 mm. longae et 2 mm. latae, loculis con- fluentibus; ovarium apice pilosum; stylus pilosulus; stigma bi- lobum ; discus annularis, integer, glaber, brevis, crassus.—CHIRIQUI: on the trail from Cerro Punta to headwaters of Rio Caldera, alt. 9250-2500 m., Jan. 14, 1939, Paul H. Allen 1449 (U.S. Nat. Herb., no. 1,749,713, TYPE). Bestera crassicaulis Morton, sp. nov. Herba 1.5-2.1 m. alta; caules crassissimi, non ramosi, са. 11 mm. diam., apice са. 9 mm. diam., teretes, densissime flavo-sericei; folia per paria aequalia, petiolata, petiolo 4-4.5 cm. longo, erassissimo, 5-9 mm. diam., densissime flavo-sericeo; lamina foliorum oblonga, magna, plus quam 30 em. longa, 10 em. lata, abrupte et breviter acuminata, basi obtusa vel late cuneata, subaequalis, integra, subcoriacea, supra viridis, glabra, subtus pallidior, praecipue in nervo mediali et venis primariis sericeo-strigosa, venis lateralibus verisimiliter ca. 12- jugis; inflorescentia axillaris, multiflora, pedunculo communi ob- soleto, pedicellis numerosissimis, 12-15 mm. longis, gracilibus, sub- hirsutis; calycis lobi orbiculares, 5 mm. longi, 4 mm. lati, herbacei, rotundati, non mucronati, venosi, integri, externe praecipue basim versus subhirsuti, intus glabri, margine valde ciliati; corolla coc- cinea, ca. 17 mm. longa, in calyce paullo obliqua, basi saccata, tubo supra basim 4 mm. diam., sursum non ventricoso, 5 mm. diam., in fauce non contracto, externe glabro, intus glabro, limbo буша, [Vor. 29 40 ANNALS OF THE MISSOURI BOTANICAL GARDEN subregulari, ca. 6 mm. lato, lobis suborbiculatis, ca. 2.5 mm. longis, subintegris, rotundatis, utrinque glabris; filamenta glabra; an- Шегае loculi parvi, са. 0.8 mm. longi, confluentes; ovarium fere glabrum, apice breviter pilosulum; stylus sparse pilosulus; stigma bilobum; discus annularis, glaber, brevis.—DpamrEN: Cana-Cuasi Trail, Chepigana District, alt. 600 m., Mar. 12, 1940, M. E. € В. А. Terry 1534 (Herb. Field Mus., no. 1,035,994, түрк). A species recalling the Colombian Besleria maxima Morton and B. tambensis Morton in its large leaves, very thick, densely hairy stem, and congested inflorescences. However, B. crassicaulis is probably not closely allied to either, but rather to B. notabilis Morton of Costa Rica, which has smaller, oblong calyx lobes only 3 mm. long and 1.5 mm. wide, and a smaller unspurred corolla. ВезтевтА obliqua Morton, sp. nov. Frutex 2 m. altus; caules ramosi, teretes, ca. 5 mm. diam., apicem versus ca. 2 mm. diam., breviter et dense scabro-pilosuli; folia per paria subaequalia, peti- olo 1-2 em. longo, scabro-pilosulo; lamina foliorum lanceolata, 4-7 em. longa, 1.2-2.1 em. lata, aeuminata, basi euneata, subaequalis, integra, chartacea, supra viridis, scabra (pilis patulis, basi valde in- flatis), subtus pallidior, ubique pilosula, venis lateralibus 6-jugis; inflorescentia axillaris, pauciflora, peduneulo communi obsoleto, pedicellis ca. 5 mm. longis, pilosulis; calyx viridis, quam corolla multo brevior, lobis ovatis, paullo inaequalibus, postico ca. 2.5 mm. longo et lato, alteris ca. 3 mm. longis et 2 mm. latis, omnibus fere liberis, obtusis, non mucronatis, non venosis, externe pilosulis, intus glabris; corolla aurantiaca, са. 17 mm. longa, in calyce subhori- zontalis, basi subealearata (2 mm.), tubo supra basim са. 4 mm. diam., sursum tubuloso, non ventrieoso, externe sparse pilosulo, intus basi glabro, sursum glanduloso-pilosulo, in fauce non vel vix contracto, ca. 5 mm. lato, limbo terminali, ca. 7 mm. lato, lobis sub- orbiculatis, subaequalibus, ca. 3 mm. longis et latis, utrinque glabris, integris, rotundatis; filamenta basi in tubum antice са. 3 mm. longum connata, glabra, partibus liberis brevibus, ca. 4 mm. longis, glabris; antherae liberae, inelusae, parvae, ca. 0.7 mm. longae, lo- culis divergentibus, apice confluentibus ; staminodium magnum, са. 4mm. longum, antheram sterilem gerens ; ovarium pilosulum ; stylus pilosulus, brevis; stigma paullo bilobum ; discus annularis, glaber, antice brevior.—cocrÉ: vicinity of La Mesa, north of El Valle de Anton, alt. 1000 m., May 12, 1941, Paul H. Allen 2396 (U. S. Nat. Herb., no. 1,793,974, TYPE). 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 41 This species is most closely related, perhaps, to B. miniata Mor- ton, and, like that, wil key out to section Neobesleria subsect. Axillares in my “Revision of Besleria.’’! I think, however, that these two species, and perhaps also B. flava Morton, are best re- ferred to Eubesleria sect. Sessiles. From B. miniata of Ecuador the present species differs strongly in pubescence, the leaves in that species being glabrous above and merely strigillose on the veins be- neath. In B. obliqua the leaves are densely scabro-pilose above, the hairs being very strongly inflated at base, and are spreading-pilosu- lous all over the lower surface. The only Central American species with which B. obliqua could be confused is В. solanoides H.B.K., but that also has leaves nearly glabrous above and merely strigose beneath, and the corolla is erect and not horizontal in the calyx. CAMPANEA CHIRIQUANA Morton, Ann. Mo. Bot. Gard. 26: 310. 1939.—cnumiqví: Valley of Río Chiriquí Viejo, near Monte Lirio, Apr. 3, 1938, Gene White 62. Previously known only from the type. САМРАХЕА Номвоготп (Kl.) Oerst. Centralamer. Gesner. 34. 1858.—cnuimüiqví: Valley of Upper Río Chiriquí Viejo, alt. 1300- 1900 m., July 17, 1937, Peggy & Gene White 20; vicinity of Bajo Chorro, alt. 1900 m., July 20—22, 1940, Woodson Ф Schery 656. Orig- inally collected in Veraguas, Panama, but not again collected in Panama until recently. CENTROSOLENI1A lineata Morton, sp. nov. Herba epiphytica, sub- caulis ; caudex brevis, erassissimus, ramosus, albidus, glaber; folia petiolata, numerosa, conferta, petiolo usque ad 5 em. longo, crassis- simo, basi parce pilosulo, sursum glabro; lamina foliorum sub- oblanceolata, 14—24 cm. longa, 3.8-6.8 cm. lata, apice longe acumi- nata, basi longe cuneata, decurrens, remote denticulata, supra viridis, subtus siccitate brunnea, utrinque glaberrima, venis lat- eralibus 9-jugis, venis secondariis obscuris; flores axillares, fas- ciculati, numerosi, pedicellis 6-12 mm. longis, gracilibus, glabris; ealyeis lobi rubro-brunnei, liberi, subulati, filiformi-attenuati, basi 1.5-2 mm. lati, medio 0.2-0.5 mm. lati, inaequales, anteriores ca. 18 mm. longi, posticus reflexus, ca. 13 mm. longus, omnes externe parce strigillosi, apicem versus ciliati ; corolla alba, externe et intus rubro- brunneo-lineata, 4 cm. longa, in calyce horizontalis, basi calearata (2 mm.), tubo basi ca. 3 mm. diam., gradatim ampliato, subinfundi- ! Contr. U. S. Nat. Herb. 26: 404. 1939. [Vor. 29 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN buliformi, non ventricoso, in fauce non contracto, ca. 15 mm. lato, externe glabro, intus glanduloso-pilosulo, limbo bilabiato, ca. 2 em. lato, lobis albis, patentibus, lobo antico majore, orbiculato, са. 1 em. longo, rotundato, externe glabro, intus valde pilosulo, glanduloso- ciliato, lobis alteris subaequalibus, minoribus, ca. 5 mm. longis, utrinque glabris, glanduloso-ciliatis; filamenta basi im corollae tubo adnata, per paria connata, glabra, apice incurva; antherae glabrae, non barbatae, inclusae, connatae, subquadratae, ca. 1.2 mm. longae et 1.5 mm. latae, connectivo crasso, loculis diseretis, di- vergentibus; ovarium pilosum; stylus crassus, glanduloso-pilos- ulus; stigma stomatomorphum; disci glandula postiea solum evo- luta, magna, crassa, glabra, integra.—cocrÉ: hills on trail to La Mesa, north of El Valle de Anton, alt. 1000 m., Aug. 31, 1941, Paul Н. Allen 2717 (U.S. Nat. Herb., no. 1,821,127, TYPE). In habit somewhat similar to Centrosolenia decurrens Morton, of Costa Rica, which may be distinguished by the following key. The genus Centrosolenia is new to Panama. * Corolla tube ventrieose at middle; anthers bearded at apex; leaves strigose on both surfaces, hirsute on veins beneath......... ccce esee nnn C. decurrens Corolla tube infundibuliform, gradually widened to throat; anthers glabrous; leaves о clic a oi cy EE eK nhs баж cers C. lineata CoLUMNEA Allenii Morton, sp. nov. Planta epiphytica, pendens; eaules vix ramosi, graciles, ca. 1.5 mm. diam., parce strigosi ; folia aequalia, breviter petiolata, petiolo ca. 3 mm. longo, strigoso; lamina foliorum oblongo-elliptica, usque ad 2 em. longa et 1.1 em. lata, breviter acuminata, basi rotundata, haud obliqua, integra, crassa, utrinque glabra, venis utrinque immersis, obscuris; flores axillares, solitarii, ebracteati, pedunculo 1.7-2 em. longo, rubro- strigoso, pilis multicellularibus, flaecidis; calyx erectus, rubro- tinctus, lobis paullo inaequalibus, 2.2-3 em. longis, са. 1 em. basim versus latis, longe et acriter acuminatis, integris, externe vix strigil- losis, ciliatis, intus basim versus longe hirsutis ; corolla coccinea, in ealyce erecta, 6.8-7.5 em. longa, basi subealearata (3 mm.), tubo calycem aequante, supra basim 4 mm. diam., sursum ampliato, fauce ca. 15 mm. lato, externe parce piloso, limbo valde bilabiato, galea 4—4.5 cm. longa, ca. 2.5 em. lata, apice truncata, lobis lateralibus cum galea alte connatis, ca. 14 mm. longis, lobo inferiore reflexo, oblongo- oblanceolato, 2.7-3 em. longo, 7-8 mm. lato; filamenta basi in tubum 4 mm. longum postice fissum connata, glabra; antherae exsertae, per paria connatae, oblongae, ca. 3 mm. longae, 1 mm. latae ; ovarium 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 43 albo-pilosum ; stylus pilosulus ; stigma bilobum.—cocr£: hills north of El Valle de Anton, alt. 1000 m., June 23, 1940, Paul H. Allen, 2179 (U.S. Nat. Herb., nos. 1,791,874—5, TYPE). CoLuMNEA arguta Morton, sp. nov. Planta epiphytica; caules pendentes, elongati, graciles, ca. 1.5 mm. diam., hornotini rubro- hispidi, pilis rigidis, patentibus; folia aequalia, brevissime peti- olata, petiolo ca. 1 mm. longo, hispido; lamina foliorum lanceolata, 1.6-2 em. longa, 6-7 mm. lata, longe acuminata, basi rotundata, paullo obliqua, exaurieulata, non amplexicaulis, crassa, integra, utrinque glabra, margine ciliata excepta, supra viridis, subtus rubescens, venis lateralibus 1- vel 2-jugis, obscuris; flores veri- similiter solitarii, breve pedunculati, pedunculo crasso, hispido; calyx erectus, lobis subaequalibus, ca. 15 mm. longis, 9 mm. latis (dentibus inclusis), utrinque hirsutis, laciniatis (dentibus elon- gatis, usque ad 3 mm. longis et 1 mm. latis, ca. 5 utroque latere) ; corolla in ealyce erecta, rubra, fauce luteo-lineata, ca. 6 em. longa, basi subealearata (3 mm.), tubo supra basim 4 mm. diam., sursum paullo ampliato, ca. 1 ст. lato, externe sparse piloso, intus basi glanduloso, limbo bilabiato, intus glabro, galea 2 cm. longa, apice truneata, latissima, ca. 2.7 em. lata, lobis lateralibus magnis, ca. 13 mm. longis, lobo inferiore elliptico, гейехо, ca. 2 ст. longo et 1.1 em. lato, rotundato; filamenta in tubum 4 mm. longum postice fissum connata, glandulosa, sursum glabra; antherae ignotae; ovarium sericeum ; stylus sparse pilosulus ; stigma truncatum ; disci glandula postica solum evoluta, glabra, crassa, bidentieulata.—cocrÉ: hills north of El Valle de Anton, vicinity of La Mesa, alt. 1000 m., Jan. 21, 1941, Paul Н. Allen 2336 (U. S. Nat. Herb., no. 1,806,120, TYPE). The form of the corolla is unusual in the subgenus Eucolumnea. The galea is short and very broad, the lateral lobes large, and the lower lobes very much broader than in the other species. The corolla approaches that of section Pentadenia, but it is not ventricose or eurved. It differs also from the Panama species of Pentadenia in its glabrous leaves. The closest relationship is apparently with Columnea flaccida Seem. (Bot. Voy. Herald 186. 1852-57), originally described from near Gualaca, Veraguas, Panama, which has been collected a few times in Costa Rica but not since found in Panama. It is a pendent epiphyte, flowering while leafless, with stems sparingly strigose (rather than coarsely hispid as in C. arguta), and corolla with a shape characteristic of Eucolumnea. [Vor. 29 44 ANNALS OF THE MISSOURI BOTANICAL GARDEN CorvuMNzA citrina Morton, sp. nov. Planta terrestris; caulis са. 60 em. longus, non ramosus, basi ca. 8 mm. diam., apice 3 mm. diam., juventute strigosus; folia per paria valde inaequalia, majora ob- longo-linearia, 20-25 cm. longa et 5 em. lata, falcata, sessilia, longe acuminata, basi inferiore auriculata, semiamplexicaulia, superiore cuneata, crassa, supra glabra, viridia, subtus substrigosa, pallidi- ora, ea. 6 em. infra apicem maculam гагат са. 8 mm. longam gerentia ; folia minora stipuliformia, sessilia, lineari-lanceolata, са. 2 em. longa et 5 mm. lata, longe acuminata, basi valde obliqua, basi inferiore auriculata et amplexicaulia, superiore rotundata; inflo- rescentia axillaris, biflora, bibracteata, bracteis flavis, basalibus, lineari-lanceolatis, са. 1.5 ст. longis et 5 mm. latis, longe acumi- natis, externe strigosis, intus glabris, pedunculis erassis, 1-1.5 em. longis, dense strigosis; calyx pallide viridi-luteus, in anthesin са. 3 em. longus, demum 3.8 em. longus, lobis aequalibus, erectis, lance- olatis, liberis, ca. 12 mm. latis, longe et acriter acuminatis, integris, externe substrigosis, intus glabris, linea mediali excepta; corolla lutea, intus roseo-lineata, са. 4.7 ст. longa, in calyce erecta, basi subealearata (са. 2 mm.), tubo 3 mm. super basi 5 mm. diam., sursum ventrieoso et 1 em. diam., apicem versus paullo contracto, fauce tubuloso, externe valde hirsuto, limbo valde bilabiato, fauce et lobis intus glabris, galea 2.5 em. longa, apice alte biloba (ca. 7 mm.), lobis lateralibus eum galea alte connatis, ca. 12 mm. longis, lobo inferiore reflexo, lineari-oblongo, 18 mm. longo, 5 mm. lato; filamenta basi in tubum 4 mm. longum connata, glabra; antherae exsertae, 3 mm. longae et 2.5 mm. latae; staminodium subulatum, 3 mm. longum; ovarium dense albo-sericeum; stylus glaber; stigma stomato- morphum; disci glandula postica solum evoluta, magna, crassa, glabra, subintegra; placentae lamellae intus solum ovuliferae.— PANAMÁ: Cerro Campana, alt. 1000 m., April 21, 1941, Paul H. Allen 2404 (U. S. Nat. Herb., nos. 1,793,9183-4, TYPE). For discussion, see under C. rubra. Cotumnea conferta Morton, sp. nov. Frutex epiphytieus, 0.6-1.2 m. altus; caules non ramosi, crassi, ca. 7 mm. diam., strigosi, mox glabri; folia apice conferta, per paria valde inaequalia, subsessilia, petiolo usque ad 4 mm. longo, crasso, strigoso; lamina foliorum majorum oblanceolata, subfaleata, 28-32 em. longa, 6.5-7 em. lata, apice breviter acuminata, basi valde obliqua, non amplexicaulis, chartacea, remote serrulata, supra viridis, glabra, subtus sparse strigosa, ca. 7 ст. infra apicem maculas 2 rubras gerens, venis 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 45 lateralibus 10—12-jugis; folia minora stipuliformia, decidua; in- florescentia axillaris, pauciflora, bracteis linearibus, ca. 2.5 cm. longis, 8 mm. latis, longe acuminatis, breviter petiolatis, integris, viridibus, pedunculis са. 9 mm. longis, dense strigosis; calycis lobi pallidi, subaequales, ovati, ca. 2.3 em. longi, basim versus ca. 1 em. lati, longe et acriter acuminati, glanduloso-serrati (dentibus minutis ca. 10 utroque latere), externe substrigosi, intus fere glabri; corolla lutea, intus postice rubro-lineata, ca. 4 cm. longa, in calyce erecta, basi parum saccata, tubo supra basim paullo contracto et 3 mm. diam., sursum ampliato, non ventricoso, ca. 9 mm. diam., externe sparse glanduloso-piloso, intus glabro, fauce vix contracto, limbo obliquo, verisimiliter paullo bilabiato, lobis subaequalibus, ca. 6 mm. longis, intus glabris ; filamenta basi in tubum 3.2 mm. longum postice fissum connata, glabra; antherae inclusae, cohaerentes, quadratae, ca. 1.8 mm. longae et latae; ovarium sparse pilosum ; stylus glaber; stigma stomatomorphum, glabrum; disci glandula postiea solum evoluta, magna, crassa, glabra, non denticulata.—paRIEN: in rain forest, on crest of Cana-Cuasi Trail, Chepigana District, alt. 1650 m., Mar. 15, 1940, М. E. @ В. A. Terry 1554 (Herb. Field Mus., no. 1,035,985, TYPE). Perhaps related to Columnea crassa, but the leaves are larger and much thinner, the stems are not densely hirsute, but strigose or glabrate, and the calyx lobes are not hirsute. CoLUMNEA CONSANGUINEA Hanst. Linnaea 34: 383. 1865-66.— CHIRIQUÍ: Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson £ Schery 651. Previously known from Costa Rica. Cotumnea crassa Morton, sp. nov. Planta epiphytica; caules erecti, crassi, 75 cm. longi, ca. 1 cm. diam., apice paullo angustiores, dense brunneo-hirsuti, pilis tenuibus, patentibus, multicellularibus ; folia per paria inaequalia, petiolata, petiolo ca. 8 mm. longo, crassis- simo, dense hirsuto; lamina foliorum majorum anguste oblance- olata, subfaleata, integra, usque ad 21 ст. longa et 5 em. lata, longe acuminata, basi valde obliqua, erassa, supra viridis, glabra, subtus viridis, maculas rubras (saepe 2) 5-6 em. infra apicem gerens, ubique subdense flavo-strigosa, venis primariis ca. 8-jugis; folia minora stipuliformia, sessilia, anguste lanceolata, ca. 2.5 em. longa, ea. 8 mm. lata, basi inferiore auriculata et amplexicaulia, viridia, supra glabra, subtus strigosa; inflorescentia axillaris, uniflora?, floribus sessilibus ; calyx flavus?, lobis erectis, paullo inaequalibus, [Vor. 29 46 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1.6-2 em. longis, 4—7 mm. latis, glanduloso-serratis (dentibus nu- merosis), acuminatis, externe appresso-hirsutis, intus subglabris, linea mediali subhirsutis; corolla ignota; disci glandula postica solum evoluta ; placentae lamellae intus solum ovuliferae.—PANAMÁ: Cerro Campana, alt. 1000 m., Apr. 21, 1941, Paul H. Allen 2423 (U. S. Nat. Herb., nos. 1,793,929-30, TYPE). Related to C. consanguinea Hanst. (Linnaea 34: 383. 1865-6), of the section Collandra, which may be distinguished as follows: Leaves lacking red spots beneath; calyx lobes entire.................. C. consanguinea Leaves with 2 or more red spots beneath about a third from the apex; calyx lobes Oil Seer К а: C. crassa Согомхел darienensis Morton, sp. nov. Frutex, 1.5-4.5 m. altus; caules vix ramosi, articulati, internodiis brevibus, dense strigosi, apice ca. 3 mm. diam.; folia per paria valde inaequalia, majora petiolata, petiolo 10-14 mm. longo, strigoso; lamina foliorum ma- jorum oblanceolata, 16-23 em. longa, 4—5.5 em. lata, acuminata, basi obliqua, late cuneata, non amplexicaulis, integra, chartacea, supra viridis, glabra, subtus pallidior, non rubro-maeulata, strigosa (praecipue in venis), venis lateralibus 7-jugis, supra obscuris ; folia minora stipuliformia, minuta, decidua, lanceolata, acuminata, supra glabra, subtus strigosa; inflorescentia axillaris, pauciflora, bracteis basalibus ovatis, ca. 1.7 em. longis, integris, acuminatis, fortasse rubris, pedicellis ca. 5 mm. longis, erassis, strigosis; calyx veri- similiter ruber, lobis subaequalibus, lanceolatis, ca. 1.7 ст. longis, 5 mm. latis, acuminatis, glanduloso-serrulatis (dentibus ca. 7 utroque latere), externe nervo mediali et marginibus strigosis, intus glabris; corolla coccineo-aurantiaca, 2.4 em. longa, in calyce erecta, tubulosa, tubo basi ca. 3 mm. diam., sursum paullo ventricoso et 5 mm. lato, faucem versus paullo contracto et 4.5 mm. lato, externe dense flavo-strigoso, limbo parvo, vix bilabiato, ca. 5 mm. lato, lobis erectis, suborbiculatis, rotundatis, 3 inferioribus ca. 1 mm. longis, 2 superioribus ca. 2 mm. longis, altius connatis; filamenta basi in tubum brevem connata, glabra; antherae quadratae, ca. 1.5 mm. longae et latae; ovarium apice strigosum; stylus glaber; disci glandula postica solum evoluta, crassa, glabra, tridenticulata.— DARIEN: Cerro de Garagará, Sambü Basin, alt. 500-974 m., Feb. 7, 1912, Н. Pittier 5660 (0. 8. Nat. Herb., по. 715,942, түрк) ; rain for- est on the crest of the Cana-Cuasi Trail, Chepigana District, alt. 1650 m., Mar. 13, 1940, M. E. € В. A. Terry 1547 (Herb. Field Mus.). From Columnea crassa this species differs in its much narrower 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 47 calyx lobes, strigose rather than hirsute stems, and absence of red spots on the lower leaf surface. It is probably more closely related to C. consanguinea, from which it differs in the serrate rather than entire calyx lobes, broader bracts, and absence of red spots on the leaves. Columnea conferta has a differently shaped corolla limb with much larger, broader, more oblique lobes, as well as larger, thinner, differently shaped, red-spotted leaves. ConLuMNEA dissimilis Morton, sp. nov. Planta epiphytica; rami usque ad 1 m. longi, internodiis brevibus, juniores rubri, hispidi, pilis patentibus, rubris, multicellularibus; folia per paria in- aequalia, petiolata, petiolo 5-9 mm. longo, dispo: lamina foliorum majorum elliptico-oblonga, usque ad 7 em. longa et 3 em. lata, apice acuminata, basi valde obliqua (latere inferiore rotundato, superiore cuneato), integra, utrinque viridis, supra pilosula, subtus praecipue in venis rubro-hirsuta; folia minora saepe subsessilia, ovata vel suborbicularia, acuta vel obtusa, basi rotundata, usque ad 3 cm. longa et 1.8 cm. lata; flores saepe 3 in axillis aggregati, pedunculati, редипешо 1-1.7 cm. longo, dense longe rubro-hirsuto; calyx ruber, erectus, lobis aequalibus, liberis, lanceolatis, 2.5-3 em. longis, acuminatis, remote glanduloso-denticulatis, herbaceis, utrinque rubro-hirsutis; corolla rubra, in calyce paullo obliqua, 3.5-4 ст. longa, basi postice subealcarata, tubo basi 4 mm. lato, tubuloso, vix ampliato, non ventricoso, externe dense rubro-hirsuto, intus glabro, fauce 8 mm. lato, non contracto, lobis 5 aequalibus, incurvatis, ovatis, vix acutis, са. 4.5 mm. longis, crassis, intus glabris, inter lobos appendiculata, appendiculis patentibus, lineari-subulatis, ca. Т mm. longis, dense hirsutis ; filamenta basi in tubum 4 mm. longum postice fissum connata, sursum libera, gracilia, glabra; antherae quadratae, ca. 2 mm. longae et latae; ovarium albo-pilosum ; stylus gracilis, glaber; stigma bilobum; disci glandula postiea solum evoluta, lata, crassa, glabra, tridentieulata.—cocrÉí: trail to Las Minas, north of El Valle de Anton, alt. 1000 m., May 10, 1941, Paul H. Allen 2483 (U. S. Nat. Herb., no. 1,808,614, туре) ; same locality, June 23, 1940, Allen 2164, and July 14, 1940, Allen 2191; PANAMÁ: hills above Campana, alt. 600—800 m., July 1, 1939, Allen 1875. СотоммЕд hirsutissima Morton, sp. nov. Planta epiphytica; caules 18-30 em. longi, non ramosi, crassi, perspicue rubro-hirsutis- simi, pilis multicellularibus, ca. 5 mm. longis; folia per paria valde inaequalia, subsessilia, petiolo crasso, 1-2 mm. longo, hirsuto; [Vor. 29 48 ANNALS OF THE MISSOURI BOTANICAL GARDEN lamina foliorum majorum oblonga vel anguste oblonga, 6-10 cm. longa, 1.7-3.5 em. lata, acuta, basi rotundata vel subcordata, non obliqua, paullo crenulata vel serrulata, utrinque viridis, non rubro- maculata, dense hirsuta, pilis rubescentibus, multicellularibus, 7-9- jugis; folia minora sessilia, ovata, ca. 1 ст. longa, decidua; flores solitarii, axillares, ebracteati, pedunculati, peduneulo 8-12 mm. longo, dense hirsuto ; calyx erectus, lobis subaequalibus, 1.7-1.8 em. longis, linearibus, ca. 2.5 mm. basim versus latis, longe acuminatis, remote glanduloso-denticulatis (dentibus 2 vel 3 utroque latere), utrinque hirsutis, pilis saepe rubris; corolla rubra, in calyce erecta, 6-7.5 em. longa, basi subealcarata, tubo supra basim са. 4 mm. lato, sursum tubuloso et gradatim ampliato, non ventricoso, externe sparse piloso (pilis multicellularibus non glandulosis), intus glanduloso, fauce ca. 8-9 mm. lato, limbo valde bilabiato, intus pilosulo, galea 2.3-2.5 em. longa, apice truncata, 1.4 em. lata, lobis lateralibus eum galea alte connatis, brevibus, ca. 6 mm. longis, lobo inferiore reflexo, lineari-oblongo, ca. 13 mm. longo, 3 mm. lato, ob- tuso; filamenta basi in tubum 4 mm. longum postice fissum connata, basim versus glandulosa, sursum glabra; antherae connatae, ex- sertae, oblongae, ca. 2 mm. longae, 1 mm. latae; ovarium albo- pilosum; stylus dense glanduloso-pilosulus ; stigma bilobum ; disci glandula postica solum evoluta, glabra, crassa, tridenticulata.— сос: hills north of El Valle de Anton, alt. 800-1000 m., Jan. 2, 1941, Раш Н. Allen 2288 (U. S. Nat. Herb., no. 1,806,116, түрк); same locality, Nov. 21, 1940, Allen 2279 (Dorothy Allen), Jan. 21, 1941, Allen 2311 and 2348. For relationships, see under C. rubra below. CoruMNzA incarnata Morton, sp. nov. Frutex epiphyticus; caules vix ramosi, suleati, apice ca. 3 mm. diam., flavo-strigosi, demum glabrescentes; folia per paria subaequalia, petiolata, petiolo 1.3- 2.3 em. longo, strigoso; lamina foliorum oblanceolata, 7-12 cm. longa, 2.3-4 ст. lata, longe acuminata, basi aequalis, cuneata, in- tegra, suberassa, supra viridis, glabra, subtus pallidior, non rubro- maculata, in margine et venis strigosa, venis lateralibus 4- vel 5- jugis, supra obscuris; inflorescentia axillaris, floribus solitariis vel binis, bracteis basalibus lineari-subulatis, ca. 7 mm. longis, basi ca. 1.5 mm. latis, acuminatis, intus glabris, externe strigosis, integris, pedicellis cernuis, 3-4 em. longis, dense flavo-strigosis ; calycis lobi virides, ovati, 3.3-3.5 em. longi, ca. 1.5 em. lati, longe et aeriter acuminati, fere integri, remote glanduloso-denticulati, utrinque 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 49 glabri, parce ciliati; corolla incarnata, 6.5-7 em. longa, in calyce erecta, basi subsaccata, tubo supra basim 5 mm. diam., abrupte ventricoso, ca. 2 cm. longo, in calyce incluso, externe puberulo, intus glanduloso, limbo curvato, valde obliquo, bilabiato, galea са. 5 em. longa, apice truncata, ca. 3.5 em. lata, lobis lateralibus cum galea alte connatis, magnis, partibus liberis ca. 2.3 em. longis, 1.3 em. latis, rotundatis, lobo inferiore patente, magno, са. 3 em. longo et 1.1 ст. lato, lobis externe pilosis, intus glabris; filamenta basi in tubum 2 mm. longum postice fissum connata, dense glanduloso-puberula ; antherae exsertae, cohaerentes, oblongae, 3 mm. longae et 1 mm. latae; ovarium albo-sericeum, apice pilosum ; stylus glaber; stigma bilobum; disci glandula postica solum evoluta, magna, crassa, glabra, emarginata.—ocniniQví: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, В. Е. Woodson ё В. W. Schery 608 (Herb. Mis- souri Bot. Gard., TYPE). This species belongs to the section Pentadenia and is perhaps re- lated remotely to C. magnifica, which also has the disk reduced to a single posterior gland, in contrast to most of the other species of Pentadenia. It may be distinguished as follows: Flowers pink; calyx glabrous, the lobes ciliate, subentire or remotely glandular- dentieuldto iirc с vs coe cree ое Wee Tans C. incarnata Flowers bright scarlet; calyx villous, the lobes conspicuously dentate...... C. magnifica ConLuMNEA LOCALIS Morton, Field Mus. Publ. Bot. 18: 1165. 1938 (Columnea microcalyx var. macrophylla Donn. Smith, Bot. Gaz. 31: 118. 1901, non C. macrophylla Kuntze).—This species has been known only from the type, collected in forests of Las Vueltas, Tucur- rique, Costa Rica. It may now be reported from: PANAMA: Cerro de Garagará, Sambü Basin, southern Darien, alt. 500-974 m., Feb. 7, 1912, Pittier 5625; and cumiquí: trail from Cerro Punta to head- waters of Río Caldera, alt. 2250-2500 m., Jan. 14, 1939, Allen 1428. In the ‘Flora of Costa Rica’ on page 1162, in the key to Columnea, the name C. microphylla is printed C. macrophylla, by an un- fortunate typographical error. CoLtuMNEA obliqua Morton, sp. nov. Planta epiphytica; caules elongati, pendentes, non ramosi, straminei, teretes, ca. 2.5 mm. diam., parce flavo-strigosi, mox glabri ; folia opposita, aequalia, sub- sessilia, lanceolata, usque ad 3.5 em. longa et 1.2 em. lata, apice longe acuminata, basi valde obliqua, rotundata, integra, crassa, supra viridia, subtus pallidiora et rubescentia, supra glabra, subtus in margine et venis strigosa ; flores solitarii, axillares, pedunculo 7-15 [Vor. 29 50 ANNALS OF THE MISSOURI BOTANICAL GARDEN mm. longo, 1 mm. crasso, substrigoso, apice paullo incrassato; calyx viridis, 1.2-1.8 cm. longus, lobis erectis, aequalibus, liberis, basi ca. 6 mm. latis, subeordatis, gradatim angustatis, longissime et acriter acuminatis, integris, externe parce strigosis, intus glabris, basi pilosula exeepta; corolla aurantiaca, 6.5-8 em. longa, in calyce erecta, basi postice paullo calcarata (1.5 mm.), tubo 3.7 mm. diam., sursum parce ampliato, non ventricoso, fauce ca. 12 mm. lato, ex- terne parce pilosulo, intus glabro, limbo valde obliquo, bilabiato, galea erecta, integra, 3.3-3.8 em. longa, lobis lateralibus eum galea alte connatis, са. 1.3 em. longis, lobo inferiore reflexo, lineari- oblongo, 1.4-1.7 em. longo; filamenta basi in tubum 2 mm. longum postice fissum connata, sursum omnino libera, didynama, glabra, antheris oblongis, 1.6 mm. longis et 1 mm. latis, per paria connatis; ovarium albo-sericeum; stylus gracilis, parce pilosulus; stigma bilobum; disci glandula postica lata, magna, emarginata, crassa, glabra, antica parva, lineari-subulata.—cniniQví: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, В. E. Woodson, Jr. Ф В. W. Schery 607 (U.S. Nat. Herb., no. 1,808,050, турв) ; same locality and date, Woodson & Schery 677. CoLUMNEA PANAMENSIS Morton, Ann. Mo. Bot. Gard. 26: 312. 1939.—cuirigui: vicinity of Finca Lérida, alt. 1750 m., July 7-11, 1940, Woodson & Schery 235. Previously known only from type. Сотоммел pectinata Morton, sp. nov. Planta epiphytica; caules pendentes, 100 em. longi, rugosi, ca. 8 mm. diam., hornotini griseo- hirsuti; folia per paria inaequalia, subsessilia, petiolo vix 2 mm. longo, hirsuto; lamina foliorum majorum oblongo-faleata, usque ad 13 em. longa et 5 em. lata, abrupte acuminata, basi valde obliqua, basi inferiore subaurieulata, crassa, apicem versus argute ser- rulata, supra viridis, hirsuta, subtus pallidior, apice sanguinea, ubique dense hirsuta, venis lateralibus 8-10-jugis; folia minora stipuliformia, sessilia, lanceolata, ca. 1.7 em. longa, hirsuta, basi inferiore auriculata; inflorescentia axillaris, pluriflora, floribus breve peduneulatis, pedunculo crasso, са. 5 mm. longo, hirsuto; calyx ruber, erectus, lobis subaequalibus, ea. 15 mm. longis, са. 3 mm. latis, peetinato-fimbriatis (dentibus 5-6 utroque latere, subulatis, usque ad 4 mm. longis), utrinque dense hirsutis, pilis hyalinis, multi- cellularibus, capitato-glandulosis; corolla aurantiaca, in calyce erecta, e calyce vix exserta, ca. 1.7 em. longa, basi subealearata, tubo supra basim ea. 4.5 mm. diam., sursum paullo ventricoso et 6.5 mm. diam., in fauce paullo contracto et 5 mm. diam., externe albo-piloso, 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 51 limbo regulari, haud bilabiato, 7 mm. lato, lobis patentibus, sub- orbiculatis, ca. 3 mm. longis, rotundatis, intus glabris; filamenta in tubum 5 mm. longum postice fissum connata, glabra, valde con- torta; antherae inclusae, per paria cohaerentes, 1.5 mm. longae et 2 mm. latae; ovarium albo-sericeum ; stylus crassus, glaber; stigma stomatomorphum; disci glandula postiea solum evoluta, glabra, tridentieulata.—cocrLÉ: vicinity of La Mesa, north of El Valle de Anton, alt. 1000 m., May 12, 1941, Paul H. Allen 2394 (U. S. Nat. Herb., nos. 1,793,972—3, түре) ; same locality, June 23, 1940, Allen 2177 and May 14, 1939, Allen 1787. Related to Columnea purpurata Hanst., which has larger leaves, not red beneath at apex, conspicuous bracts, and a larger calyx and corolla. CoLuMNEA perpulchra Morton, sp. nov. Planta epiphytica; caules non ramosi, crassi, basi са. 7 mm. diam., apicem versus са. 2.5 mm. diam., densissime brunneo-hirsuti, pilis saepe 6 mm. longis; folia per paria valde inaequalia, subsessilia, petiolo vix 2 mm. longo; lamina foliorum majorum oblongo-oblanceolata, usque ad 16 cm. longa et 4.7 em. lata, apice breviter et abrupte acuminata, basi valde obliqua sed non auriculata vel amplexicaulis, serrulata, herbacea, supra viridis, pilosa, subtus apice rubra, vel interdum ubique rubra vel rubro-maculata, hirsuta, venis lateralibus ca. 11-jugis, subtus prominulis; folia minora stipuliformia, sessilia, ovata, usque ad 2 em. longa et 1 em. lata, longe et acriter acuminata, basi valde obli- qua, basi inferiore auriculata, subamplexicaulia, utrinque hirsuta, subtus apice rubra; inflorescentia biflora, axillaris, basi bibracteata, bracteis lanceolato-subulatis, са. 7 mm. longis, integris, acuminatis, hirsutis, viridibus, pedicellis gracilibus, ca. 2 cm. longis, hirsutis; ealyx pallidus, erectus, lobis subaequalibus, ca. 2 em. longis, 3 mm. latis (dentibus exclusis), acuminatis, laciniatis (laciniis subulatis, usque ad 3 mm. longis, ca. 6 utroque latere), externe pilosis, intus glabris ; corolla lutea (lobis basi coccineis), ca. 4 cm. longa, basi sub- ealearata (3.5 mm.), tubo externe glaberrimo, intus glabro, supra basim contracto et 3.5 mm. diam., abrupte deflexo et ventricoso et са. 8 mm. lato, faucem versus paullo contracto, fauce са. 7 mm. lato, limbo subregulari, lobis reflexis, externe parce strigosis, lobis 2 superioribus ca. 2 mm. connatis, alteris liberis, suborbiculatis, ro- tundatis, omnibus ca. 5 mm. longis, intus glabris, eglandulosis ; fila- menta basi in tubum 8 mm. longum postice fissum connata, glabra, plus minusve contorta; antherae inclusae, connatae, quadratae, ca. [Vor. 29 52 ANNALS OF THE MISSOURI BOTANICAL GARDEN 2 mm. longae et latae; ovarium fere glabrum; stylus glaber ; stigma bilobum ; disci glandula postica solum evoluta, glabra, crassa, emar- ginata.—cocLÉ: hills north of El Valle de Anton, vicinity of La Mesa, alt. 1000 m., Jan. 21, 1941, Paul H. Allen 2305 (U. S. Nat. Herb., no. 1,806,117, TYPE) ; соох: Dos Bocas, Río Fató Valley, alt. 40-80 m., Aug. 16, 1911, H. Pittier 4209. The related species C. sil- varum, described below, may be distinguished as follows: Corolla tube entirely glabrous; leaves red beneath at apex, or sometimes all over. C. perpulchra Corolla tube pilose externally, pilosulous at base within; leaves green, not at all red- spotted C. silvarum ооо ооо ооо ово ооо ооо CoLUMNEA rubra Morton, sp. nov. Planta epiphytica; caulis stra- mineus, vix ramosus, strigosus, mox glaber; folia per paria in- aequalia, majora brevissime petiolata, petiolo ca. 2 mm. longo, crasso, са. 4 mm. diam. ; lamina foliorum majorum anguste oblonga vel oblanceolata, usque ad 14 ст. longa et 4.7 ст. lata, acuta, basi subaequalis, rotundata, erassa, supra pallide viridis, glabra, subtus ubique rubra, strigosa, integra, nervo mediali valde incrassato, venis lateralibus ca. 8-jugis, supra obscuris, immersis, subtus prominulis; folia minora decidua; inflorescentia axillaris, biflora, bibracteata, bracteis lineari-lanceolatis, ca. 5 mm. longis, integris, externe rubro-strigosis, pedunculis ca. 1 em. longis, dense rubro- strigosis; calyx ruber, са. 1.9 em. longus, lobis erectis, aequalibus, lineari-lanceolatis, ca. 5 mm. latis, longissime et acriter acuminatis, remote glanduloso-serratis (dentibus ca. 4 utroque latere), utrinque dense rubro-strigosis; corolla lutea, 7 ст. longa, in calyce erecta, basi subealcarata (са. 2 mm.), tubo supra basim З em. lato, sursum gradatim ampliato, non ventricoso, fauce 11 mm. lato, externe piloso, pilis paucicellularibus, glanduloso-capitatis, galea 2.7 em. longa, 1.4 ст. lata, integra, apice apiculata, lobis lateralibus са. 1 em. longis, cum galea longe connatis, lobo inferiore reflexo, lineari, 1.7 em. longo, 4 mm. lato, lobis omnibus utrinque glanduloso-pilosu- lis; filamenta basi in tubum 5 mm. longum connata, glabra; antherae connatae, oblongae, са. 2.2 mm. longae, 1.6 mm. latae; ovarium cylin- drieum, sericeum; stylus omnino glanduloso-pilosulus; stigma bi- lobum, glanduloso-pilosulum ; disci glandula postiea solum evoluta, glabra, crassa, integra.—cocrÉ: trail to Las Minas, north of El Valle de Anton, alt. 1000 m., May 10, 1941, Paul H. Allen 2469 (U. S. Nat. Herb., no. 1,808,603, түрк). The above measurements for the corolla are taken from flowers preserved in liquid. In drying, the corollas contract a great deal. 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 53 The following measurements of dried corollas are given for com- parison with herbarium specimens: Length 6-6.5 cm., tube above base 2 mm. broad, throat 7 mm. broad, galea 2.5 cm. long, ca. 7 mm. broad, lateral lobes ca. 7 mm. long, lower lobe ca. 1.4 em. long. The section Cryptocolumnea is a peculiar small one in which the corolla has the bilabiate shape of Eucolumnea, but the leaves are similar to those of Collandra. The only previously known North American species is C. praetexta Hanst. (Linnaea 34: 394. 1865- 66), originally collected at an undesignated locality in Costa Rica by Warscewicz, which has never been found again. Columnea citrina, described above, is certainly a member of the section, and the present species, C. rubra, will probably be found to belong there also, the corolla being that of Eucolumnea, but the leaves, by their shape and red coloration, suggesting those of Collandra. However, the characteristic stipuliform leaves are lacking on the specimen at hand. Another very distinctive species, C. hirsutissima, is de- scribed above. These species may be separated as follows: Stigma stomatomorphie; ealyx lobes entire or nearly so; braets large (2 em. long or more) ; leaves very oblique at base, with the lower side auriculate and amplexicaul, green beneath with a purple spot toward apex Braets orbieular, erenulate; leaves villous above; stipuliform leaves 2.5-5 em. lon Braets lanceolate, entire; leaves glabrous above; stipuliform leaves about 2 em. long. Style and stigma glabrous; ealyx lobes glabrous within exeept on median line; corolla lobes. glabrous within CR ON ee eT Nan sis C. citrina Stigma bilobed, Died M calyx lobes serrate or denticulate, strongly ubescent on both sides; braets small (not over 5 mm. lon absent; leaves ng) o equal at base, iin киы or amplexieaul. Style E bis. pilosulous through- out; corolla lobes hairy within Stems strigose; leaves glabrous N strigose Rte red over whole lower urface; filaments glabrous throughout; corolla yellow.................. C. rubra n densely long-hirsute; leaves hirsute on both p^ not red or red-spotted beneath; filaments glandular toward base; corolla red............ C. hirsutissima CorvMNEA silvarum Morton, sp. nov. Herba vel frutex, 0.6-3.6 m. altus; caules crassi, articulati, dense rubro-hispidi; folia per paria inaequalia, subsessilia, petiolo crasso, ca. 2 mm. longo; lamina foliorum majorum oblonga, usque ad 16 em. longa et 5.5 em. lata, breviter et abrupte acuminata, basi valde inaequalis et obliqua, non auriculata, non amplexicaulis, herbacea, glanduloso-denticulata, non rubro-maculata, utrinque viridis, pilosula, subtus in venis hirsuta, venis lateralibus ca. 10-jugis, prominulis; folia minora stipuliformia, lanceolata, ca. 1 cm. longa, 4 mm. lata, acuminata, basi obliqua, viridia; inflorescentia axillaris, 2- vel 3-flora, bracteis basalibus lineari-subulatis, ca. 8 mm. longis, 1 mm. latis, integris, [Vor. 29 54 ANNALS OF THE MISSOURI BOTANICAL GARDEN pedicellis 1.5-2.5 cm. longis, gracilibus, dense rubro-hirsutis ; calycis lobi lineari-subulati, subaequales, 2-2.5 em. longi, са. 3 mm. basi lati, longissime acuminati, remote laciniati (dentibus subulatis, ca. 2 mm. longis, ca. 3 utroque latere), utrinque rubro-hirsuti; corolla lutea (lobis basi purpureis), in calyce erecta, 4—4.5 em. longa, basi subealearata (3 mm.), tubo supra basim contracto et 4 mm. diam., tubuloso, sursum ampliato et paullo ventricoso, 1 ст. diam., fauce contracto et 7 mm. lato, externe parce piloso, intus basi pilosulo, limbo subregulari, non bilabiato, са. 1 em. lato, lobis patentibus, sub- orbiculatis, ca. 3 mm. longis, rotundatis, 2 superioribus altius con- natis, intus glabris; filamenta basi in tubum 5 mm. longum postice fissum connata, glabra; antherae exsertae, cohaerentes, quadratae, ca. 2 mm. longae et latae; ovarium sericeum ; stylus glaber; stigma breviter bilobum, parce glanduloso-pilosulum; disci glandula postica solum evoluta, crassa, glabra, bidenticulata.—paAnRrEN: in rain forest on the crest of the Cana-Cuasi Trail, Chepigana District, alt. 1500 m., Mar. 15, 1940, M. E. and R. A. Terry 1566 (Herb. Field Mus., no. 1,035,998, түрк). A second specimen in the Field Museum was collected near the same locality at 600 meters elevation, Mar. 11, 1940, Terry & Terry 1499. This species and C. perpulchra differ strongly from other Panama species of the section Collandra in their elongate corollas, about twice as long as the calyces. In the other species the corolla is in- cluded in the calyx or is only very slightly exserted. DRYMONIA ALLOPLECTOIDES Hanst. var. vallicola Morton, var. nov. A var. typica foliis minoribus, calycis lobis angustioribus, minus hirsutis, corolla minus obliqua, minus hirsuta, limbo minus obliquo а1 егі. —сосі : north rim of El Valle de Anton, near Cerro Turega, alt. 650—700 m., June 30, 1940, В. E. Woodson Ф В. W. Schery 203 (U. S. Nat. Herb., no. 1,808,024, түрк); same locality and date, Woodson & Schery 162. The typical form of this species is confined to Costa Rica. Drymonta brevipes Morton, sp. nov. Frutex epiphyticus, scan- dens, 3-5 m. longus, ramosus; rami i^retes, glaberrimi, apicem versus са. 4 mm. diam. ; folia per paria no inaequalia, petiolata, petiolo foliorum majorum 1.5-2 em. longo, glabro; lamina foliorum majorum anguste oblonga, 20-24 cm. longa, 4.7-6 ст. lata, acumi- nata, basi cuneata, aequalis, coriacea, integra, utrinque glaberrima, viridis, non rubro-maculata, venis lateralibus supra obscuris, sub- 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 55 tus elevatis ; folia minora similia, sed са. 9.5 cm. longa et 3 cm. lata; flores solitarii, axillares, brevipedunculati, pedunculo ca. 13 mm. longo, glabro, apice incrassato; calyx verisimiliter coloratus, valde obliquus, lobis 3, valde inaequalibus, lateralibus late ovatis, ca. 2.4 em. longis, 2.2 cm. latis, apice bilobatis (ca. 4 mm.), lobo postico fere libero, multo minore, ovato-lanceolato, ca. 18 mm. longo, 12 mm. lato, obtuso, omnibus glaberrimis; corolla rosea, ca. 7 em. longa, in calyce obliqua, basi obtuse calcarata (6 mm.), tubo supra basim 11 mm. lato, gradatim ampliato, non ventricoso, fauce ca. 13 mm. lato, utrinque glaberrimo, limbo subregulari, ca. 3.5 cm. lato, paullo obliquo, lobis patentibus, subaequalibus, suborbieulatis, ca. 13 mm. longis, rotundatis, subintegris, non fimbriatis; filamenta basi in tubum 9 mm. longum erassum postice fissum connata, glabra, con- torta; antherae connatae, inclusae, oblongae, ca. 6.5 mm. longae, 2 mm. latae; ovarium parce strigillosum; stylus glaber; stigma stomatomorphum; disei glandula postica solum evoluta, glabra, crassa, acuta.—cocrÉ: hills in vicinity of La Mesa, north of El Valle de Anton, alt. 1000 m., Jan. 21, 1941, Paul H. Allen 2330 (U. S. Nat. Herb., no. 1,806,119, TYPE). Evidently a rather close relative of Дгутота conchocalyx Hanst. of Costa Rica, but with a quite different calyx. In D. conchocalyx the calyx is obviously 5-lobed, the 2 anterior lobes being connate high up. In D. brevipes the calyx is apparently 3-lobed, the 2 an- terior lobes being only slightly connate, but each anterior lobe is almost wholly connate with a lateral lobe. In both species the pos- terior lobe is nearly free and much smaller than the others. Dry- monia brevipes may be distinguished also by the very short, thick peduncles. | DRYMONIA LANCEOLATA (Hanst.) Morton, Field Mus. Publ. Bot. 18: 1173. 1938.—In the ‘Flora of Costa Rica,’ I reported this species from Panama, but I do not now know on what basis. A defi- nite record is: CHIRIQUÍ: Quebrada Velo, alt. 1800 m., July 8, 1940, Woodson & Schery 268. DRYMONIA PARVIFLORA Hanst. Linnaea 34: 352. 1865-66.—снтвт- 901: vicinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July 12, 1940, Woodson & Schery 566. This distinctive species, which has the general appearance of the genus T'ussacia, was originally collected at Candelaria, Costa Rica, by Hoffmann, and has since been known only from the type. The [Vor. 29 56 ANNALS OF THE MISSOURI BOTANICAL GARDEN specimen recently collected agrees perfectly with Hanstein’s de- scription. Drymonta parvifolia Morton, sp. nov. Frutex parvus, epiphyti- cus; caules са. 30 em. longi, sulcati, glabri, apice са. 4 mm. diam. ; folia per paria aequalia, petiolata, petiolo 7-19 mm. longo, glabro, gracili ; lamina foliorum lanceolata, 4—5.5 em. longa, 1.3-2 em. lata, obtuse acuminata, basi aequalis, cuneata, integra, papyracea, utrinque glaberrima, venis lateralibus 4- vel 5-jugis, subtus vix ele- vatis; inflorescentia axillaris, uniflora, braeteis nullis, peduneulo ca. 9 mm. longo, glabro; calyx irregularis, lobis liberis, 2 anticis anguste oblongis, 11 mm. longis et 3 mm. latis, obtusis, 2 lateralibus ovato-lanceolatis, 13 mm. longis et 5 mm. latis, acutis, postico ovato- lanceolato, 8 mm. longo et З mm. lato, acuto, omnibus integris, gla- berrimis ; corolla alba, са. 2.5 em. longa, in calyce horizontalis, basi postice calcarata (1.5 mm.), tubo basi ca. 3 mm. lato, sursum ampli- ato et in fauce non contracto, ca. 1 em. diam., externe glabro, limbo terminali, subbilabiato, lobis magnis, externe pilosis, lacerato- dentatis; filamenta glabra; antherae cohaerentes, inclusae, ob- longae, sagittatae, ca. 3 mm. longae; ovarium apice sericeum ; stylus glaber; disci glandula postica solum evoluta, bidenticulata, glabra, crassa.—cHIRIQUÍ: trail from San Felix to Cerro Flor, alt. 100-850 m., Aug. 13-14, 1939, Paul H. Allen 1949 (Herb. Missouri Bot. Gard., TYPE). This species is another of those intermediate between Drymonia and Alloplectus sect. Erythranthus. It is not closely related to any other species, being distinguished by its entirely glabrous stems, leaves and calyces, very small and strongly unequal calyx lobes, and white corollas. Drymonia rosea Morton, sp. nov. Planta epiphytica, erecta; eaules ca. 30 ст. longi, subtetragoni, crassi, ca. 1 cm. diam., vix strigillosi; folia per paria subaequalia, biévitér petiolata, potídiis erasso, vix 1 em. longo, dense rubro-sericeo ; lamina foliorum oblique ovata vel oblonga, subfalcata, usque ad 22 em. longa et 8.5 em. lata, acuta, basi obliqua, rotundata, subchartacea, apicem versus ser- rulata, supra viridis, scabro-strigillosa, subtus ubique rosea, praecipue in venis strigillosa, venis lateralibus 8-10-jugis, venulis ultimis subtus valde prominulis, reticulatis ; inflorescentia axillaris, uniflora, bracteis minutis, subulatis, pedicello ca. 7 mm. longo; calyx viridis, lobis inaequalibus, acutis, anterioribus et lateralibus ob- longis, 2.5-2.8 ст. longis, 1 cm. latis, postico ovato, 2.2 em. longo et 1942] MORTON—NEW GESNERIACEAE FROM PANAMA 57 12 mm. lato, omnibus acriter dentatis (dentibus subulatis, ca. 1 mm. longis, ca. 8 utroque latere), utrinque strigillosis, pilis minutis, rubris, glanduliferis; corolla lutea, in calyce valde obliqua, basi postice ealearata (2 mm.), tubo ca. 3.2 em. longo, supra basim ca. 5 mm. diam., sursum ampliato, non ventricoso, in fauce non con- tracto, ca. 13 mm. lato, externe rubro-pilosulo, intus basi dense glanduloso-pilosulo, limbo obliquo, paullo bilabiato, lobo antico ma- jore, са. 9 mm. longo et 10 mm. lato, lobis lateralibus ca. 9 mm. longis et 7 mm. latis, lobis posticis ca. 5 mm. longis et 10 mm. latis, omnibus breviter Hot denas (dentibus ca. 0.5 mm. longis), intus glabris; filamenta basi in tubum 7 mm. longum postice fissum con- nata, glabra, vix contorta ; antherae liberae, inclusae, oblanceolatae, ea. 5 mm. longae, basi ca. 1 mm. latae, apice ca. 2 mm. latae, basi sag- ittatae, basi solum dehiscentes; ovarium sericeum; stylus dense pilosulus, brevis, crassus; stigma stomatomorphum; disci glandula postiea solum evoluta, magna, erassa, glabra, paullo emarginata.— cocLÉ: vieinity of La Mesa, hills north of El Valle de Anton, alt. 1000 m., May 11, 1941, Paul H. Allen 2489 (U. S. Nat. Herb., nos. 1,808,619-20, TYPE). Perhaps related to D. marmorata Hook. and D. turrialbae Hanst., which also have leaves deep rose color beneath, but in these species the calyx lobes are entire or inconspicuously and irregularly crenu- late; in D. rosea the calyx lobes are sharply and deeply subulate- toothed. In D. rosea the ultimate veinlets are elevated and very prominent beneath, giving a reticulation suggesting that of some genus of Rubiaceae such as Sommera. I have never seen any sim- ilar venation in the Gesneriaceae. DrYMONIA TURRIALBAE Hanst. Linnaea 34: 359. 1865-66.—In the ‘Flora of Costa Rica’ I reported this from Panama. The record on which this was based is Pittier 5671 from Cerro de Garagara, Sambu Basin, Darien, alt. 500-974 m., Feb. 7, 1912. I have also examined another specimen in the Field Museum, Terry & Terry 1421, col- E on the Cana-Cuasi Trail, Chepigana District, Darien, alt. 1200 „ Мат. 9, 1940. Neither of these collections has flowers in good con- dition for study, so it is possible that this Darien plant is different from the Costa Rican species. GLOXINIA PALLIDIFLORA Hook.—caNaL ZONE: on large boulders in bed of waterfall, Government Forest Preserve, Madden Road, alt. 50 m., Nov. 29, 1940, Paul H. Allen 2280. Not previously collected wild in Panama. [Vor. 29, 1942] 58 ANNALS OF THE MISSOURI BOTANICAL GARDEN Мокорутв panamensis Morton, sp. nov. Herba terrestris; caules non ramosi, usque ad 70 cm. longi, puberuli; folia per paria in- aequalia, longe petiolata, petiolo foliorum majorum usque ad 3.8 cm. longo, pilosulo; lamina foliorum majorum elliptica vel anguste el- liptiea, usque ad 20 cm. longa et 8.2 cm. lata, acuminata, basi valde obliqua, rotundata, tenuiter herbacea, grosse dentata, supra hirsutula, subtus puberula, venis lateralibus ca. 11-jugis, subtus prominulis; folia minora brevius petiolata, similia sed multo mi- nora, ca. 4 em. longa; inflorescentia terminalis, paniculata, 6-18 em. longa, pedunculo communi 3.5-9 em. longo, pilosulo, eglanduloso, apiee bibraeteato (bracteis subfoliaceis, viridibus, petiolatis), in- floreseentiae ramis primariis brevissimis, usque ad 7 mm. longis, apice bibraeteolatis (braeteolis subulatis, flavo-pilosulis), pedicellis 2-4, terminalibus, subumbellatis, inaequalibus, 2-22 mm. longis; ealyeis tubus cylindricus, ad anthesin ca. 5 mm. longus et 2 mm. latus, basi vix eurvatus, ubique dense glanduloso-pilosulus ; calycis lobi virides, oblongi, ca. 5 mm. longi et 1.5 mm. lati, obtusi, integri, crassi, externe pilosuli, intus apicem versus pilosuli, non reeurvati ; corolla basi alba, sursum eaerulea, 1.2-1.6 em. longa, late campanu- lata, in calyce obliqua, basi ecalearata, non gibbosa, tubo basi ca. 4 mm. diam., sursum 11-13 mm. diam., externe pilosulo, intus glabro, limbo subregulari, ca. 1.7 em. diam., lobis latis, rotundatis, intus glabris ; filamenta libera, glabra; antherae cohaerentes, inclusae, ca. 1.5 mm. longae, loculis discretis, subdivergentibus; ovarium fere omnino inferum, pars libera conica, pilosula; stylus glaber; stigma stomatomorphum, glabrum; discus nullus; fructus capsularis, ca. 1.5 em. longus, 3 mm. latus.—PaANAMÁ: Cerro Campana, alt. 1000 m., Apr. 21, 1941, Paul H. Allen 2413 (U. S. Nat. Herb., no. 1,793,920, TYPE) ; same locality, July 1, 1939, Allen 1871. Related to Monopyle puberula Morton, of Costa Riea and Guate- mala, but differing as follows: Leaves long-petiolate (1.5-3.8 em.) ; inflorescence (rhachis, pedicels and ydp tube) c асан вий чана ealyx and lobes hairy within; corolla 6 em. n llam mE d o PPP" t panamensis Leaves «Ая -petiolate pies 5 mm.) ; inflorescence eglandular- Кы ealyx and lobes glabrous within; corolla 2-2.8 em. long; stigma pilose............ M. puberula Monopyle Махопи is related also, but differs as follows: Infloreseenee branehes and pedieels glabrate; hairs of calyx tube with elongate, uneinate gland; calyx lobes red-purple, recurved at tip, glabrous within; 1 leaf surfaee glabrous; eorolla usually 2-3 em. long; stigma pilose...... M. Mazonii Inflorescence branches densely pilosulous; hairs of calyx tube with small, capitate gland; calyx lobes green, not recurved, hairy within; lower leaf surface strigose- puberulous; eorolla 1.2-1.6 em. long; stigma glabrous.............. M. panamensis Annals of the Missouri Botanical Garden Vol. 29 APRIL, 1942 No. 2 CONTRIBUTIONS TO OUR KNOWLEDGE OF AMERICAN CARBONIFEROUS FLORAS III. SrrPrroPTERIS L. WAYNE LENZ University Scholar, Henry Shaw School of Botany of Washington University The fossil described in this paper, a fern petiole fragment ref- erable to the form genus Stipitopteris, was discovered in a coal- ball collected at the Pyramid Mine of the Binkley Coal Company, located three miles south of Pinckneyville in Perry County, Illinois. The coal in this mine is described as belonging to the lower Mc- Leansboro series and is known as coal number 6. A more detailed discussion of the occurrence of coal-balls at this locality has been given in the first of this group of contributions (Andrews, 42). Materials and Procedure.—The single specimen of this fern con- sists of a fragment of a petiole approximately 4 ст. long. The coal- ball containing it, numbered WBC170 in the collections of the Henry Shaw School of Botany of Washington University, was cut into six sections, each being approximately 13 mm. in thickness. The fern petiole was well preserved in the two center sections, but rather badly crushed and distorted in the two adjoining ones. Of the cen- ter sections, one was used to prepare a series of transverse peels while the other was used for longitudinal sections. It was possible in this way to observe the nature of the tracheid pitting as well as the distinctive transitional anatomy of the trace as a whole. The nitrocellulose peel method was used in obtaining prepara- tions. This method produced much more satisfactory preparations for study and photographic purposes than ground sections, chiefly Issued April 20, 1942. (59) [Vor. 29 60 ANNALS OF THE MISSOURI BOTANICAL GARDEN because the dark contents in the parenchyma cells of the latter made it very difficult to grind sections sufficiently thin. Description.—The petiole is more or less oval in outline and meas- ures approximately 10 by 13 mm. in diameter. The most distinctive feature is the transitional organization of the xylem of the leaf trace. At one end of the specimen (pl. 9, fig. 1) the trace is in the form of an elongated ‘‘horseshoe’’ with the free ends involuted, while at the opposite end (fig. 4) it forms a continuous band enelos- ing an isolated W-shaped segment. A selection from the peel prep- arations shows these extremities as well as two intermediate stages (figs. 3, 4) in the isolation of the W-shaped strand. Text-figures 1 and 2 show the organization of the xylem tissue of the leaf trace as reconstructed from the series of peel preparations. The actual distance through which the illustrated transition takes place is ap- proximately 10 mm. The longitudinal dimension has been exag- gerated somewhat in the text-figures in order to show the transition clearly. The xylem of the trace, which is shown as the rather prominent white band in figs. 1-4, consists of primary metaxylem 2-4 cells wide with scattered groups of protoxylem located on the inner side. The metaxylem cells are uniformly scalariform (pl. 10, fig. 7) and vary from 15 to 91p in transverse section, the average being about 47y. The protoxylem groups are endarch and numerous, being ar- ranged around the entire inner periphery of the metaxylem, al- though they are more numerous in the involuted portion of the trace (fig. 1). It is difficult to ascertain the exact number of proto- xylem groups, partly because of faulty preservation in spots and partly because the size distinction is not always great between meta- xylem and supposed protoxylem. The number is approximately 25- 30 and the average size of a clearly defined protoxylem cell is 21p. In the isolated W-shaped strand they are found on the upper side, as the petiole is illustrated in fig. 3. This corresponds to the inner side of the surrounding cylinder. Scott (200), in describing the protoxylem of Caulopteris, does not mention the number of protoxylem groups present but only says that they occur, ‘‘as a number of small groups of cells on the inner edge of the xylem, which is thus endarch’’ [page 302]. This agrees with the specimen described here which has been shown to have centrally located protoxylem groups. Phloem, pericycle and endodermis are not preserved. In all our 1942] LENZ— AMERICAN CARBONIFEROUS FLORAS. III 61 preparations the inner cortex is rather closely appressed to the xylem although there is some evidence that a tissue (or tissues) was formerly present between them. If such were the case it must have been a very narrow layer. The cortex presents a number of distinctive features. Taken as a whole, it is composed of two sharply defined zones. The outer one reaches a thickness of a little more than 1 mm. and consists of cells which are large, thin-walled and rather badly crushed due to their apparently delicate nature in life. The inner zone averages about 1.5 mm. thick and consists of somewhat smaller cells partially filled with an opaque substance which may represent metamorphosed tan- nin or gum deposits. These two zones are separated by a narrow band of dark brown, closely compacted cells suggestive of meriste- matic activity. The pith or central parenchymatous mass consists of isodiametric cells with some dark contents but these are not so abundant as in the inner cortex. Taxonomy.—There is little doubt that this petiole belongs to a stem of one of the Carboniferous tree ferns known from struc- tually preserved remains as Psaronius and it evinces leaf scars of the Caulopteris! type. This is evident from a comparison of the Il- linois specimen as shown in figs. 1-4 with Caulopteris varians Renault and Zeiller (288) (pl. 10, fig. 5). The portion of the stem of C. varians in Renault and Zeiller’s figure shows leaf scars at slightly different levels, and since the transition of the leaf trace takes place rapidly the comparison is possible. One of the bundle scars shown in their figure presents the elongated horseshoe type of trace similar to that in the Illinois specimen shown in pl. 9, fig. 1. Another bundle scar on the same portion of the stem shows the trace as a continuous band with the isolated W-shaped segment within the outer band (c.f. pl. 9, fig. 4). Since the petiole described here was not found attached to the stem of Caulopteris, though there is little doubt that it was origi- nally, it becomes necessary to describe it under the name Stipitop- teris. That genus was introduced by Grand'Eury in 1877 for petioles of ferns whose structure in cross-sections corresponds to the marks on the leaf scars of Caulopteris, but which were not known in organic connection with the latter. The leaf scars of certain species of Stemmatopteris Corda, de- 1 The name Caulopteris was first used by Lindley and Hutton (’32) for supposed stems of the tree fern Psaronius which had been preserved as casts showing only surface features. [Vor. 29 62 ANNALS OF THE MISSOURI BOTANICAL GARDEN seribed by Lesquereux ('80) from the Carboniferous formation of Pennsylvania, show a similarity to a cross-section of the petiole described in this paper. Stemmatopteris and Caulopteris are un- doubtedly closely related or even congeneric, and Lesquereux him- self wrote that ‘‘the whole number of our species could have been described without inconvenience under the name of Caulopteris."' Hirmer (727) considers the two congeneric. There is also a striking similarity between the petiole deseribed here and species of the genus Anachoropteris Corda. Darrah (741), in his recent account of the Coenopterid ferns of the American coal- balls, points out that the taxonomic limits of Anachoropteris are in considerable doubt and that some of the forms included in the genus are now known to be portions of plants really belonging to other genera. He states, however, that the involute stele of most of the species seems to constitute a generie charaeter. Corsin (737), in his treatment of Anachoropteris, speaks of the protoxylem as oc- eurring on the outside of the metaxylem rather than on the inside as in Stipitopteris. This apparent difference, in conjunction with the similarities between both of the aforementioned genera, brings up the question of the taxonomic value of the location of the proto- xylem. Tansley and Lulham (’05) considered that in many cases it is of considerable significance, and they point out the universal exarchy of the steles of the Sphenophyllales and Lycopodiales and the endarchy of the Angiosperms. However, they conclude that the Filicales are rather different in that the exact position of the spiral protoxylem in relation to the metaxylem of the vascular strands of the stem is decidedly variable. They found exarchy, endarchy and mesarchy all within comparatively small groups, and concluded that the actual course of evolution seems to affect the position of the protoxylem much more freely and rapidly than in the other great groups of vascular plants. From this evidence it would seem that the use of the location of the protoxylem to separate genera in this group of ferns might be questioned. It is, however, evident that the stability of this character varies in different groups. It is probable that an extensive taxonomic revision of the genera Anachoropteris, Caulopteris, Stipitopteris and Psaronius will be necessary in the future. This paper is intended as a further contri- bution to the apparent relationship between petioles described as Anachoropteris and Stipitopteris and fern stems described as Psaronius and Caulopteris. 1942] LENZ—AMERICAN CARBONIFEROUS FLORAS. III 63 STIPITOPTERIS Grand'Eury (’77) described four species of Stipitopteris from the Carboniferous of France. His descriptions are very brief, and his only illustration is the cross-section of a petiole which is merely labeled Stipitopteris. It is interesting to note, however, that the isolated strand of xylem in his specimen is turned in the exact op- posite direction from the Illinois specimen. Since the Illinois speci- men is quite distinct from any described species of Stipitopteris it is described here as a new species. Stipitopteris americana Lenz, sp. nov.—Petioles somewhat oval, 10 by 13 mm. in diameter; cortex consisting of two distinct zones, the outer approximately 1 mm. in thickness and formed of cells which are large, thin-walled and rather badly crushed, the inner about 1.5 mm. in thickness and consisting of somewhat smaller cells partially filled with an opaque substance; zones separated by a narrow band of dark closely compacted cells; leaf trace in the form of an elongated horseshoe with the ends involuted and fusing to form an isolated W within a continuous xylem cylinder; proto- xylems endarch, of spirally thickened tracheids in small scattered groups; metaxylem 2—4 cells thick, consisting of scalariform-pitted tracheids, 15-91 y in diameter. Locality: Pyramid Mine of the Binkley Coal Company, Perry County, Illinois. Type: Coal-ball WCB170 in the collections of the Henry Shaw School of Botany of Washington University. Acknowledgment.—The author wishes to express his apprecia- tion to Dr. H. N. Andrews who suggested the problem and under whose direction the work was carried out; to Dr. A. H. Blickle, for helpful suggestions relative to the natural affinities of the fossil described; and also to the Binkley Coal Co., for its continued kind- ness and cooperation in extending the privilege of collecting coal- balls from the mine. LITERATURE CITED Andrews, H. N. (1942). Contributions to our knowledge of American Carboniferous ras. I. Scleropteris gen. nov., Mesoxrylon and Amyelon. Ann. Mo. Bot. Gard. 29: 1-18. Corsin, Paul (1937). Contribution à l'étude des fougères anciennes du groupe des In- versieatenales. Lille Darrah, W. C. (1941). The Coenopterid ferns in American coal-balls. Am. Midl. Nat. 25: 233-269. [Vor. 29 64 ANNALS OF THE MISSOURI BOTANICAL GARDEN Grand’Eury, F. C. (1877). Flore Carbonifére du Département de la Loire et du centre де la France. Acad. Sci. de 1’Inst. de France, Mem. 24: 624 pp. Hirmer, M. (1927). Handbuch der Palüobotanik. Berlin Lesquereux, Leo (1880). Coal flora of Pennsylvania. Верона Geol. Survey of Pa. Report of progre Lindley, John, and William Hutton (1832). Fossil flora of Great Britain. London. Renault, B., et R. Zeiller (1888). Études sur le terrain houiller de Commentry. Flore fossile. Soc. Ind. Min., Bull. II, 2. слее D. H. ipud Studies in fossil botany. 2nd ed. Tansley, A. G., and Е. В. 8. pu EU A study of the vascular system of Matonia АА "A Bot. 19: 475-51 1942] Ш CARBONIFEROUS FLORAS. AMERICAN LENZ Text-figs. 1, 2. Stipitopteris americana. Reconstruction from a series of peels showing the transition of the xylem tissue from the horseshoe за shown ii isolated W- shaped segment, shown in fig. 2 1 fig. 1, to the eylinder of xylem enclosing the [ VoL. 29, 1942] 66 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 9 Figs. 1-4. Stipitopteris americana, Part of a series of peel preparations showin transitional anatomy of leaf trace; fig. g the 1, WCB170D.B9, fig. 2, WCB170D,.T40a; fig. 3, WCBI170D.T25; fig. 4, WCB170D.T8. All figures x 5. ANN. Mo. Bor. Garp., Vor. 29, 1942 PLATE 9 LENZ—AMERICAN CARBONIFEROUS FLORAS. III [Vor. 29, 1942] 68 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 10 Fig. 5. Caulopteris varians Renault and Zeiller, From Renault and Zeiller (’88). Fig. 6. Stipitopteris americana Lenz. Longitudin - section showing the spirally thiekened tracheids of the itio dus WCB1700C.818, x 110. Fig. tipitopteris americana Lenz. Langttadinal section showing the scalariform- pitted кейде of the ыл WCB170C.821, x 110. Ра.Ате 10 ANN. Mo. Bor. Garb., Vor. 29, 1942 AS ROUS FLOR N CARBONIF RICA NZ—AME LE RACES OF ZEA MAYS: I. THEIR RECOGNITION AND CLASSIFICATION! EDGAR ANDERSON Geneticist to the Missouri Botanical Garden Engelmann Professor in the Henry Shaw School of Botany of Washington University AND HUGH C. CUTLER Research Fellow in the Henry Shaw School of Botany of Washington University It is now half a century since E. Lewis Sturtevant ('85, '86, '87, 294, 299) reviewed the variability of Zea Mays and catalogued it in six main groups, five of which were based upon the composition of the kernel. There are now available two new kinds of facts in addi- tion to those which were at his disposal. In the first place, archeo- logieal research has made great advances and the museums of this country have numerous collections of prehistoric corn, the best of them in a remarkable state of preservation, their age approximately dated by dendrochronology (Guernsey and Kidder, '21). In the sec- ond place, the rise and development of corn genetics have provided us with a large body of technical information concerning the rela- tionships of different kinds of corn. We know, for instance, that the change from flint corn to flour corn is controlled by a single gene difference, whereas the change from a flint to a dent is the result of many genes, so that in working out the relationships of corn the difference between flint and flour is trivial compared to the differ- ence between flint and dent. Mangelsdorf and Reeves have recently (38, ’39) advanced a new theory as to the relationships between corn and its closest relatives. For the examination of this hypothesis, or of any hypothesis deal- ing with the history of corn, it is important that the classification of the varieties of maize be re-examined in the light of our present knowledge of their history and relationships. For this purpose Sturtevant's classification has another disadvantage in addition to the fact that it does not incorporate the modern evidence. It is largely artificial rather than natural, as he himself was well aware (294, p. 320). 1 Mueh of the work reported in this paper was made possible by two grants from the Penrose Fund of the American Philosophical Society, to whom grateful acknowledgment is made. (69) [Vor. 29 70 ANNALS OF THE MISSOURI BOTANICAL GARDEN It is not only in the classification of maize that one has to choose between these two methods, the natural and the artificial. Which to choose is one of the fundamental problems of all attempts at classification. Hach method has its advantages and disadvantages. An artificial classification is an efficient catalogue and nothing more; it sets up a comprehensive series of pigeon-holes. A natural classification attempts in addition to put closely similar objects in adjacent pigeon-holes. An artificial classification may be simple, objective, rapid and complete, but it tells us little more than the pigeon-hole number of the catalogued object. A natural classifica- tion is often difficult, incomplete, and more or less subjective, but in addition to cataloguing an object it tells something about what kind of an object it is. In most taxonomic work natural classifications are seldom at- tempted within the limits of a species or sub-species. Though a monographer may often realize that the individuals which he is lumping together as one species are grouped into more or less well- defined natural races, nevertheless he seldom has the time or the spe- cial training for attempting intra-specifie classification. But though the problem may be diffieult it eannot be ignored by those who work with man or domesticated plants and animals. While, for such prob- lems, a wholly natural classification would be desirable, a more or less artificial one usually has to be adopted for practical reasons. In dealing with the people of a community, for instance, the tele- phone company lists them alphabetically. Like any good artificial classification, the telephone book is complete, objective, and requires no special training either to make or to use. It is convenient but it does not put similar people in adjacent positions. The O’Connors find themselves next to the Ochsners and the Cabots to the Cabril- liacs. When a sociologist studies the same community he might like to have a complete natural classification in which the O’Connors would go near the Kellys, and the Cabrilliaes with the Oldanis. In practice he must be satisfied with a few large natural groupings such as nationality, occupation, income, ete. Anything more elab- orate than this is seldom put down on paper because it is too difficult. Sturtevant’s classification of corn was both artificial and natural, principally the former. It had six main subdivisions. One was pod corns. The other five considered only the composition of the kernel : the pop corns, the flint corns, the dent corns, the flour corns, and the sweet corns. To the extent that any two corns with the same kind 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 71 of kernel are by that very fact somewhat alike, it did put similar kinds of maize together. On the other hand, as Sturtevant himself pointed out, by ignoring other features of the maize plant, it often separated varieties which naturally belonged together. This sys- tem had exactly the advantages and the disadvantages of a classifi- cation of mankind based entirely on hair color. Such a classification would be rapid and complete and would, to a certain extent, group like peoples together, but it would separate a black-haired Nor- wegian from his fair-haired relatives and put him in the same class with Sicilians and gypsies. The problem of races and their recognition is indeed almost the same in Zea Mays as in mankind. In both cases it is not easy to work out the racial composition of the whole and it is difficult to give a precise definition to the term ‘‘race.’’ The latter problem has been discussed by Coon (’39) in the introduction to his book on the races of Europe: ‘í Since man is the oldest domestic animal ... any attempt to classify him by a rigid scheme is из difficult and the scheme must be elastic if it is to work at E: Hence the term ‘race’ must also be elastic. We may recognize if we like, cer- n major raees of the Old World such as the Bushman-Hottentot, the Pygmy, the ted the Negro, the Mongoloid, and the White. Within each of these major racial groups there are, or have been, smaller entities, which may deserve the desig- nation of race in a lesser sense. These smaller entities consist, for the most part, of groups of people reasonably isolated, and developing into local physical enclaves. . At sha border point such an entity becomes a major race it is not always НЫЕ % Ray: For the classification of Zea Mays we shall define the word race as loosely as possible, and say that a race is a group of related in- dividuals with enough characteristics in common to permit their rec- ognition as a group. The last three words should be emphasized. As Hooton has said in his discussion (726) of racial analysis, ‘‘races are great groups and any analysis of racial elements must be pri- marily an analysis of groups, not of separate individuals. One must conceive of race not as the combination of features which gives to each person his individual appearance, but rather as a vague physi- eal background, usually more or less obscured or overlaid by indi- vidual variations in single subjects and realized best in a composite picture." From the standpoint of geneties a race is a group of individuals with a significant number of genes in common, major races having a smaller number in common than do sub-races. The degree to [Vor. 29 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN which a species can be divided into recognizable races and sub-races will depend upon the degree to which it has been divided into dis- tinet intra-breeding groups with distinetive combinations of genes. In the recognition of race in Zea Mays it is therefore essential that we rely upon characters with a broad genetic background (such as cob shape and kernel size) rather than those which are indicators of but a single locus (as starchy vs. sweet or flint vs. flour). A natural classification of the races of Zea based upon characters like cob shape, kernel size, and tassel type will be incomplete, and one must expect disagreement as to details, even among authorities. For some time to come it must be a goal to work towards rather than a project to be completed. As a basis for understanding the history and genetics of corn, it is worthy of such long-continued effort. Since it is an attempt at a classification based upon the entire germ-plasm, it will integrate the maximum number of genetie faets. This does not mean that the races defined in this paper will sup- plant Sturtevant's classification as a cataloguing device. Any large enumeration of the varieties of corn must still be indexed by his system or by some such artificial system, though eventually some of the features of a natural system should be worth incorporating in a new comprehensive classifieation. For that reason no Latin names or precise diagnoses are given to the races of corn tentatively recognized in these papers. We have instead given short descriptive names to some of those groups of varieties in our colleetions with a recognizable core of common features. A person who is acquainted with Zea Mays only in one section of the world will be amazed at a collection of varieties from another region. In the United States, for instance, maize is so variable that one who has not seen it elsewhere would think that it could not ever be much more variable. Yet a collection from a single Mexican state will show at least one and sometimes two or three new races. Each will have a variability of its own and will differ from the maize of the American corn belt. A discussion of the North American races of Zea Mays must therefore begin with a description of its total variability, with special reference to those features which dis- tinguish the maize of one region from that of another. A SURVEY OF INTER-RACIAL VARIABILITY IN MAIZE For purposes of description (see Weatherwax, Chapters у-тх, хту-ху) we may divide a plant of Zea Mays into root and shoot. The 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 73 shoot consists of a primary axis, which often gives rise to secondary axes atits base. The primary axis is composed of a variable number of nodes with a leaf at each, and is prolonged into a compound male inflorescence, commonly called the tassel. At one or more of the nodes, short secondary branches arise with short wide foliar or- gans (the husks) and terminated by a condensed compound female inflorescence (the ear). Smaller ears arise in the axils of these foliar organs but they seldom develop to a functional stage. Root.— Though variation in the root is not as conspicuous as that in the shoot, it certainly exists (see Collins, 1918). Some Mexican and southwestern races of corn have much longer mesocotyls and many of the varieties of the Mexican plateau have poorly developed root systems as reported by Kuleshov (230). We have not as yet been able to work out any standardized measurement of these characters which can be used as a criterion of race. Branching of the Primary Aais.— The degree of branching of the primary axis and the relative development of the primary and secondary axes vary greatly in different races of corn. They are so unlike in the same variety of corn when it is planted in widely separated localities or at different times of year that we have not found it practicable to use this character as a criterion. The proportionate development of pri- mary and secondary axes (stalk and tiller) is a more stable char- acter and, with a little more study, might be found useful. Node Number and Internode Length.— Races of corn differ significantly in node number, but it is too affected by the environment to be useful as a racial criterion. The fundamental pattern of internode length is perhaps more stable and with a little further study might be found useful. Figure 1 shows the pattern of internode lengths in four different races of corn. It will be seen that there are considerable differences in the degree to which the internodes lengthen immediately above the up- permost ear. Leaves (Blade and Sheath).— The leaves of the various races differ greatly not only in their absolute size but in their proportions, as has already been com- mented upon by Kuleshov (loc. cit.). There is so much variabil- [Vor. 29 74 ANNALS OF THE MISSOURI BOTANICAL GARDEN ity upon each plant that we have not been able to determine an efficient method of scoring leaf shape. The leaf sheaths differ in their hairiness from race to race, and this we have been able to score in three qualitative grades, slight or none, evident, and heavy. Male Inflorescence.— The tassel of Zea Mays presents us with more easily measured characters than all the rest of the plant combined. While the homologues of its variation are present in the ear, the mature cob is so grown together and lignified that it is difficult to dissect and recognize the component parts. 304 ere PUEBLO MO. D. l. Successive internode lengths for representative plants of “Mexican Pyramidal, Pueblo, a dent corn from Missouri, and Pima-Papago. All grown out-of-doors at St. Louis. Circles repre- sent tassels, and triangles represent ears. HEIGHT IN CM. The tassel is composed of a central axis and a varying number of secondary axes, the lowest of which are themselves compound (Weatherwax, pp. 100-102). The spikelets tend very strongly to be arranged in pairs, one sessile and one pedicellate. The spikelet pairs are arranged distichously upon the secondary axes and poly- stichously upon the central spike. Although all these features are subject to variation, the general plan remains remarkably constant in different races of corn. The number of tassel branches is a sur- prisingly stable character (Weatherwax, p. 102) and is one of the most useful criteria of race. To be sure, it is a character which is affected by the vigor of the plant and for that reason we did not con- sider using it until experience forced its general stability upon our attention. It was first brought to our notice in connection with Cutler’s collections of maize from the Mexican Plateau for which we had (1) tassels collected in the Mexican corn fields, (2) tassels 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 15 grown in the greenhouse at St. Louis, (3) tassels from corn started in the greenhouse and ripened out of doors, (4) corn planted out of doors in St. Louis in May. While for each of these four lots the material was variable, the variation in each was from 0 to 4 branches. At the same time we also had in cultivation corn from Guatemala which bore a very large number of tassel branches though planted at the same time and in adjacent rows. It was also found to be generally true that the range of varia- tion in number of tassel branches was roughly the same for those cultures grown in triplicate at College Station, Cuba, and St. Louis (see pl. 11, tassels D and E). In one characteristic Mexican race there are no branches or only a few, while in South American corns there may be 30 or more branches. The stiffness of the entire tassel and the angle at which its branches depart from the primary axis is useful in the field but we have not been able to score it effectively on herbarium specimens. The size of the outer glumes of the spikelet is also important. It varies from 8 or 9 mm. in some South American corns to 15 or 16 mm. in corn from the Mexican plateau. The length of the sterile zone at the base of the branches is also significant. This varies from 1 or 2 mm. in some Mexican corn to 10 or 20 mm. in certain South American varieties. Two technical characters are highly diagnostic of certain races, the condensation of internodes and the degree to which the pedicel on the upper spikelet is developed. In Peruvian corn successive internodes are equal in length and exhibit a characteristic, regular zigzag arrangement. In many North Amer- ican varieties the internodes vary in length and are often so short that the spikelets appear as if borne four at a node. Two successive nodes may be so close together that they appear as one. In the corns of the Mexican plateau this condensation is sometimes quite regular, every second internode being condensed (fig. 2). The spikelets, as has been said, occur in pairs, the upper spikelet of each pair being more or less pedicellate. The relative length of this pedicel varies greatly, however. In some South American corns it may be as long as the spikelet. In many Mexican corns it is sub- pedicellate and may even be quite as sessile as the lower spikelet. Plate 11 A shows an extreme example, a corn from Xochimileo in which all the spikelets are sessile and the alternate internodes con- densed so that it has the appearance of bearing spikelets in whorls of four. [Vor. 29 76 ANNALS OF THE MISSOURI BOTANICAL GARDEN Female Inflorescence.— Next to the tassel the most important criteria of race are to be found in the female inflorescence (the ear). The number of husks and their shape are highly diagnostie but we have not used it because ears with husks attached are difficult to store and to study. Ear and kernel were the basis of Sturtevant’s classification and we have merely selected those of his criteria which seemed most diagnostic ; Fig. 2. Portions of tassel branches, greatly enlarged: left "теь а 128882, from Quinta, Peru; right, Cutler 45, from Xochimileo, Mexico row number, ear length, kernel width, thickness and length, and cob width. The composition of the kernel is an important racial eriterion only to the extent that it is based upon a large number of genes. Starchy vs. sweet or flint vs. flour are simple gene differences and therefore of minor importance. Denting and capping, on the other hand, are the results of larger numbers of genes and are there- fore more important though they are difficult to measure. Color differences, for the most part, are the results of a few gene differ- ences and are somewhat superficial. However, it will be seen that when the races of maize have been separated on other grounds that each has its own characteristic color range. 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 74. In addition to these readily scored features, there are a number of general differences in ear shape, kernel pattern and kernel texture which are easily perceived by the trained eye but which are difficult to score objectively and commensurately (ie., in a graded series). These include: (1) the general shape of the cob, cylindrical, long-tapering, short-tapering, (2) the shape of the butt, appressed, rounded or enlarged, (3) the extent to which the kernels are arranged in evident rows, (4) the sulci between the rows, (5) the regularity of the kernels (i.e., the amount of variation in size and shape from kernel to kernel), (6) the degree to which the kernels have been compressed by the husks. The general problem of recognizing and measuring races has been discussed by Hooton (726) in one of his classic contributions to the subject. Though he was considering mankind rather than Zea Mays, the problem is fundamentally the same and our techniques differ from his only in detail. Our material, however, had the ad- vantage that we had two sets of complicated organs for analysis rather than one. His work was mainly limited to skulls. Ours had not only the ear but also the associated tassel. We followed his method in sorting out the ears by general perception, using the total impression given to the trained observer by its totality of char- acters. Having grouped his crania by this method, Hooton then measured them and computed averages for each recognized race, sub-race, etc. We had a somewhat better check on our conclusions sinee we first sorted the ears and then measured and averaged the associated tassels. In every case the exact data derived from the tassels confirmed our judgments with regard to the ears. Both of these conclusions received additional confirmation when the distri- butions of the races and sub-races were determined and were found to agree with recognized geographical and ethnic barriers. We had an added advantage in that much of our material was grown in triplicate at St. Louis, Mo., College Station, Texas, and Cienfuegos, Cuba, so that we had a rough idea of what differences were purely environmental. We are indebted to Dr. Paul C. Mangelsdorf and the Arnold Arboretum for the Cuban collections and to Dr. R. G. Reeves and the Texas A. and M. College for that grown at College Station. The general method of work can be illustrated by our procedure with the collections from the southwestern United States. (These are to be discussed in greater detail in a forthcoming article by An- derson and Carter). For this area we had about fifty collections [Vor. 29 78 ANNALS OF THE MISSOURI BOTANICAL GARDEN made in the Southwest over a series of years by the junior author and by George F. Carter, to whom we are greatly indebted not only for a comprehensive collection but for much pertinent information. The tassels were collected and numbered by an assistant to re- duce the personal equation to a minimum. The ears (some of them collected in the Southwest, others grown from seed collected there) were spread out on a large table and were carefully examined. It GLUME LENGTH IN MM. o NO. TASSEL BRANCHES ^ L L р # SUB-SESSILE SPIKELETS \ edere \ (Ж % CONDENSED INTERNODE t—i— © LENGTH STERILE ZONE c——- SMALL PIMA- PAPAGO E a п АВС INTERMEDIATES PUEBLO Figure 3 was almost immediately apparent that there were at least two centers of variation: (1) the large-cobbed, straight-rowed varieties grown by the Pueblo Indians, and (2) the small-cobbed mosaic- seeded type grown by the Pima and Papago Indians. We therefore sorted the material into three piles, ‘‘Pueblo,’’ ‘‘Pima-Papago,”’ and unknowns. By study it was possible to reduce the third pile little by little and we soon came to the conclusion that, aside from a few obvious recent admixtures such as pop corns and dents, there were only these two basic races of corn in the Southwest. The ma- 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 79 terial lent itself to classification in four groups: (1) Pueblo, (2) ob- viously intermediate, (3) Pima-Papago, but somewhat intermedi- ate, and (4) Pima-Papago. We then studied the numbered tassels and measured five features previously chosen as racially diagnostic. Averages for each of the four groups were then computed and are shown diagrammatically for all 5 characters in fig. 3. It will be seen that five measured characters of the tassels agree with our total impression of the ear and kernel. The Pima-Papago and Pueblo are extreme for each of the five tassel characters, the intermediates oc- cupy an intermediate position, and the averages of class 3 depart only slightly from those of the Pima-Papago. The distribution of these two races among the southwestern In- dians was then examined and found to agree with geographical and ethnic barriers. The Pueblo varieties are grown by the pueblo- dwelling Indians of the plateaus, the Pima-Papago by the Indians inhabiting the intensely hot deserts near the delta of the Colorado. The Pima-Papago varieties are not grown at all by the eastern Pueblos and are found only among those western Pueblos (most particularly the Hopi) which have for some centuries been iso- lated from the main group of Pueblo Indians and which are geo- graphically much closer to the territory of the Pima and the Papago. After the classification of the southwestern maize had been com- pleted the Mexican and Guatemalan collections were studied in the same way. For these areas, in addition to the material collected by Cutler, we had extensive collections which came to us through the courtesy of Prof. Carl Sauer and Prof. Ralph Beals of the Univer- sity of California, and a small but extremely useful collection of tassels and associated ears made in Mexico by our colleague, Mr. R. W. Schery. For these areas our collections are not so compre- hensive and the variation of Zea Mays is much more complex. There are undoubtedly numerous sub-races which we have not yet recog- nized and some of those which we are now recognizing may ulti- mately prove to be intermediates. They are put forward tentatively as a basis for the discussion that may eventually give us a compre- hensive yet natural classification of the races of Zea Mays. The three Central American races (or sub-races) which we are describing below under the names of ‘‘Guatemalan Big Grains," ** Mexican Pyramidal"' and ‘‘Guatemalan Tropical Flints’’ are ad- vanced merely as a basis for study. Their relations to each other and to the other races of Central and South America remain to be [Vor. 29 80 ANNALS OF THE MISSOURI BOTANICAL GARDEN determined. They do not constitute the only races in our collections and it is not even suggested that the three groups are of the same rank. It is quite possible that when the racial composition of corn has been thoroughly worked out one of these three might constitute a major race and the other two would be of subordinate rank. Though all three are variable each one has enough of a core of com- mon features and a definite enough distribution in space and time to be thought of as an entity. In the southwestern United States the corn situation is simpler and we have much more comprehensive collections. The two races, **Pima-Papago'' and ‘‘Pueblo,’’? which we are describing below, are certainly the two main entities in the maize grown by the Indians of that region. Aside from the obvious intermediates discussed above, the only other perceptible influences are admixtures derived from the American and Spanish communities and various Mexican types brought in by Apache raids. M exican Pyramidal (fig. 4; pl. 11 A, B, C).— This is the common corn of the region about Mexico City and was probably therefore closely associated with the Aztecs. It is so distinctive in leaf and habit that it was named Zea hirta over a hundred years ago by Bonafous (1836) and was described in detail by Kuleshov (730), who called it the ‘‘central Mexican type’’ and deseribed its pubescent leaf sheaths, drooping leaves, poorly branched tassels, and weak root systems. He also called attention to the fact that this type might have kernels which were pop or flour or dent **as well as any other form of maize." In addition to the characters described by Kuleshov, the Mexican Pyramidals have long harsh glumes in the tassel and a high percentage of condensed internodes and subsessile upper spikelets (pl. 11 and fig. 4). The ears are equally distinctive. Though varying greatly in size they all have the same general shape. They are relatively short and taper sharply and regularly, hence the name ''Pyramidal." Row pat- tern varies from regular to irregular with mixtures of both being common. They most generally show little husk compression; there are wide spaces between the rows and often between individual grains. The kernels are prevailingly long though they vary in shape, usually on the same ear, pointed to dented to rounded ker- nels often occurring together. The kernel is nearly always elon- gated, however, and the denting is usually a slight depression towards the tip of the grain. White is the commonest color, accom- 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 81 panied by heavy white capping, but dark reds and blue-blacks are also met with. This race is of particular importance not only because of its prob- able association with Aztee civilization but because it apparently has entered into the make-up of our corn-belt dents. Any one who will examine large fields of non-inbred dent corn can find (in a greatly reduced degree) much of the same variation which char- acterizes the ears and tassels of the Mexican Pyramidals. Many of the inbreds derived from dent corns exhibit in an exaggerated form certain of the characters to be found in Mexican Pyramidals though in any one inbred they are usually combined with other quite different features. The Mexican pop corns apparently form a sub- race of the Mexican Pyramidals since in our material their tasscls are indistinguishable and their cob shapes (as distinct from cob size) and kernel shapes are practically the same. More material must be examined before this conclusion can be accepted. Guatemalan Tropical Flints (fig. 5; pl. 11 D, E, F).— Apparently a considerable number of tropical varieties have been classified under the general category of ‘‘Tropical Flints." For the present we are restricting the term Guatemalan Tropical Flints to a fairly homogeneous race of which we have one collection from Mexico, several from Cuba, and a number from Guatemala, princi- pally on the Caribbean side. The cobs are short to medium but pro- portionately broad and the seeds are small. The kernels are reg- ular in size and shape and the rows are very straight and sulci evi- dent though not very large. The tassel is much larger than in most North American varieties and there are more branches (the average number is over 20) but the rachises are more slender and the spike- lets smaller. Condensed internodes are very rare as are also sub- sessile spikelets, and the tassel branches by comparison with those of most North American corn look more delicate and more regular. The race is characteristically flint but slight indications of dent- ing are frequently seen. A small cap of white or light yellow is char- acteristic and there is variation in color. Guatemalan Big Grains (fig. 6; pl. 11 G, H, I).— The outstanding characteristic of this race is the shape of the cob, which is large, sub-cylindrical with a more or less conspicuously enlarged base. The rows are usually straight in the cylindrical por- tion and too irregular to be recognizable in the enlarged portion. In color the race varies greatly, bright colors being the rule. The [Vor. 29 82 ANNALS OF THE MISSOURI BOTANICAL GARDEN Figs. 4-7. Ear and tassel characters for 5 races of maize: fig. 4, Mexican Pyramidal; fig. 5, Guatemalan Tropical Flint; fig. 6, Guatemalan Big Grain; fig. 7, Racial averages (GBG, Guatemalan Big Grain, PP, Pima-Papago, GTF, Guatemalan Tropical Flint MP, Mexican Pyramidal). , 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS # А 12 1 в го t [e 16 L| D 0 E 6 F 60 U © 100 6 A |2 в 20 с 16 D б Е 0 F 60 G 100 83 In figs. 4—6 each line represents an individual plant and shows its value for 7 different characters as follows: А, median cob width in em.; B, kernel width x thickness in sq. mm. ; C, kernel length in mm.; D, number of tassel branches; E, length of sterile zone in mm.; F, % greatly condensed internodes; G, % sub-sessile spikelets. [Vor. 29 84 ANNALS OF THE MISSOURI BOTANICAL GARDEN majority are flints though we have some flour varieties and a num- ber of semi-dented flints. The tassels are similar to those of the Tropical Flints but are even larger and, in our material at least, were characterized by their lax drooping habit. They tend to have very long sterile zones at the base of the branches even up to 10 or 20 cm. in length. The only plant to show subsessile spikelets had already been suspected on ear characters of having been crossed with a Mexican Pyramidal. Condensed internodes are occasionally seen, but, on the whole, the tassels look very much like those of the Tropical Flints as defined above. Pima-Papago.— With very few exceptions these varieties are either white or a bright light yellow. Flour corn is the rule, though flints are ос- casionally found, while denting is unknown. The kernels are small but the cob is proportionately even smaller. This is particularly noticeable at the base. The ear tapers slightly, usually toward both ends, and there is evidence of compression in the narrow base and in lengthwise striations across the kernels. The inner husks are pro- portionately wide. While the rows are often straight, the individual seeds vary so much in shape that when looking at an ear one notices the kernels rather than the rows. For this reason they look more or less like tiles in a pavement and we have used the term ‘‘mosaic- seeded?’ to describe this effect. The tassels are stiff and harsh and though of medium size have very large glumes. While sub-sessile spikelets are not as common as in Mexican Pyramidal corn, about a tenth of the internodes usually show them. Condensed internodes also occur, though less frequently. As was stated above, these vari- eties are grown by the agricultural Indians of the deserts near the delta of the Colorado, the Pima, Papago, Yuma, etc. To archeolo- gists this maize is of particular interest because it is very similar to the prehistoric Basketmaker corn which is the oldest recognized type of corn in the Southwest. The habit of the plant is distinctive. Compared to corn-belt dents the stem is slender and the internodes are long for their width, as are also the leaves. Tillers are usually sub-equal with the main stalk. Pueblo.— This race is very different from the above and is apparently al- lied to the Big Grains of Central America. It is characteristically big-cobbed and big-shanked. The ears are long and straight-rowed and the sulci are distinct. While the kernels are by no means so 1942] ANDERSON AND CUTLER—RACES OF ZEA MAYS 85 wide as those of the Central American Big Grains, they are larger than the Pima-Papago and more regular. The base of the ear is usually square or sightly rounded though an enlarged base is oc- easionally seen. The kernel is either flour or flint and a very faint semi-dent is quite common. The kernel is usually colored, and most white varieties show obvious influence of Pima-Papago in other characters beside color. The tassel is larger than that of the Pima-Papago race with nearly twice as many tassel branches on an average but with slightly smaller glumes. Sub-sessile spikelets are rare but nearly every tassel shows irregular internodes and about one out of ten is obviously condensed. Some plants have a peculiar growth habit. The internodes tend to be short and the leaves wide. The tassels are not greatly exserted and the ears are long, giving the whole plant a curious ‘‘squatty’’ appearance. SUMMARY The need of a modern natural classification for the kinds of maize is discussed. Natural and artificial classifications are contrasted and illustrated by simple examples. It is pointed out that each has its advantages and disadvantages; artificial classifications are simple and comprehensive but do not integrate a large number of facts ; natural classifications are often difficult to make and cumber- some to use but when they can be achieved they integrate the maxi- mum number of facts. The difficulties of grouping Zea Mays into natural races and sub- races are discussed. For the purpose of this paper a race or sub- race is defined as a number of varieties with enough characters in common to permit their recognition as a group; in genetical terms it is a group with a significant number of genes in common. The external morphology of the corn plant is outlined with ref- erence to those characters which might be used as criteria for the recognition and description of race, and the application of these criteria to the classification of maize is illustrated by examples. Indian varieties of maize from the Southwest belong to two main races, the Pueblo and the Pima-Papago, with a few recent admix- tures and some obvious intermediates. The first of these races is grown chiefly by the Pueblo-dwelling Indians of the plateaus, the second by desert peoples near the delta of the Colorado River. The latter is very similar to the corn of the prehistoric Basketmakers. Three Central American groups are described and illustrated, [Vou. 29, 1942] 86 ANNALS OF THE MISSOURI BOTANICAL GARDEN though their exact relationships are as yet uncertain: (1) the Guate- malan Big Grains, (2) the Guatemalan Tropical Flints, and (3) the Mexican Pyramidals. The last is of particular interest because of its association with the territory of the Aztecs and its relation to modern corn-belt varieties. BIBLIOGRAPHY Bonafous, M. (1836). Histoire naturelle, agricole et économique du Mais. Paris. Collins, G. N. (1918). Tropical varieties of maize. Jour. Hered. 9: 147—154. Coon, C. S. (1939). The races of Europe. pp. vii-xvi, 1-739. New York. Guernsey, S. J., and A. V. Kidder (1921). Basketmaker caves of northeastern Arizona. Peabody Mos. Papers 8?: › 1-121. Hooton, E. A. (1926). Methods КА беда analysis. Science N.S. 63: 75-81. P ы x (1930). The maize of Mexico, Guatemala, Cuba, Panama and Colombia. l. In Bukasov, S. M. The cultivated plants of Mexico, Guatemala and алан. Bull. Appl. Bot., Genet. & Plant Breed. Suppl. 47: English summary, 470—553. Mangelsdorf, P. C., and В. С. Reeves (1938). The origin of maize. Nat. Acad. Sei. Proc. 24: 303-312. , (1939). The origin of Indian corn and its relatives. Tex. Agr. Exp. Sta. Bull. 74: 1-315. Sturtevant, E. L. (1885). N. Y. Agr. Exp. Sta. Ann. Rept. 3: 124-188. — — ———, (1886). Ibid. 4: 73-89. — — ——, (1887). Ibid. 5: 58-66. речна Notes on maize. Bull. Torrey Bot. Club 21: 319—343, 503-523. — — ———, (1899). Varieties of corn. U. S. Dept. Agr. Off. Exp. Sta. Bull. 57: 1-108. Митин, P. (1923). The story of the maize plant. pp. i-xv, 1-247. Chicago. [Vor. 29, 1942] 88 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 11 Fog eats of ears and tassels (to different scales) of representative plants of three aize, as follo upper row, Mexican Pyramidal; middle row, nei alan Tropical Flint, bottom row, Guatemalan Big Grain. The ears were collected by Cutler at the points listed below, and the tassels are from seeds taken from the ears = grown at College Station, Texas, and Cienfuegos, Cuba. Except in one instance (E on the plate), the tassel associated with each ear was grown from seed taken from that particular ear. The ears have been somewhat mutilated in measuring mid-cob width. The scales in the center of the figure each represent 5 em.; the scale to the left is for the ears, the one to the right for the tassels. Cutler's Number of ear Place at Figure collection Source of ear from which tas- which tassel sel was pot was grown A 42 Xochimileo, Mex. 42 College Station B 45 Xochimileo, Mex. 45 College Station 0 2 Toluca, Mex. 2 College Station D 61 El Progreso, аа 18 College Station E 18 Joyabaj, Gua 18 Cienfuegos F 24 Chica], 24 College Station G 16 ина, Е 16 College Station H 5 Quiché uat. 5 College Station I 6 а, Guat. 6 College Station ANN. Mo. Bor. GARD., Vor. 29, 1942 docete "mo ж Vu ям МИ rF р | 5 AZN д "y “ \ H ПРИЕ pr liane AURI С "s «= № eras [I7 еги 2 SW e i # +. in 3 = Bh 2329. Ec 4 Ek $4 „= 2-5 эъ H var 4 КАК | + 2 DET i3 2285: ЕЕ ---I $ та Y т T3252 4- JESS 5 жу 2 a 2= аа 2s за: газ = е: тъ“ = «12 x E ize “г =» r أ ЧҮ] ло, Фу ANDERSON AND CUTLER—RACES OF ZEA MAYS PLATE z ‚№ L4 $ 4 Y a g * s К ў ” Ё у Я у щат UNE а 11 DOLOMITIC GLADES OF EAST-CENTRAL MISSOURI RALPH O. ERICKSON University Fellow, Henry Shaw School of Botany of Washington University LOUIS G. BRENNER Formerly Graduate Apprentice, Henry Shaw School of Botany of Washington University AnD JOSEPH WRAIGHT Formerly Assistant in Geography, Washington University A characteristic feature of the landscape of the Ozarks is the oc- currence on south- and west-facing hillsides of barrens which are locally known as ‘‘glades’’ or ‘‘cedar glades.’’ Such glades, oc- curring on outcrops of thin-bedded dolomite or dolomitic limestone, principally of lower Ordovician age, have been studied in the area shown on the map (fig. 1). It includes about three-fourths of Jefferson County and small parts of adjacent Franklin, Washing- ton and Ste. Genevieve Counties, Missouri, and lies from 25 to 50 miles west, southwest and south of the city of St. Louis, on the north- eastern border of the Ozark Plateau. The unique character of the Ozark glades has long been recog- nized by botanists, though no extended accounts have been pub- lished. Discussions may be found in the papers on the Ozark flora by Palmer (’21) and by Steyermark (’34; ’40), who consider them as relics of the former plains flora of the region. References to the glades may also be found in the geological and geographical literature (e. g., Weller and St. Clair, ’28; Pike, ’28; Sauer, ’20; Cozzens, ’37; 39). The present study is largely a by-product of two other investiga- tions, by Brenner and by Erickson. Brenner’s interest in the glades grew out of his study of the environmental variables at the Gray Summit Arboretum of the Missouri Botanical Garden (Brenner, '49). Erickson is engaged in a study of the distribution of Clematis Fremontwu in Missouri, and the detailed mapping of the glades was done primarily as preliminary work for that study. Wraight be- came interested in the problem for its geographical and ecological implications. The analysis of the physical characteristics of the glades which follows is largely due to the field work of Brenner and Wraight. This paper is primarily descriptive, but an attempt has also been made to analyze the factors which are responsible for the occur- (89) [Vor. 29 90 ANNALS OF THE MISSOURI BOTANICAL GARDEN rence of the glades and for their characteristic physical conditions. Although a detailed treatment of the flora is beyond the scope of the present paper, the following remarks may be made. The glades are sometimes completely treeless, but more often there are ‘‘is- lands’’ of small trees surrounding gullies or ledges of rock. The species most frequently found are:! Juniperus virginiana, Quercus Muhlenbergii, Q. stellata, Celtis pumila, C. pumila var. georgiana, Acer saccharum var. Schneckii, Rhamnus caroliniana, Vitis Lince- сити var. glauca, Cornus florida and Bumelia lanuginosa. Most characteristic of the glades, however, are the herbaceous species. A partial list includes: Andropogon provincialis, A. scoparius, Panicum virgatum, Sporobolus heterolepis, Bouteloua curtipendula, Allium stellatum, Nothoscordum bivalve, Agave vir- ginica, Arenaria patula, Clematis Fremontii, Draba cuneif olia, Leavenworthia uniflora, Baptisia vespertina, Psoralea esculenta, Petalostemum purpureum, Croton capitatus, Euphorbia corollata, Viola pedata, Oenothera missouriensis, Polytaenia Nuttall, As- clepiodora viridis, Acerates viridiflora, Heliotropium tenellum, Scutellaria parvula, Houstonia angustifolia, Solidago Gattingeri, Aster oblongifolius, Silphium laciniatum, S. terebinthinaceum, Rudbeckia missouriensis, Echinacea pallida and Coreopsis lance- olata. Although the grass cover of the glades is rather sparse, usually consisting largely of separated clumps of Andropogon, the grass is the most conspicuous of the herbaceous plants (fig. 4). Despite the grassy appearance of the glades during most of the year, there are a number of other plants which, during their blooming periods, cover the glades with solid masses of flowers. Such plants include both annuals like Leavenworthia uniflora, which flowers in April, and perennials such as Rudbeckia missouriensis, which carpets the glades in late July and early August. The glade species are all more or less xerophytic, and a number of them are remarkable for their linear or finely divided leaves. In mapping the glades, use was made of the aerial photographs on file at the offices of the county Agricultural Conservation Associa- tions. The photographs were taken at an altitude of about 13,000 feet and are printed to a scale of 8 inches to a mile. Most of them * The list of species given here agrees with Palmer and Steyermark’s (735) eatalogue, although the authors realize that, in some instances, that work is not up-to-date in nomenclature, 1942] ERICKSON, BRENNER & WRAIGHT—DOLOMITIC GLADES 91 / RE 7 RE КАЕ ? Gray Summito х. / P Im ( V. poLoMrric j /Pacifi “psi GLADES~ 3 IN z 4, JEFFERSON COUNTY MISSOURI AND PORTIONS OF поза . ADJOINING COUNTIES N СР ==... SCALE=1:350,000 та Фе etel uj T EUM д X چ“‎ mt . = , pas Morse Mi 1 ^ РА + C) | 3 / ^ T.39N @ йш menm [и >. ~ ERA ТЕНЕ СКЫ LS Б = Г 7 v gi ON СЯ, al dh RSF, На ар: E Bm "rg NGEL. Цена: LET TAI ча 1 T ВВ E а КЕҮ МАР RISE FIGURE 1 [Vor. 29 92 ANNALS OF THE MISSOURI BOTANICAL GARDEN were taken in September, 1937. Glades can be recognized rather easily on the photographs by their whiteness, and they can be distinguished from other white areas, such as bottomland pasture, by their characteristic outline, by the islands of red cedar which occur in most of them, and, in the larger glades, by ledges of resist- ant rock which give the appearance of contour lines. For each square mile within the area, a tracing was made showing the out- lines of glades and some other features such as roads and creeks. The map (fig. 1) was drawn from the photo tracings on a base taken from the county highway maps published by the Missouri State Highway Department. All the glades were drawn as ac- curately to scale as possible. The drainage was drawn in from the photo tracings and the topographie maps of the U. S. Geologieal Survey. The map has been checked by field work to some extent throughout the area and rather completely in a few places. The errors are confined to the smallest glades, since it is somewhat diffieult to make a distinetion between very small glades and hill- sides with a rather open cover of red cedars. The glades in this region almost invariably oecur on south- and west-facing slopes. They rarely extend over the top of a hill or ridge and are never found in a valley bottom. The soil mantle is extremely thin, very often less than one inch deep, and there are always numerous flat fragments of the bed rock strewn on the sur- face (fig. 4). The ground-water relations of the glades are char- acteristic. During late autumn, winter and spring they are typically so saturated that there is almost continual seepage, and such plants as Nostoc terrestris and Isoetes Butleri, members of normally aquatic genera, flourish. This saturation leads to a great deal of disturbance from frost heavage when the ground thaws in the spring. During the summer months, on the other hand, the glades are extremely xeric, and for a period in the middle of the summer very little green may appear. The conditions determining the topographic position of the glades, their unusual physical characteristics and their peculiar flora represent a rather complex interplay of factors such as angle of slope, thin soil mantle, attitude to the sun’s rays, and meteoro- logical conditions, as well as the composition and physical char- acteristics of the underlying rock. The geological relations of the glades are quite evident from a study of their distribution. Formations of upper Cambrian and 1942] ERICKSON, BRENNER & WRAIGHT—DOLOMITIC GLADES 93 middle and lower Ordovician age outcrop within the area of this study, the geological column being as follows :? f ion Kimmswick Mohawkian < Decorah Plattin Big Buffalo E Ordovician ¬ St. Peter Everton а m Canadian Jefferson City lane ica Roub e glade [o "this E er 3asconade Ozarkian Eminence Potosi Cambrian Derby-Doerun avis Bonneterre These formations present some lithologic contrasts, including a variety of sandstones, limestones, shales and dolomites. The dolomites are of particular importance, since the presence of thin- bedded, relatively non-cherty dolomite (fig. 5) appears to be the decisive factor in the production of glades. The largest glades oc- cur on the outcrop belt of the lower part of the Powell formation, which includes the greatest thickness of such rock in the column. Similar dolomitic strata are found in the Joachim formation, the upper portion of the Everton, the Cotter, the Jefferson City and perhaps in all the other formations listed above except the St. Peter sandstone, the Roubidoux formation and those formations above the Joachim. The largest and most numerous glades, as stated above, occur on outcrops of the lower Powell. The Cotter and Jef- 2 The geological column is ee from the VM Lo e Map of Missouri, with the omission of formations absent from the area of this study. з The formations of the ps Series are notoriously | difficult to rl in the field. They vary greatly in lithology within short distances, both vertically and hori- zontally, and fossils are difficult to find. The geological Lc dd A the glades made in this paper are based upon the geological map accompanying Pike’s (728) dissertation on the Crystal City ds oa which agrees in interpretation with the ma Genevieve County of Weller and St. Clair (228); on personal conversation with Mr. James . Cullison; and on such field entiation by the i as was possible. Cozzens (737; 739) has correlated the glades with the Cotter formation. While the Powell is the most important pu in the southern part of the area described [Vor. 29 94 ANNALS OF THE MISSOURI BOTANICAL GARDEN ferson City, which, with the Powell, constitute the Jefferson City group of the state geological map, are next in importance as glade- producing formations. Small glades are found frequently on the Joachim; the location of the St. Peter-Joachim escarpment (Sand Ridge) is indicated in fig. 1 by the glades on this formation. The Everton is of slight importance in this study. The Gasconade, Emi- nence and Potosi formations are not important, but small glades apparently occur occasionally. The numerous small glades in northeastern Washington County (T.39N, R.3E) are found on the Derby-Doerun. Glades are reported to occur on the Bonneterre formation (Sauer, '28, pl. III a), but its outerop lies outside of the area investigated. In the area of the Plattin and Kimmswick outcrops, south-facing slopes have a sparse tree cover in which red cedars often predomi- nate, and in some places there are few, if any, trees. However, such slopes in the Plattin and Kimmswick country are much steeper than in the area of this study, and there is no thin-bedded dolomite. Where the hillsides are barren enough to justify the term glade, the appearance is quite different both floristically and lithologically from the dolomitic glades. Glades occur on the St. Peter sandstone which also support a different flora from that of the dolomitic glades. Besides the lithologic character of the rocks, a second factor de- termining the areal distribution of glades is the amount of local relief. The glades lie almost wholly within the Mississippi and Missouri River Border Provinces defined by Sauer (720, fig. 18), which are equivalent to the River Border Region of Cozzens (739, fig. 4). The region as a whole is one of moderate relief for the Ozarks, values of 300 to 400 feet being given by Cozzens, but there is, of course, considerable variation within the region. Glades seem to occur only where the proper lithologie conditions are found in conjunction with relatively high local relief. For instance, refer- ence to the map will show that there are practically no glades be- tween Hillsboro and Sand Ridge. This area is underlain by the same thin-bedded dolomite which elsewhere produces glades, but the relief is not sufficient to allow their development here. In the southeastern part of T.40N, R.5E, one to three miles north of the here, the only part for which a large-scale geological map is available, the glades are by no means limited to the Powell. Considering the variability of the Jefferson City group from place to place, the Cotter formation may very well be the most important glade-former elsewhere. 1942] = = E AH Y a > . . a . 1TA Y пое - = үл © Fig. 2. Glade on lower part of the Powell formation, one mile north of Plattin (SE 14, S, 36, T. 40N, В. 5E). АП the photographs E were taken in January, 1942. = a NE E og 3. Severely grazed glade, 215 miles southeast of Hillsboro (NE И, S. 14, T. 40N, Е. 4E). ANNALS OF THE MISSOURI BOTANICAL GARDEN л cent road eut near Hillsboro, showing thin- "ug Re bedded E rit of the kind on which glades occur, 1942] ERICKSON, BRENNER & WRAIGHT—DOLOMITIC GLADES ЗЕ village of Plattin, on Ше other hand, where rock of Ше same lithol- ogy underlies land dissected by tributaries of Plattin Creek, glades occur on every ridge. Much of the rock scattered on the surface of the glades is soft and friable, and is called **cotton rock?! by residents of the Ozarks. Samples from several localities were tested for dolomite content by the method described by Brenner (742, p. 104), with results indicat- ing a very high proportion of dolomite erystals. Porosity tests showed about 25 per cent pore space, as compared with values of 3 to 9 per cent for massive dolomite and limestone from the same lo- calities, and 10 to 12 per cent for sandstone. Soil tests made on the glades at Gray Summit (S.17, T.43N, R.2E) indicated a slight acid reaction and a high organic content. There is very little material corresponding to the leaf litter of the adjacent woods, and the high organic content should probably be ascribed to slow decomposition rather than abundant supply of organie ma- terial. This is related to the meager soil flora and fauna, which is in turn due to the extreme variations in temperature and moisture conditions of the glades. It is probable that the amount of available plant nutrients in the glade soil is low, though no attempt has been made to investigate this point. Probably more important than the chemical composition of either the rock or the soil is the thin-bedding of the dolomite (fig. 5). The beds vary from a fraction to three or four inches in thick- ness, and contain numerous vertical joints. The presence of bed- ding planes and joints, along with the extreme porosity of some of the beds, results in ample space for subsurface water. The thin- bedded strata whieh outerop in the glades are typically underlain by relatively impermeable massive strata, usually dolomitie, which check the downward percolation of water. This water moves lat- erally along the numerous bedding planes and, since there is no soil mantle to diffuse it, there is abundant seepage at the outerops from late autumn to early summer. Pluvial eonditions are also important in the saturation of the glades during the winter and their desiecation during the summer. The mean annual precipitation in this region is about 43 inches. On the whole, the greatest part of this falls during the summer, and the least during the three winter months. At Union, which can be taken as typical of the glade area, since it is only about three miles west of the area included in the map, there are two peaks of [Vor. 29 98 ANNALS OF THE MISSOURI BOTANICAL GARDEN 4.54 inches in May and 4.70 in August, and a minimum of 2.20 inches for February. The effective precipitation for the glades, however, presents a somewhat different picture. During the win- ter and spring months, the rhythmic succession of cyclonic storms brings some sort of precipitation at least once a week. This is usually in the form of a slow drizzle resulting in relatively high infiltration and low run-off, hence much water passes underground. Furthermore, little water is lost by evaporation because of the relatively low temperature and high humidity during such periods of precipitation. This saturation may be an important factor in excluding species of the upland elimax from the glades by encourag- ing winter rotting. During the summer, on the other hand, precipitation is very ir- regular and in the form of violent, erratie thunderstorms, sepa- rated by longer or shorter periods of drought. During such storms the volume of water falling in a given period is great, runoff is ex- cessive and infiltration is low, hence little water goes underground, particularly on the glades where the soil mantle is thin. The high temperature and low humidity which often follow such thunder- storms further reduce the effective precipitation. Again, the thin soil of the glades contributes to their xeric condition in the summer. A thick layer of soil normally forms a mulch which retards evapora- tion of subsurface water. The lack of such a mulch on the glades makes for rapid desiccation during warm weather. The angle of slope of the glades is relatively steep and it appears to be an important factor in determining their characteristics. The slope of a few large glades has been measured by hand-leveling and pacing. There is considerable local variation in gradient at different levels, which can be ascribed to the presence of strata of different resistance, but the average slope appears to vary from 15 to 20 per cent, while that of adjacent north-facing wooded slopes approxi- mates 10 per cent. This steep gradient permits even small volumes of water washing down the slope to carry a considerable load, and is probably one of the principal reasons for the thin soil cover. The steep slope of the glades, coupled with their southern and western attitude, gives them a high angle of incidence to the sun’s rays dur- ing the warmest part of the day. The attitude of the glades also ex- poses them to the drying effect of the prevailingly southern winds of the Ozark region. In the opinion of the writers the dolomitic glades of this area 1942] ERICKSON, BRENNER & ERGE GLADES 99 are permanent ecological entities, entitled to be called an edaphic climax, rather than the first stage in the succession leading to a sugar maple-white oak climax, as they have been regarded by Steyermark (740, p. 372 et seq.). The fact that open grassland was formerly much more extensive in the Ozarks, and that, with the white settlement of the country and the accompanying reduction of fires, there has been a widespread invasien of the grassland by forest, is too well documented to be doubted. However, the extensive Ozark grasslands of the early nineteenth century occupied hill and ridge tops, where glades rarely occur, and there is no reason for be- lieving that they owed their existence to any more permanent factor than the annual fires which were encouraged by the Indians and early settlers. The dolomitie glades described here, however, seem to present too extreme a set of environmental factors to permit their invasion by trees of the upland climax of this region. The glades seem to be so largely determined by the character of the geological substrate that it would appear to require a rather large- scale change in climatic conditions to obliterate them. Many of the glades are subjected to grazing by cattle, and to a lesser extent by hogs. Grazing naturally aggravates the barren- ness, and occasionally the overgrazing is so severe that practically no plant cover remains. At best the glades produce poor pasture, and a considerable proportion shows no evidence of grazing. It does not seem probable, from a comparison of grazed and apparently ungrazed glades, that they owe their existence to grazing. On the contrary, moderate grazing seems to encourage the appearance of red cedars. It has been pointed out above that glades occurring on the St. Peter sandstone and on the Plattin and Kimmswick formations dif- fer considerably in character from the dolomitie glades. Barrens are also found on the La Motte sandstone, and in granitic, porphy- ritie, and highly cherty areas of the Ozarks (Steyermark, '40; Palmer, 710). None of these have been investigated by the present writers, but they should probably all be distinguished from the dolo- mitie glades of this study. The glades of this area bear resemblances to the glades and bald knobs of the White River region of Missouri and Arkansas, and to the cedar glades of the Nashville Basin in Tennessee (Harper, '26 ; Freeman, '33). Relations may also be pointed out between the flora of the glades and that of the Great Plains. The shale barrens [Vor. 29 100 ANNALS OF THE MISSOURI BOTANICAL GARDEN of Virginia and West Virginia (Wherry, '30; Core, '40) appear to represent a similar geologically determined habitat, and they have some floristie similarities to the glades in this area. Acknowledgments are due the following persons for valuable assistance and criticism: Prof. Edgar Anderson, Prof. L. F. Thomas and Prof. Courtney Werner, of Washington University, Dr. Julian A. Steyermark of the Field Museum of Natural History, Prof. James S. Cullison of the Missouri School of Mines, Mr. H. 8. McQueen of the Missouri Geological Survey and Dr. Norman Hinchey, formerly of the Missouri Geological Survey. Mr. Me- Queen has generously given us access to valuable unpublished maps. The writers are also indebted to the staffs of the Jefferson, Frank- lin, Washington and Ste. Genevieve County Agricultural Conserva- tion Associations, particularly Mr. Wayne O. Killion of Jefferson County, for their generosity in permitting study of the aerial photo- graphs. SUMMARY This paper is a description and analysis of a characteristic type of Ozark vegetation, known locally as **glades,'' which occur on out- crops of thin-bedded dolomite. A brief enumeration of the flora and a detailed map of these glades for an area in east-central Mis- souri are presented. The areal distribution of the glades is determined by geological factors and local relief. Their environmental characteristics, most important of which are winter saturation and summer desiccation, are ascribed to a complex of factors, including thin-bedding of the dolomite, thin soil mantle, steepness of slope, and attitude to the sun’s rays. The glades are regarded as an edaphic climax rather than as a stage in the succession to forest climax. The relation of the glades to similar vegetational areas elsewhere is discussed. LITERATURE CITED Brenner, Louis G. (1942). The environmental variables of the Missouri Botanical Gar- Wildflower Reservation. Ann. Mo. Bot. Gard. 29: 103—135. * Core, Earl L. (1940). The shale barren flora of West Virginia. W. Va. Acad. Sci. Proc. 14: 27-36. Cozzens, Arthur B. (1937). The natural regions of the Ozark provinee. Doct. Diss., Wash- ington Univ. , (1939). Analyzing and mapping natural landseape factors of the Ozark prov- ince. Acad. Sci. St. Louis Trans. 30: 37-63. 1942] ERICKSON, BRENNER & WRAIGHT—DOLOMITIC GLADES 101 Freeman, pcd P. (1933). Ecology of the cedar glade vegetation near Nashville, Ten- n. Aead. Sci. Jour. 8: 143-228. Harper, eur x (1926). The cedar glades of middle Tennessee. Ecology 7: 48-54. Palmer, Ernest J. та” Flora of the Grand Falls ehert barrens. Асай. Sei. St. Louis Trans. p 97-112, 21). The forest flora of the Ozark region. Jour. Arn. Arb. 2: 216-232. д s Julian A. d (1935). An annotated catalogue of the flowering plants of Missouri. Ann. Mo. Bot. Gard. 22: 375-758. Pike, R. W. (1928). The кос of a portion of the Crystal City quadrangle, Missouri. iss., Univ. of Chi Sauer, pum O. (1920). The piod of the Ozark highland of Missouri. Chieago. pp. 1-2 Steyermazk, yep А. ү Some features of the flora of the Ozark region in Missouri. 214—233. Я P "m Studies of the vegetation of Missouri. I. Natural plant associations and sueeession in the Ozarks of Missouri. Field Mus. Nat. Hist. Bot. Ser. 9: 349—475 Weller, ibd and Stuart St. Clair Seins Geology of Ste. Genevieve County, Missouri. . Geol. & Mines, II, 22: 1—352. Wherry, ES Tu (193 0). Plants P the Appalachian shale-barrens. Wash. Acad. Sci. Jour. 20: 43-52. THE ENVIRONMENTAL VARIABLES OF THE MISSOURI BOTANICAL GARDEN WILDFLOWER RESERVATION AT GRAY SUMMIT! LOUIS G. BRENNER, JR. Formerly Graduate Apprentice, Henry Shaw School of Botany of Washington University INTRODUCTION Included in the 1,296-acre tract representing the Arboretum of the Missouri Botanical Garden is a parcel of land, roughly 300 acres in extent, which has been set aside to be developed as a wildflower reservation. Here the native trees, shrubs and herbs peculiar to the various natural habitats will be brought together where the public can readily become acquainted with them. Here also rare or otherwise uncommon indigenous wild plants will be held safe from extinction by the hands of the predatory tourist. It is the purpose of this paper to present a preliminary survey of some of the more important factors conditioning the various en- vironments represented in the Wildflower Reservation of the Mis- souri Botanical Garden in order that its development may be саг- ried on in a systematic manner. It is hoped that this investigation, despite its strictly elementary nature, will indicate the more im- portant factors to be studied specifically and critically at some future date. An attempt will be made also to correlate the extant arboreal vegetation with the various environmental factors in order to arrive at bases for interpreting the distribution of the native flora. In the course of this study the geology, physiography, soils, and climatology of the area were investigated and the alteration of the flora by the activities of man was reviewed. For the purpose of this study an area was selected which em- braced the various habitats characteristic of the reservation as a whole. The area chosen is bounded on the north by a service road commonly known as the Ridge Trail, on the south by a line connect- ing the face of the cliff terminating Cliff Ridge to the east with the ledge terminating Ledge Ridge to the west. A line drawn down the 1 An investigation carried out in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University, under the direction of Prof. Lewis F. Thomas and Dr. Edgar Anderson, and submitted as a thesis in partial fulfillment of the requirements for the degree of master of science in the Henry Shaw School of Botany of Washington University. И (103) [Vor. 29 104 ANNALS OF THE MISSOURI BOTANICAL GARDEN erest of Cliff Ridge forms the eastern boundary of the area. A similar line drawn down the crest of Ledge Ridge bounds the area to the west. GEOLOGY The area lies on the northern border of the region of dolomitic limestone rocks which compose the greater part of the northern Ozark plateau. The rocks belong to the Canadian Series of the Lower Ordovician Period.? Two formations are represented, the uppermost stratum of the Jefferson City formation, and the Cotter formation. METHOD OF STUDY The rugged land-form of the area and the numerous outcrops of the bed rock made the study and accurate mapping of these outcrops comparatively easy. The cliff terminating Cliff Ridge on the east side of the area offered an excellent exposure for studying the stratigraphy. The thickness of the formations and their different phases were accurately measured with tape and plumb line. The rocks outcropping in the ledge that terminates Ledge Ridge on the west and the numerous outcrops of bed rock on either side of the valley dividing the area provided a sufficient number of stations for correlating the areal geology on a topographic map. A transect line was established up the valley, and from this line shorter lateral lines, at right angles, were made to the various points where out- crops occurred. The relative elevations of the different outcrops were obtained by means of a Brunton compass which was also used to measure the angles of the various slopes. Distances were paced off. Since the two formations represented on the area are essentially dolomitic in nature, the examination of the chemical properties of the rocks was limited to a determination of the percentage of the dolomite crystals in the various formations or in such phases of these as had an apparent bearing on the vegetation.? Early in this investigation it seemed advisable to study the phys- ical properties of the various rocks, particularly those of the glades. з Weller, Stuart, and Stuart St. Clair. Geology of Ste. Genevieve County, Missouri. Mo. Bur. Geol. and Mines, II, 22: 30—31. 1928. ' By polishing a smooth surface on the roek, then staining it with Lemberg's solution or molar copper-nitrate solution Cu(NO;),, the dolomite crystals were easily differentiated from the calcite and other calcareous minerals. 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 105 Perhaps in such features as porosity and permeability might lie the basic cause for the development of the unique glade flora. The rela- tive porosities of the various rocks were determined by gravimetric methods, using a procedure outlined by Melcher.* The permeabili- ties of the rocks were tested by comparing the depths to which a dye would penetrate in a given period of time. This procedure, though simple and providing little quantitative data, did, however, show the relative extent to which the rocks were permeable to the infiltration of ground waters. STRATIGRAPH Y* The Jefferson City Formation.—'The Jefferson City formation forms the lowest rock outcrop of the strata exposed in the area (fig. 1). The uppermost seven feet of this formation are exposed at only two stations, one in the mouth of the valley, the second at the base of the east end of the ledge terminating Ledge Ridge. The fact that this is the lowest of the rocks outeropping on the area has caused it to be covered with talus debris over the greater part of its range. The rock is a dolomitie limestone, for the most part mas- sive, oolitic, and light buff to gray-brown in color. It has a compara- tively low porosity, approximately 5 per cent, is quite permeable, and contains 30 per cent of dolomite erystals. No unconformities are evidenced between the Jefferson City formation and the over- lying Cotter formation. The Cotter Formation.—The remaining rocks outeropping on the area belong to the Cotter formation. Unlike the underlying Jefferson City formation, deeply overlain with talus debris, it presents a great areal exposure of bare rock surface. It was sus- pected that these various outcroppings, because of their differ- ent physical and chemical natures, were apt to be of direct influ- ence on the flora. The formation is composed of phases of sand- stones and dolomitic limestones. The limestones appear in both massive and thin-bedded phases, the latter being essentially ‘‘cot- ton rock.’’ The basal phase of the Cotter formation is one of gray to gray- red sandstone 37 feet in thickness. This sandstone, like the final * Melcher, A. F. Determination of pore space of oil and gas sands. Trans. Amer, Inst. Met. Eng. 65: 496—497. 1921. * The geological correlations made in this paper are based on a well log which is on file at the offices of the Missouri Geological Survey. [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN 106 потувшлоу Ао Е orjruro[op “әлтвв®ш—әпоўвәштү “т Kv1d—ouojspuvg "FT {әтұпшогор *oArssvur—ouojsourt'T ет f£wid—ouojspugg ‘GI é ‘oruo[op “алтввеш--9пофв -eur T ‘тү ‘fvis—ouoyspurg "01! fopspsuop “Фдтевъш -опофвашт 1 °6 1,,93?01 по}300),, `8 *9HTUIO[OD “Фатвевш- -опофвоштт 'L fQ,99091 п03300),, `9 :'pei—euojspugg се (отш -отор /feArssvur—ouojsourr] °F Égo X901 103400 » 76 *orruro[op ‘oayeseur—ouoqsoulyy % UOIjEULI0J 19300 tt cct Фра шво[ 408 uortuf)’ PO OP POT د‎ ee ulna коа `+ адво 318—108 “Т "(в "Зу 998) q-y Bore ou} qugnoiqj поцоов oinjonijg '[ "Sud p Š REE ASAR =. (аг эк ат ыу ат жы ас В ЗЕ э эу лы ЯВНИ ВЕ ЩЕ Жо д эш ян Ыр жу т= ик ДР ш лш в в дир жи ат ло ты ли ди яу Ж ЖЕ ЛЕ эк Эу аг НЕ ЖК BS E ر‎ Ra y V NV ж AF а к ыг AU ок ж ат NN жы RN жє чн эу 28: DU B ДР МЕ эк ИЕ Ял Ли эр шр аг эш чк э HUS эш ARDOR Фи ват. атш: e тт яс оте жс oe ar ча m за юш AR аш сме Ая МУ Ry е МУ л РЕ ЖР бшу у лы ыш ат SII лаг Щи = 89 мє ЛИ МУ ди ба зш жє Б ёт = © г/т ч зае: oak - sugar maple maple Oak-hickory association -hickory association Closed glade associes Open glade associes 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 107 phase of limestone of the Jefferson City formation which it over- lies, is almost entirely covered with a mantle of talus, though num- erous small outcrops occur on the valley slopes and in the bed of the wash draining the valley. Overlying this sandstone phase is one of massive limestone 34 feet in thickness and having an inter- ealation of sandstone about 2 feet thick near its center. Like the higher massive dolomitie phases of the formation, it has a rela- tively small proportion of pore spaces, 5 per cent, is quite highly permeable, and contains 40 to 50 per cent of dolomite crystals. Chert nodules and lenses, white, brown, to pinkish-blue in color, commonly occur. This phase outcrops in entirety on the face of the cliff at Cliff Ridge, forming a series of high step-like ledges. On the valley walls it has been smoothed off by erosive agents and is overlain by a mantle of chert and float material from strata above. A fine-grained, gray sandstone overlies this massive phase to a depth of two feet and this in turn is overlain by a phase of massive dolomite 9% feet thick and having a six inch phase of sandstone interealated four feet above its base. Three feet of thin-bedded dolomite, ‘‘cotton rock,’’ overlie this massive phase and in turn are overlain by another five feet of similarly massive dolomite and seven feet of thin-bedded dolomite. The last is typical ‘‘cotton rock,’’ pink or buff to gray in color, easily fractured, and in beds from 1% to 4 inches thick. Exhibiting the highest degree of porosity, 24 per cent, of all the rocks occurring on the area, and composed of pure dolomite, CaMg(COs)2, together with a very high degree of permeability, it is indeed a most interesting rock. The thin beds are interealated with very thin lenses of sandstone and a blue-white or pink chert. Rarely, minute intercalations of shale occur between thin beds of slightly argillaceous ‘‘cotton rock.’ The small remaining portion of the formation has, for the most part, been covered by the Union silt loam which caps the ridges. Outerops of these higher and final rocks of the formation are not abundant within the area. A few occur at the head of the valley, and some have been laid bare in the construction of the trail on the crest of Cliff Ridge. The latter are thin-bedded, coarse-grained dolomite with thin lenses and nodules of chert. A phase of putty- gray shale six inches in thickness overlies the ‘‘cotton rock’’ and is overlain by a medium-grained, loosely cemented, gray or red- brown sandstone one to three feet in thickness. A five-foot phase of massive faintly oolitic dolomite overlies the sandstone. 108 [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN TOPOGRAPHY AND RELIEF The area, essentially rugged, is in the mature stage of the ero- sional cycle, a feature typical of the hills and ridges bordering the southern face of the area to the north of the Meramec River.* Two ridges, Cliff Ridge and Ledge Ridge, truncated by an ancient meander of the Meramec River, rise 140 feet above the flood plain of that river (fig. 2). The cliff terminating Cliff Ridge, together Scale Contour interval loft. e е бр” А, 32 77 L 2 Lye T e e 4, =, дур, m, 22 2 Фе, AS "nr ТА, ure АА 422 A ү зетя) Cotter formation-thin-bedded T АШ Ice ‘dolomite. , P —— Cotter formation-massive dolomite. Ь beu TO: сег ти у Cotter formation-sandstone. ЕЁ Jefferson City formation- massive dolomite Fig. 2. Geologie map. A-B, location of geologic structure section. 1900. eforested 1880; cultivated 9-9; pas- tured 1900-1925. Fig. 3. Showing land use of the area 1880-1935. with the high ledge facing truncated Ledge Ridge, and the numerous ledges, shelves and other outcrops enhance this rugged nature. A deep valley with a youthful ‘У’? profile bisects the area (fig. 1). During the rainier seasons of the year a rivulet occupies the acute trough of the valley, carrying the waters drained from the adjacent uplands to the flood-plain where they are soon lost in the gravels underlying the silt. The ridges are, for the most part, narrow and appear as lateral spurs from the higher ridge serving locally as the * Marbut, C. F. Soil reconnaissance of the Ozark region of Missouri and Arkansas U. В. Dept. Agr., Bur. Soils and Field Operations, Rept. 13: 1727-1873. 1911. 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 109 drainage divide between Brush Creek to the north and Meramec River. The area, as a whole, offered an interesting study in angles of slope? resulting from differential weathering of the formations and their included phases. A profile of the valley, taken at points A-B (figs. 1 and 2), shows the massive limestone of the Jefferson City formation overlain by the weaker, basal sandstone phase of the Cotter formation and this in turn by the resistant massive lime- stone. Here the angle of slope is quite high, 15 to 18 degrees, and few outcrops occur since the rock is deeply overlain by soils creep- ing down from above. This steep angle of slope persists through the massive phases of the Cotter formation where it is broken by the thin-bedded phase of that formation. Here the slope is about 10 degrees. The steeper angle is resumed by the recurrence of the second massive phase of the Cotter, represented in part by a low ledge. The second, thicker of the thin-bedded phases of the Cotter formation causes a conspicuous break in the angle of slope. Here the weakly resistant, thin-bedded rocks have been rapidly beveled off, and the angle of slope is more gentle, 10 degrees. A low ledge marks the thin sandstone phase overlying the **cotton rock,” after which an even lower angle of slope, 5 degrees, is assumed and re- tained to the crest of the ridge. боп, SuRVEY The soils of the area have been classified by the United States De- partment of Agriculture, Bureau of Soils, as belonging to the sandy subsoil phase of the Union silt loam.’ The report did not consider the smaller, more specialized types of soil, particularly those classi- fied as “rough stony land.’’ Such soils lie on the steep slopes of the ridges immediately north of the Meramec River in the vicinity of Gray Summit. Since a great portion of the area under investigation is covered by such soil, it was found necessary to study its physical and chemical properties to aid in the interpretation of the flora. METHOD OF STUDY Samples, weighing about 100 grams, of the upper six inches of the soil were systematically taken at 100-foot intervals across the val- ' Cozzens, Arthur B. Analyzing and mapping natural landseape faetors of the Ozark provinee. Acad. Sci. St. Louis Trans. 30: 37—63. 1939. * Vanatta, E. S., and H. G. Lewis. Soil survey of Franklin County, Missouri, U. S. Dept. Agr., Bur. Soils and Field Operations, Rept. 13: 1603-1633. 1911. [Vor. 29 110 ANNALS OF THE MISSOURI BOTANICAL GARDEN ley and at 200-foot intervals up the valley. After drying thoroughly in the air they were examined for content of organic matter, rela- tive proportions of different sizes of soil particles, as recognized by the Bureau of Soils, and soil reaction or degree of acidity. Analysis of Organic Content.—Ten-gram portions of the samples were taken along line A-B (fig. 2). After being weighed, they were heated in a erucible for 30 minutes over a hot Bunsen burner. They were weighed again after cooling, and any loss of weight was con- sidered to be organie material removed by combustion. The results of these operations appear in TABLE I. TABLEI MECHANICAL AND ORGANIC-CONTENT ANALYSIS OF THE SOIL ALONG BELT TRANSECT B (see fig. 11) % 0 % % % % % Фо Sample | Rock & Fine Coarse | Medium | Fine | Very йпе| Silt & | Organic gravel gravel sand sand sand sand elay matter 1 0 | 0 | .3 284 | 439] 199 | 8041 | 95 2 | 783 | 285 | 227 | 317 | 437| 365 | 6458 | 108 3 15.62 7.20 3.34 | 481 | 540 | 639 | 5406 L7 4 | 59.87 1.91 116 | 208 | 341| 241 | 4331 | 170 - Glade 5 35.57 694 | 220 | 7.13 | 1016 | 439 | 1694 | 1588 .— 6 34.30 157 | 127 | 246 | 634| 5.87 | 15.77 | 314 7 39.24 | 2.16 | 3.0 | 272 | та | 5.04 | 19.22 | 199 8 | 1485 | 461 | 462 | 5.99 | 6.08 | 531 | 4385 | 143 — 9 | 0 179 | 1.61 | 458 | 458 | 5.024 | 73.71 | 101 - 10 0 93 1.58 | 3.26 | 437 | 3.25 | 79.05 | TO — T 0 о | 8 ET | 4.32 EI 84.56 | 39 | | Mechanical Analysis.—An analysis was made of samples taken along line А-В (fig. 2) to show the percentages of the different sizes of soil particles. Fifty-gram portions of the air-dried samples were weighed out. They were then dispersed in a mortar by gently rub- bing with a pestle, after which they were shaken successively through a series of screens recommended for mechanical analysis of soils. Soil Reaction (degree of acidity ).—The samples taken over the whole area were tested for soil reaction with Brom Cresol Green, 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES LT Brom Cresol Purple, and Phenol Red indicators in a method de- scribed by Moore.? The results are compiled in fig. 6. Soil Horizons.—Several test pits were dug at stations covered with a typical flora and suggesting a typical soil type. Where pos- sible the pits were dug down to bed rock, but when the soil mantle was so deep as to make this impractical the approximate depth of the soil was tested by probing into the bottom of the pit with a crow- bar. The thicknesses of the various soil horizons were measured and their eolor, texture, consistence, structure and porosity noted 5$ leaves leaves — leaves - mo А mold о _ mold — brown д cogn silt loam; numerous —bewn- block, Sandy. loom: nA like: — gray-brown silt loom; 0 pH. 6 NT crumb-like; many root 4 —gray-b fin silt loa ef of trees. pH. 5. bd P im oot ю gern silt, clay 10 crumb-like. | oam. рН. 5.0 _ Cotter dolomitic limestone 5 15 A m Vien d silty, el pan: occasio кА ed o . 2 gH. 4.0 sl. ragments; t hike, pH. 6.1 o 25 25 ЖЕ — chert frogments abundont; 9.43 gments o ~ | ¢ |" yellow-brown, motHed › Га e|- hert fragments; yellow Ма with red; li 1351, o? brown, m ottled with red; $5 |o Ы pH. 5.4 ° gr —ploty T — Cotter dolomitic limestone. nod - platy Г |- Cotter dolomihc limestone a. Union silt loam, sandy b. Union silt loam, c. Rough stony land, lower subsoil phase, Main Ridge subsoil phase, [idus та у Hae of Ledge Ridge. Fig. 4. Soil profiles. in accordance with recommendations made by Kellogg.'^ Sample profiles of the typieal soils are shown drawn to scale in fig. 4. THE SOILS As was previously stated, the soils of the area have been classified by the Bureau of Soils into two distinct groups: the sandy-subsoil phase of the Union silt loam, and the ‘‘rough stony land."' The Union Silt Loam, Sandy Subsoil Phase.— The Union silt loam, sandy subsoil phase, is a brown or grayish-brown mellow silt loam 6-8 inches deep, which changes at depths between 8 and 20 inches into a light brown or yellowish-brown friable silt loam or ° Moore, W. An improved E for the determination of the soil reaction. Brit. Golf Unions Jour. 4: 136-138. 193 ? Kellogg, Charles Е. Soil survey manual. U. S. Dept. Agr. Misc. Publ. 274. 1937. [Vor. 29 112 ANNALS OF THE MISSOURI BOTANICAL GARDEN a silty elay loam. In structure the soil is erumb-like, grading into a nut-like subsoil (fig. 4a & b). Its origin is not perfectly under- stood, but its uniform nature over broad areas seems to suggest that it is loessial. An interesting character of this soil is the high proportion of fine sand grains (TABLE т). This soil lies as a cap, about 3-3% feet deep, upon the crests of the ridges. Its loose sandy nature, together with the relatively high angles at which it lies on the slopes, has made it an easy prey to agents of erosion. In places X жэ CM OX 652525 <> б D А % $ ü > RES OK) x x 0:0:0 e Se Ф. таг. — 767059292 05955 с оо УХ А, Хо X Ох BS XIX XIX) ЖКХ < X o е Unionsilt loam, sandy subsoil phase | | pH.4.0-5:4- very acid. Г] Rough stony land. Aun отобу, Rough stony land, glade phase) а pH. 71 75 пие ива alkaline Fig. 5. Soil map. Fig. 6. Showing distribution of different acid-alkaline soil types. where the plant cover is sparse, as on the east slope of Ledge Ridge, many deep gullies have been eroded into the soil and sub- soil horizons. This soil, when treated with indicators, is found to be consistently very acid, from pH 4.0 to pH 5.4, the degree of acidity inereasing as one proceeds to the lower soil horizons (fig. 4a & b). The low pH is perhaps partly due to its sandy, porous nature which has facili- tated leaching to a moderate degree. There is also a decided ten- dency for vegetable litter to collect on the forest floor. It is in those places where the litter has collected to a considerable depth and 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 113 where it is predominantly of oaks and hickories, which is relatively slow to decay, that the lowest pH ranges are reached. The soil shows a low content of organie matter, 1.7 to 10.8 per cent (TABLE I). The majority of the trees are of tap-root nature. Fibrous roots would tend to inerease the organie content of the soil, but few such root systems are formed and they are necessarily short-lived. For the most part, few herbs are found on the forest floor on the ridges, and such plants are an important source of humus. Perhaps the poor herbaceous flora may be due to the thick layer of forest litter which prevents seeds and other propagules from reaching a suitable growth medium, particularly since the surface of this litter is dry for the greater part of the time. Also the forest floor is densely shaded during most of the growing sea- son. Animal action in this soil is not particularly great. Compara- tively few ‘‘signs,’’ castings, burrows, or actual specimens of the more common subterranean animals were found in the course of sampling. The high degree of acidity and the low organic content of the soil might account for the poor fauna. “Rough Stony Land.’’—The soil of the ‘‘rough stony land" is essentially residual and very shallow, from a mere film to 12 or 14 inches in depth. It is a very dark, brown-black to black sandy loam, plastic when wet, very brittle when dry, and essentially nut-like in structure (fig. 4c). The very shallow depth of the soil has permitted it to mature relatively early, which is indicated by the lack of defi- nite horizons and the abundance of small insoluble fragments of chert. This soil may be considered as a vertically uniform mass. It contains a relatively high per cent of fine sand grains which in- crease as one progresses towards the valley (TABLE т). It is inter- esting to note the abundance of oolite composing these sand grains, a feature offering good evidence of the residual origin of this soil, for they have weathered out of the underlying oolitic dolomite rock strata. This soil differs markedly from the Union silt loam of the ridges in being essentially a basic or lime soil. Soil tests show a pH range from very slightly acid, 6.8 at the uppermost limit, to alkaline, 7.5, in the valley (fig. 6). This soil contains a large amount of organic matter, 14.3 to 31.4 per cent, which increases steadily as one pro- gresses down the slope and reaches its maximum in the valley (TABLE 1). The high organic content of this soil, its shallow depth, and the massiveness of several of the underlying rock formations or their phases cause it to be exploited to a high degree by the roots [Vor. 29 114 ANNALS OF THE MISSOURI BOTANICAL GARDEN of trees and herbs. 'The high organie content makes this soil a valu- able reservoir of infiltrated water and as a consequence able to support a rich herbaceous and arborescent flora. Animals are quite active here, numerous burrows, castings, puppae and specimens be- ing noted. “Rough Stony Land,’’ Glade Phase.—This soil is limited to the glade, formed by the thin-bedded phase of the Cotter formation, on the west slope of Cliff Ridge (fig. 1). It is a very thin (from a mere film to 4 inches in depth) gray-brown silt loam. Soil reaction tests show it to have a pH range from 6.8 to 7.1, an essentially neutral re- action. An interesting feature of the glade soil is its high humus content, approximately 17 per cent. This can easily be understood when one remembers that the glades are of a xeric nature and there- fore unfavorable to conditions of decay. The shallow depth of the glade soil and the generally xeric conditions tend too to discourage occupancy by subterranean animals, also important agents in the reduetion of organie material. CLIMATOLOGY OF THE AREA The geomatie position of the area, between longitudes 90? and 91? West, and latitudes 38? and 39? North, places it in the temper- ate zone. Here the weather is governed by cyclonic conditions in winter. Local continental thunder-storms prevail in summer. The region has an annual rainfall of approximately 39 inches. The mean annual temperature is approximately 55? К. The last killing frost in the spring usually occurs about April 16, and the first killing frost in the autumn about October 22. Accurate records of the precipitation on the area were not kept. However, a weather station at Pacific, Missouri, seven miles away, offers data which may be applied with reasonable reliability.’ Relatively accurate temperature records were kept on the area continuously for one year, at stations selected for extremes in en- vironment. Thermometers were housed in specially designed shelters three feet above the ground. Comparisons were made of temperatures recorded on the ridge (Sta. A-1) and in the valley (Sta. А-2); оп a west-facing slope (Sta. B-1) and an east-facing slope (Sta. B-2) ; on the open glade (Sta. C-1) and closed glade (Sta. C-2) (figs. 8-10). Temperatures of two-week periods were selected as representative of each season of the year. Unfortu- " Climatological data, Missouri section. U. S. Dept. Agr., Weather Bureau. 1940—1941. 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 115 nately, with only two recording thermometers available, continuous records at any one station were not possible, and in order to obtain a recording representative of each season the instruments had to be moved to another station every third week. This fact has pre- vented the direct comparison of most of the stations at a given time. Comparison of Temperatures on Ridge (Sta. A-1) and in Valley (Sta. A-2).—In spring the ridge tends to be cooler during the day- y —— 1940 e JN ---- 1939 pus П] \ 7 4 ^ h N | \ ES. Ue a’ [ \ 6 B Y | \ I \ | \ i ` 5 / \ H \ | \ ! \ ` І \ t \ \ 1 t \ \ П] \ i 1 4 J 1 \ ` 1 ~ I ` \ 1 NS г \ ` П] м, N г \ 3 4 \ i \ N 1 \ PN i \ AS H \ И Fin 2 J \ Р ~~ еще 1001 .75 4 504 25 9 T T T T T | Г T T | | Jan. Feb Mar. Apr. May June July Aug Sept Oct Nov. Dec. Fig. 7. owing monthly precipitation in inches, at Pacific, Missouri, during 1939 and 1940. light hours and warmer at night (fig. 8B). The valley tends to be from 2 to 5° warmer between the hours of 10 a. m. and 2 p. m., and about 1-7? cooler from 6 p. m. to 6 a. m. In summer the ridge is consistently warmer than the valley (fig. 8C) ; in the night and very early morning, from 11 p. m. to about ба. m., 16 is usually only 2-4? warmer. For the remainder of the day, the ridge is about 5-8? warmer than the valley, except from 11 a. m. until 2 p. m., when it is only slightly warmer, 0 to 4°. The maximum mid-day temperature occurs at noon in the valley, due perhaps to the fact that after the sun drops several degrees past a vertieal position the valley is again shaded by the foliage of the trees. The maximum mid-day temperature occurs at about 2 p. m. [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN 116 ‘(apes peso[?) 5-0 рив (ерерд uado) т-у 54014838 подмуод SUOT}VIIBA ро so[dures вв popioo ‘(adojs Зшоврвво) g-q pue (edops Зитовр-вем) Т-Я suonezg uo94j9q SUOT}VIIVA [8008898 ро so[durvs ве popuioo ° (Кецел) 5-У рие (э9ры) [-y 81014818 u99A3eq SUOHBIIEA [8105898 ро so[dures se рәрлоә -әл ветгпувлодшер јо syde ‘oT ‘ALF -эт $o1njeioduioj jo sydein ‘6 “SLT -91 > волпуваофшор jo sydvin "8 ‘814 * e" “ а ” чет x 7 o Е ووو وود‎ з аза з дага з а 3 поропощоповодощод "ي‎ Ели LT TE ew etsy о Фор om ч o на lot he» 9 “а EN 7“ N м ha ое ошл " E » ” »- =’ w Hn ony ГЛ r7 ino: вэ Bo? opor pos pos тэп 29 وو ووی‎ ааз РУ وا و ر و بو وو او ج و کک واو وک “وو ووو ود‎ a АР" id 5 ved " [or „” [05.2 los 2 swung РА] 09 ва lo "1 ee юг e „Ке Е е дее los os ч poor T сег LE 2229 2.9% з юа" 9 no c'e э т 9 =. nstanscstusmoncsasadámnji" 135. pee اور ر وک ووو‎ ә و پر‎ EI xp er lez ot op rg е? ны |”? os oor e2 ыы - ка ет К ot 3 эм o ES os у Ру " = m Los 3-2* -—- E og ارهد‎ 2. — loi - 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 117 on the ridge, where even after the sun has fallen from its vertieal position it still penetrates through the less dense foliage. With the approach of autumn and winter, the gradual loss of leaves by the trees permits a proportionately greater penetration of the sun's rays into the woods. It was noticed then that the maxi- mum mid-day temperature in the valley was reached at noon as it was in the spring (fig. 8A and D). Similiarly, as the day tempera- tures in the valley increased, the night temperatures decreased. This marked difference in nocturnal temperatures is caused by the cooler, denser night air settling in the valley, a phenomenon known as air drainage. Day temperatures here tend to be 5-9? higher than those on the ridge, while night temperatures are 2-4? lower. Comparison of Temperatures on West-facing Slope (Sta. B-1) and East-facing Slope (Sta. B-2).—In the early spring, autumn, and winter, when the trees are bare, the west-facing slope is strik- ingly cooler in the morning and very early afternoon, from 6 a. m. to 1 p. m. (fig. 9A, B, D). At this time the west slope was from 2 to 10° cooler than the east slope. During the late afternoon, evening and night the west slope is 2-6? warmer than the east slope. The maximum temperature of the day occurs at about 4 p. m. This rather odd distribution of temperature is caused by the attitude of the western slope to the rays of the sun. Exposed as it is, it is shaded in the morning and it is not until late afternoon that the rays of the sun strike it at a high angle and raise the temperature to the max- imum point. At night this slope is somewhat warmer than the east- facing slope. This is because the sparse vegetation, the relatively large amount of exposed rock, and thin covering of mold and forest debris on the west-facing slope permit a great degree of heat radi- ation at night. During late spring and summer the west-facing slope is consis- tently warmer than the east-facing one (fig. 9B and С). In general, the difference is small, only 1-4°, this being due to the relatively greater amount of heat-consuming moisture transpired by the more dense vegetation on the east-facing slope. As in winter, spring, and autumn the maximum day temperatures occur at about 11 a. m. on the east slope and at about 4 p. m. on the west slope. Comparison of Temperatures on Open Glade (Sta. C-1) and Closed Glade (Sta. C-2).—In winter, early spring, and autumn the open glade is usually 3-8? warmer during the day than the closed glade (fig. 10A, B, D), an obvious condition when one remembers [Vor. 29 118 ANNALS OF THE MISSOURI BOTANICAL GARDEN the barren nature of the open glade. At night during these sea- sons it is from 1 to 4? cooler than the closed glade where the con- siderable abundance of junipers tends to make the air more quiet. A. denser covering of mold and litter on the forest floor prevents rapid heat radiation which takes place on the barren open glade. In summer, however, the open glade is constantly warmer than the elosed glade (fig. 10C), being from 3 to 8? warmer between 9 a. m. and 7 p. m. and usually 1-4? warmer during the night and early morning. Normally it would be expected that the barren nature of the open glade would favor rapid heat radiation and that it would be cooler than the closed glade at night. However, a canopy of broad-leaved trees cover the closed glade during the growing season, which, together with numerous herbaceous plants on the forest floor, create a more equable climate. HISTORY or THE LAND USE or THE AREA In order to interpret the relatively complex flora inhabiting the area at present it was found necessary to learn the use of the land from faets revealed by the area itself. The history was begun as near as possible from the time when the land was first occupied by the white settlers emigrating from the East up to its present use as a wildflower reservation. Use was made of data obtained from field notes taken on the area, and of documentary evidence offered by an abstract to the title of the land. Old trees were important sources of information concern- ing the aspects of the original flora. The younger trees served as indicators of changes in environment during the time that the area has been oceupied by white man. A survey and map of tree stumps revealed the age of living trees of like diameter and also provided a definite record of the former use of the land. The abstract to the land title proved valuable by supplying the dates on which the par- cels were granted to the first settlers. It also indicated the dates on which the land fell into the hands of the various families and finally the present institution, the Missouri Botanical Garden. The uses to which the various portions of the area were put, together with the time and intensity of use, were compiled in fig. 3. The area is composed of two parcels of land which have had sep- arate title histories up to the time that the Missouri Botanical Gar- den acquired them (fig. 3). A small portion of the area, on Cliff Ridge, represents part of ап eighth section granted by the United 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 119 States to Edward J. Roberts on July 13, 1853. In 1881 it became the property of James B. Miles. The larger portion of the area is a part of the quarter section granted by the United States to William M. McPherson on June 19, 1851. Later the property was transferred to Powell; from Powell to Knapp in 1854; from Knapp to Crews in 1858; and from Crews to North in 1880. On March 12, 1925, both parcels of land were included in the area bought by the Missouri Botanical Garden. An inspection of the area shows that it has been used in the fol- lowing four ways, arranged progressively according to the de- gree to which the practice has altered the vegetation: (1) grazing, (2) light cutting over and grazing, (3) deforestation and grazing, (4) deforestation, cultivation, and grazing. Grazing.—A survey of the area as a whole revealed it to have been generally and quite intensely grazed over. Though certain of the more ancient typical forest trees, oaks, hickories and maples, show decided preferences in their distribution, many younger speci- mens have a general distribution suggesting uniform conditions throughout the area. Such trees as red cedar (Juniperus virgin- iana), slippery elm (Ulmus fulva) and redbud (Cercis canadensis ) are generally distributed and have been noted to invade local for- ested areas upon the advent of grazing. Although it is evident that grazing is only one of the factors which have altered the vegetation of the area as a whole, there is a por- tion, representing slightly less than half the total expanse, on which grazing has been the only altering factor (fig. 3a). This fact is expressed by the total absence of stumps in a tract which contains an appreciable number of very old trees, obviously members of the virgin flora for their trunk diameters are comparable to stumps of trees calculated to be 175 to 260 years of age. These old trees have tall straight trunks and form a high canopy which, however, is broken in many places where individuals have died. Numerous dead specimens and standing ‘‘snags’’ are to be found, and many de- caying trunks lie on the forest floor. Judging from the habit of the old trees, the aspect of the virgin forest was, for the most part, more closed and densely shaded. The present light open aspect of the forest suggests that a change of environment has taken place in a relatively short time. Throughout this plot there is a marked pre- dominance of the group of trees referred to earlier in this account as indicative of intensive grazing. Counts of the annual growth [Vor. 29 120 ANNALS OF THE MISSOURI BOTANICAL GARDEN rings of red cedars are perhaps alone the best indicator of condi- tions accompanying grazing of local forest areas. Several counts were made on trees from different points of the plot, and 30—40 years was found to be the average age. Allowing 10-20 years for grazing to alter the habitat sufficiently to encourage invasion by the red cedar and other heliophilous trees, it becomes evident that grazing was begun here shortly after the land was acquired by Joseph North in 1880. Grazing was practiced over the whole area up to the time that it was bought by the Missouri Botanical Garden. Light Cutting-over and Grazing.—A small portion of the area, that immediately bordering the old pasture on the east slope of Ledge Ridge and extending a short distance down the slope toward the ravine, was lightly eut over during the period that the Arbore- tum was maintained by government-supervised transient labor. These recently eut stumps have supplied data concerning the ages of the forest patriarchs, and have offered some idea as to the ap- proximate ages of stumps and trees found elsewhere on the area. The removal of many old trees has produced a poor canopy of foli- age and has made the forest here very open, almost park-like. The rich mold on the forest floor has proved to be an excellent seed bed, and a dense thicket-like growth of shrubs and saplings forms a lower stratum of vegetation. In addition to the older trees of the group which have invaded under grazing conditions, many seedlings of these species originated during the period following cutting-over and the resultant increased light on the forest floor (see fig. 3b). Deforestation and Grazing—Two portions of the area, which when combined represent approximately half of the whole, have ob- viously been deforested (fig. 3, с, d). Many stumps аге to be found here, all of a uniform age and none from old trees. The cut-over portion represented on Cliff Ridge (fig. 3d) is divisible into two parts, each having been deforested at a different date. The portion representing the land formerly belonging to James Miles was ap- parently deforested about 60 years ago. Though no trees were cut to obtain counts of annual growth rings, comparison of the trunk diameters with those of other young trees of known age warrants this statement. The stump remains found here are in poorer state of preservation than those of more recently deforested portions, but the diameters of the stumps place the entire former vegetation in the class of virgin flora. When the approximate age of the trees was compared with the title to the land it was found that this por- 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 121 tion had been eut over about the time that the land was bought by James Miles. | The second, smaller portion of the deforested area on Cliff Ridge formerly belonged to Joseph North. The annual rings of several of the young trees now growing there were counted and showed the trees to be approximately 40 years old. The stumps, an undetermin- able species of oak, found in this portion are in a better state of preservation than those of the trees cut from the Miles property. The age of the trees, together with the better state of preservation of the stumps, indicates this portion to have been more recently cut over than that on the Miles property, some 10-20 years after it was acquired by Joseph North. It is evident that both portions have been grazed over quite in- tensely, for numerous islands of red cedar, slippery elm and redbud occur where the live stock succeeded in checking the rapid growth of the encroaching second growth of the forest. Deforestation, Cultivation and Grazing.—A study of the plot on the east slope of Ledge Ridge (fig. 3c) has shown it to have been deforested and afterwards probably cultivated, for no stumps have been found there. Later on, and evidently for quite a period of time, it was apparently grazed and burned at intervals. A great number of the oaks and hickories now growing on the site exist as clumps composed of two to six trunks. This feature suggests that an old pasture already having a fair growth of seedling trees had been trampled by the hoof of livestock, or burned off, causing the original stem to die and numerous basal sprouts to spring up. The annual growth rings of several of these trees were counted and showed an average of 21. It is known that the site was still main- tained as a pasture at the time of its purchase by the Missouri Bo- tanical Garden in 1925. Relatively older trees, 40-60 years old, which border the plot on either side, have low limbs spreading into the clearing, indicating development as marginal trees. This fact places the clearing of this plot shortly after its purchase by Joseph North in 1880. VEGETATION The arborescent flora is a complex one composed of 28 genera and a total of 40 species. As noted previously, the large number of spe- cies and the relatively wide distribution of many of them over the area may, perhaps, be best attributed to human occupation rather than to natural causes. [Vor. 29 122 ANNALS OF THE MISSOURI BOTANICAL GARDEN METHOD OF STUDY The complexity of the vegetation suggested that rather intensive methods should be applied to its study in order to secure a complete pieture of the distribution of all the component genera and their species. A series of five 150-foot belt transects was run across the area at 150-foot intervals in order to obtain a complete and detailed map of the trees and shrubs. Because of the wide range of genera and species represented in the flora, and because of the presence among them of many small trees and shrubs considered as being valuable indicators of various abnormal conditions, practically no attempt was made to divide the flora into size classes on the basis of actual field measurements. However, record of the relative sizes of the various consistently large specimens was kept and is dis- eussed below. Data obtained from these individual maps were ar- ranged in graphieal form. 'The survey of stumps used in the study of the land use of the area has also served in interpreting the relies of the virgin forest flora. Belt Transects.— The transects were mapped by plotting, in linear fashion, a series of 50-foot quadrats across the area. Three such lines of quadrats were plotted adjacent to one another, thus forming a belt transect 150 feet in width across the area (fig. 11). Individual quadrats were marked with flagged staffs, oriented transverse to the area with a compass. Distances were paced off. 'The number of each species represented in the quadrat was counted in turn and noted in corresponding quadrats on individual maps. Figures 13-17 show species in each belt transect, together with their respective distributions aeross the area. Species having a similar distribution have been grouped together. On the basis of the belt-transect maps and graphs showing the density of each species in the different belt transects, a composite map has been prepared showing the grouping of the species into associations and associes and their areal extent (fig. 12). The terms, association, associes, and dominant, used in discussion of the vegetation in this study, have been used in the sense originally applied by Weaver and Clements.'? DISCUSSION As stated above, the complex nature of the vegetation of the area is essentially a condition resultant of intensive grazing. Likewise, п Weaver, J. E., and Е. E. Clements. Plant ecology, pp. 43-54. New York. 1939. 1942 BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 123 mention was made of a class of trees locally indicative of pastured forest land. It seems essential at this point to review the succession of conditions leading to the present state of the vegetation of the area. Judging from the habit of the various scattered ancient forest trees, it is apparent that the aspect of the local forests before the advent of man was, for the most part, open. The trees were widely spaced. The canopy of foliage was quite well developed and gen- пишишсчалиа ПИ Be aS Bs ча AAD pn moro RT IN SERN BIY Б ИЙ} А e : И : 1 + TAHA Ara bie j CIBCE АЛЯ bl Wel ЛЛ SSS Vi A | AAA SCRE ^ (4 ГАЗИ СИ. a Я И 2 S LA TL ОаК- hickory association. White oak — sugar maple association. [7 Z Chinquapin oak — sugar maple association., Open glade associes. Closed glade associes, Fig. 11. Showing location of belt Fig. 12. Showing the tree associa- transects. tions of the area. erally excluded the growth of a rich herbaceous flora on the forest floor. The forest floor in all probability was carpeted with a thick mat of litter, and a thick layer of humus lay on the soil surface. Such forests, though they may be poor in herbaceous plant material dur- ing the summer months, do, however, contain a rich herbaceous spring flora. It is not unlikely that the early landowners took ad- vantage of this fact and turned their stock into the forests in the spring to forage on the fresh, succulent herbs. They were probably also quick to learn that by burning away the thick mat of litter on [Vor. 29 124 ANNALS OF THE MISSOURI BOTANICAL GARDEN the forest floor, greater numbers of herbs could be induced to grow and that even certain grasses entered burned-off areas. Fire as an agent soon destroyed the layer of humus protecting the forest floor. What little remained after the ravages of fire was stirred up by the cloven hoofs of the cattle, and, becoming an easy prey to the erosive power of heavy spring and summer showers, it was washed from the ridges and the slopes into the ravines and thence down to the river. Lacking the blanket of humus and litter the forest soils rapidly lost their moisture during the common late summer droughts. Competition for moisture among the trees became keen, many less sturdy individuals succumbing in order to balance this new deficit in water supply. The death of individuals left relatively large open ‘‘islands’’ in the forest canopy through which the rays of the sun could readily penetrate and thus intensify the xeric con- ditions already initiated. Continued grazing in the forest brought about even more erosion. More trees died, being unable to compete with their more sturdy kin for the necessary water with their roots bare. A vegetational adjustment became imminent to prevent the com- plete denudation of the land. At this stage there was a marked in- flux of heliophilous plants formerly excluded from the forest. These plants, predominately red cedar (Juniperus virginiana), slippery elm (Ulmus fulva), and redbud (Cercis canadensis), per- sisted. They were able to endure the late summer droughts and rigorous winters when the soil cover was scant. When the area be- came the property of the Missouri Botanical Garden grazing of forested areas was discontinued. The soil cover was no longer dis- turbed, and forest fires no longer reduced the protective layer of leaves and litter deposited by the trees each autumn. Ina short time the heliophilous population matured, forming a lesser canopy of its own. Since the forest floor was shaded, a new carpet of humus and litter developed. The soil was once more cool and moist, per- mitting the origin and establishment of a new generation of the vir- gin flora. With this condition well in mind the complexity of the vegetation, previously referred to, becomes obvious when coupled with the wide range of habitats created by the geology, physiog- raphy and soils of the area. An attempt was made to sort the various species into associations and associes on the basis of data secured from the graphs of the belt transects. However, the group of three heliophilous plants, previously referred to, appeared again as one to be dealt with be- 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 125 fore the true picture of the vegetational distribution could be com- posed. The density of the individuals comprising this group tends to mask any suggestions of association between the remaining spe- cies. The grazing conditions under which they have invaded the area and have become established have been discussed in consider- able detail in a previous paragraph. In addition to these three spe- cies there may also be added: honey locust (Gleditsia triacanthus), wild plum (Prunus americana), black haw (Viburnum rufidulum), and prickly ash (Zanthoxylum americanum). Three members of the heliophilous group tend to confine them- selves to characteristic habitats, though, for the most part, they have a general distribution. Red cedar exhibits a decided prefer- ence for alkaline soils, and the oldest and healthiest specimens are to be found on the west slope of Cliff Ridge where the thin soil cover causes it to be influenced by the underlying dolomitie lime- stone rock strata. The black haw is essentially a small tree, loving open, rocky situations. The belt-transect graphs indicate it as be- ing partieularly distributed on the west slope of Cliff Ridge, where it has been shaded out by the more rapid growth of such large heliophilous trees as red cedar and slippery elm. Its present dis- tribution is due to the open, xerie nature of the west slope of Cliff Ridge, maintained naturally by the southwestern exposure to the more or less drying, prevailing winds. It is probable that this area served as a center of distribution of the black haw to the east slope of Ledge Ridge and likewise to the pasture areas on either ridge. Priekly ash tends to prefer the more equable climate offered by habitats on the east slope of Ledge Ridge for it is here that fine mature specimens have developed. Like other members of this group, however, it will endure a wide variation in habitat and is rather generally distributed over the area. With the above group of heliophilous plants withdrawn from consideration, the task of grouping the species into associations be- comes comparatively easy. Inspection of the belt-transect graphs shows that it is possible to arrange the various genera and species of trees into three principal groups, two of which contain sub- ordinate phases. A group, predominately oaks and hickories, shows decided preference for the ridges where it forms an oak-hickory (Quercus-Carya) association. Two developmentary phases are represented within this group. A second group dominated by white oaks and sugar maples, forming a white oak-sugar maple (Quercus alba-Acer saccharum) association, shows preference for the lower (Vou. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN 11.5-16.4 plants; Showing distribution and density of species ДШ | . a Ы " E = g а + 2 S. ' эў 43 37 44 23 1z 536 3 3 $3.95 Заб - ки. tí535 - 5. ЗЕЕ 3 48 5 < فع چ‎ Ё Se 2* t$? o$32 SH 55 обо. 5 S322 Е hd 29 822 ^ 3 $2 + ETE is "bi 33 55252 $3. Ее сеза + ti Silii: 58 Еее н 55 92999 > Se ЕЕ A © бехоцш uo 259935040 22.5—29.4 plants. Showing distribution and density of species in belt transect A (see fig. 11). Unit 1 = 0-1.4 plants; 2 = 1.5-2.4 plants; 3 = 2.5-4.4 plants; 4 = 4.5-7.4 plants; 5 = 7.5-11.4 plants; 6 16.5-22.4 plants; 8 | [ E m Ё BSE s. m: 35“ $13.95 $28 | int TURAN 2255 mi a23t-9 02999 ы ` SEPS GHIBHB 85.38 Jérsó3350é 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES ругоз virginiana assafras officinale 5 Dios A Diospyros virginiana Sassafras of ficinale EJ Bumelio lanuginosa ; | 19 £0 2 1$ 16 17 18 19 £0 М 1213 14 ig ie 17 iB 7 8 9 On Ann IZ т е one Е 45 2 QW o m “a + E © © тесе 3 = $335 X св +. $133 се - е? ИЧ RI та ЗЕ 8 на Е 5.2 5 а ' ane озо НЕ Ро ЕЕ 23 ЕЕ 7.9 = 58 #532 2= 3 عت‎ v س‎ о > "Orr са оа Showing distribution Fig. 17. and density of species in belt tran- Showing distribution and den- sect E. Fig. 16. sity of species in belt transect D. and density of Showing distribution Species in belt transect C. Fig. 15. 127 [Vor. 29 128 ANNALS OF THE MISSOURI BOTANICAL GARDEN slopes of Ledge Ridge. A third association is distinet in preferring the barren, rocky exposure of the west slope of Cliff Ridge. It is dominated by a growth of chinquapin oak and sugar maple. This association has a subordinate phase, a group of trees dominated by red cedar which prefers the more xerophytie environment of the glade. The Oak-Hiekory (Quercus-Carya) Association.—Grouped into this association are: post oak (Quercus stellata), Ozark hickory (Carya Buckleyi var. arkansana), black oak (Quercus velutina), black jack oak (Quercus marilandica), and the shingle oak (Quercus imbricaria) (figs. 13-17). The status of these trees within the as- sociation is not one of equal importance. A study of the belt-tran- sect graphs indicates that the post oak and Ozark hickory are dom- inant. As stated above, these graphs are purely a study of density and distribution, and give no indications of size and age. On Ledge Ridge and that portion of the ridge at the head of the valley, large trees are consistently white oak, black oak, and Ozark hickory. Judging from the ages of stumps of similar diameter else- where in the area their age varies between 150 and 200 years. This places them in that group comprising the relies of the virgin flora in which they may serve as indicators of the former climax. On Cliff Ridge the conditions are not so easily read. The total deforestation of this portion (fig. 3c) has left no relie trees of the virgin forest. Although numerous stumps of trees, approximately 150-200 years old, are to be found, they were in a rather poor state of preservation and all that could be identified were oaks. The rela- tively great intervals at which the stumps oceur seem to suggest that the trees were of stout, spreading habit such as that commonly developed by ridge-top post oaks. The probability of the former climax being composed essentially of post oak is supported by the faet that this ridge is for the most part very narrow. It receives little protection from hot dry winds by the surrounding ridges, and its high isolated position does not permit water to linger long either on or in the soil. Temperature recordings show that in summer this ridge is consistently warmer than the valley or lower habitats, and that during midday and early afternoon it is usually 5-8? warmer than a comparable station in the valley (fig. 8C). The present vege- tation is predominantly of post oaks approximately 30 years old. Black oaks, white oaks, and Ozark hickories occur in minor num- bers (figs. 13-17). Seedling trees are absent for the most part. 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 129 These facts seem to indicate the revival of a former post oak climax on this ridge. Although post oaks also occur in great numbers on Ledge Ridge (figs. 13-17), their distribution is different from that observed on Cliff Ridge. On Ledge Creek it is almost totally limited to the boundaries of the old deforested and pastured plot (figs. 3, 12), and exists as an associes, a stage leading to the development of the oak-hickory climax association. The belt-transect graphs show a relatively high per cent of black oak, white oak, and Ozark hickory also composing this old pasture flora (figs. 15-16). These are, for the most part, young trees 5-10 years old, while the older post oaks are 21 years old, a fact which indicates that the oak-hick- ory climax common to Ozark ridges is rapidly becoming dominant. An even earlier stage of the development in the succession from cleared ridge lands to an oak-hickory climax association is repre- sented by the pioneer plants, now almost completely dominated by the post oak associes (fig. 12). Included in this ‘‘old field’’ associes are persimmon (Diospyros virgimana), sassafras (Sassafras of- ficinale), and shinning sumac (Rhus copallina). With the exception of the persimmon, which has a more general distribution, these oc- cur as small ‘‘islands’’ only partially invaded by the more advanced stage of the post oak associes. A comparison of the composite map of the vegetation (fig. 12) with the map of the areal distribution of the various degrees of soil acidity (fig. 6) shows a very close corre- lation between the environment of these ridge-top trees and areas of high degree of soil acidity. The white and black oaks of this as- sociation have an indifferent attitude toward degree of soil acidity and appear quite generally over the entire area. The remaining members have distributional boundaries almost identical with the areas of acid soils. These correlations support Steyermark’s con- clusions concerning the vegetation of the Ozarks in Missouri.'* No correlation could be found between this association and the soil types (fig. 5), nor could it be identified with any particular outerop- ping of the various rock strata (fig. 2) which at this point are deeply overlain with soil (fig. 4, a, b). The White Oak-Sugar Maple (Quercus alba—Acer saccharum) Association.—The belt-transect graphs reveal a second prominent group of trees which conspicuously prefer the exposure of the east 13 Steyermark, J. A. Studies of the vegetation of Missouri—-I. Bot. Series, Field Mus. Nat. Hist. 9: 351—475. 1940. M Ibid., p. 405. [Vor. 29 130 ANNALS OF THE MISSOURI BOTANICAL GARDEN slope of Ledge Ridge (fig. 12). Unlike the neighboring ridge, this association 1s represented by a large group of species: white oak (Quercus alba), sugar maple (Acer saccharum), white ash ( Frazi- nus americana), black oak (Quercus velutina), shagbark hickory (Carya ovata), hackberry (Celtis occidentalis), burning bush ( Evonymus atropurpurous), black walnut (Juglans nigra), shad- bush (Amelanchier canadensis), and bladder-nut (Staphylea tri- folia) (figs. 13-16). As indicated in the discussion of the preceding association, it is not possible to assign dominance to any several spe- cies solely on the basis of facts represented in the belt-transect graphs. This association is fortunate in having a considerable num- ber of virgin trees. A comparison of their trunk diameters with those of stumps of readable age elsewhere in the area places all of these trees between 160 and 200 years old, in that age class which is conspicuously composed of virgin trees. They are all straight- boled and form a relatively high canopy broken where many of them died out when the environment was altered by grazing. Of the large and older trees, white oaks and sugar maples are dominant. Aged, living sugar maples are not common in the area, probably because they were unable to endure the conditions brought about by graz- ing, namely, the depletion of the thick layer of litter and humus on the forest floor. The association has a well-developed understory of bladder-nut (Staphylea trifolia), spice bush (Benzoin aestivale), prickly ash (Zanthoxylum americanum) and burning bush ( Evon- ymus atropurpureus). Seedlings and saplings of white oaks and sugar maples abound. An attempt to correlate this association with the various environ- mental factors reveals that it corresponds most closely with the limits of alkaline soils of the lower portion of the east slope of Cliff Ridge (fig. 6). This association prefers a pH range from 6.8 to 8.5. Here the degree of acidity of the soil is affected to a greater extent by the underlying dolomitic limestone because the soil layer is con- siderably thinner, varying from 8 to 14 inches in depth. The associ- ation may be further correlated with the limits of the ‘‘rough stony land" on this slope (fig. 5). The eastern exposure of the slope on which this association has developed has caused it to assume a mesophytie aspect. Being exposed to the east as it is, the slope re- ceives the direct rays of the sun only in the morning before the air has been heated up. Thus the sun has a less burning effect upon the foliage of the trees during the summer drought periods. The slope is shaded from the rays of the sun during the heat of the day, from 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 191 2to4p.m. The temperatures recorded on the slope are consistently from 1 to 4? cooler than those of a comparable position on the west- ern slope (fig. 9C). The greater equability of this environment as compared with the distinctly xeric habitat on the west slope is man- ifest in a more succulent herbaceous flora. Humus and litter have collected in a thicker carpet on the forest floor, thus protecting it from excessive erosion and likewise serving as a moisture-conserv- ing mulch. The Chinquapin Oak-Sugar Maple (Quercus Muhlenbergu—Acer saccharum) Association.—As previously stated, this association inhabits the western slope of Cliff Ridge (fig. 12). The dominant trees are chinquapin oak (Quercus Muhlenbergii) and sugar maple (Acer saccharum). Other important members of the association are: red cedar (Juniperus virginiana), chittim-wood (Витейа lanuginosa), black oak (Quercus velutina), and blue ash (Fraxinus quadrangulata). Trees playing a minor part in the association are white ash (Fraxinus americana) and black walnut (Juglans nigra). Trees and shrubs forming the rich understory are: dwarf hackberry (Celtis pumila), redbud (Cercis canadensis), Indian cherry (Rhamnus caroliniana), black haw (Viburnum rufidulum ), buckthorn (Rhamnus lanceolata), flowering dogwood (Cornus flor- ida) (figs. 13-16). The association has alight, open aspect, no canopy in a true sense having been developed. Relatively few of the dominant trees are actually dead though many exist in a poor state of health. A com- parison of the trunk diameters with stumps of known age places the majority of the older specimens in the group of virgin trees. The habit of these trees seems to suggest that the aspect of this as- sociation has not been appreciably altered by grazing conditions. The fact that these trees have a lower and more spreading branch- ing habit than that of the typical forest specimens in the associa- tions previously described indicates that they developed under con- ditions very similar to the semi-xerophytic ones existing today, and may perhaps account for the slight damage suffered from grazing. It is probable that even prior to the advent of human activities in the area, the soil cover lay thin on the bed rock beneath, and that though a rather rich herbaceous flora had developed it was neces- sarily xerophytic in nature. These facts are in accord when one remembers that the western exposure of this habitat leaves it rela- tively unprotected from hot, dry south and southwestern summer (Vou, 29 132 ANNALS OF THE MISSOURI BOTANICAL GARDEN winds, and likewise raw, cold winter winds. The attitude of the slope causes it to have a high angle of incidence to the rays of the sun during the heat of the day, that is, from 2 p. m. to 4 p. m. This results in a temperature of 1-4? over that recorded at a comparable station on the east slope of Ledge Ridge (fig. 9). This xeric condi- tion is intensified when the numerous thin-bedded phases and sand lenses incorporated in the rock strata are taken in consideration (fig. 1). These serve as drains through which percolating ground waters are rapidly conducted off, leaving little water in reserve for periods of drought despite the high organie eontent of the soil. As previously mentioned, the shallow depth of the soil has caused it to be exploited to a high degree by roots of trees and herbs, which necessarily means that it would be rapidly depleted of moisture re- gardless of the high content of organic matter. The relatively thin depth of the soil layer, together with its residual origin from the underlying dolomitie limestones, has favored the development of a typieally alkaline soil (fig. 6) which is reflected markedly in the flora. Steyermark includes the dominant plants of this association in his list of plant indicators of alkaline soils.!* The Glade Associes.—The glade is a plant environment distinct from all other habitats presented by the area (fig. 1). The soil is very shallow, and numerous rock shelves are exposed, which, to- gether with the conspicuous lack of trees over the greater part of its extent, give the glade a barren open aspect. Further inspection shows the position of the glade to be correlated with the extent of the outcropping of the upper, thin-bedded phase of the Cotter forma- tion on this western slope. This suggests that the origin of the glade lies within the properties of this rock. Ina previous paragraph this thin-bedded phase of the Cotter formation was characterized as being a highly porous and permeable rock. These features and the comparatively great depth, 7 feet, of this phase cause it to present a plant habitat with a wide range of seasonal conditions. The high porosity and permeability of the rock permit free movement of water. Early spring rains rapidly percolate through the rock or water may be drawn from it by rapid evaporation, having been brought to the surface by capillary action. This means that the rock ereates an arid substratum during summer. In autumn and late winter another extreme is experienced. With the increase in precipitation and accompanying decrease in evaporative powers of * loc. cit., p. 407. 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL VARIABLES 189 the atmosphere, ће rock becomes literally saturated with percolat- ing ground water. The sparse plant cover does not bind the soil in place against spring and later summer rains (fig. 7). The surplus of water in the bedrock beneath the soil in winter forces its way to the surface of the rock, where, acting as a lubricant and erosive agent, it also carries the soil away. 'The western exposure of this unique environment only tends to intensify the already extreme edaphic conditions. In summer the hot, dry, southern and south- western winds, and the high angle of incidence of this slope to the rays of the sun during the heat of the day, create intensely desiccat- ing conditions. The Open Glade Associes.—Almost the entire area of the glade is open (fig. 12), appearing like a barren, rocky, upland meadow. These conditions have resulted in almost total exclusion of an ar- borescent flora. Over the greater part of its extent an herbaceous associes prevails, Missouri black-eyed Susan-bluestem (Rud- beckia missouriensis-Andropogon scoparius). Trees are excluded from the open glade, the arid summer conditions preventing the establishment of seedlings. However, several small ‘‘islands’’ of red cedars are trying to exist here. The Closed Glade Associes.—The small closed glade is founded on essentially the same environmental conditions as those control- ling the open glade. However, it differs from the open glade in be- ing forested with a red cedar (Juniperus virginiana) associes. Other important trees are: chinquapin oak (Quercus Muhlenbergu) and chittim-wood ( Bumelia lanuginosa). A well-developed under- story is composed of Indian cherry (Rhamnus carolimiana), buck- thorn (Rhamnus lanceolata), dwarf hackberry (Celtis pumila), black haw (Viburnum rufidulum), redbud (Cercis canadensis), and flowering dogwood (Cornus florida). This associes is distributed as a narrow band bordering the upper edge of the open glade as- socies (figs. 1, 12), and may be correlated with the uppermost por- tion of the thin-bedded phase of the Cotter formation and the basal sandy phase of the Powell formation (figs. 2, 12). Here the sub- stratum of rock remains more moist since it lies adjacent to the mas- sive rocks of the Powell formation which retains percolating ground waters considerably longer than the thin-bedded cotton rock of the Cotter formation. Humus is washed down upon the glade from the oak woods above it and is an important factor in increasing the soil moisture since it serves as both a reservoir of [Vor. 29 134 ANNALS OF THE MISSOURI BOTANICAL GARDEN moisture and as a mulch. This increase in soil moisture has per- mitted the invasion of trees upon the glade edge. The extremely xeric conditions prevailing during the summer months, however, limits the habitat to oceupation by drought-resistant species such as red cedar, chittim-wood, and chinquapin oak. Though a canopy of interlacing cedar boughs is well developed it is characteristically of a gauze-like texture, and permits abundant light to encourage the rich understory of small trees and shrubs. These glade associes have, in all probability, been the least in- jured by grazing. The xerophytic nature of the open glade neces- sarily produces plants of a tough, leathery texture which offers little encouragement to cropping by animals. A thick mat of slowly decaying cedar needles prevents seeds of herbs from reaching a suitable stratum for germination. Thus it is evident that this as- socies was undisturbed by livestock. A portion of the associes was deforested (fig. 3c), but this caused little alteration of the flora aside from the removal of the virgin trees. Steyermark has shown how similar glades and rocky barrens in the Ozarks are invaded and dominated in part by red cedar and chittim-wood among other trees. Dead or dying specimens are almost totally absent. It is evident that this associes exists today much as it must have before settlement of the region by white man. BIBLIOGRAPHY Adams, С. L. (1901). Physiography and geology of the Ozark region. U. 8. Geol. Surv. Ann. Rept. 22: 69-94. Bates, C. G., and R. Zon (1923). Research methods in the study of forest environment. U. 8. D "di Agr. Bull. 1059. Cain, Stanley A. (1939). The nature and complexities of the elimax. Amer. Midl. Nat. 21: 146-181. Kellogg, Charles E. (1937). Soil survey manual. U. S. Dept. Agr. Mise. Publ. 274. Krumbein, W. C., and F. C. Pettijohn (1938). Manual of sedimentary petrography. New York. Marbut, C. F. (1911). Soil reeonnaissance of the Ozark region of Missouri and Arkansas. U. S. Dept. Agr. Bur. Soils & Field Operations Rept. 13: 1727-1873. Moore, W. (1935). An improved method for the determination of the soil reaction. British Unions Jour. 4: 136-138. Palmer, E. J. (1921). The forest flora of the Ozark region. Jour. Arnold Arb. 2: 229- 232. Raup, H. M. (1938). Botanical studies in the Blaek Rock Forest. Black Rock Forest 1.7, (1941). The history of land use in the Harvard Forest. Harvard Forest Bull. 20. 1 loc. cit., р. 372-377. 1942] BRENNER—GRAY SUMMIT ENVIRONMENTAL, VARIABLES 135 EC. John (1940). Distinction between calcite and dolomite on polished surfaces. mer. Jour. Sci. 11: 791-795. ends fult A. (1931). Notes on Missouri plants found on certain geologieal for- mations. Mo. Bot. Gard. Bull. 19: 52-56. - (1933). An annotated catalogue of the flowering plants of Missouri. Ann. Mo. Bot. Gard. 22: 375-788. (1934). Some features of the flora of the Ozark region in Missouri. Rhodora (1937). Natural plant suecessions in the Clark National Forest. U. S. Dept. Agr. Forest Serv. Grazing Report, Sections C & D. (1940). Spring flora of Missouri. Mo. Bot. Gard. and Field Mus. Nat. Hist. (1940a). Studies e the vegetation of Missouri. I. Natural plant associations d suceession in the Ozarks of Missouri. Field Mus. Nat. Hist. Bot. Ser. 9: 349—475. Tickell F. C. (1931). The examination of fragmental rocks, Stanford University, Calif. rev. ed. 1939. Twenhofel, W. W. (1932). Treatise on sedimentation. 2nd ed. Baltimor Vanatta, E. S., and Н. С. Lewis (1911). Soil survey of Franklin iE Milone: ШЕМ; Dept. Agr. Field Operations Bur. of Soils Rept. 13: 1603. Weller, E wt, and Stuart St. Clair (1928). Geology of Ste. Genevieve County, Missouri. Mo. Bur. Geol. and Mines, II, 22: Wherry, Я т, (1927). E E. acidity and alkalinity by indieators in the field. Wash. Aead. Sei. Jour. 10: HELMINTHOSPORIUM SPOT OF CITRONELLA AND LEMON GRASS IN GUATEMALA CARROLL WILLIAM DODGE U. S. Exchange Professor, Faculdad de Ciencias aturales y Farmacia de la Universidad Nacional de Guatemala Mycologist to the Missouri Di. Garden Professor in the Henry Shaw School of Botany of Washington Universita? During August, 1941, a severe epiphytotie developed on citronella (Cymbopogon Nardus (L.) Rendle subspec. genuinus) and lemon grass (Cymbopogon citratus (DC.) Stapf) at Los Cerritos, a very large plantation near the upper edge of the coastal plain below Escuintla in Guatemala. The weather had been somewhat ab- normal, with earlier light rains and higher humidities at the end of the preceding dry season, although the rainfall for the whole sub- sequent rainy season was not conspicuously high. Owing to the inereased demand for essential oils, the plantation had been irri- gated throughout the dry season instead of cutting off the water for the last two or three months as in previous years. This kept the trash and marcescent leaves at the bases of the plants moist for the produetion of a larger number of spores while the higher hu- midities probably aided in securing more abundant germination and infection. By the end of August all but the very young leaves had died back about half their length, and the distilling plant was closed for about six weeks for the first time since the plantation came into production about ten years before. An examination of a large number of plants showed that the principal damage resulted from a leaf spot, although an occasional plant of citronella showed a rot associated with Fusarium sp. at the bases of the stems a short distance above the soil. The leaf spot begins as a small yellowish area between the veins. It develops more rapidly between the veins than across the smaller ones, resulting in an elliptie to nearly linear area of necrosis with a reddish margin, quite similar in appearance to the Helminthosporium leaf spot of sugar cane. The central portion soon becomes brown and dry, but does not drop out. As the spots increase in number, evidently the water supply is reduced and the distal portion of the leaf slowly ! On leave of absence, 1941-42. Issued April 20, 1942 (137) [Vor. 29 138 ANNALS OF THE MISSOURI BOTANICAL GARDEN wilts and dies. Transverse sections of the leaf through a leaf spot apparently show eross-sections of thick-walled, brownish hyphae in the phloem of the bundle next the leaf spot, with the whole phloem area brown and necrotic. The dead tissue may then be invaded by saprophytes such as Hormodendrum. The Helminthosporium was easily isolated on Thaxter's potato- glucose agar and grows equally well on Sabouraud's glucose agar, although on both media the spores are somewhat smaller (24-35 x 8 15.5 и, average 29.5 x 12.4 и) than when developed on the host (46—54 x 18-24 и, average 49 x 20 и). The strain on lemon grass has somewhat smaller spores (19.5-28.6 x 7-15 и, average 25.8 x 10 u) than those of the citronella strain, but further study will be neces- sary to determine whether the difference is significant as no morpho- logical differences have been noticed. On the host, the conidia develop at the tip of a short, stiff conidio- phore. As the first spore develops, the supporting cell proliferates dichotomously and produces a second spore. This may be repeated, giving about three spores in a compact group at the tip of the conidiophore as in Acrothecium (pl. 13, fig. 8). The mature spore is asymmetric, flattened on one side or somewhat curved, with one cell (usually the subterminal) much larger than the others. In agar colonies, conidia are borne singly, both terminally and laterally (pl. 13, figs. 5, 10, 11), as well as ina terminal group on short lateral branches corresponding to the conidiophores on the host (pl. 13, figs. 6, 11). In the original cultures on potato-glucose agar, the large subterminal cell of the conidium often proliferated laterally, producing a mature spore suggestive of the staurospore found in Triposporium Corda (pl. 13, figs. 4, 6, 11). This type of spore has not been seen in subcultures on Sabouraud’s glucose agar, prepared from the original cultures after six months. The systematic position of this organism is not altogether clear from the limited literature at my disposal. The staurospores are considered as abnormal since I have found them only in cultures on potato-glucose agar, and may result from some morphogenetic stimulus in the medium. The grouping of spores at the tip of the conidiophore suggests Acrothecium Preuss but is believed to result from a different ontogeny, similar to that of Helminthosporium Sacchari (Breda da Haan) Butler, if the internodes between spores were greatly shortened and the spores were less eaducous. Brachy- sporium Saccardo and Napicladium Thuemen resemble our or- ganism in many respects, but their distinction from Helmintho- 1942] DODGE—HELMI ORIUM CYMBOPOGI 139 sporium Link is not clear. Hence in the absence of further information regarding this group of genera, I have preferred to describe the species in the oldest genus. HELMI овом Cymbopogi Dodge, sp. nov. Conidiophori rigidi, erecti, fumosi vel obscure brunnei, septati neque ad septa constricti neque ramificati; conidia 3-5-вер аа, terminalia, singula vel in capitulis parvis, asymmetrica, uno latere applanata aut leviter curvata, 46-54 х 18-24 y. Conidiophores rigid, erect, smoky or dark brown, septate, not constricted at the septa nor branched; conidia with 3-5 thick, trans- verse septa, terminal, single or in small groups, asymmetric, flat- tened on one side or slightly curved, thick-walled, usually with the penultimate cell much larger than the others, 46-54 х 18-24 и. Guatemala: Escuintla, Los Cerritos, Dodge, on citronella, TYPE ; same locality and collector, on lemon grass. The spore germinates from the basal cell by a tube which pushes between the epidermal cells until it is below the thicker portion of the wall of the host cell, then penetrates the epidermal cell (pl. 13, fig. 7). If the tube fails to penetrate, it forms a dichotomously branched mycelium (pl. 13, figs. 1, 9). Preliminary experiments indicate that this disease may be con- trolled by spraying with either Bordeaux mixture or lime-sulfur. A promising beginning has been made in the selection of resistant clones. Two clones with a relatively small amount of infection in a very heavily infected area were selected, divided into the usual seed bits and planted in a freshly prepared section of the plantation. As they mature for the first cutting, about half of them are rela- tively free from infection. The other half and the surrounding area are again heavily infected, as this area has not yet been sprayed owing to lack of adequate equipment for spraying such a large plantation. It is hoped to carry this selection further, since in normal times the cost of spraying is so great that it will be impos- sible for Guatemalan farmers to compete with those in other regions with much lower labor costs. [Vor. 29, 1942] 140 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 13 Fig. 1. Germinating spore of Helminthosporium Cymbopogi Dodge, strain on cit- ronella. Fig. 2. Terminal group of conidia, strain on lemon grass, from cultures. ig. unger stage of the above. The older spore on the right has already боа its walls, while the younger one on the left has just produced transverse septa. Fig. 4. Conidia of the strain on lemon grass from potato-glucose agar eultur Fig. 5. Single terminal eonidium of the strain on lemon grass from potato- ой agar eulture. Fig. 6. Group of conidia, strain on lemon grass from je iet agar culture. This hypha is more moniliform than the usual hyphae in eultur Fig. 7. Germinating conidium, showing method of erer, of the citronella leaf epidermis. ig. 8. Conidiophore and group of conidia on leaf of eitronella. Fig. 9. Germinating conidium of lemon grass strain from culture. Fig. 10. Very young terminal group of conidia of lemon grass strain; only the first spore has thickened its walls, the second spore has divided only once, while the two youngest spores have not begun to divide. A young spore is forming laterally below. Fig. 11. Various types and groupings of conidia of the citronella thain from potato- glucose agar culture along a single hypha. ANN. Mo. Bor. GARD., Vor. 29, 1942 PLATE 13 DODGE—HELMINTHOSPORIUM CYMBOPOGI Annals of the Missouri Botanical Garden Vol. 29 SEPTEMBER, 1942 No. 3 COMMENTARY ON THE NORTH AMERICAN GENERA OF COMMELINACEAE ROBERT E. WOODSON, JR. Assistant Curator of the Herbarium, Missouri Botanical Garden Associate Professor in the Henry Shaw School of Botany of Washington University During the course of a revision of the species of Tradescantia indigenous to the United States, published several years ago," I found it advisable to make some correlative observations upon the genus as a whole and the American representation of the family generally. These rather casual excursions afield were both com- forting and disquieting, for they showed that, although the Trades- cantias of the United States are relatively homogeneous phylo- genetically, those of the tropics are extremely heterogeneous, and also that the systematics of the family, at least in North America, is very precarious indeed. Such being the case, I fixed my atten- tion upon the limited job in hand with a profound sense of thanks to Providence for my lot at that time, and a nebulous vow of propit- iation by a revision of the tropical Spiderworts in the indefinite fu- ture. The ‘‘future’’ has arrived rather unexpectedly at last, for I find my vow exacted by the needs of the *Flora of Panama? upon which I have been working for some years past. The Commelinaceae always have been difficult subjects for her- barium study because of their deliquescent flowers. It is not easy to understand, therefore, why previous systematists of the family have focused almost their whole attention upon floral structure in the delimitation of subfamilies, tribes, and genera. In his aecount of the family for de Candolle's *Monographiae', C. B. Clarke? 1 Anderson, E., and В. E. Woodson, Jr. Contr. Arnold Arb. 9: 1-132. 1935. * C. B. Clarke, in A. & C. DC. Monogr. 3: 115-324. 1881. Issued September 18, 1942. (141) [Vor. 29 142 ANNALS OF THE MISSOURI BOTANICAL GARDEN erected three tribes, Pollieae, with fruit indehiscent, Commelineae, with dehiscent fruit, fertile stamens 3-2, sterile stamens 0-4, an Tradeseantieae, with dehiscent fruit, fertile stamens 6-5. In es- tablishing these groups, Clarke found it prudent, in the case of the last two tribes, to call attention to exceptions amongst a number of genera obviously included arbitrarily within either tribe in spite of divergence from the characters of diagnosis. The modern reader probably will find it difficult to understand why Phaeosphaerion ( Athyrocarpus) was placed within the Pollieae, whilst Commelina remained with several discrepant genera in the Commelineae, as well as why Callisia was placed within the Tradescantieae, upon the characters provided, since the greater number of its species (as interpreted by Clarke himself) have only 1 to 3 stamens. Numerous other instances of ambiguity and inconsistency could be mentioned. The most recent general system of Commelinaceae is that by Brückner? for Engler’s ‘Natiirlichen Pflanzenfamilien’, in which two subfamilies are provided: Tradescantieae, with aetinomorphie flowers, and Commelineae, with flowers zygomorphic. With such distinctive terms employed in the general key, the reader is prone to remember the regular flowers of Tradescantia and the strongly irregular flowers of Commelina (as he is apt to be acquainted with them), and to proceed on his way until tripped by the deliquescent flowers from the herbarium before him or stopped dead in his tracks by a reference to the text of the generic descriptions of Com- melineae. These usually side-step the issue of zygomorphy entire- ly ог admit inconsistency (as for Aneilema, p.175: **Pet. frei, das üussere mitunter kleiner’’ [italics mine]). Reference to herbarium specimens and to standard icones shows that in most genera the zygomorphy of the corolla is either absent or so slight that it is highly impractical. Brückner further divides the Tradescantieae into two tribes, Hexandrae and Triandrae, upon the basis of **6 fer- tile Stam.’’ and ‘‘3 fertile Stam., 3 oder O sterile’’, respectively. The latter, of course, is highly embarrassing to such a genus as Callisia, the fertile stamens of which vary from 1 to 6, without ac- companying staminodes. Generic characters used by both Clarke and Briickner, as well as by less prominent authors, have tended to accentuate the stamen constituency, as number of fertile stamens, presence of staminodia, bearding, etc. My general impression, based upon observations of з Brückn. in Engl. & Prantl, Nat. Pflanzenfam. 15a: 159—181. 1930. 1942] WOODSON—COMMENTARY ON COMMELINACEAE 143 such genera as Callisia and Tripogandra (Descantaria), is that the stamens of the family show great variation frequently amongst spe- cies of a single genus. An outstanding example of the impractical- ity of Briickner’s application of staminal characters is provided by his inclusion of Descantaria within the Hexandrae and Neodon- nellia within the Triandrae. Even a casual examination of repre- sentative species shows that the outer stamens of both are essen- tially alike. In making a major subdivision of the Commelinaceae, I would base my separation upon inflorescence structure. In the entire fam- ily the basic inflorescence design is the scorpioid cyme. But in the Commelineae, as I distinguish the tribe, the ultimate branches or units of the inflorescence are composed of individual scorpioid cymes which appear 1-sided superficially ; these may be solitary or variously compounded even in a given species, or very rarely re- duced to a solitary flower. In my interpretation of the Tradescan- tieae, on the other hand, the basic structure of the inflorescence has been modified so that the ultimate branches or units of the inflores- cence are paired sessile scorpioid cymes which appear as a 2-sided unit superficially, rarely reduced to a solitary flower. Very little experience is necessary both to distinguish these types of inflorescence and to appreciate their validity. At first, the observer may confuse occurrences of separate but superficially paired inflorescences of the Commelineae type with the more high- ly evolved type characteristic of the Tradescentieae. But closer examination of such specimens (as in Tinantia leiocalyx Clarke) will show that pairing here is merely a numerical chance and not a concrete unit of structure as in the whole tribe Tradescantieae. The paired cymes of Tradescantia and its near relatives, also, are like- ly to cause trouble at first, because they are so condensed. But su- perficially they usually will appear 2-sided or ‘‘parted in the mid- dle’’, which a dissection or closer examination will confirm. The paired cymes of the Tradescantieae are considered to be an integral condensation derived from the individual cymes of the Commelineae. Amongst the genera of this tribe condensation pro- ceeds within the limits of the paired суше. In the genus T'ripogan- dra the paired cymes are borne at the tip of a common naked pe- dunele which usually is elongate, although greatly foreshortened in T. Warscewicziana, a species somewhat transitional to Callisia. These pedunculate cymes may be terminal or axillary to a stem [Vor. 29 144 ANNALS OF THE MISSOURI BOTANICAL GARDEN leaf and solitary or in clusters. The few to many flowers of each cyme are subtended by rather inconspicuous bracteoles which ordi- narily are hyaline or somewhat petalaceous, only rarely slightly foliaceous, and are disposed in two imbricated series upon either side of the axis. In the remaining genera of Tradescantieae, typified most famil- iarly by T'radescantia (sensu stricto), the paired cymes are sessile, and are borne in the axils of bracts which are more or less leaf-like. These bracts, therefore, are oriented at a divergence approaching 90° to the position of the bracteoles of the cymes proper. It appears fairly obvious, then, that the bracts of Tradescantia and its closest relatives are merely subterminal foliage leaves which subtend the sessile umbelliform cymes through the complete reduction of the peduncle so manifest in Tripogandra. These remarks pertain particularly to the terminal or peduncu- late lateral inflorescences of the Tradescantias. Still more pro- nounced modification is found in further reductions in sessile lat- eral inflorescences, as may be seen in such a common species as T. canaliculata. In this species the bracts are reduced proportion- ally to the shortening of the lateral peduncule until they are demon- strated only with difficulty in completely sessile cymes; thus the single foliage leaf of the main stem comes to subtend the paired cymes previously terminal to a lateral, 2-bracted branch. In the same species such condensations of lateral inflorescences frequent- ly, if not almost invariably, occur in conjunction with the terminal inflorescences of the main stem, resulting in a congested flower head capable of resolution into basically paired cymes only with rather careful dissection and balanced interpretation. The conclusion of this trend of reduction is attained in T. nana Mart. & Gal. in which both the terminal and lateral inflorescences consist of a solitary, sessile flower. As separated on the inflorescence structure, the Commelineae in- clude genera of both hemispheres, whilst the Tradescantieae are entirely confined to the Americas. I view the latter as being the more recent, not only upon the basis of distribution and morpho- logical philosophy of the inflorescence, but also for a reason that my readers may soon appreciate: the differentiation of the genera is much the more flexible, so much so, in fact, that not infrequently it will be necessary to invoke a combination of characters, as is often done for families and orders. These ‘‘genera’’ of Tradescan- 1942] WOODSON—COMMENTARY ON COMMELINACEAE 145 tieae are more to be eonstrued as evolutionary tendencies still in the process of differentiation than as distinct entities in sharp fo- cus. Nevertheless, I am reasonably confident of the validity and convenience of the groups as they stand here. In separating genera, I usually have found that there are suffi- cient characters without employing some of the more difficult eri- teria of my eminent predecessors which, if applied consistently, would increase the number of groups, obscure phylogenetic lines, and render identification even more difficult, at least in the her- barium. In short, my view of the North American genera of Com- melinaceae is that expressed in the key which follows. Although designed particularly for North and Central America, including the Antilles, the key will be found of use in the southern continent, where nearly the same representation occurs, with the addition of two or three other groups which are not likely to cause trouble. In order to clarify my position the better, I have appended a few comments on the revised generic lines, as well as some of the more obvious nomenclatural adjustments necessary. These changes are based only upon species with which I am familiar at present, and are not complete compilations. With the advent of better World political conditions, a monographic treatment of the whole group will be in order. KEY TO THE GENERA . Ultimate branches of the inflorescence composed of individual seorpioid cymes Оре 1-sided о solitary or variously clustered, very rarely re- uced to a solitary terminal flower; corolla regular or irregular (Commelineae ). i Cymes variously di Or iN P rarely solitary, but never enelosed by a spathaceous braet. e. Fertile anthers иа plants caulescent, terrestrial. d. Anthers large, with an inconspicuous connective, гари by apical pores; seeds ut A fleshy TILES к ето CT . DICHORISANDRA dd. E small, but with a conspicuous sterile connective, long- inally; seeds dry. е: Це ary and capsules 3-celled. f. Flowers regular or essentially во....................... II. ANEILEMA ff; Flowers very strongly irregular. seese Teee зз не: III. TINANTIA ee: Ovary and capsules 2-celled- CR ЫТ: V. FLOSCOPA ee, Fertile anthers fused into a cochleate hood; large acaulescent epiphytes А м ты ЖЕ Comet: е V. COCHLIOSTEMA bb. Cymes solitary, enclosed by a conspicuous spathaceous braet. e. Fruits ook capsular; sterile stamens (when present) ies eruciate dccus P ее eNO AMI О Che . COMMELINA ee. и indehiscent, pergamentaceous; sterile stamens with нам tria Ie SIQUID UD UE OU TI DUO VLE Кайыгы [Vor. 29 146 ANNALS OF THE MISSOURI BOTANICAL GARDEN aa. Ultimate branches of the inflorescence composed of paired sessile seorpioid cymes ppearing as a 2-sided unit superficially, rarely reduced to a solitary flower; corolla regular ( Tradescantieae ). b. Corolla apopetalous, the petals free to the base e. Paired cymes distinctly реда apes never sessile and subtended by leafy bracts (but the bracteoles rarely somewhat foliaceous in part); stamens 6, usually in 2 very dissimilar series, тый outer occasionally sterile, rarely all ertile and essentially similar; sepals foliaceous or petalaceous............ ll p даи таа ааа каза сак TUE IT LT VIII. TRIPOGANDRA ec. Paired cymes sessile and subtended by more or less conspicuons leafy bracts, rarely appearing pedunculate and the bracts greatly reduced, but the sta- mens usually 1-3, rarely 6, all fertile, and the sepals paleaceous....IX. CALLISIA essentially similar; sepals foliaeeous or petalaceous, d. Cymes on slender peduncles lateral to the main e. Flowering peduncles elongate, vinis branched; — foliaceous ; sepals becoming fleshy in fruit............cccccceecesccecs . CAMPELIA ee. Flowering peduncles very short, simple; bracts appearing as coriace- ous spathes; sepals drying in Їгїїї.......................... XI. RHOEO dd. Cymes terminal to the main stem, occasionally also lateral in the upper leaf axils; lateral cymes very rarely reduced to a solitary flower........ TREY TREC е CCE ES PKT PORTER CET COTE eee Te XII. TRADESCANTIA bb. Corolla gamopetalous, the petals united at the bas e. Flowers borne in leafy-bracted cymes; corolla ик relatively short; plants with extensive creeping stems. d. Sepals separate, foliaceous; plants tumid............... XIII. SETCREASEA dd. Sepals united into an ^ aoa та tube, hyaline; plants somewhat —— but not АИ АЯ XIV. ZEBRINA ee. Flowers solitary and кке in the axils of the congested upper leaves; corolla tube long and slender; semiacaulescent alpines........ XV. WELDENIA I. Dichorisanpra Mikan, Del. Fl. & Faun. Bras. pl. 3. 1820; C. B. Clarke in A. & C. DC. Monogr. 3: 272. 1881; Brückn. in Engl. & Prantl, Nat. Pflanzenfam. 15a: 170. 1930, nom. conserv. ? Stickmannia Neck. Elem. 3: 171. 1791, nom. гейе. Petalowis Raf. Fl. Tellur. 2: 83. 1836 [1837]. This is one of the most distinctive genera of Commelinaceae, as is shown particularly by the anthers and arillate fruit. The petals, also, are much more resistant to deliquescence than those of other Spiderworts. The genus is best represented in Brazil, from whence nearly 30 species have been described, mostly spurious, I suspect upon the basis of familiarity with the common and highly variable D. hexandra (Aubl.) Standl. in Panama. II. Ахепемл К. Br. Prodr. 270. 1810; C. B. Clarke, loc. cit. 195. 1881; Brückn. loc. cit. 174. 1930 1942] WOODSON—COMMENTARY ON COMMELINACEAE 147 Murdannia Royle, Illustr. Bot. Himal. 403. pl. 95. 1839 [1840]. Briickn. loe. cit. 173. 1930; also numerous other Asiatic and African synonyms for both genera enumerated by Clarke and Brückner. I am placing within Ancilema the erstwhile species of Trades- cantia having simple scorpioid cymes as component units. This is an entirely natural procedure as anyone who examines a suite of the large Asiatic and African genus will see. Why these plants were ever placed within Tradescantia in the first place is hard to understand. The Old World species of Апейета present much variation in inflorescence modification, and it is significant to find that those of America follow much the same system of variation. Although Aneilema (sensu stricto) was placed in the Commelineae and Murdannia in the Tradescantieae by Brückner, it seems quite obvious to me that the two are congeneric. The irregularity of the corolla of the former, as Brückner himself confesses, is only осса- sional (‘‘Pet. frei, das äussere mitunter kleiner", p.175) ; while the same, oddly enough, can be said for the regularity in Murdannia (**Pet. frei, mitunter das äussere wenig anders gestaltet", p.179): ANEILEMA Chihuahuensis (Standl.) Woodson, comb. nov. Tradescantia chihuahuensis Standl. Field Mus. Publ. Bot. 17: TE МЕТА ANEILEMA geniculata (Jacq.) Woodson, comb. nov. Tradescantia geniculata Jacq. Select. Stirp. Amer. 94. pl. 64. 1763; C. B. Clarke, loc. cit. 300. 1881. ANEILEMA Greenmanii Woodson, nom. nov. Tradescantia macrophylla Greenm. Proc. Amer. Acad. 33: 472. 1898, non Aneilema macrophylla В. Br. AxNkEILEMA holosericea (Kunth) Woodson, comb. nov. Dichorisandra longifolia Mart. & Gal. Bull. Acad. Brux. 92: 378. 1842, non Aneilema longifolia Wall. nec Hook. Tradescantia floribunda Mart. & Gal. loc. cit. 377. 1842, non Aneilema floribunda Hook. & Arn. Tradescantia holosericea Kunth, Enum. 4: 92. 1843; C. B. Clarke, loe. cit. 302. 1881. Tradescantia holosericea Kunth В. dracaenoides C. B. Clarke, loc. cit. 1881. Tradescantia longifolia (Mart. & Gal.) Greenm. Proc. Amer. Acad. 33: 471. 1898. [Vor. 29 148 ANNALS OF THE MISSOURI BOTANICAL GARDEN Tradescantia dracaenoides (Clarke) Greenm. Proc. Amer. Acad. 39: 70. 1903. Certain of the several varieties of this species, as enumerated by Clarke, may well be entitled to specific rank, others appear ques- tionable even as varieties. This is a question that must await mon- ographie study. ANEILEMA Karwinskyana (К. & S.) Woodson, comb. nov. Tradescantia Karwinskyana R. & S. Syst. 7: 1165. 1830; C. B. Clarke, loc. cit. 299. 1881. ANEILEMA linearis (Benth.) Woodson, comb. nov. Tradescantia linearis Benth. Pl. Hartweg. 27. 1839; C. B. Clarke, loc. eit. 298. 1881. Tradescantia graminifolia Mart. & Gal. loc. cit. 378. 1842. Tradescantia venustula Kunth, Enum. 4: 87. 1843; C. B. Clarke, loc. eit. 298. 1881. Tradescantia rhodantha Torr. Bot. Mex. Bound. Surv. 225. 1859. Tradescantia linearis Benth. Q. graminifolia (Mart. & Gal.) C. B. Clarke, loe. eit. 299. 1881. Амепжма pulchella (HBK.) Woodson, comb. nov. Tradescantia pulchella HBK. Nov. Gen. & Sp. 1: 262. 1815 [1816] ; 7: pl. 673. 1825; C. B. Clarke, loc, cit. 997. 1881. ПІ. Tinantia Scheidw. in Otto & Dietr. Allgem. Gartenzeit. 7: 364. 1839; C. B. Clarke, loe. cit. 285. 1881; Brückn. loc. eit. 175. 1930, nom. conserv. Pogomesia Raf. loc. eit. 67. 1836 [1837]. nom. гејіе. I have criticized Brückner in his use of regularity or irregularity of the corolla so severely in preceding paragraphs, that I find it necessary to apologize for my use of the same character in distin- guishing Aneilema and T'inantia. As I have explained (p. 147), how- ever, the irregularity of the corolla of Aneilema and the irregular- ity of Murdannia as applied by Brückner is merley a variable com- parison of size of the petals. The flowers of Tinantia, on the other hand, are as incontestably irregular as are those of Commelina, not only in size of the petals, but in their coloration and structure of the stamens. The species of Tinantia are extremely variable in the size of the plants and the degree of compounding of their inflores- cences. 1942] WOODSON—COMMENTARY ON COMMELINACEAE 149 ТУ. Froscopa Lour. Fl. Cochinch. 192. 1790; С. B. Clarke, loc. cit. 265. 1881; Brückn. loc. cit. 176. 1930. Dithyrocarpus Kunth, loc. cit. 76. 1843. I regard Floscopa as very closely related to Tinantia, with which I would unite it were it not for the 2-celled ovary and capsules. The common speeies of Central America, F. robusta (Seub.) Clarke, sets seeds with the most astonishing abundance and rapidity. V. СоснтлозтемА Lem. Illustr. Hort. 6: Mise. 70. pl. 217. 1859; C. B. Clarke, loc. cit. 231. 1881; Briickn. loc. cit. 180. 1930. This magnifieent monotypie genus has recently been discovered in Panama. The thick indument of the petal margins is remarkable, as it is composed of beaded hairs similar to those of the staminal filaments of the family generally. VI. СоммеммАа І. Sp. РІ. 60. 1753; С. В. Clarke, loc. cit. 138. 1881; Briickn. loc. eit. 177. 1930, also numerous synonyms supplied by Clarke and Brückn. Sauvallea Wright in Sauv. Fl. Cub. 156. 1873. Commelinantia 'Tharp, Bull. Torrey Club 49: 269. 1922. The genus Commelina is represented by more species in the Old than in the New World. Amongst the Old World species are found many startling morphological departures, as judged from the spe- cies of the United States, including forms with 2- or 3-loculed ovaries, others with variable numbers of flowers or even stamens, wide variations in bearding of stamens, and in relative size of the petals. Such being the case, it seems 1n every way better to com- bine Sauvallea and Commelinantia within the inclusive genus. CoMMELINA anomala (Тотг.) Woodson, comb. nov. Tradescantia anomala 'Torr. Bot. Mex. Bound, Surv. 225. 1859. Tinantia anomala (Torr.) C. B. Clarke, loc. cit. 287. 1881. Commelinantia anomala (Torr.) Tharp, Bull. Torrey Club 49: 269. 1922; Briickn. loc. cit. 176. 1930. The interested reader should refer to Dr. Tharp’s full account of the reasons for regarding this species as a distinct genus. I do not think it necessary to answer his arguments in detail, for that would entail an extended discussion of morphology with particular regard to the Asiatic species of Commelina beyond immediately profitable ends. The student of Commelinaceae who reads this inadequate paragraph, however, may be sufficiently familiar already with the [Vor. 29 150 ANNALS OF THE MISSOURI BOTANICAL GARDEN variable species of the Old World and with the perplexing variabil- ity of the whole family, for that matter, and probably will well un- derstand the taxonomic confusion in the family that would be caused by over-evaluation of such characters as bearding of the stamens, ete., ete. In the Gray Herbarium Card Index there is already an entry for “Commelina anomala Torrey’’ ex Tharp, loc. cit., so а word of ex- planation for the new combination here is necessary. The source of the Card Index entry is a footnote to Dr. Tharp’s paper on Com- melinantia (Bull. Torrey Club 49: 269. 1922), which explains that Dr. Torrey first determined the type specimen as ‘‘Commelyna anomala Torr.’’, later striking out the ‘‘Commelyna’’ and substi- tuting **Tradescantia"'. It seems quite clear that this publication in Dr. Tharp's footnote cannot be maintained as valid, as it vio- lates Article 40 of the International Rules, being patently citation as a synonym. COMMELINA Blainii (Wright) Woodson, comb. nov. Sauvallea Ваши Wright, loc. cit. 157. 1873; C. B. Clarke, loc. cit. 315. 1881; Briickn. loc. eit. 171. 1930. The distinctive characters of Sauvallea are the 6 fertile stamens and the solitary flowers, which, however, are enclosed within the charaeteristie spathe of the familiar Commelinas. The petals are subequal. These variants from the general run of Commelinas, however, are of the sort not found to be good generie criteria in the Commelinaceae, and even Clarke called attention to the similarity to Commelina. VII. PHAEOSPHAERION Hassk. Flora 49: 212. 1866; С. B. Clarke, loc. cit. 135. 1881. Athyrocarpus Schlecht. Linnaea 26: 454. 1853; Brückn. loc. cit. 179. 1930. Phaeosphaerion is recognized here since there can be no doubt that Athyrocarpus was merely a suggested segregation from Com- melina as far as Schlechtendal was concerned, and no formal trans- fer was made under Athyrocarpus. VIII. Trreocanpra Raf. Fl. Tellur. 2: 16. 1836 [1837], emend. Heminema Raf. loc. cit. 17. 1837. Descantaria Schlecht. Linnaea 26: 140. 1853; Briickn. loc. cit. 171. 1930, nom. subnud. provis. 1942] WOODSON—COMMENTARY ON COMMELINACEAE 151 Disgrega Hassk. Flora 49: 215. 1866, nom subnud. Leptorhoeo C. B. Clarke in Hemsl. Diagn. Pl. Nov. 55. 1880; C. B. Clarke, loc. cit. 317. 1881; Briickn. loc. cit. 167. 1930. Cuthbertia Small, Fl. Southeast. U. S. 237. 1903. Donnellia C. B. Clarke, Bot. Gaz. 33: 261. 1902. Neodonnellia Rose, Proc. Biol. Soc. Wash. 19: 96. 1906; Brückn. loc. cit. 174. 1930. Tripogandra, Heminema, Descantaria, Disgrega, Donnellia, and Neodonnellia all were erected to include species, previously placed in Tradescantia, having two greatly dissimilar stamen series, the outer with shorter filaments and anthers (occasionally sterile) quite different from those of the inner. Species of Leptorhoeo and Cuth- bertia have essentially similar stamens, but without the foliaceous bracts of Tradescantia (sensu stricto). In addition, the former is supposed to be distinguished by having the seeds solitary in the locules of the capsules. Tripogandra is emended here to include these numerous elements for the following reasons: (1) It is obvious to any student of the Tradescantia complex that there is a marked tendency towards in- equality of the stamens throughout; even in the Descantaria group such characters as the bearding of the stamens are quite variable, and could be made the basis of further dubious generic segrega- tions, which are devoutly to be avoided; (2) seed number in the locules of the capsules varies amongst individual plants of a single species (cf. Anderson & Woodson, Contr. Arnold Arb. 9: 27-29. pl. 7. 1935). In my revision with Anderson, to which reference has been made, Cuthbertia was included with Tradescantia, for it was not at that time realized that the bracts of the latter were wholly lacking in the former. It is now recognized that such is the case (cf. p. 143, paragraph 3, of this paper), and that the foliaceous structures at first interpreted as bracts are in reality the lower bracteoles of the paired cymes, as is shown by their orientation. I do not understand why Descantaria was taken up by Brückner, as Schlechtendal published it merely as a suggested segregation from T'radescantia, without a description, as a matter of fact with- out even indieating valid reasons for separating it, and without making any formal transfers of species. Tripogandra and Hemi- nema, on the other hand, are quite well described as Rafinesquian genera go, and there ean be no doubt of their application. Char- acteristically, both genera consist of the same species, T'radescantia [Vor. 29 152 ANNALS OF THE MISSOURI BOTANICAL GARDEN multiflora, although Rafinesque credits one to Schwarz and the other to Jacquin. T'RIPOGANDRA amplexicaulis (Kl.) Woodson, comb. nov. Tradescantia amplexicaulis Kl. ex C. B. Clarke, loe. cit. 304. 1881. Descantaria amplexicaulis (Kl.) Briickn. loc. eit. 1927; loc. cit. 1930. TnrPocANpRA angustifolia (Rob.) Woodson, comb. nov. Tradescantia angustifolia Rob. Proc. Amer. Acad. 27: 185. 1892. Descantaria angustifolia (Rob.) Briickn. Notizblatt 10: 56. 1927; Brückn. in Engl. & Prantl, loc. cit. 171. 1930. TRIPOGANDRA cumanensis (Kunth) Woodson, comb. nov. Tradescantia cumanensis Kunth, Enum. 4: 96. 1843; C. B. Clarke, loc. eit. 306. 1881. Descantaria cumanensis (Kunth) Schlecht. ex Brückn. loc. cit. 1927 ; loc. cit. 1930. TniPocANDRA Disgrega (Kunth) Woodson, comb. nov. Tradescantia Disgrega Kunth, loc. cit. 97. 1843; C. B. Clarke, loc. cit. 305. 1881. Disgrega mexicana Hassk. ex C. B. Clarke, loc. cit. 1881, nom. nud. in synon. Descantaria Disgrega (Kl.) Brückn. loc. cit. 1927 ; loc. cit. 1930. Trrpocanpra elongata (С. Е. W. Meyer) Woodson, comb. nov. Tradescantia elongata G. F. W. Meyer, Fl. Esseq. 146. 1818; C. B. Clarke, loc. eit. 303. 1881. Descantaria elongata (G. F. W. Meyer) Briickn. loc. cit. 1927; loc. eit. 1930. TiPoGANDRA floribunda (Hook. & Arn.) Woodson, comb. nov. Aneilema floribunda Hook. & Arn. Bot. Beechey Voy. 311. 1840. Tradescantia filiformis Mart. & Gal. Bull. Acad. Brux. 92: 276. 1842. Leptorhoeo filiformis (Mart. & Gal.) C. B. Clarke in Hemsl. Diagn. Pl. Nov. 55. 1880; C. B. Clarke, loc. cit. 317. 1881; Briickn. loc. cit. 167. 1930. Leptorhoeo floribunda (Hook. & Arn.) Baill. Hist. Pl. 13: 218. 1894. 1942] WOODSON—COMMENTARY ON COMMELINACEAE 153 Three errors are widespread in citing this species under Leptor- hoeo, including two variants in spelling the generic name ( Leptor- rhoeo, Leptorhoés ), and the combination itself is frequently accred- ited to Hemsley (even the citation of Clarke himself in DC. Monogr.). Trreocanpra grandiflora (Donn. Sm.) Woodson, comb. nov. Callisia grandiflora Donn. Sm. Bot. Gaz. 31: 125. 1901. Donnellia grandiflora (Donn. Sm.) Clarke, Bot. Gaz. 33: 261. 1902. Neodonnellia grandiflora (Donn. Sm.) Rose, Proc. Biol. Soc. Wash. 19: 96. 1906; Brückn. loc. cit. 174. 1930. Trreocanpra Lundellii (Standl.) Woodson, comb. nov. Tradescantia Lundellii Standl. Field Mus. Publ. Bot. 22: 5. 1940. TrreoGaNpRA minuta (C. D. Clarke) Woodson, comb. nov. Tradescantia minuta C. B. Clarke, loc. cit. 307. 1881. Descantaria minuta (C. B. Clarke) Brückn. loc. cit. 56: 1927; Brückn. in Engl. & Prantl, loc. cit. 171. 1930. TRIPOGANDRA MULTIFLORA (Sw.) Raf. loc, cit. 16. 1836 [1837]. Tradescantia multiflora Sw. Prodr. 57. 1789; С. B. Clarke, loc. cit. 305. 1881. Tradescantia procumbens Willd. Sp. Pl. 2": 19. 1800. Heminema multiflora (Sw.) Raf. loc. cit. 17. 1836 [1837]. Descantaria procumbens (Willd.) Hassk. ex С. B. Clarke, loc. cit. 1881, nom. nud. in synon. Descantaria multiflora (Sw.) Brückn. loc. cit. 56. 1927 ; Brückn. in Engl. & Prantl, loc. cit. 172. 1930. TRIPOGANDRA Palmeri (Rose) Woodson, comb. nov. Tradescantia Palmeri Rose, Contr. U. S. Nat. Herb. 1: 113. 1891. Descantaria Palmeri (Rose) Brückn. loc. cit. 56. 1927. TRIPOGANDRA rosea (Vent.) Woodson, comb. nov. Tradescantia rosea Vent. Jard. Cels. pl. 24. 1800; Anders. & Woods. loc. cit. 112. 1935; C. B. Clarke, loc. eit. 298. 1881; Brückn. loc. cit. 167. 1930. Cuthbertia rosea (Vent.) Small. Fl. Southeast. U. S. 237. 1903. I am not making adjustments under Tripogandra for the two varieties which Anderson and I recognized under T. rosea in our [Vor. 29, 1942] 154 ANNALS OF THE MISSOURI BOTANICAL GARDEN earlier revision (loc. cit. 1935), as I am not certain whether they should be interpreted as varieties or species. This is a question that can be settled only by extensive field study. TRIPOGANDRA saxicola (Greenm.) Woodson, comb. nov. Tradescantia saxicola Greenm. Proc. Amer. Acad. 39: 70. 1903. Descantaria saxicola (Greenm.) Briickn. loc. cit. 56. 1927. Trreocanpra Warscewicziana (Kunth & Bouché) Woodson, comb. nov. Tradescantia Warscewicziana Kunth & Bouché, Ind. Sem. Hort. Berol. 11: 1847; C. B. Clarke, loe. cit. 302. 1881. Spironema Warscewiczianum Hassk. ex. C. B. Clarke, loc. cit. 1881, nom. nud. in synon. Spironema Warscewiczianum (Kunth & Bouché) Brückn. loc. eit. 171. 1930. IX. Cauuista L. in Loefl. It. Hisp. 305. 1758; C. B. Clarke, loc. eit. 309. 1881; Briickn. loc. eit. 173. 1930. Hapalanthus Jacq. Select. Stirp. Amer. 11. pl. 11. 1763. Spironema Lindl. Edwards's Bot. Reg. N. S. 3: pl. 47, Miscel. 26. 1840; C. B. Clarke, loe. eit. 313. 1881; Brückn. loc. cit. 171. 1930. Tradescantella Small. loc. eit. 238. 1903. Rectanthera Degener, Fl. Hawai. 1: 62. 1932. CauuisiA fragrans (Lindl.) Woodson, comb. nov. Spironema fragrans Lindl. loc. cit. 1840; C. B. Clarke, loc. cit. 1881; Brückn. loc. cit. 1930. Rectanthera fragrans (Lindl) Degener, Fl. Hawai. 1: 62. 1932. Aside from its gigantie size, the relationship of this species to the small ereeping plants of Callisia should be suffieiently clear. The long runners by which C. fragrans propagates vegetatively probably are related phylogenetieally to the ereeping stems of the latter. The remaining genera require little comment, as my views co- ineide with eurrent interpretations of them. "These studies still leave Tradescantia a large and rather complieated genus, although not as much so as formerly. Future studies may well restrict the genus even further, devoting particular attention to such tropical and subtropical groups as T. micrantha Torr., T. commelinoides В. & S. and their relatives, and the T. fluminensis Vell. complex. A REVISION OF THE GENUS BUMELIA IN THE UNITED STATES! ROBERT BROWN CLARK Formerly Garden Apprentice, Missouri Botanical Garden The genus Витейа is the hardiest representative of the Sapota- сеае in North America, occurring as far north as the lower valley of the Missouri River. It therefore differs from other members of this tropical and subtropical family in its adaptation to a temper- ate climate. The majority of species grow in mesophytie situations, although certain species have become well established in xerophytic habitats. GENERAL MonPHOLOGY Bumelia is distinguished morphologically from other genera of the Sapotaceae in the structure of the flower and fruit, for example: (a) a pair of lateral appendages flank each corolla-lobe, (b) stam- inodia replace a cycle of stamens, and (c) the seeds are without al- bumen and exhibit a circular, basal hilum.? Stem: The stems are sometimes armed with simple or branched spines. The bark is reddish-brown, appearing gray in age, while the branchlets in certain species are somewhat pubescent. The wood is singularly tough, becoming brittle in B. angustifolia, invariably yellowish, and very close-grained.? ‘‘Short-shoots,’’ or brachy- blasts, suggestive of certain gymnosperms, are frequently found in the leaf axils. Corolla: The genus is characterized in part by the structure of its corolla, which is sympetalous through basal coalescence of the whitish petals to form a short tube. Each corolla-lobe bears a pair of lateral, petaloid appendages, thus constituting a pentamerous eorolla with trifid lobes.* * An investigation carried out in the Graduate Laboratories of the Henry Shaw School of Botany of Washington University and submitted as a thesis in partial fulfillment of the Е for the degree of master of scienee in the Henry Shaw School of Botany of Washington University. d C., in m 7: 424. 1938. d, S. J., in Trop. Woods 59: 26-27. 1939. “А completely satisfactory evaluation of the various components of the flower of Bumelia, especially the lateral appendages of the eorolla-lobes, has not become established. Students of the Sapotaceae have merely suggested the probable origin and relationship of the pie а Hartog, M. M., On the floral structure and affinities of Sapotaceae. Jour. Bot. 16: 66. 1878; Lam, H. i. On the system of the Sapotaceae. Recueil Trav. Bot. Neerland. 36: ph 1939. Issued September 18, 1942. (155) [Vor. 29 156 ANNALS OF THE MISSOURI BOTANICAL GARDEN Androecium : The five fertile stamens are epipetalous and oppo- site the lobes of the corolla. The anthers are attached to slender filaments and dehisce extrorsely by means of longitudinal slits. The staminodia, alternating with the fertile stamens, are petaloid and at an early phase of anthesis form a tube through which the fili- form style emerges while the rapidly maturing stamens are en- closed by the yet unexpanded corolla. Gynoecium: The pistil is pentacarpellary. The superior ovary bears a simple, erect, relatively short style surmounted by an in- conspicuous stigma that becomes receptive before the corolla is completely unfolded. Normally, the outer wall of the ovary is hir- sute, but this character is not constant throughout the genus, e. g., B. reclinata. Although the ovary bears five anatropous ovules (one in each locule), it is seldom that more than one develops. Fruit: The fruit is a single-seeded berry, enclosed in the fleshy wall of the ovary and often tipped by the persistent style. The ber- ries are rather small, generally purplish-black when ripe, and edible. Seed: The seeds are small with a hard, shiny, and oily testa which is generally brown though sometimes becoming variegated. The embryo is large, filling the cavity, while the radicle is barely evi- dent; the cotyledons are thick, fleshy, usually consolidated, and al- bumen is wanting. GEOGRAPHICAL DISTRIBUTION Bumelia proves to be not only the hardiest genus of the Sapota- ceae but also the most extensive. It ranges from the southeastern section of the country as far north as central Missouri and as far west as southeastern Arizona. The region of greatest diversity of species is centered in Florida, where eleven species occur, five be- ing endemic. The greatest diversity of habitat is in Texas, where one species may be found growing in moist woodlands, while an- other inhabits only dry, sandy situations. There are three natural groups of Bumelia in the United States. The first group consists of those species with lanuginose or serice- ous pubescence on the branchlets, the under-surface of the leaves, the petioles, and the inflorescence, and includes B. lanuginosa, B. tenax, B. lacuum, B. rigida, ete. Except for the xerophytie spe- cies B. rigida, which occurs from southern Texas to southeastern Arizona, members of this group are chiefly mesophytie and inhabit sandy uplands, though not far removed from a source of water. 1942] 157 CLARK— REVISION OF BUMELIA "593839 рәји ду 3104810143 риши zo soroods jo uorjnqujstp Sutmoys Фед rex Риъх?) хри? Deus PSOJUDUOJOJNA PIJOJIUISSPI PjemouP prnoyysnbue а т ae n Я А ят а o g x g V я + бж п a g в e о go g A g + я ¥ ae | а [Vor. 29 158 ANNALS OF THE MISSOURI BOTANICAL GARDEN They range from coastal South Carolina westward to southeastern Arizona, their distribution being interrupted by the lower Missis- sippi River Valley. Most widespread of these species is B. lanugi- nosa, which occurs in southern Georgia with two varieties extend- ing from central Missouri to east-central Texas. The second group is characterized by a tomentulose evanescent pubescence, and comprises such species as В. lycioides, B. Зтайи, B. reclinata, ete. This group, except for B. texana which is found in the mountains or the dry, rocky plains of the Southwest, frequents more moist localities, occurring in the lower Mississippi River Val- ley, the Gulf and Atlantic coastal regions, and along water-courses as far north as southern Illinois and southeastern Virginia. The third group is entirely glabrous or with only a few scattered hairs, and in this country B. angustifolia is its only representative. Plants of this species abound in the lower Rio Grande Valley of southern Texas, and are again found on the Keys and Gulf coast of Florida. ACKNOWLEDGMENTS The writer takes this opportunity to express his appreciation to those who have so generously helped him in this study, especially to Mr. H. B. Parks, of the Texas Agricultural Experiment Station, for cooperation in making special collections, and to Mr. E. J. Palmer, of the Arnold Arboretum, for advice and criticism. Thanks are also due the curators of the herbaria from which specimens have been borrowed. These herbaria, with the abbreviations used in this paper, are as follows: A—Arnold Arboretum, Harvard University. G—Gray Herbarium, Harvard University. L—Louisiana State University. M—Missouri Botanical Garden. NY—New York Botanical Garden. P—Academy of Natural Sciences of Philadelphia. Tr—s. M. Tracy Herbarium, Texas Agricultural Experiment Station. Tu—Tulane University. UO—University of Oklahoma. US—United States National Herbarium. UT—University of Texas. 1942] CLARK—REVISION OF BUMELIA 159 TAXONOMY Bumelia® Sw. Prodr. Veg. Ind. Осе. 49. 1788; Benth. & Hook. Gen. Pl. 2: 660. 1876; Engler & Prantl, Nat. Pflanzenfam. 41: 145. 1891; Sarg. Silva 5: 167. 1893. Lycioides L. Hort. Cliff. 488. 1738. Robertia Scop. Introd. Hist. Nat. 154. 1777. Sclerocladus Raf. Sylva Tellur. 35. 1838. Sclerozus Raf. Aut. Bot. 2: 73. 1840. Lyciodes O. Ktze. Rev. Gen. 2: 406. 1891. Shrubs or small trees frequently with milky sap and spinescent branches. Leaves alternate or often fasciculate, simple, obovate to elliptie, entire, penninerved, short-petiolate, pubeseent or gla- brous beneath, usually coriaceous, deciduous or persistent. Flow- ers small, on rather short pedicels from the axils of scarious braets, in many-flowered clusters from leaf-axils or from leafless nodes of the previous year, hermaphrodite, pentamerous, actinomorphic, hypogynous, proterogynous; calyx with 5 dimorphie, free sepals which are imbricate in aestivation and decurrent on the pedicel, persistent ; corolla whitish, sympetalous, forming a short tube, the 9 lobes with pairs of lateral appendages; stamens epipetalous, 10 in 2 isomerous whorls—the fertile cycle opposite and equal in length to the corolla-lobes, the alternipetalous one reduced to peta- loid staminodia; ovary 5-celled, generally hirsute, gradually or abruptly contracted into a filiform style stigmatic at apex, each locule with a solitary, anatropous ovule from the base of an axile placenta. Fruit a berry, oblong-ellipsoid or subglobose, blackish, often tipped by a persistent style; seed solitary with crustaceous, smooth, shiny, brownish testa, without albumen; embryo large, fill- ing the cavity; cotyledons thick and fleshy. Витейа is confined to the New World, where it is distributed from central United States southward to Mexico, Central America, the West Indies, and South America as far as northern Argentina. Of the 35-40 species, 14 are found in the United States. TYPE SPECIES: Bumelia retusa Sw. Prodr. Veg. Ind. Occ. 49. 1788. KEY ro THE SPECIES A. Branchlets (except in B. anomala), petioles, lower surface of leaves, and in- florescence conspicuously pubescent. B. Pubeseenee woolly and lusterless. ° Internat. Bot. Kongr. in Wien, Verhandl. No. 6374 in list of Nomina Conservanda. Jena, 1906. [Vor. 29 160 ANNALS OF THE MISSOURI BOTANICAL GARDEN C. Leaves 2-10 em. long, 1-4 em. broad; flowers rather numerou D. Pubescenee dense, rusty-tomentose; pedicels slender; ii oblong-ob- ved, vàlegaled о и нее hr 1. B. lanuginosa DD. Pube escence dense, whitish-tomentose; pedicels slender; seed ellipsoid, ‚ЖАУ К К TE E s vet a. var. albicans DDD. Pubescence sparser than in NN whitish becoming rq a pediee e seed obovoid, DOWIE 1b. var. sap rt CC. Leaves 1-4 em. long, 0.5-1.5 em. Мани, flowerg ЧЕ 2. B. rigida ny. BB. Чечения sericeous and shin C. Branches stout and spiny; leaves often small, not more than 4 em. long; pubescence rufous. D. Branchlets and міц surface of leaves densely tomentose; fruit obo- void, 8-1 ДИЙ. арка ааа К ГС ОГ ЛҮ ТУЛУС а ves 3. В. lacuum DD. Branehlets ула tomentose, lower surface of leaves gi only o midrib and veins; fruit subglobose, 4—6 mm. long...... 4. аи CC. Branches slender and nearly spineless; leaves rather large, жебе more than em. long; pubescence "tria -white to tawny. D. Leaves oblong-obovate to oblaneeolate, euneate at base; pedieels ex- tremely slender, is rim. 10ng 22535059559» ESS 5. B. tenax DD. — oblong-obovate to elliptic, obtuse at base; pedicels stout, 3—5 . long. E. peces glabrous, pale yellowish-gray ; leaves shiny and pale green above, silvery benoskh.;.... cevacccs as ити» 6. B. anomala EE. Branchlets very tomentose, red or reddish-brown; leaves dull and dark green above, tawny ђепеаёћ.......................... 7. B. rufa AA. — ры petioles, lower surfaee of leaves, and infloreseenee with few incon- spicuous hairs generally vide. along midrib; or entirely glabrous. B. Leave puberulous or glabrate; fruit жоем or subglobose (except in B. lycioides var. ellipsoidalis). C. Leaves си «re broadest at middle р. 1 —13.5 long, obtuse to айлам at apex (exeept var. ре oe yer ore Te ee re Cee ee T ре 8. B. lycioides E. Leaves rounded at apex; corolla-tube shorter than in species; fruit 7 MET sled ad fee kev аз EPA. Жиза n 8а. var. pe be EE. ibn acute to acuminate at apex; corolla-tube same length as in species; fruit ФО. ‚кавер аа дека у, 8b. var. ellipsoidalis DD. Leaves 2—5 em. long, aeute to occasionally rounded at apex. .9. B. Smallii CC. Leaves obovate or spatulate, mainly broadest above middle. . Leaves Apurques corolla-lobes erose; seed wise not variegated. E. "in t 4-6 mm, long; seed occasionally variegated; ovary glabrous TP CE T TITRE QU EIL e 10. B. reclinata DD. Leaves coriaceous; corolla-lobes entire; seed майы 3 2h eaves 1-7 em. pid fruit obovoid, 8-10 mm. long...... 12. B. texana (kag adel Cees cue К ERR OUR DRESS Sao eG ЛК ТУ megacocca BB. Leaves glabrous; fruit ellipsoid-eylindrical.................. ng ч, алек, 1. Bumelia lanuginosa (Michx.) Pers. Syn. 1: 237. 1805; DC. Prodr. 8: 190. 1844; Nutt. Sylva 3: 36. 1849; Sarg. Silva 5: 171, pl. 247. 1893, pro parte. 1942] CLARK—REVISION OF BUMELIA 161 Sideroxylon tenax Walt. Fl. Car. 100. 1788, поп L. Sideroxylon lanuginosum Michx. Fl. Bor. Am. 1: 122. 1803. ? Chrysophyllum ludovicianum Raf. Fl. Ludovic. 53. 1817. Lyciodes lanuginosum O. Ktz. Rev. Gen. 2: 406. 1891. Shrub or small tree, 3-7 m. high ; branchlets, petioles, lower sur- face of leaves, and inflorescence covered with lusterless tomentum at first pale, becoming reddish-brown; stems with reddish-gray bark, armed occasionally with short stout spines; leaves narrowly oblong-obovate, rarely elliptic, 2-10 em. long, 1-4 em. broad, round to acute or sometimes mucronate at apex, cuneate at base, dark green and lustrous above, thin and firm at maturity; petioles 2-15 mm. long; flowers disposed in few- to many-flowered fascicles; pedicels 8-12 mm. long, greatly enlarged towards apex; sepals broadly obovate to suborbicular, rounded at apex; corolla-lobes broadly ovate to elliptic, appendages ovate-lanceolate, acute ; stam- inodia ovate-elliptic, obtuse at apex, erose, nearly as long as the petals; fruit purplish-black, oblong-ellipsoid, 6-10 mm. long, fre- quently tipped by a persistent style; seed oblong-obovoid, 5-8 mm. long, variegated tan and brown Distribution: in sandy uplands and along streams, southwestern Georgia and adjacent Florida, westward to southeastern Louisiana. Specimens examined: GEORGIA: precise data lacking, Michaux s.n. (Mus, Paris, TYPE; G, fragment). DECATUR Co.: near Bainbridge: Curtiss 1762, in part (A), Harbison 15, 1127 (A). DOUGHERTY Co. along Flint R. at Albany: May 24-28, 1895, Small (A), Harbison 14 (A); Dr. Gillespie’ 8 Poeoson Place, Gillespie 4956 (А). RANDOLPH Co.: Cuthbert, Harbison 6 (А). AIR : Lumber City, Harbison 16 (A). TERRELL Co.: banks of Kinekafoonee Cr., {= ун nis (G, M). А. exact locality unknown, Chapman (С, М, Р); “ХАМ. Fla.’’, Curtiss 35 (С). ALACHUA Co.: Gainesville, Seibert 1412, 1412a (M). COLUMBIA Co.: PA. City: Nash 2167 (A, G, М), Wiegand $ n 2486 (G). DIXIE Co.: A FRANKLIN Co.: banks of Chattahoochee R., above e 0 Biltmore Hb. 1688, coll. by Chapman ae само E (P); Carrabelle, Harbison 23, 27 (A). JACKSON Со.: Marianna: кок 14, 29 (А), ni 35301 (A, M, P). LEVY Co.: Cedar Keys, Tracy 7464 (A, G, M). WAKULLA Co.: lla Springs, Harbison 1216 (A, M). TAYLOR Co.: Ham шр Springs: Palmer 2 /n ^0: ALABAMA: locality not given: Beck (P), Cabell (G). BALDWIN Co.: Fairhope, Jack 3026 (A); Perdido, Mohr 47 (A). DALE Co.: near Ozark, Palmer 38653 (A, M). DALLAS Co.: Selma, Harbison 96, $29, 1095, 1111 (A). MOBILE Co.: Mt. Vernon, Harbison ? (A). Mcgee: HARRISON Co.: Mississippi City, Harbison 3 (A); Stabilized Beach, Cat Isl, Apr. 19, 1931, Penfound (Tu). LOUISIANA: ST. TAMMANY Parish: Slidell, Aug. 1899, Cocks (A). B. lanuginosa is limited in its distribution; it seems to be related to B. tenax but is distinguished by its sparser, woolly pubescence [Vor. 29 162 ANNALS OF THE MISSOURI BOTANICAL GARDEN (which in B. tenax is sericeous), its generally larger and broader leaves, and its fewer-flowered inflorescence. 1a. Bumelia lanuginosa var. albicans Sarg. in Jour. Arn. Arb. 2: 168. 1922. Bumelia arborea Buckl. in Proc. Phil. Acad. 13: 461. 1861. Small tree to 20 m. high, with branches rarely armed ; pubescence dense, whitish, lusterless, appressed at first, becoming woolly; leaves somewhat narrower than in species; petioles 5-15 mm. long; inflorescence many-flowered; pedicels slender, 5-13 mm. long; se- pals reniform to elliptic; corolla-lobes oblong-elliptic; margins of appendages crenulate; staminodia oblong-elliptic, rounded and somewhat cucullate at apex; ovary subglobose, abruptly contracted into an elongated style; fruit ellipsoid to oblong-obovoid, 8-12 mm. long; seed ellipsoid, brown with tuberculate apex. Distribution: in sandy river-bottoms and rocky uplands, east- central Texas, and adjacent Mexico. "Mr examined: Texas: ''S. Tex.’’, Buckley (P, type of B. arborea, G, fragment). ATASCOSA Co.: s. of San Antonio on -— to Corpus Christi, Buckley (A); Cam n Palmer 11242 (M). BEXAR Co.: San Antonio: Bush 802 s M), Hildebrandt Rd., Clark $ Parks 567 (M); Swearingen Ranch, Schulz 49 (A). BOSQUE Co.: Meridian, К 29, 1927, Tharp (UT). BRAZORIA Co.: Brazos R., Brazoria, Palmer 6741 (A, M); Puce Bush 1, 2, 8, 210, 226, 877, 1427, 1524 (M), Palmer 5050 (A); Velasco, Pine 13131 (A, M). BRAZOS Co.: Bryan, Palmer 11739 (A, M); near Neleva, Palmer 13446 (A, M). CALHOUN Co.: Indianola, ону 11618 (A); Wolf Pt., Tharp 1431 (UT). CHAMBERS Co.: Арг. 8, 1936, Le (UT). сомль Co.: Brazos В. bottoms, Lindheimer 90 td New Braunfels, юды heimer 269, in distribution as no. 979 (А, 9, DALLAS Co.: Dal beds 384 (G), 590 (A, M), June 22, 1899, Жоро" (M), “a 21, 1913, ore (A). ERATH Co. EN Palmer 14176 (A, M). rr. BEND Co.: Brazos R., Richmond, Palmer 4943 (A). LVESTON Co.: Galveston Isl., podia 1848, ecol 00. GOLIAD Co.: Goliad: Apr. ч pum реве (M), along San Antonio R., aeneo Clark § Parks 574 (M). GONZALES Co.: Най Ranch, Mas 8, 1940, ie e ; ; Ottine wamp, Cory 5770 (A). HARRIS Co.: Houston: June 1842, Aug. 1843, "ent eli es June 16, 1903, Reverchon (M), June 4, 1915, Fisher (M), Palmer 11456 (A, M). JACKSON Co.: E. Karankawa Pt., Tharp 1424 (UT). LAMPASOS Co.: Lampasos, Mar. 21, 1911, Sargent (A). LLANO Co.: Enchanted Rock, June 11, 1930, Tharp (UT). MATAGORDA Co.: Peyton's Cr., Palmer 9734 (A, M). NAvARRO Co.: Corsicana, Reverchon 3194 (M); ; near Dawson, Reverchon 3867 (M). NUECES Co.: Nuecestown, Apr. 27, 1896, Marlatt (US). PALO PINTO Co.: Strawn, Palmer 14266 (A, M). RUSK Co.: Henderson, July 30, 1927, Tharp (UT). SAN SABA Co.: San Saba, Palmer 11844 (A). SOMERVILLE One Squaw Cr., above Glen Rose, Oct. 9, 1891, Ward (US). TARRANT Co.: Caddo Cr, bottoms, near Hand- ley, Sept. 23, 1902, sit (M); near Ft. Worth, Ruth 202 (С, M, P). TRAVIS Co.: Austin: Apr. 1860, Buckley (M), Mohr 12 (A), Mohr 623 (US), Mar. 29, 1885, Sargent (A), Schostag 3002 очы McKee $ Wesley 3834 (UT), Parker 4791 (UT), №. Y. State College of Forestry, Proj. I, 8269, coll. by Tharp (A), Schulz 2434 (A). vicTORIA Co.: Guadalupe R. bottoms, Victoria: Apr. 9, 1915, Sargent (A, TYPE), Palmer 9104 (A, M). 1942] CLARK—REVISION OF BUMELIA 163 WALKER Co.: Huntsville, Dizon 403 (G). WALLER Co.: Hempstead, Hall 394 (С, M). WHARTON Co.: banks of Colorado R., Wharton, Palmer 4902 (A). WILSON Co.: Parks & Cory 7791 (Tr); Sutherland Springs Oet. 1910, Mackenson (A), Apr. 8, 1915, Sargent (A), Palmer 9209 (A, M), Cutler 858 (M). MEXICO: NUEVO LEON: near реа Apr. 6, 1887, Sargent (А). Tallest of the Bumelias in the United States, B. lanuginosa var. albicans is limited in distribution and is easily recognized by its rather dense, whitish pubescence. 1b. Bumelia lanuginosa var. oblongifolia (Nutt.) Clark, comb. nov. Bumelia oblongifolia Nutt. Gen. 1: 135. 1818; Sylva 3: 33. 1849. Bumelia arachnoidea Raf. New Fl. N. Am. 3: 28. 1836. Bumelia tomentosa A. DC. in DC. Prodr. 8: 190. 1844. Armed shrub or small tree; pubescence rather sparse, dull, whit- ish, becoming tawny ; leaves obtuse to rounded or somewhat retuse at apex, broadly euneate at base; petioles 2-10 mm. long; flowers numerous; pedicels 5-8 mm. long; sepals suborbicular to oblong- ovate or elliptic; corolla-lobes oblong-ovate; ovary subglobose, ab- ruptly contracted into the style; fruit broadly elliptic-ovate; seed obovoid, brown. Distribution: in dry, rocky uplands, southern Missouri and southeastern Kansas to Louisiana and eastern Texas. Speeimens examined: MISSOURI: BARRY Co.: Eagle Rock: Bush 182, 3130 (M); Roaring R., Trelease 1117 (M). BENTON Co.: along Cole Camp Cr.: Demetrio 47 (G), Trelease 430 (M), Palmer 30067 (A, M); along Osage R.: near Fredonia, Palmer 35956 (A € near Warsaw, Palmer 36769 (A, M). BOONE Co.: Missouri R. bluffs above Wilton . O., Jeffrey 307 (M). CALLAWAY Co.: along Stinson Cr., of Fulton, 2 E (M). CARTER Co.: Club House, Trelease 427 (M); К ар Ваггеп Сг., of Fremont, Steyermark 11888 (M). cEDAR Co.: along Sac R., of Stoekton, Steyrmart 13479 (M). CHRISTIAN Co.: Swan Cr., s.e. of Chadwick, Fea 22991 (M). COLE Co.: near Brazito, Palmer 89227 (A, 2 near Jefferson City, Palmer 39231 (A). cooPER Co.: Bush 13654, 13893 (A). DADE Co. r Turnback, Palmer 35595 (A, M). DALLAS Co.: along Niangua R., в. of еще: Steyrmare 13737 (M). DOUGLAS Co.: between Roosevelt and Richville, Steyermark 14693 (M). NKLIN Co.: July-Aug. 1835, ic (M); Catawissa, July , 1887, Eggert (G, а 26 Greenman 3898 (M); along Meramee К., Meramec State Pk., Sullivan, Steyermark 1492 (M). GREENE Co.: near Springfield, Palmer 30148 А, M). HENRY Co.: along Grand R., n.e. of Finey, Steyermark 15964 ey HICKORY Co.: along Pomme de Terre R., near Horuitage Palmer 35983 (A, M). N Co.: near Iron- ton, Palmer 22642 (A). JASPER Co.: Carterville, Palmer 2655, 18318 T ; Carthage, Bush 10390 (A, G, М); Muddy Cr., Golden City, Palmer 4585 (A, М); near Jasper, Palmer 28931 (A, M); near Joplin: Trelease 433 (M), Palmer 22730, 23366, 26316 (A); Neck, Palmer 15723 (A); Spring R., Trelease 1116 (M) ; near Waco, Palmer 18483 (A); Webb City, Palmer 28, 927, 14291 (M). JEFFERSON Co.: De Soto, July 25, 1887, Hasse (М); s.e. of Pacific, Steyermark 1270 (M); Victoria, May 10 and July 8, 1890, Hitchcock (M). [Vor. 29 164 ANNALS OF THE MISSOURI BOTANICAL GARDEN LAWRENCE Co.: near Red Oak, Palmer 26980 (A, M). MADISON Co.: near Fredericktown, Palmer 31622, 31639 (А). MARIES Co.: Lanes Prairie, Bush 13470 (A). MCDONALD Co.: Bush 236 (G, М); Noel: Bush 4978, Palmer 4264, 5491, = 14664 (A Co.: Bagnell, Trelease 426 (M) ; near Tuscumbia, Palmer M). Mo AU Co.: between California and Jamestown, Steyermark 14758 tin MORGAN Co.: Bush 13596, 13691 (A, М); along Proctor Cr., Steyermark 13181 (M). osAGE Co.: along Maries R., w. of Westphalia, July 14, 1933, Jeffr ey (A), OZARK Co.: near Bakersfield, Palmer 32819 (A, М); ““ Bald Jesse’’ near Gainesville, fuis 34761 (A, M). PETTIS Co.: Bush 13628 (A, М). PHELPS Co.: ‘‘Slaughter Sink'' near Arlington, Palmer $ Steyermark 41399 (A, M) ; Jerome, get 388 (A, M). POLK Co.: along Pomme de Terre R.: n. of Burns, Steyermark 13630 (M), Buzzards Den, n.w. of Pleasant Hope, Steyermark 24108 (M). PULASKI Co.: Big Piney R., Trelease 429 (M). RIPLEY Co.: gn Mills, Mackenzie 392 (M). ST. CHARLES Co.: Meramec R. at St. Pauls, 25 mi. w. of St. Louis, Nov. 2, 1854, ot роо ). ST. CLAIR Co.: along Osage R.: near Osceola, Palmer 35646 (A Py, n. of fe, Steyermark 20237 (M). вт. FRANCOIS Co.: Bismarck, Palmer 18055. d (M); Flat River, Trelease 431, 432 (M). вт. Louis Co.: July 26, 1886, Eggert (A, US); Allenton: 1876-1911, Letterman (M, P, US), July 29, 1887, Eggert (M); Big Bend, Meramee R., ex Glatfelter Hb. 308 (M); near Glencoe and Valley Park, July 18, 1879, Eggert (M); Meramee Highlands: 1874, Letterman (M), July 28 and Sept. 6, 1886, Eggert (M, UT), June 25, 1904, Gleason (G), Bartram 2425 (P) ; Pleasant Grove, Bush 262 (M). SHANNON Co.: Bush 956 (М); Montier: Bush 171, 704, 1138 (M), Palmer 19390 (A, M). STODDARD Co. f Puxieo, Steyermark 11641 (M). STONE Co.: near Galena: Palmer 4641, 5772, 5860, pis, 14360, 22860, 24601, 26136 (A); '*White Rock Bluff," White R., Palmer 5857 (A, M). T о. : n Cr., Trelease 428 (M). WASHINGTON Co.: near Irondale, Palmer 30172 с, Ж, WAYNE Co.: Williamsville, Palmer 6113 (А, М). wEBSTER Co.: headwaters of James R., Engelmann 53 (M); near Niangua, Palmer 39504 (А, М.Р), E BENTON Co.: Twin Mt. and Osage R., Demaree 4563, 4614 (М); near Monte Ne, vc 24749 (A, M). BOONE Co.: Harrison, Palmer 6917 (A). CARROLL Co.: Кагуты Palmer 6330 (A, M); near Eureka Springs, Palmer 20472 (A). CLARK Co.: i^ Demaree 10017 (M); Arkadelphia, Palmer 10536 (A); Okolona, Demaree 16100 T EW Со.: Tillar, Demaree 19281 (M). FAULKNER Co.: Demaree 99 (M); Ans air yh Demaree 5294 (M). GARLAND Co.: near Hot vim Palmer 23083, 29075 (A, M). HoT SPRINGS Co.: Magnet Cave, Palmer 26565 (A). 1zARD Co.: along White R., near Calico ock, Palmer 35553 (A, M). JACKSON Co.: “Айкөлү May 5, 1881, Letterman (А, M), Palmer 35533 (A, M). LOGAN Co.: Magazine Mt., Palmer 24165 (A, M. 36 MARION Со.: bluffs of White R., Cotter, Palmer 5915, 14328 (М). MONTGOMERY Co.: bottoms of Caddo Cr., Norman, рр 9561 (M). POPE Co.: near Хоро, Merrill 88 (А, M, NY). PULASKI Co.: Little Rock: 2 Harvey (M), banks of Arkansas R., Palmer 22929 (A, M); Palarm, Demare 6 (A); Maumelle са уу Pulser 28015 (A). SALINE Co.: T + of Saline dl near Benton, Demaree 8496 (A). SEVIER Co.: Horatio, Palmer 8389 M). VAN BUREN Co.: near о. ied 24314 (A, M) ean, exact locality lacking, Hale (С, P). CADDO Parish Shreveport, July 1909, Dickson (Tu). cALCASIEU Parish: L n Charles: Apr. 13, 1915, Sargent (A), Palmer 7699 (A, M). CAMERON Parish: Cameron, Palmer 8536 (A, M). LA SALLE Parish: Stand- ard, Sept. 1, 1927, Whitehead (L). NATCHITOCHES Parish: Boe cera Palmer 7468, 8022, 8763 (A, M). ORLEANS Parish: New Orleans, 1832, Drummond 207, part (С, co- type of B. tomentosa). TANGIPAHOA Parish: Hammond, Мау 1910, Cocks p^ VERNON Parish: Leesville, Brows 6869, 6872 (L). KANSAS: CHAUTAUQUA Co.: along Middle Caney Cr., near Sedan, Palmer 41809 (A, M). CHEROKEE Co.: Black 757 (G, M) ; near Galena, Sime 20884, 21982, 25339 (A). COWLEY 1942] CLARK—REVISION OF BUMELIA 165 Co.: Arkansas City, Palmer 21244, 22072 (A). CRAWFORD Co.: near Pittsburg, Palmer 20826 (А). LA BETTE Co.: у. of Oswego, Rydberg $ Imler 317 (M). NEOSHO Co.: La- dore, Nov. 15, 1893, Hart s WILSON Co.: near Neodesha, Palmer 21160, 22030 (A). OKLAHOMA: BLAINE Co.: Canton, Andrews 20 (A). CANADIAN Co.: Devils Canyon, e. of Hinton: Little 3911 (A, Sor Hopkins 2108 (UO). CHEROKEE Co.: e. of Tahlequah, Little 477 (UO). cuocrAw Co.: Ft. Towson, Palmer 8308 (A, M). CLEVELAND Co.: near Norman: Aug. 20, 1924, Braner (A, US), Felkner 11 (UO), Staey 107 (UO). COMAN- HE Co.: Boulder Camp, Wichita Nat. Forest, n 18180 (UO). CUSTER Co.: Clin- ton Palmer 12559 (A, M). GARFIELD Co.: n.w. Enid, Brigham 11 (UO). GREER Co.: e. of Granite, Bull 248 (UO). JOHNSON Co.: etur Palmer 6416 (A, M). KIOWA N near Mountain Pk., Stevens 126115 (A, С). LE FLORE Co.: near Page: Stevens 2628 G), Palmer 20556 (A); Poteau, Palmer 8266, 8267 (A, M). MAJOR Co.: Cleo, d 1720 (С, M). MecuRTIN Co.: s. of Bethel, Hopkins $ Cross 2543 (UO). MCLAIN Co.: w. of Purcell, Hopkins 1187 (UO). MURRAY Co.: Arbuckle Mts., Hopkins 2750 (UO); Davis, Emig 658 (M); Platte Nat. Pk., near Sulphur, Merrill 908 (A), 1724 (M). MUSKOGEE Co.: Muskogee, Palmer 11198 (^ M). OKLAHOMA Co.: Oklahoma sd Palmer da 22107 (A). oTTAWA Co.: Commerce, Bush 9266, 10148 (A). ar Perkins, Oct. 3, 1896, Bogue (P) ; s.w. of Ens ater, Spears 113 (UO). ROGER MILLS Со. пеаг pet Hills, Bigelow 678 EN TULSA Co.: along Arkansas R., Bush 491 (A, G, M); near Fisher, Stevens 2960-E (G). WASHINGTON Co.: Copan, Stevens 2154 (А, С). woops Co.: s.w. of Waynoka, Jackson 340 (UO). WOODWARD Co.: near Wood- ward, Palmer 41927 (A, M). TEXAS: ANGELINA Co.: Pine Island, May 5, 1903, Reverchon (M); Jack Cr., July 25, and Haw ee Aug. 12, 1934, Boon (UT). ARCHER Co.: Diversion Lake, Cory 13227 (A). BOWIE Co.: near Texarkana, Heller $ Heller 4242 (A, M, NY, US). GREGG Co.: n. о ARE June 7, 1899, Eggert (M). DALLAS Co.: Dallas, Bush 1619 (A, M). GRAYSON Denison, Palmer 14273 (A, M). HARDEMAN Co.: n. of Chillicothe, Parks ф Cory 13368 (Tr), Cory 13369 (A). HARRIS Co.: Houston, Palmer 11927 (A, M). HARRISON Co.: Marshall, Palmer 13226 (A, M). LLANO Co.: Enchanted voa. Sept. 1, 1930, White- house (UT). K Co.: Livingston, Pama 5262 (A, M). SK Co.: Tatum, Oct. 10, io. £e t. SAN AUGUSTINE Co.: San Augustine, DE: 7892, 7894 (A,M). The most widespread element in the B. lanuginosa complex, var. oblongifolia differs from the species in its rather sparse pubescence and shorter pedicels. 2. Bumelia rigida (Gray) Small in Bull. N. Y. Bot. Gard. 1: 444. 1900; Britt. N. Am. Trees, 780, fig. 711. 1908; Wooton & Standl. in Contrib. U. S. Nat. Herb. 19: 495. 1915. Bumelia spinosa S. Wats. in Proc. Am. Acad. 18: 112. 1885, non A. DC Bumelia lanuginosa var. rigida Gray, Syn. Fl. N. Am., ed. 2, 2': 68. 1886. Bumelia pauciflora Engelm. ex Gray, Syn. Fl. N. Am., ed. 2': 68. 1886, non Roem. & Schult. Armed shrub or small tree, 6-8 m. tall; branches numerous, short, spinescent, clothed in reddish or grayish bark; branchlets, lower surface of leaves, petioles, and inflorescence densely covered with [Vor. 29 166 ANNALS OF THE MISSOURI BOTANICAL GARDEN pale tomentum; leaves numerous, coriaceous, small, obovate to cuneate-oblanceolate, 1-4 em. long, 0.5-1.5 em. broad; flowers few; pedicels stout, 2-5 mm. long; sepals oblong-elliptic to suborbicular ; eorolla-lobes suborbicular, truncate at base, margins erose; ap- pendages lanceolate to ovate-lanceolate, acute at apex, margins erose; staminodia ovate-lanceolate, obtuse at apex, margins erose; ovary cylindrical, abruptly contracted into a slender, elongated style; fruit obovoid to subglobose, 6-9 mm. long, purplish-black, short style often persistent; seed obovoid to subglobose, 3-6 mm. long, frequently narrowed at base, variegated, dark brown and pale tan. Distribution: near streams on dry, gravelly mountain slopes, south-central Texas, westward to southeastern Arizona, and ad- jacent Mexico. х аата арау : San Antonio, Clemens $ Clemens 889 (M). DUVAL Co.: San Diego, сащ m posing (A). KERR Co.: Lacey’s Ranch, Schulz 36 (US). LLANO Co.: Harris, Parks § Cory 6271 (Tr). NUECES Co.: Corpus Christi, Mar. 26, 1907, York (UT). UVALDE Co.: Uvalde, July 1879, Edw. Palmer 156 (G, TYPE, US). VALVERDE Co.: Devils R., Palmer 1136? (A, M). : Dier Cr., June 20, 1906, Wooton (US); Dog Spring, Dog Mts., Mearns РАД, pow (08у, m Bernardino R., Mearns 2545 (US). HIDALGO Со.: Дебиан УаПеу, в. of Animas, Wolf 2580 (С). ARIZONA: COCHISE 2 Camp Bowie, Lemmon 200 (G); Guadalupe Canyon, e. of ر‎ ад Peebles 11698 (US); Ft. Bowie: June 25, 1894, Fisher (US), Blumer 2298 (NY); drcum face 5382 (A, G, US); Chiricahua Mts., near Paradise: McKel- vey 605 ү" , US), y Cr., Мау 27, 1906, Holmes e Чезена Cr., ÉL 1908, Sudworth (US); sine Mts., July 21, 1894, Towmey (A, NY). A Co.: Santa Cata- lina Mts., June 16, 1881, Pringle s.n. (M, type of B. pauciflora, ie G, US); аб Cr., Tucson, Eastwood 8164 (A); Baboquivari Canyon, Peebles, Harrison $ Kearney 388 (US). MEXICO: COAHUILA: Saltillo, Palmer 232 (A). B. rigida seems to be composed of two elements. One of these may possibly be a variety of B. lanuginosa that has adapted itself to the conditions of soil and climate of the Southwest: in its pubes- cence it resembles B. lanuginosa var. albicans, while in its fruit and pedicels it looks like B. texana. However, it seems advisable not to attempt any segregation until more material is available. 3. Bumelia lacuum Small, Man. S. E. Fl. 1034. 1933. Armed evergreen shrub, 0.5-3 m. tall; branchlets, petioles, lower surface of leaves, and inflorescence densely covered with lustrous, appressed, rufous pubescence; leaves oblong-obovate to spatulate, 1-4 em. long, 1-3 em. broad, obtuse at apex, cuneate at base, mar- 1942] CLARK—REVISION OF BUMELIA 167 gins slightly revolute; petioles 1-4 mm. long; sepals and corolla- lobes oblong-elliptic; corolla-tube short, lobes scarious, appendages ovate-lanceolate, acute at apex; staminodia ovate-elliptic, obtuse at apex, margins slightly crenulate; ovary subglobose, abruptly contracted into the style; fruit oblong- to elliptic-obovoid, 8-10 mm. long; seed elliptic-obovoid, 6-8 mm. long, pale tan. Distribution: sandhills, usually in scrub, central peninsular Florida. Specimens examined: FLORIDA: exact locality unknown, Rugel 161 (M); “Е. on ", Buckley (С). BREVARD Co.: Indian R., Palmer 326 (G, M). COLLIER Co.: Caxambas Tals Simpson 578 (G). COLUMBIA Co.: е ke City, Rolfs 46 (M). DUVAL Co.: near DUE Curtiss 1762' (M). HIGHLANDS Co.: between Avon Park and Sebring: Small $ DeWinkeler 9042 (NY), Aug. 30-31, 1922, Small, Small $ DeWinkeler 10666° (NY, fr. part of TYPE), 10755 (NY); n. of Kuhlman, Apr. 25, 1921, Small $ DeWinkeler 9965 (NY, fl. part of TYPE, С); near Sebring: Harbison 29, 87 (A), pe 27438 (A, M). LAKE Co.: near Eustis: Hitchcock (M), Nash 818 (A, G, M, NY), Harbison 9 (A). ORANGE Co.: e. of porri, Baker 513 (A); edge of L. Apopka hammock, w. of MeDonald, Baker 504, 505, 506, 508, 509, 510, 511а,0,2, 512a,b,2, 514, 515, 525, 544 (A); near Plymouth, PM 38350 (A, M). OSCEOLA Co.: Kissimmee, Aug. 2, 1929, O'Neill (US). POLK Co.: near Bartow, Palmer 38366 (A), near Frostproof, Palmer 27385 (A); Winter Haven, McFar- lin 6678 (UT). SEMINOLE Co.: Oviedo, Harbison 8, 9 (A). B. lacuum is segregated from B. tenax because of its dwarfness. Also the pubescence is rufous and less lustrous. 4. Bumelia rufotomentosa Small in Bull. N. Y. Bot. Gard. 1: 440. 1900, as ruf tosa; Man. S. E. Fl. 1033. 1933. Armed shrub less than 1 m. high; branchlets and petioles densely covered with rufous tomentum ; leaves coriaceous, obovate or oval, 1.5-4 em. long, 0.5-3 em. broad, rounded to slightly retuse at apex, euneate at base, sparingly pubescent beneath; petioles very short, 1-3 mm. long; flowers few; pedicels 3-5 mm. long; sepals oblong- elliptic to suborbicular; corolla-lobes suborbicular, appendages ovate-lanceolate, obtuse at apex; staminodia ovate, acute at apex; ovary subglobose, gradually contracted into a stout style; fruit subglobose, 4-6 mm. in diameter. Distribution: in pine woods, peninsular Florida. Specimens examined: FLORIDA: ALACHUA Co.: Gainesville, Seibert 1411 (M). HILLSBOROUGH Co.: Tampa, May 1876, Garber (NY, ТУРЕ, US). ORANGE Co.: Zellwood, Baker 508a,b,c,z (А). SUMTER Co.: Sumterville, Curtiss 1764 (A, в Small, Small ф Гей inkeler 10666 distributed as ‘‘ Bumelia tenax Willd.’’ is apparently Dipholis salicifolia (L.) DC [Vor. 29 168 ANNALS OF THE MISSOURI BOTANICAL GARDEN This speeies is possibly a pubescent form of B. reclinata, but it is recognized here as a separate species because of its rufous-tomen- tose branchlets and its villous ovary. 5. Bumelia tenax (L) Willd. in L. Sp. Pl. 12: 1085. 1798; Loud. Arb. & Frut. Brit. 2: 1193, fig. 1017. 1838; DC. Prodr. 8: 190. 1844; Nutt. Sylva 3: 35, pl. 92. 1849; Sarg. Silva 5: 169, pl. 246. 1893; Small in Bull. N. Y. Bot. Gard. 1: 446. 1900. Sideroxylon tenax L. Mant. 48. 1767, non Walt. Chrysophyllum carolinense Jacq. Obs. 3: 3, pl. 54. 1768. Sideroxylon sericeum Walt. Fl. Car. 100. 1788. Sideroxylon chrysophylloides Michx. Fl. Bor. Am. 1: 123. 1803. Bumelia chrysophylloides Pursh, Fl. Am. Sept. 1: 155. 1814; P. W. Wats. Dendr. Brit. 1: 10, pl. 10. 1825. Sclerocladus tenax Raf. Sylva Tellur. 35. 1838. Sclerozus tenax Raf. Aut. Bot. 2: 73. 1840. Bumelia reclinata Chapm. Fl. Southeast. U. S. 275. 1860, non Vent. nee Torr. Lyciodes tenax O. Ktz. Rev. Gen. 2: 406. 1891. Shrub or small tree, 2-10 m. high; branchlets, petioles, lower surface of leaves and inflorescenee densely covered with a tawny pubescence; branchlets slender, with reddish-brown bark, осса- sionally armed with stout spines, 2-2.5 ст. long; leaves oblong- obovate to oblanceolate, 2-7 em. long, 0.5—3.5 em. broad, obtuse at apex, cuneate at base, margins slightly revolute, midvein promi- nent; petioles 2-9 mm. long; inflorescence many-flowered ; pedicels slender, 8-13 mm. long, slightly enlarged toward apex; sepals ob- long-obovate to subglobose; corolla-lobes oblong-elliptic to sub- globose, margins erose, appendages lanceolate, acute at apex, cren- ulate; staminodia broadly ovate, obtuse to rounded at apex, mar- gins erose; ovary narrowly ovoid, gradually contracted into an elongated style; fruit oblong-ellipsoid, 10-14 mm. long, often tipped by a persistent style; seed oblong-obovoid, 10-12 mm. long, apex tuberculate, variegated. Distribution: dry, sandy soil along sea-coast, South Carolina to Florida. Specimens examined: CAROLINA: precise data lacking, Walter s.n. (С, fragment of type of Siderorylon sericeum). SOUTH CAROLINA: BEAUFORT Co.: Bluffton, 1871-99, радвате. (А, С, M); St. Helena Ravenel (С). Isl., Harbison 10 (А). CHARLESTON Co.: Sullivan Isl., GEORGIA: CHATHAM Co.: near Savannah, Beyrich (P). GLYNN Co.: Brunswiek, Harbi- 1942] CLARK—REVISION OF BUMELIA 169 son 4, 5, 1222 (A). LIBERTY Co.: Colonels Isl, Harbison 26 (A); St. Catherines Isl., Harbison 9, 10 (А). MCINTOSH Co. Hurban s, 3,4 (A FLORIDA: exact locality RA Rugel 320 (M). BREVARD Co.: along Indian R., Tropic, Merritts Isl, N. Y. State College of Forestry, Proj. I, 8334, coll. a Rhoads (A). CLAY Co.: Magnolia Springs, Harbison 7051a (A). DUVAL Co.: Fredholm 5345 (G); near Jacksonville, Curtiss 1762, in part, 4362, 5678 (M). KE Co o.: Eustis, Nash 1662 (A, P). вт. JoHNS Co.: near Matanzas, Harbison 6 ma VOLUSIA Co.: Seabreeze, Webber 467 (M). B. tenaz is a coastal species characterized by dense, tawny seri- ceous pubescence. 6. Bumelia anomala (Sarg.) Clark, comb. nov. Bumelia lanuginosa var. anomala Sarg. in Jour. Arn. Arb. 2: 169. 1922 Small tree; branches glabrous, armed with occasional stout spines, bark yellowish-gray; leaves coriaceous, oblong-obovate to broadly elliptic, 1.5-6.5 em. long, 1-3 em. broad, rounded to obtuse at apex, obtuse at base, lower surface densely covered with lus- trous, appressed, silvery pubescence, margins slightly revolute; petioles 2-6 mm. long; inflorescence few-flowered ; pedicels 3-5 mm. long, slightly enlarged toward apex; sepals oblong-elliptic to sub- globose, rounded at apex; corolla-lobes oblong-elliptic to subglo- bose, rounded at base and apex, erose, appendages linear-lanceo- late, crenulate; staminodia broadly ovate, obtuse at apex, crenu- late; ovary broadly ovoid, abruptly contracted into an elongated style; fruit not observed. Die bation in sandy or rocky upland woods, Gainesville and Orlando, Florida. Specimens examined: FLORIDA: ALACHUA Co.: Gainesville: June 17, 1917, Harbison 47 (A, TYPE), 61, 64, 97 (A). ORANGE Co.: near Orlando, Harbison 51 (A) A close examination of this entity reveals that it is not an anoma- lous form of B. lanuginosa, but rather a distinct species, perhaps more closely related to B. tenax. It is distinguished from B. tenax by its broad leaves and silvery pubescence. In shape of leaf it re- sembles B. rufa, but its pubescence is not at all reddish. 7. Bumelia rufa Raf. New Fl. N. Am.3: 29. 1836; Aut. Bot. 2: 74. 1840. Bumelia ferruginea Nutt. Sylva 3: 34. 1849. Branchlets, petioles, lower surface of leaves and inflorescence densely covered with a tawny pubescence; leaves elliptie, about 5 em. long, about 3 em. broad, obtuse or acutish at apex, obtuse or [Vor. 29 170 ANNALS OF THE MISSOURI BOTANICAL GARDEN rounded at base; flowers glomerate or in many-flowered fascicles; pedicels stout, 3.5 mm. long; sepals reniform to oblong-elliptic, rounded at apex; corolla-tube short, lobes suborbicular, truncate at base, appendages lanceolate, acute at apex; staminodia broadly obovate, obtuse at apex; ovary ovoid, gradually contracted into a stout style; fruit not observed. Distribution: western Florida. Specimen examined: FLORIDA: ‘№, Fla."", Ware (P, TYPE). Although it is not unlikely that the type specimen represents a teratological form of the B. lanuginosa complex, convincing evi- dence for this supposition has not been found. It therefore seems advisable to maintain this species, pending further collections in УГ. Fia.” 8. Bumelia lycioides (L) Pers. Syn. 1: 237. 1805; Gaertn. f. in Gaertn. Fruct. & Sem. 3': 127, pl. 202, fig. 3. 1805; DC. Prodr. 8: 189. 1844; Nutt. Sylva 3: 31, pl. 91. 1849; Small in Bull. N. Y. Bot. Gard. 1: 444. 1900. Sideroxilon spinosum Duham. Arb. 2:260, pl. 68. 1755. Sideroxylon lycioides L. Sp. PL, ed. 2, 279. 1762. Sideroxylon decandrum L. Mant. 48. 1767. Sideroxylon laeve Walt. Fl. Car. 100. 1788. Lyciodes spinosum О. Ktz. Rev. Gen. 2: 406. 1891. Shrub or small tree, 1-8 m. high; branches clothed in gray bark and armed occasionally with short spines; leaves oblong-elliptic to oblong-obovate, 3.5-13 em. long, 1.5-5 em. broad, obtuse to acumi- nate at apex, pale green and prominently reticulate on both sur- faces ; petioles 5-15 mm. long; flowers numerous; pedicels 2-12 mm. long; perianth scarious, glabrous; sepals and corolla-lobes oblong- elliptic, appendages lanceolate to ovate-lanceolate; staminodia ob- long-ovate, obtuse at apex; ovary subglobose, rather abruptly con- tracted into an elongated style; fruit obovoid to oblong-obovoid, 8-10 mm. long; seed obovoid, 6-8 mm. long. Distribution: in moist situations, southeastern Virginia to north- ern Florida, westward to southern Illinois and Louisiana. Specimens examined: the TYPE specimen, without exact locality and date of collec- tion, is preserved in the Herbarium of the Linnean Society, London is Mee VIRGINIA: ISLE OF WIGHT Co.: near Smithfield, Oct. 1933, Stott (U CAROLINA : precise data lacking, Walter s.n. (С, fragment of type of ЗИМ laeve). NoRTH CAROLINA: Churchs Isl, McAtee 1212 (US). NEW HANOVER Co.: Wrightsville, 1942] CLARK—REVISION OF BUMELIA 171 Williamson 111 (P); near Erie how Wilmington, Harbison 34 (A). ROWAN Co.: near Salisbury, Biltmore Hb. 1689с (А). SouTH CAROLINA: ABBEVILLE Co.: ond Falls, Harbison 36 (A). BEAUFORT Co.: Bluffton, Mellichamp (A). HORRY Co.: Little В. Inlet, Harris C19598 (US). RGIA: exact locality unknown, Boykin (G, M). CAMDEN Co.: St. prit R., Harris од (US). caToosa Co.: near Ringgold, Aug. 10, 1895, Small (A, US). RICHMOND Co.: Augusta, Harbison 10 (A). STEWART Co.: along Chat ERST R., Harper 1 G, M). WALKER Co.: summit of Bluebird Gap, Pigeon Mt., Harper 365 (NY FLORIDA: near Mosquito Inlet, Curtiss 1761, in part (M). GADSDEN Co.: River Junc- tion, Harbison 140 (A). JACKSON Co.: Aspa R., near Aspalaga, July 1843, Rugel (M). JEFFERSON Co: near E Mieeosukee, Palmer 38459 (A). LAKE Co.: Wekiwa Springs, Aug. 12, 1929, O'Neill (U LABAMA: exaet Dm unknown, Drake (G), Buckley 19 (P). DALLAS Co.: Ala- bama R., Selma, Harbison 94 (A). ETOWAH Co.: n. of Attalla, July 9, 1898, Eggert (M). JACKSON Co.: Seottsboro, June 30, 1899, Earle (NY). JEFFERSON Co.: Powderly, Pal- mer 35353 (A, М); along Little Cohaba Cr., Palmer 38935 (A, M). MARENGO Co.: near Demopolis, Pune 27205 (А, M). ушсох oe June 1841, Buckley (G, M, US). MISSISSIPPI: CLARK Co.: Enterprise, Tracy 3289 (NY). HINDS Co.: s. of Jackson, E 95 г MARION Co.: Columbia, Oct. 28, 1894, Mohr (US). ох Co.: near Nashville, Biltmore Hb. 1689b (A, С, M). DYER Co.: DT poa 255 (A, G, M). FRANKLIN Co.: Cumberland Mt., Cowan, July 21, 1897, Eggert (M). MONTGOMERY Co.: bluffs of Cumberland R., Clarksville, Palmer 17579 (A, M). RUTHERFORD Co.: near Murfreesboro, Palmer 35496 (A, M) KENTUCKY: exact loeality unknown, 1842, Short b M, Tu). BALLARD Co.: near Swan Pond, McFarland, "Bhaoklett еф Plymale 43 (M). LOGAN Co.: Russellville, Palmer 17749 (A, M). LYON Co.: Eddyville, Palmer 17863, Pu 23703 (A, M); near Star Lime Works, Shacklette 440-b (С). MUHLENBERG Co.: Central City, Palmer 17731 (А, С, M). WARREN Co.: Jenning's Cr., Bowling Green, May-Aug. 1895, Price (A M) L S: ALEXANDER бс airo, Palmer 14930 (A). DIN Co.: ‘Car ve-in Rock, Palmer 15473 (A, M). РОРЕ Co.: near Ohio R., Goleonda, TUR 15376, 28776 (A, M). PULASKI Co.: Mound City: June 1861, Vasey (G, M), Palmer 15078, 16583 (A, M). LOUISIANA: W. FELICIANA Parish: Wakefield, June 1907, Cocks (A). Though infrequently collected, B. lycioides 1s à variable and widespread species. Like B. Smallit, it grows near water-courses. 8а. Bumelia lycioides var. virginiana Fernald in Rhodora 38: 439. 1936; Sarg. Silva 5: pl. 248. 1893. Shrub or small tree to height of З m.; leaves 1.5-13.5 em. long, 1-3.5 em. broad generally rounded at apex, bright to pale green; petioles stout, 2-12 mm. long; inflorescence rather few-flowered; pedicels 3-9 mm. long; fruit obovoid ; seed pale. Distribution: dry wooded slopes and margins of streams, south- eastern Virginia to Georgia; also southern Indiana. Specimens examined: VIRGINIA: ELIZABETH CITY Co.: Ft. Monroe, May 30, 1878, каны sor NORFOLK 11 Co.: Sewell’s Pt., Jun , 1872, rau d M). PRINCE GEORGE Co.: r James R L Upper Brandon, nu n wo 9393 (С); байт! s Cr., York 9 и e (QNS): PRINCESS ANNE Co.: near Third St. bridge, Great Neck: Fernald $ Griscom 4492 (С), [Vor. 29 172 ANNALS OF THE MISSOURI BOTANICAL GARDEN Fernald, Griscom & Long 4688 (С), Fernald 4 Long 4987 (С, TYPE, M); Little Neck, Fernald § Long 4988 (G, M). Мовтн CAROLINA: NEW HANOVER Co.: near Wrightsville, Biltmore Hb. 1689a (A, С, SouTH CAROLINA: BEAUFORT Co.: Bluffton, ета (А). GEORGIA: FRANKLIN Co.: near Royston, Palmer 42448 (M). INDIANA: PERRY Co.: Ohio R. bluffs, above Cannelton, Deam 16602 (A). 8b. Bumelia lycioides var. ellipsoidalis Clark, var. nov." Small tree, 8-10 m. high; leaves elliptie to oblong-elliptie, 5-12 em. long, 2-4 em. broad; petioles slender, 4—5 mm. long; inflores- cence many-flowered ; pedicels 4-8 mm. long, only slightly, if at all, enlarged toward apex; sepals broadly ovate; fruit ellipsoid, 10-15 mm. long, 6-8 mm. broad, black and fleshy ; seed 8-10 mm. long, 4-5 mm. broad, tan. Distribution: in open woods on loess hillsides, Tennessee, south to northern Alabama, Mississippi, and eastern Arkansas. Specimens examined: ера DAVIDSON Co.: Nashville: July-Sept. 1879, Gattinger s.n. (M, TYPE), ga -Aug. 1879, Gattinger (A), Curtiss 1761, in part, coll. by Gattinger (G, M), Oct. , 1887, Gattinger (A, M). sy amid CALHOUN Co.: Tarsus, Cocks (A).FRANKLIN Co.: Russellville, Sept. 1893, Mohr MISSISSIPPI: OKTIBBEHA Co.: Agricultural College, Pollard 1336 (G, M); Starkville, Tracy 2009 (A). ARKANSAS: PHILLIPS Co.: Crowley's Ridge near Helena, Palmer 26649 (A, M). This variety is characterized by its ellipsoid fruit and its limited distribution. 9. Bumelia Smallii Clark, nom. nov. Bumelia lucida Small in Bull. N. Y. Bot. Gard. 1: 443. 1900; Britt. N. Am. Trees, 778, fig. 709. 1908, non Roem. & Schult. Glabrate shrub or small tree, 2-7 m. high; branches slender, red- dish-brown, armed with short spines; leaves numerous, coriaceous, elliptic to elliptic-ovate, 2-5 em. long, 1-2 em. broad, acute to осса- sionally rounded at apex, cuneate at base, dark green and lucid above, paler and dull beneath, reticulate on both surfaces; petioles slender, 3-8 mm. long; flowers rather numerous; pedicels glabrous, slender, 3-7 mm. long; perianth searious; sepals oblong-elliptic to suborbicular; corolla-lobes oblong-elliptic, appendages ovate- lanceolate, acute at apex, erose; staminodia broadly ovate, obtuse at apex; ovary ovoid; fruit subglobose, 7-10 mm. in diameter, dark © Bumelia lycioides var. ellipsoidalis a specie fructu ellipsoideo differt. 1942] CLARK—REVISION OF BUMELIA PO bluish-black ; seed subglobose, 6-9 mm. in diameter, variegated pale tan and brown. Distribution: in low wet woods, southeastern Missouri, south to Louisiana and eastern coastal Texas. Specimens examined: uo BUTLER Co.: Neelyville: Bush 6866 (M), Sargent 6866 (A). NEW MADRID Co.: Oct. 18, 1856, Swallow (M). iint BRADLEY Co.: Warren, Demaree 18967 (M). CLAY Co.: Corning, Demaree 4235 (M). CRAIGHEAD Co.: bottoms of St. Francis R., Bono, Demaree 3532 (M). CRIT- 1 TENDEN Co.: Turrell, Demaree 3728 (M). HEMPSTEAD Co.: along Yellow Cr., near Me- Nab, Palmer 26725 (A, M). SON Co.: near Newport, Palmer 35532 (A, M). LOUISIANA: exact station mor Riddell (NY), Featherman 95 (NY), 1836, Hale (G, Age no MA Hale (NY). ARCADIA Parish: Esterwood, Brown 6287 (L). E. BATON ROUGE Parish: Baton Rouge, Apr. 3, 1899, Anders (L). FELICIANA Parish: а Car- penter 19 (NY, TYPE). ORLEANS Parish: New Orleans: Drummond 207, in part (G), near L. Ponehartrain, May 1838, Riddell (US), Apr. 16, 1846, Fendler (M); Audubon Pk., Oct. 6, 1936, Apr. 5, 1937, Penfound (Tu). OUACHITA Parish: Monroe, Brown 6041 (L). PLAQUEMINES Parish: English Turn on Mississippi R., Sept. 1879, Langlois (NY). RAPIDES Parish: Alexandria, Ball 528 (G, M). sT. LANDRY Parish: Opelousas, 1880—83, Letterman (M). ST. MARTIN Parish: near St. Martinsville, Oct. 16, 1893, Langlois (US), (M). TERREBONNE Parish: n. of Houma, Oct. 28, 1912, May 22, 1913, Wurzlow (NY). W. CARROLL Parish: Oak Grove, Copes (L). TEXAS: BRAZORIA Co.: Columbia: Bush 141, 945 (M), Brazos R., Palmer 5067, 6690 (M). CHAMBERS Co.: White’s Ranch, Tharp 3218 (UT). ERR SUN Co.: Beaumont, Palmer 12722, 13084 (A, М). The specific designation of this plant, being preoccupied by Roemer and Schultes’ binomial, must be changed. I therefore take pleasure in dedicating this species to the late Dr. John K. Small. B. Smallii is the complement of B. lycioides, inhabiting for the most part the river banks west of the Mississippl River. Its leaves are consistently smaller than in В. lycioides, are smooth and shiny, and the pedicels are gracefully cuneate. The fruit is more or less globose. 10. Bumelia reclinata (Michx.) Vent. Choix des Plantes, pl. 22. 1803; DC. Prodr. 8: 190. 1844; Small in Bull. N. Y. Bot. Gard. 16 441. 1900, non Chapm. nee Torr. Sideroxylon reclinatum Michx. Fl. Bor. Am. 1: 122. 1805. Bumelia macrocarpa Nutt. Sylva 3: 34. 1849. Bumelia lycioides var. reclinata (Vent.) Gray, Syn. Fl. N. Am. 2': 68. 1878. Bumelia microcarpa Small in Bull. N. Y. Bot. Gard. 1: 440. 1900. Decumbent or ascending shrub, 1-2 m. high; branches slightly geniculate, spinescent; leaves numerous, thin, obovate-spatulate to oblanceolate, 2-7 cm. long, 0.5-4 em. broad, rounded or retuse at [Vor. 29 174 ANNALS OF THE MISSOURI BOTANICAL GARDEN apex, euneate at base, dark green and shiny above, paler and dull beneath, rather prominently reticulate; petioles 2-4 mm. long; in- florescence few-flowered; pedicels 3-5 mm. long, enlarged toward apex; sepals oblong-obovate to suborbicular; corolla-lobes sub- orbicular, slightly erose, appendages lanceolate, slightly erose; staminodia lanceolate, not plicate, slightly erose; ovary glabrous, suborbicular to ovoid, gradually contracted into a stout style; fruit subglobose, 4—6 mm. in diameter, frequently tipped by a persistent style; seed subglobose, 3-6 mm. in diameter, brown, occasionally variegated. Distribution: in sandy soil, Georgia to Florida and Louisiana. Specimens examined: GEORGIA: exact place and date of collection unknown, Michaux s.n. (С, fragment of TYPE); complete data lacking, Nuttall s.n. (P, type of B. macrocarpa). THOMAS Co.: gen Ochlockonee R., near Thomasville, May 28-June 6, 1895, Small (NY). FLORIDA: exact моет unknown: Zugel 101 (M, ща s 24 (G, M). ALACHUA Co.: енна 104, 108 (A); Gainesville: Mar.-June arber s.n. (NY, type of B. тістосатра, US), Palmer 38429 (A, NY). COLLIER а TI y Cypress, head of Chokoloskee R., Small 7751 (NY). DADE Co.: near Homestead Ra. iain Cutler and Longview Camp, Small 4: Carter 1469 (NY); Homestead to Big Hammock рабан Small, Carter ф Small 3406 (N); Camp Jackson to Camp Longview, Small, er $ Small 3500 (NY); Paradise Key: Bessey 2 (A), Harper 110 (A), Rehd P p Royal Palm mock, Small ф Mosier 6400 (NY). DIXIE Co.: near Cross City, Palmer 27301 (А). FRANKLIN Co.: — 1896, Chapman (M). GADSDEN Co.: River Junction, Har- bison 119, 123 (A). HIGHLANDS Co.: Kissimmee Prairie, e. of L. Istokpoga, Small, Mosier $ DeWinkeler e are, US). HILLSBOROUGH Co.: Tampa, June-July 1898 Ferguson (M). JACKSON Co.: near Grand Ridge, Palmer 35266 (A, M, NY); d arbison 2, 28 (A). LAKE Co.: Eustis, Nash 1261 (С, NY). LEE Co.: n. of Bonita Springs, qc 1386 (M); Estero, Jan. 28, 1920, Ames (A); Ft. Myers: Standley 154 (А, С, M, NY), Hitchcock 199 (С, M, NY); Samville, June 6, 1912, Harshberger (NY). LEON Co.: s. of Tallahassee, Palmer 38484 (A). MANATEE Co.: Manatee, June 1878, Garber (G). MONROE Co.: Long Key: Small g Wilson 1852 (NY, US), Small $ Carter 2912, 2986, 2987, 2988, 2988a (NY); Flamingo, Cape Sable Region, Small, € ф De- Winkeler 11537 (G, NY, ). OKEECHOBEE Co.: Okeechobee Prairie, e. of F gg Small, Mosier ф DeWinkeler 10944 (G, M, NY). oscEOLA Co.: Fredholm 5832 (С, NY). SARASOTA Co.: s. of Venice, Seibert 1392 (M). SUWANNEE Co.: banks of oa R., Branford, Curtiss 6656 (A, G, M, NY). TAYLOR Co.: near de: Springs, Palmer 88471 (A, M, NY); ‘ee Harbison 3 (A). VOLUSIA Co.: near Mosquito Inlet, Curtiss ie in part (A, G NY); near Seville, Curtiss 6690 (А, G, M, NY); edge of causeway, n. of St. sole E beide. Baker 516 (A). WAKULLA Co.: St. Marks, Harbi- son 1503 (A). LOUISIANA: exact locality unknown, Hale (G, NY). 11. Bumelia cassinifolia Small in Bull. N. Y. Bot. Gard. 1: 442. 1900; Britt. N. Am. Trees, 778, fig. 708. 1908. Glabrous shrub or small tree, 5-10 m. high; branches slender, frequently armed with short, stout spines, clothed in reddish-brown 1942] CLARK—REVISION OF BUMELIA 170 bark; leaves thinnish, oblong-obovate, 2.5-10 em. long, 1-4.5 em. broad, acute to rounded at apex, cuneate at base, margins slightly revolute, dark green and reticulate above, paler and puberulent be- neath with midvein and secondary nerves prominent; petioles slen- der, 6-8 mm. long; inflorescence few-flowered, glabrous; pedicels slender, 4-11 mm. long, slightly enlarged toward apex; sepals and corolla-lobes suborbieular, appendages lanceolate or ovate-lance- olate, obtuse to acuminate at apex, erose; staminodia broadly lanceolate or ovate, obtuse at apex; ovary ovoid, gradually con- traeted into an elongated style; fruit subglobose, 8-12 mm. in di- ameter; seed subglobose to broadly ovoid, 6-10 mm. long. Distribution: in molst, sandy soil, southwestern Georgia to Louisiana. Specimens examined: GEORGIA: DECATUR Co.: along Flint R., near Bainbridge, Harbison 1357, 1358 (A). FLORIDA: JACKSON Co.: n. of а, Wiegand 4 Manning 2487 (G). LOUISIANA: ST. LANDRY Parish: alt. 60 ft., Opelousas, May.-Aug. 11, 1883, Letterman 287 (NY, TYPE) B. cassinifolia is a rare plant having the foliage characters of a very vigorous В. reclinata and the fruiting characters of B. Smallu. It may possibly prove to be of hybrid origin. 12. Bumelia texana Buckl. in Bull. Torr. Bot. Club 10: 91. 1883, as Texana; Small in Bull. N. Y. Bot. Gard. 1: 443. 1900. Bumelia reclinata Torr. Bot. Mex. Bound. Surv. 109. 1859, pro parte, non Vent. nec Chapm. Bumelia monticola Buckl. in Bull. Torr. Bot. Club 10: 91. 1883; Sarg. Man., ed. 2, 814, fig. 725. 1922. Armed shrub or small tree, 1-7 m. high, with reddish-brown bark ; lower surface of leaves, petioles, and inflorescence glabrescent; leaves coriaceous, generally obovate-spatulate, 1-7 em. long, 0.5-8 em. broad, rounded to retuse or even apiculate at apex, often slight- ly revolute, obtuse to cuneate at base, prominently reticulate and pale green on both surfaces; petioles slender, 3-10 mm. long; in- florescence few-flowered; pedicels 1-6 mm. long, enlarged toward apex; sepals subreniform to suborbicular; corolla-lobes oblong- elliptic, appendages ovate-lanceolate; staminodia ovate-lanceolate, obtuse to rounded at apex; ovary subglobose, abruptly contracted into an elongated style; fruit obovoid, 8-10 mm. long, often tipped by a persistent style; seed oblong-ellipsoid, 6-8 mm. long, varie- gated, red-brown and tan. [Vor. 29 176 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: in dry limestone soil, Arbuckle Mountains in Okla- homa, and west-central Texas. Specimens examined: OKLAHOMA: MURRAY Co.: Arbuckle Mts., Clark ф Williams 510 (M). TEXAS: mountains near lower erossing of Pecos R., W. Tex., 1875, Buckley (NY, TYPE) ; С. Tex.-N. Mex.’’, Wright 422 (С, NY, US); Live Oak Cr.: Wright 1434 (А, С, NY), Bigelow s. n. (NY, tyne of B. reclinata Torr.). BEXAR Co.: Jermy 83 (NY), 117 (M); Leon Springs, Clemens $ Clemens 891, 893 (M); Blanco Rd. at Cibolo Cr. crossing, n. of San Antonio, Metz 419 (NY). BLANCO Co.: Blaneo: Reverchon 1555 (A, M, US), Cory 15629 (A). BREWSTER Co.: Dog Gap, Santiago Mts.: Apr. 14, 1936, Cory (A), Parks $ Cory 18702 (Tr.). BROWN Co.: near Brownwood, Palmer 29539, 29567 (A). BURNET Co.: Burnet, Palmer 10257 (A, M). CALLAHAN Co.: Baird, Palmer 13678, 13679 (A, M). COLEMAN Co.: Santa дъба Palmer 10386 (A, M). соке Co.: Bronte, Palmer 10355 (A, M). сомль Co.: New Braunfels, Lindheimer 199 (M); ‘‘Devil’s Backbone,’’ Palmer 12187, 12207 (A); Fischer's Store, Palmer 12208, 12208a (A). соксно Co.: n.w. of Paint Rock: Cory 9761, 9763 I Parks $ Cory 9762 (Tr). CROCKETT Co.: Reverchon 97 (6. EDWARDS Co.: Palmer 10975 (A). EL PASO Co.: mountains, 1875, Buckley s.n. (NY, type of B. monticola). GILLESPIE Co.: Bear Mt.: Cory 12925, 12926 (A), Parks ф Cory 12926 (Tr); Fredericksburg, Palmer 10057 (A, M). HARDIN Co.: s. of Wee E о 11279 (A). HOWARD Co.: ний м. — 12483, 13066 (A, M). ON Mer ‚ Palmer 12426 (A, M). к , Palmer 9840, 10818, dE bero pi 18648 (A, M); eA fuc bud Do Pre KERR Co.: Kerrville: Heller 1938 (NY, US), Mackenson 5 (A), Palmer 9937, 12210 (A, M); L wid 's Raneh, Palmer 10007, 11226, 12229a (A). KIMBLE Co.: Junetion, Pane 109 M). MENARD Co.: Menard, Palmer 11888, 11899 a MITCHELL Co.: Colorado: Е 9, 1900, Eggert (M), Palmer 13787 (A, M). NOLAN Co.: Sweetwater: Palmer 12431, 18051, 14522, 14528, 88978 P M), Aug. 3, 1934, Barkley (M, UO). PALO PINTO Co.: Strawn, Palmer 14246 € 1, US). Pecos Co.: near Sheffield, Ferris $ Duncan 2919 Ба REAL о.: Barksdal aw 18525 (А, x ird inr 10148 (A, M). SAN SABA Co.: San Saba, Palme r 11810 (A, M). в N Co.: head of Llano R, between мие and Sawyer Spring, Eggleston 16729 Un в.е. ^ اة‎ Cory 15315 (A), Parks ф Cory 15314 (Tr). TERRELL Co.: near Sanderson, Palmer 33442 (A). TOM GREEN Co.: banks of Concho R., San Angelo: Reverchon 3868 (M), Palmer 10322, 10860 (M). UVALDE Co.: bluffs of Frio R., Concan, Palmer 10203, 11545 (А, M); Uvalde, Palmer 11321, 11347 (A); Leona R., Parks § Cory 23871 (Tr). VALVERDE Co.: Comstock, Palmer 11063 (A); near Shumla, Palmer 38489 (A, M, NY). A glabrous or glabrescent shrub or small tree, B. texana has leaves which vary from long and spatulate to short and obovate. There may possibly be two elements here; nevertheless B. texana and B. monticola are conspecific. 13. Bumelia megacocca Small in Bull. N. Y. Bot. Gard. 1: 441. 1900, as megococca. Evergreen, glabrous, thorny shrub with spreading or procum- bent branches and with pale bark in angular, eorky ridges broken by numerous lenticels; leaves coriaceous, obovate or oblong-ob- lanceolate, 1—3.5 em. long, 0.5-1.2 em. broad, rounded or retuse at apex, deep green and somewhat shiny above, paler and prominent- 1942] CLARK—REVISION OF BUMELIA IU ly reticulate beneath, slightly revolute, short-petiolate; pedicels stout, 1-3 mm. long; flowers not observed; fruit subglobose, 10-13 mm. in diameter, black; seed 9-10 mm. in diameter, pale, smooth and shiny, variegated. Distribution: in sandy soil, Florida. Specimen examined: FLORIDA: HILLSBOROUGH Co.: Tampa, Oct. 1877, Garber s.n. (NY, fragment of TYPE). The only available material is a fragmentary fruiting specimen which seems to be a Витейа. 14. Bumelia angustifolia Nutt. Sylva 3: 98, pl. 93. 1849; Sarg. Silva 5: 175, pl. 249. 1893; Small in Bull. N. Y. Bot. Gard. 1: 439. 1900. Bumelia reclinata Torr. Bot. Mex. Bound. Surv. 109. 1859, pro parte, non Vent. nec Chapm. Bumelia cuneata Gray, Syn. Fl. №. Am. 21: 68. 1878, non Sw. Lyciodes angustifolium О. Ktz. Rev. Gen. 2: 406. 1891. Витейа parvifolia Chapm. Fl. Southeast. U. S., ed. 3, 295. 1897, non A. DC. Bumelia Schotti Britt. N. Am. Trees, 777. 1908. Glabrous shrub or small tree, 2-8 m. high; branchlets unarmed and slender or spinescent and stout; leaves numerous, variable, persistent, coriaceous, narrowly oblanceolate-spatulate to obovate, 1-1.5 em. long, 0.5-2 em. broad, rounded to acutish at apex, cuneate, pale green or grayish and smooth above, brighter green and ob- scurely reticulate beneath, margins slightly revolute; petioles 1-10 mm. long; inflorescence few- to many-flowered; pedicels 3-11 mm. long; sepals oblong-elliptic, erose, appendages ovate-lanceolate, acute at apex, erose; staminodia ovate, acute at apex, erose; ovary ovoid, gradually contracted into a slender, elongated style; fruit oblong- to ellipsoid-cylindrical, 8-10 mm. long, purplish-black, fleshy, usually tipped by a long, persistent style; seed oblong-ellip- soid, 6-8 mm. long, narrowed at base, pale. Distribution: in sandy soil, Florida, Texas and adjacent Mexico, also in Bahama Islands. Specimens examined: FLORIDA: BREVARD Co.: Merritts Isl, Harris C19999 (US). CHARLOTTE Co.: Воса- grande, Jan.-Feb. 1920, Ames (A). COLLIER Co.: Ten Thousand Isls., Simpson 364 (G). DADE Co. o.: Planter, Key Largo, Eaton 444 (A); lower end of Old Rhoades Key, Apr. I 1911, Simpson (A). HILLSBOROUGH Co.: Tampa Bay: Leavenworth (G, NY); Paradise Key, Maxon 10926 (US). LEE Co.: Coconut, Moldenke 5780 (NY); Ft. Myers, Chapman 14 (US); Marco, Standley 12808 (US); Sanibel Isl, Apr. 10, 1886, Sargent (A). LEVY [Vor. 29 178 ANNALS OF THE MISSOURI BOTANICAL GARDEN Co.: Cedar Keys, Garber 25 (G, US). MANATEE Co.: Terra Ceia Isl., peo 276 (G, US). MONROE Co.: Bahia Honda Key, ae 1881, Curtiss 1765 (G, M, US); s. border of everglades, Jan, 1882, Curtiss 1765 (A M, US); ; Big Pine Key, Seibert tie. 1311 (M); e. of Cape Sable, Nov. 1912, Dinda (A); ; Key West: Blodgett s.n. (NY, part of TYPE, та Chapman s.n. (NY, co-type of B. parvifolia), € 86, 8? (G), 132, 174 (NY), 8, I Sargent (A, С), Britton 508, 509 (NY), Small ф Small 4909 (С), Seibert и (M); Lignum-vitae Key, Apr. 1911, Simpson (A); Long Key, Curtiss 503 (A); Marquesas Key, Apr. 12, 1886, Sargent (A); Palm Key near Panini, Howell 1146 (A); Sugarloaf Key, Pennell 9581 (M, NY); Upper Metacombe Key, Nov. 15, Curtiss (A). TEXAS: Torrecillas, Griffiths 6424 (M). ARANSAS Co.: Copano Bay, Tharp 1550, 1664 (US, UT). ATASCOSA Co.: Schulz 108, 482 (US); Campbelton, Palmer 11230 (А, С). CALHOUN Co.: Indianola: Cory 11617 (A), Parks § Cory 11617 (Tr); Magnolia Beach, May 22, 1930, Tharp (UT). CAMERON Co.: Brownsville: Hanson 451 (US), Runyon 252 я Козе у Russell 24190 (A, US), 24190a (NY); Rio Hondo, Chandler 7088 (G, M, NY); Santa Maria, May 13, 1889, Tucker (A, US). иммат Co.: El Jardin, Small $ Wherry 11889 (NY). FRIO Co.: Dilley, Пес. 16, 1932, Parks (Tr). HIDALGO Co.: n.w. of Edinburg, Clover 1599 (UT); La Joya, Mar. 2, 1940, Parks (Tr); near Mis- : Hanson 336 (С, M, NY, US, UT), Clover 181 (M), Walker 17819 Pei — Clover 1372 (A). Jackson Co.: E. Karankawa Pt., Tharp 1419 (UT). SALLE Co. near Encinal, May 1, 1882, Buckley (A); near Nueces, Aug. 1881, Pdl (A, NY), Oct. 1881, Buckley (US). MATAGORDA D s w. of gulf: Cory 11555 (A), Parks $ Cory 11555 (Тт). NUECES Co.: п.м. of eite Christi, Cory 17076 (A); Osa Bay: Cory 20566 (A), Parks $ Cory 20565 (Tr). STARR Co. of gessi m o 5936 (UT); near Rio Grande City, Sept. 10-11, a Phe (UT). B Co. redo: Rose 18047 (NY, US), Apr. 12, 1936, Penfou à (te), June en MAR ee "NY, type of B. Schottii); s. of Milo: Cory 16901 ү Parks § Cory 16900 (Tr MEXICO: NUEVO LEON: near Monterrey, Pringle 2787 (A). BAHAMA ISLANDS: Hog Isl, New Providence, Wilson 8317 (M). A distinct, though variable species, B. angustifolia is easily rec- ognized by its entirely glabrous foliage. DOUBTFUL AND EXCLUDED SPECIES Bumelia ambigua Ten. Sem. Hort. Neap. 1827. Bumelia confertiflora Nutt. in Fras. Cat. No. 12. 1813, nomen nudum. Bumelia denticulata Raf. New Fl. N. Am. 3: 29. 1836. Rafi- nesque later stated, in Aut. Bot. 2: 74. 1840, that this is ‘‘certainly no Bumelia.’’ Bumelia lucida Roem. & Schult. Syst. 4: 499. 1819 = SIDEROXYLON LUCIDUM Soland. ex Lam., fide DC. Prodr. 8: 194. 1844. Bumelia pubescens Ten. Sem. Hort. Neap. 1827. Bumelia serrata Pursh, Fl. Am. Sept. 1: 155. 1814 = Prunus CAROLINIANA Ait., fide Nutt. Gen. 1: 136. 1818; S. Wats. Bibl. Ind. N. Am. Bot. 304. 1878. Bumelia serrulata Raf. New Fl. N. Am. 3: 29. 1836. 1942] CLARK—REVISION OF BUMELIA 179 Bumelia syderoxyloides Hort. ex Lavallée, Arb. Segrez. 160. 1877, nomen nudum. Витейа undulata Raf. New Fl. №. Am. 3: 28. 1336. List or COLLECTORS AND EXXSICCATAE Collector ’s numbers are printed in italics; Numerals in parentheses represent numbers assigned to species are indicated by a dash. in this revision. Ames, J. S. — (10); —, —, —, — (14) Anders, W. S Jr. — (9). Andrews, D. 20 (1b). Baker, C. H. 504, 505, 506, 507, 508, 509, 510, 511a,b,z, 512a,b,2, 513, 514, 515, 503 (4); 516 (10). 9 = (оў 2425 (1b). "s 2 (10). и Bigelow, F M. 678 (1b); — (12). 1688 (1); 1689b,c (8); Blodgett, Lh E: 1b 218, 219 (10); 508, 509 Brown, n A. 6869, 6872 (1b); 6041, 6287 (9). Buckley, S. B. —, — Don (1а); ГЕТ (2); mio 19, А а е ща И =: — (14). m R. 248 (1b). Bush, B. F. 1, 2, 3, 210, 226, 802, 877, B. 1524 (1a); 171, 182, 236, 262, 491, 704, 905, 956, 1138, 1619, 3130, 4978, 9266, 10148, 10390, 10390a, 13470, 1 13628, 13654, 13691, 13893 (1b); 141, 945, 6866 (9). Cabell, P. H. — (1). Carpenter, W. M. 19 (9). Chandler, H. P. 7088 (14). or, when specimens are not numbered, they Chapman, A. W. — (1); — (10); 14, — (14). TS J. W., Jr. — (8a). . & H. B. Parks. 567, 574 (1a). c J. W. Williams. 510 (12). Clemens, Mr. & Mrs. J. 889 (2); 891, 893 1 (12). Clover, E. U. 181, 1372, 1599 (14). Coeks, R. S. — (1); — (1b); — (8); — (8b). Copes, Mrs. H., Jr. — (9). Cory, p 11618 (1a); 13227, 13369 CB s 9761, 9763, 11278, 11279, 12925, 12926, 15815, 15629, 19348, (12) ; 11555, 11617, 16901, 17076, 20566 (14). Cutler, H. C. 858 (1a). Curtiss, A. H. 35, Tu in part (1); 1762’ (3); 1764, — (4); 1762, in part, 4362, 5678 (5); 1761, in LO (8); — (8a); 1761, in part (8b); 1761, in part, 6656, 6690 (10); 508, 1765, 1765, — (14) Deam, C. C. 16602 (8a). 99, 4563, 4614, 5294, 8496, (1b); 3532, 3728, 4235, 18967 (9) Demetrio, C. H. 47 (1b). Dickson, D. — (1b). Dixon, В. А. 403 (1а). Drake, D. — (8). Drummond, T. 207, in part (1b); 207, in part (9). Earle, F. 8. — (8) Eastwood, A. 8164 (2). Eggi AsO an s ues e жог ›); ر‎ (8); — (12 Eggleston, W. и 16729 (12). Emig, W. H. 658 (1b). Engelmann, E- 58, — (1b). 180 Featherman, ч H (9). Felkner, W 1 (1b). eine, А. — ү ‚М. — (10). Fernald, M. L. & L. Griseom. —— & ng. 4688 4492 (8a). , (8a & B. odd. 4987, 4988, 9393 (8a). Ferris, R. S. & C, D. Duncan. 2919 (12). Fisher, A. E — (2 Fisher, G. — (1a); 4 (10). Fr азоби, T 5345 ДА, 9832 (10). Garber, A. P. — (13); 25 — (4); 24, —, — (10); (14). Gattinger, A. —, — (8b). Gillespie, J. W. 4956 (1). Glatfelter Hb. 308 (1b). Gleason, H. A. — (1b). Gree . M. 8898 (1b). Griffiths, D. 6424 (14). Grimes, E. J. 4269 (8a). Hal ,J. = С b); —, — (9; — (10). Hall E. 394 (la). Hanson, H. C. 336, 451 (14). Harbison, T. С. 3, 6, 7, 12, 14, 14, 15, 16, 23, 27, 29, 96, 829, 1095, 1111, 1127, 1216 ; 2, 8, 4, 4, 5, 6, , 1222, 7051a (5); 47, 51, 61, 64, 97 (6); 10, 34, 36, 94, 95, 140 (8); 2, 3, 28, 104, 108, 119, 123, 1503 (10); 1357, 1858 (11). Harper, R. M. 1153 (1); 365, 1100 (8); 110 (10). arris, J. А. С19598, C21149A (8); С19999 (14). Harshberger, J. W. — (10). Hart, F. P b). Harvey, F. L. — (1b) se, H. E. (1b). Heller, A. A. 1938 (12). & E. G. Heller. 4242 (1b). Hiteheoek, A. S (10). Holmes, J. S. — (2). Hopkins, M. 1187, 2108, 2750 (1b). & а ross. 2543 (1b). Howell, A. H. 1146 (14). Jaek, J. G. 3026 (1). Jackson, H. H. I. 340 (1b). Jeffrey, A. A. 30? (1b). ANNALS OF THE MISSOURI BOTANICAL GARDEN 1b); — (3); 199 [Vor. 29 Jeffrey, L. — (1b). Jermy, G. 83, 117 (12). Kellogg, J. H. 388 (1b). anglois, A. B. 45, — (9). M. — (14) Leavenworth, Lemmon, J. G. 200 (2). — G. W. —, —, — (1b); — (9); ? (11). амат, F. 90, 269 or 979, —, — (1а); 9 (12). RT A 477 (1b). Mae PS B. — (1a); 5 an. Macken ак OW. р 1212 605 À n 2351, 2356, 2545 (2). Mellichamp, J. H. — (5); — (8); — же ). Merrill, G. M. 88, 908, 1724 (1b). Mets, Sister M. C. 419 (12). Meyer, T. (1b). Michaux, A. — (1); 10). Mohr, C. 47 (1); 12, 623 (la); — (8); — (8b). Moldenke, H. N. 5780 (14). Nash, С. V. 2167 (1); 818 (3); 1662 (5); 1261 (10). Nuttall, T. — (1b); — (10). N. Y. State College of Forestry, Proj. I. 269 (1a); 8334 (5). O’Neill, H. — (3); — (8). Palmer, E. J. 35301, ии 38653 (1); ibat 4943, 5050, 6741, 4, 9209, 9734, TOi 14. eee Fue AERE 14176, 14266 (1a); 28, 927, 2655, 4264, 8 1 11927, 12559, 13226, 14278, Е unt 14360, 1942] CLARK—REVISION OF BUMELIA 14610, 14660, 14664, 15723, 18055, 18318, 18483, 19390, 20472, 20556, 20826, 20884, 23366, 24165, 24314, 24601, 24749, 25339, 26136, 26316, 26565, 26980, 28931, 29075, 30067, 30148, 30172, 31622, 31639, 32819, 34761, 35533, 35553, 35595, 35646, 35956, 35983, 36769, 39220, 39227, 39231, 39504, 41809, 41927 (1b); 11367 (2); 27385, 27438, 38350, 38366 (3); 14930, 15078, 15376, 15473, 16583, 17263, 17579, 17731, (8a); 26649 (8b); 5067, 6690, 12722, 13084, 26725, dd (9); 27301, 35266, 38429, 35471, 384 40, 9937, 14528, 29539, 29567, 88442, 33489, 33978 са 11230 (14). . A. Steyermark. 41399 (1b). mu cr 156, 232 (2); 326 (3). Parker, C. M. 4791 (1a) Parks, H. B. — (1a) ; —, — (14). . L. Cory. 7791 (1a) ; 18868 (1b); 6271 (2); 9762, 12926, 15314, 18702, 23871 (12); 11555, 11617, 16900, 20565 (14). Peebles, R. H. 11698 (2). ‚ а. J. Harrison & T. H. Kearney. 388 (2). Penfound, W. T. — (1); — (9); — (14). Pennell, F. W. 9581 (14). Pollard, C. L. 1336 (8b). Price, 8. — (8). Pringle, C. в. — (2); 2787 (14). Ravenel, H. W. — (5) Rehder, A. 898 (10). Reverchon, J. 384, 590, 3194, а); —, — A 97, A ps (12). in — (9). ) D E 24190, 24190a (14). 181 Rugel, F. 161 (3); ee » — (8); 101 (10) ; 86, 87, 132, 174 (14) Runyon, R. 252 (14). Ruth, A. 202 (1a). Rydberg, P. A. & Е. Imler. 317 (1b). Sargent, C. S. (1b) ; 6866 (9); Saurman, B. F. — (1). Schostag, E. L. 3002 (la). Schott, A. 69% (14). Sehulz, E. D. 49, 2434 (1a) ; 36 (2); 108, 482 (14). Seibert, R. J. 1412 , 1412a (1); 1411 (4); 1386, 1392 (10); 1236, 1276, 1311 (14). Shaeklette, H. T. 440-b (8) Short, C. W. (8) Shreve, F. 5382 (2) Simpson, C. T. —, —, — (14). Simpson, J. Н. 278 (3); 276, 364 (14). Small, J. K. — (1); — (8); 7761, — (10). с J. J. Carter. 1469, 2912, 2986, 2987, 2988, 2988a. (10). — —, —— & G. К. Small. —— — & J. B. DeWinkeler. 3406, 8500 (10). 9042, 9965 (3). — — & С. A. Mosier. 6400 ———, ——— & J. B. DeWinkeler. 10897, 10944 (10). —— & E. W. Small. 4909 (14). , K. W. Small & J. B. DeWinkeler. 10755 (3). —— —, J. W. Small & J. B. DeWinkeler. 10666 (3); 11537 (10). — — & E. T. Wherry. 11889 be & P. Wilson. 1852 (10) Spears, W. ~ Eo (1b). сасу, НТ. 2 (15). Standley, J. E у (10: Standley, P. C. 12808 (14). Stevens, G. W. 126114, 1720, 2154, 2628, 2960-E (1b). Steyermark, J. A. 7270, 1492, 11641, 11888, 13181, 18479, E 18737, 14693, 14758, 15964, 20237, 22991, 24108, 26076 (1b). Stott, A — (8). Sudworth, G. B. Swallow, G. C. — (la); — (2). = (0) Tharp, B. C. 1424, 1431, ү 3218 (9); 1419, 1550, 1664, 5936, c (4 182 Toumey, J. W. — (2). re S. M. 7464 (1); 3289 (8); 2009 W. 426, 427, 428, 429, 430, 431,‏ ا (1b)‏ 1117 ,1116 ,433 ,432 Tucker, J. G. — (14 Vasey, G. — (8). Walker, Mrs. E. J. 17819 (14). Walter, T. — (5); — (8). Ward, L. F. — (1a) Ware, N. A. — (7). Wiegand, K. M [Vor. 29, 1942] ANNALS OF THE MISSOURI BOTANICAL GARDEN Webber, H. J. 467 (5). Whitehead, C. R. — (1b). hitehouse, E. — i & W. E. Manning. 2486 (1); 2487 (11). Williamson, Сй. АА (B. Wright, с. 422, 1434 (12). Wurzlow, Е. C. ) York, H. H. — (2). INDEX TO SPECIES New names, varieties and combinations are printed in bold-face type, synonyms are in italics, while valid names appear in Roman type Page Bumelia: ETENE IREE E E A ET 178 И РГ 169 о. ie ЗАПОР 177 arachnoidea ..................... 163 ре. E is bass bo ТОК 162 оз езда аа ERRARE RECO 174 chrysophylloides ................. 168 БОТОВ о.е екъ оне 178 И T ГГ 177 наи Solace varsa aa 178 ONS 66550 0 00 0h ens a 169 О TT TT 166 Кассе ве веба 160 РРО О 162 ЧАР, ЭНОШ 2.06... tr bp 169 var. EG rer REVUES 163 ТЪКАНИ еее pee 1 ома Se OTE IR Lee ка 172, 178 ac Ir CI ъв ала Дл ОТИ 170 var. ellipsoidalis rere eee cr 172 WOT: TOREM асове ry vices «ss 173 ES ПРИЦЕЛА ‚р, у» 171 macro УТЕ ЕТ 173 а ....................› 176 Ob S О ада 173 LSU TITLET ала пее 175 O OMG. Sili eser waste e's PR 163 е ЗАРРАИ 177 ЛОР ....................... 165 а ....................... 173 В ааа 173 о. 168, 175, id om ж жж ж o m ж ж ж ж 9] 9 o9 o9 9 ж 9 э ж э n э ө Page he uix dH me cers FETA ee yee re 169 rufotomentosa .......... eot 167 ПОЛОВИ ео E T E ото свака 177 БИТИЕ. РЕСУРСИ C E 178 VOTER оен сина те EE 178 UM Lois ТОЧИРИ 172 м, 5. бика ди ална Ta 165 fibrosis rrr puro 179 LEXICO ORO аа ки 168 Ces VoM ania a feo e ou 175 TORIA. Veo EX X cnn Vra n 163 ONES uite e руни өз 179 Praha НН COROUMNONEE civinan ê 168 ludoviociamum .................... 161 Ly es: angustifolium ................... 177 lamuginosum oc ас ьоевовонврввво 161 ПОКОРНО LU SSA aha FA VAR еа 170 ПИТИ ГАЛИН RUM S CU e 168 Prunus caroliniana ................. 178 Sclerocladus tenaz ................. 168 BSoelerosus Темава... 168 Biderowilon v ал 170 xylon Falc atqiie E bie Rak a ba Rare V aom 168 СОФИЯ curru kx nt» kun) a v 170 ШЕВА а 170 lanuginosum .................... 161 ID Т, ЕО 178 МИДА аи sees кан Е 170 reoclinalum аа ое a а" a re era LU E TT. 168 gui es ave vagy Hees DO ORI 161, 168 MONOGRAPH OF MALVAVISCUS! ROBERT WALTER SCHERY Research Assistant, Missouri Botanical Garden Instructor in Botany in the Henry Shaw School of Botany of Washington University INTRODUCTION In the plant kingdom there are certain families which are quite distinetive but with very indefinite generie bounds, the Malvaceae being one such family. For example, two species have been in- cluded in Malvaviscus which intergrade with Рагота and which would better be placed in the latter genus. The same is true in other genera of the family as Kearney? found to be the case in Sphaeral- cea. And when a taxonomist comes to consider specifie and sub- specifie delimitation within Malvaviscus, he finds himself in the midst of an unmanageable and indefinable maze of incipient taxo- nomic entities. This can be attested to by the fact that no compre- hensive work has been done in the genus since its establishment almost 200 years ago. The situation is exemplified by the Malvaviscus arboreus com- plex where there is no constant character of taxonomic value. There is tremendous variation, even with separate branches on the same plant. In Panama, within an area of a few hundred meters, plants with 2, 3, 4, or 5 calyx lobes were found; with short or long involucre lobes; with exeurrent rows of hair on the stem or without hair; ete. As a result of such manifest variation, innumerable species have been deseribed, with descriptions based on a single specimen. If this were to be the general practice there would be almost as many species of Malvaviscus as there are specimens, since almost every specimen collected differs from all others at least to some slight degree. Thus in a taxonomic treatment of the genus two alternatives exist: excessive splitting into innumerable indistinet and undefinable * species," or lumping into few well-marked species with varietal classification for those groups which are distinct at their extremes but do intergrade with all or most other groups. 1 An investigation carried out at the Missouri Botanical Garden in the d Lab- oratory 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 ied ophy. ? Kearney, T. H., in Univ. Calif, Pub. Bot. 19: 1-128, pl. 1-12, 1 fig. 1935. Issued September 18, 1942. (183) [Vor. 29 184 ANNALS OF THE MISSOURI BOTANICAL GARDEN For the conscientious taxonomist there is no middle course in this genus, for if these intergrading types are to be considered as worthy of specific rank, surely the ‘‘good’’ or distinct species, ob- viously of higher degree difference even to the most casual ob- server, merit generic rank. It is my opinion that the second or ‘lumping’’ alternative is the only acceptable one, both from a scientific and a practical standpoint. History oF THE Genus Being a showy and conspicuous plant, Malvaviscus has been recognized, under one name or another, for centuries. One of the first names under which this genus was included was Alcea of Plukenet. Sloane, Tournefort, Plumier, and others placed it in Malva, and Linnaeus in Hibiscus. Earlier, Hernandez suggested a close relationship with Althaea, but used the Aztec name ** Atlat Zopillin’’ in his writing; however, Petiver in 1704 used Althaea in a polynomial deseription for a species of Malvaviscus (fide Dillenius). In 1732 Dillenius proposed the name Malvaviscus as distinct from Malva, but still used it in a polynomial form as: ‘‘Malvaviscus arborescens, flore miniato clauso." He recognized that this new name was synonymous with Alcea (in part) of Plukenet, Petiver's ** Althaea americana coccinea, flore clauso,” and the Aztec ** Atlat Zopillin."' Although Linnaeus in 1753 included Malvaviscus in Hibiscus, Adanson in 1763 and Cavanilles in 1787 accepted Malvaviscus as a valid genus in the binomial sense. But in 1788 Swartz proposed to rename the genus Achania, a name recognized only by a few authors of the early 19th eentury. In 1824 A. P. de Candolle divided Malvaviscus into two sections, one of which he called Achania, the other, Anotea; the latter was raised to generic rank by Kunth in 1846, but evidently it should be included in Pavonia. During the last century Malvaviscus has been generally recognized as a valid genus synonymous with the section Achania of de Candolle and doubtfully synonymous with de Candolle's other section, Anotea. Previous Work ох THE MALVACEAE In the Malvaceae, as a whole, little recent taxonomic work has been done. Kearney? monographed the North American species * Joe, eit. 1935. 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 185 of Sphaeralcea. Roush* monographed Sidalcea and also published a synopsis of Robinsonella.^ Earlier in the century Fries? pre- sented a monographie treatment of Wissadula and Pseudabutilon, Hill? published on the acaulescent species of Malvastrum, and Hochreutiner? offered a monograph of Anoda. In 1907 Watt? pub- lished a book which included a revision of Gossypium, and in 1900 appeared Hochreutiner's revision of Hibiscus.!? However, previous to the turn of the century, monographie treat- ment of genera in the Malvaceae seems scarcely to have been at- tempted, although numerous non-monographie publieations had appeared, some of first-rank importance. Cavanilles published on Sida; Сагске on Pavonia, Abutilon, Malvastrum, ete. ; Schumann and Gürcke on Malvaceae in Martius’ ‘Flora Brasiliensis’ and else- where; E. G. Baker on a synopsis of the Malveae; Bentham on a key to the Malvaceae and Stereuliaceae; Gray and Robinson on Malvaceae of North America. References to these and many other important systematic publications on the Malvaceae can be found in the ‘Bradley Bibliography’! and in other bibliographies. Non-taxonomie studies on Malvaceous genera have appeared rather frequently. Because of its economic importance, Gossypium has been the subject of many investigations, especially along cyto- genetic and anatomical lines, and probably ten times as much litera- ture has been published concerning this genus as all other genera of the Malvaceae combined. However, the bast fiber production in Sida and other genera, the virus transmission in Abutilon, photo- periodism and embryological development in several genera, and ecological studies involving certain Malvaceae have been the sub- jeets of numerous recent investigations. 'Then too, chromosome counts have been made in almost all genera of the family and about 15 per eent of the species. Nor can we fail to notice the continuous use of certain species in horticultural work. * Roush, Eva M., in Ann, Mo. Bot. Gard. 18: 117-244. 1931. 5 Jour. Arnold Arb. 12: 49-59, 7 figs. 1931. в Fries, К. E., in Kgl. Svenska Vet. Handl. 43: 1-114, 10 pls. 1908. 7 Hill, Н. W., in Jour. Linn. бос. Bot. 39: 216—230. 1909. * Hochreutiner, B. P. G., in Ann. Cons. & Jard. Bot. Genéve 20: 29-68. 1916. ° Watt, С. The Wild and Cultivated Cotton Plants of the World. 406 pp. 1907. 1 Ann. Cons. & Јата. Bot. Genève 4: 28—191, 9 figs. 1900. п Rehder, A. The Bradley Bibliography. 2°: 518-528. Arnold Arb. Pub. 3. 1912. [Vor. 29 186 ANNALS OF THE MISSOURI BOTANICAL GARDEN MORPHOLOGY ок MALVAVISCUS The genus Malvaviscus consists of woody perennial plants that may be shrub-like or vine-like, clambering or suberect. Roots are of the tap-root type, fairly thick, often twisted and distorted. Stems are terete, woody, and, although moderately stout, are unable to support the mature plant erect; they branch freely, bear many leaves, are green and generally stellate-pubescent when young, but become glabrous or subglabrous and gray-brown with age. Branches neither twine nor possess tendrils, vine-like specimens merely clambering upon available support. The leaves are alternate, stipulate, petiolate. The blade is mod- erately thick, broadly ovate-cordate to linear-lanceolate, unlobed to palmatifid, deeply serrate to subentire. The pubescence of the blade varies, consisting of stellate or straight hairs or a mixture of the two on the upper surface, of more abundant stellate hairs on the lower surface; or the blade may be glabrous in age (pl. 16, figs. 2, 3, 4). Stellate hairs may be appressed to erect, few or many- rayed, large and coarse or small and fine, or a mixture of the types may occur. Venation is prominent, reticulate, usually consisting of five larger veins with numerous anastomosing lateral veinlets. The petiole is variable in length, uniformly pubescent, with a ridge of hair (often decurrent on the stem), or glabrous. Stipules are linear-lanceolate, about 7 mm. long, caducous and usually missing in herbarium material. The inflorescence is a few-flowered, terminal or subterminal cyme, or flowers may occur singly or in groups in the axils of leaves. Usually there is a combination of these types, the young branch be- ing itself a leafy ‘‘inflorescence.’’ The pedicels are relatively short, scarcely ever as long as the flower, pubescent like the petiole, and usually more or less aggregated. Flowers are small to large in the Malvaviscus arboreus complex (1.5-5.5 em. long),'? very large in Malvaviscus candidus (about 8 ст. long), and are generally very showy, but not fragrant. They never open fully but remain as a con- torted tube, each petal overlapping the next. The mature staminal column is exserted about one third its length beyond the petals. The involucre (pl. 17, fig. 4), often designated as subealyx or in- volucel, is conspicuous, more or less enclosing and hiding the calyx. It is usually densely pubescent with small stellate hairs, sometimes 1 In this monograph flower length indicates the distance from the base of the calyx to the tip of the petals, disregarding the exserted staminal column. 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 187 subglabrous. The involucral lobes are entire, linear to broadly lanceolate or obovate, varying in number (8 is a common number), and usually about as long as the calyx. They are attached to the pedicel immediately below the calyx and are persistent until ma- turity of the fruit. The calyx (pl. 17, fig. 1) is campanulate, usually uniformly and densely stellate-pubescent without, sometimes subglabrous or with longitudinal ridges of hairs. On the inside it is short-lanose near the tips of the lobes, sub-lepidote below. The lobes are variable in number (generally five), essentially deltoid-lanceolate, and are usu- ally unequal in size. The persistent calyx may or more often may not entirely enclose the mature fruit. The petals (pl. 17, fig. 2) are five in number, asymmetrically ob- ovate-cuneate, usually emarginate at the top, unguiculate toward the base and uncinate-auriculate on one side of the claw. On the outside they are slightly stellate-pubescent basally or subglabrous. The thin, glabrous, staminal column (pl. 17, fig. 3) is usually one- third longer than the corolla and is five-lobed apically. The many unilocular, oval anthers are borne towards the tip on short fila- ments. From the top of the staminal column are exserted ten style branches, each with a capitate stigma. The ovary appears as a swell- ing at the base of the staminal column and is sessile, five-carpellate, five-ovulate. The fruit (pl. 17, fig. 5) is depressed-globose, medium-sized in the Malvaviscus arboreus complex, but larger in Malvaviscus candidus. It consists of five stone-like, one-seeded carpels enelosed by a fleshy covering which soon dries, allowing the carpels to sep- arate easily. A small reniform, basally attached seed is found in each earpel. Often only two or three of the carpels of the fruit mature. Of the morphological characters, few seem to be of diagnostic value. The habit of the plant seems to vary with the environment, as does to some extent the type of inflorescence. The structure of branch, petiole, and pedicel is essentially uniform in the genus, as is the type of stipule. The internal structure of the flower is also essentially uniform throughout the Malvaviscus arboreus complex, just as it is, for example, in the Cruciferae, and is of little value for taxonomie delimitation within Malvaviscus arboreus. In addition, many herbarium sheets of Malvaviscus are with but one or two flowers, the removal and boiling of which would leave little critical flower material for the next observer. Also the type of fruit differs [Vor. 29 188 ANNALS OF THE MISSOURI BOTANICAL GARDEN significantly only between the Malvaviscus arboreus complex and Malvaviscus candidus, two species which are already amply sep- arated on other characters. Moreover, herbarium specimens of Malvaviscus rarely contain fruit, and it would be impractical and unwise to base taxonomic units on this character unless absolutely necessary for a natural classification. The form of the involucre and of calyx is variable even on flowers of the same plant and is of taxonomic value only in a limited way. Similarly, type of leaf margin and structure of the staminal column offer little critical evidence to the taxonomist. However, the fundamental difference in leaf shape, the type and degree of pubescence (in part), and the size of the flower seem to offer a broad basis for taxonomic segregation. Yet there is not a single character in the genus that by itself is of constant value for critical delimitation within the Malvaviscus arboreus complex. Taxonomic RELATIONSHIPS AND SUGGESTED GENERIC CHANGES Delimitation of taxonomic groups: The Malvaceae belong, of course, to the order Malvales of the Polypetalae. There has been great diversity of opinion as to how many families should be recog- nized in this order, from two to nine distinct families having been accepted by various authorities. However, since the early eighteen- hundreds the family Malvaceae has been distinct and generally recognized as an entity with essentially its modern limits, defined especially by the presence of unilocular anthers and monadelphous staminal column. Within the Malvaceae subfamilies, tribes, and subtribes have been recognized chiefly on the basis of fruit characters. Perhaps the most useful and generally accepted division of the family is that of Gray,'? in which he recognized, on the basis of number and arrangement of carpels, mode of dehiscence of the fruit, and morphology of the staminal column, four tribes: the Malopeae, Malveae, Ureneae, and Hibisceae. The tribe Ureneae, of which Malvaviscus is a member, is distinguished by having five uniovulate сагре!в, ten style-branches, Ше staminal column antheriferous along the upper part (but not at the truncate or five-toothed sum- mit), and the seeds ascending, with a superior radicle. Among the genera included in this tribe are Malachra, Urena, Pavonia, and Malvaviscus. Malvaviscus is distinguished from the other genera ? Syn. Fl. М. Am. 1": 294-338. 1897. 1942] | SCHERY— MONOGRAPH ОЕ MALVAVISCUS 189 in having a ‘‘fleshy’’ or ‘‘drupaceous’’ fruit. However, the genus is apparently as closely related to Pavomia as are the ‘‘dry-fruit’’ genera of the Ureneae to one another. Proposed generic changes: The dividing line between Malvavis- cus and Pavonia has been diffieult to recognize. Just where does the “Агу? fruit of Pavonia stop and the ‘‘fleshy’’ fruit of Malvavis- cus start? In the mature fruit of Malvaviscus the outer **fleshy "' covering dries and finally ruptures exposing the carpels. In Pa- vonia the **dry"' outer wall of the ovary holds Ше carpels together until maturity. The difference is in the thickness of the outer ovary wall which, however, in both genera becomes dry and rup- tures at maturity of fruit. An additional and more distinct divi- sion between the two genera is possible by using the old Candollean character for subdivision of Malvaviscus into the sections Anotea and Achana, namely whether or not the petals are auriculate at the base. Even de Candolle questioned whether the section Anotea belonged in Malvaviscus. It is here proposed to transfer this sec- tion (raised by Kunth to generic rank) with its few and in some cases unrecognizable species to Pavonia. Thus Pavonia will include species without auriculate petals while Malvaviscus will include only those species with auriculate petals. This will necessitate the transfer of but two undoubted species of Malvaviscus to Pavonia, both of which evidently should be transferred anyway on fruit character alone. Thus Malvaviscus Palmeri Baker f. (Malvavis- cus cinereus Baker ex Robins. & Greenm., Pavonia amplifolia Standl.) should be Pavonia Palmer? (Baker) Schery, and specimens heretofore incorrectly determined ‘‘ex char." as Malvaviscus acerifolius Presl should be Pavonia firmiflora Schery n.sp. Both of these species have а ‘‘dry’’ fruit as well as non-auriculate petals. They are treated in this monograph as excluded species. GEOGRAPHIC AND CLIMATIC RANGE OF THE GENUS Malvaviscus is native to the Western Hemisphere, although a few plants have evidently been introduced into the Philippines and Malaya and are growing there as escapes. The genus occurs in- digenously from Peru and northern Brazil to the southern United States and also in the West Indies. Malvaviscus arboreus and its varieties are distributed generally throughout this range. The other species are found only in limited areas, one in central Mexico, another in Brazil. Pavonia firmiflora and Pavonia Palmeri (here- tofore included in Malvaviscus) are confined to western Mexico. 190 ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vor. 29 Maps 1-6 show the present known distribution of the species and varieties of Malvaviscus and of the two above-mentioned species of Pavonia. It is interesting to note that certain groups have be- Map 1. Showing distribution of M. arboreus B ч HE candidus — vat. bribondus arbor. ar. arboreus var. Hintoni arboreus var. palmanus arboreus var. sepium Map 2. Showing distribution of M. ca and M. ر‎ hiha vars. brihondus, cubensis, Drummondii, Hintoni, ge ai and sepi 191 1942] SCHERY—MONOGRAPH OF MALVAVISCUS come locally segregated and divergent enough to warrant varietal classification (vars. palmanus, brihondus, sepium, Hinton, ейс.), while others are very widespread, overlapping, and more or less 5 . palmatus . arboreus var longifolins | ‚ arboreus var. Williamsit Map 3. Northern South America с: рыш of M. palmatus and M. arboreus vars. longifolius and Willia `, . Е: Sv " | ik г А $ a s саса N T Y " | Copt e Ms | | | Le | e 3 | | | E В KEAN Map 4. Showing distribution of M. arboreus var. penduliflorus. artificially delimited, evidently the parent stock from which local segregations are occurring (Malvaviscus arboreus and vars. pen- duliflorus, mexicanus, ete.). Malvaviscus is quite tolerant of geographic-climatie differences. It is found growing from near sea-level to highlands of 2500 meters [Vor. 29 192 ANNALS OF THE MISSOURI BOTANICAL GARDEN as, for example, in Panama, where plants ean be found within sight of mangrove swamps (Puerto Armuelles) while about a hundred kilometers towards the interior it is abundant near the lower slopes ^ x à wh { à A “э LT I ити - “Nie . cot Bd. ‚Бу ди ра с а al о Рой. ое а A ме: РС ДД orm СИИ ОСН denote САРА Ва Bhs ss x Н ve jt ^ 5 Y ^ x у Г "T Є а ™, ел N tie К. я E / i А 2 wy 3 ав dee ; TN ge x , | Se i 3 $) Xv. 4 e ( де i) ^. ч ———_. EEE E г С БА . ха » р ko ! 4 \ г! `` > - X e / S. J А f d 5 ^ T t А à У! 23. СА { : A jeu ( рам 3 R pr. ка e. n 1 й EI У 4 ( | | A A ) има Ды. íi cu P d Map 5. Showing distribution of M. arboreus var. mexicanus ROE \ \ x \ е Е ^ ^ \ t N 1 МА pry I n M ; ý М \ т Ae хо 1 7 — деи.” T V ә TON m t oí ENCE T l = a wr ki. "db oe P ot WES „ | اي‎ A 7 ао 7 РЧА К во s d. 2 2 у= Фр "a Ф ^ А А (boa? со, v. М 4 ге РЕР { j ye eg TS ч wi ( 4 o > te „№ Фл ^. „М MN crx ул d e ) И Аи: 7 4 : о м Wart? Gs IS К ut ` es: i 2 А ( ко? ^ : э =ч Чу ~ Pavonia firmiflora / i Ф ~- Pavonia Palmeri | e | 07 P^ БЕСТ фи, Betty ФА Map. 6. Map of Mexico showing distribution of P. firmiflora and P. Palmeri. of Volean de Chiriqui. In Mexico the genus is found in the hot, steaming forests of the Tehuantepec Peninsula and the high, cool plateau region near Mexico City. Plants of northwest Mexico and Texas grow in open semi-arid regions (damp river bottoms, ete.) 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 193 while those of Costa Rica and Panama may occur in dense rain- forests. Apparently there are ample geographic-climatic forces pressing for speciation. That these forces have had relatively lit- tle effect (in the sense of forming distinet, isolated species) testifies again as to the plasticity and variability of the genus. EVIDENCE ron TAXONOMIC CONCLUSIONS Study of natural populations: In Malvaviscus, populations, in the sense that the term is used for dense, isolated groupings or clusters of herbs and shrubs in the United States, do not occur. Rather, whole regions, perhaps several kilometers in extent, are found in the tropics where the plants are relatively abundant. Scarcely ever more than three or four separate plants grow side by side, and numerous single individuals occur frequently. The plants trail over shrubbery, are more or less shrub-like in open areas, and frequently in the forest can be found liana-like reaching almost to the tops of the trees. A single plant may spread over an area seven or eight meters in diameter and perhaps overlap partly the area occupied by another Malvaviscus. In no case have I found hundreds of Malvaviscus plants within a small area, able to be collected, analyzed and counted, as are many North American plants used in population studies. Never- theless, in areas even kilometers in extent, as, for example, moist stream banks, it seems plausible to assume that all the plants of this genus are apt to be rather closely related inasmuch as it may be tens or even hundreds of kilometers before another such area is encountered. Examination of a young stem in flower from every accessible plant in such an area in Chiriqui, Panama, showed that a general similarity in certain characters existed (viz. general leaf shape; broader structural features of calyx, corolla, etc.), but that tre- mendous individual variation in certain specific characters (viz. number, length, and shape of calyx and involucre lobes, type of serration of leaf, continuity of pubescence, ete.) was also evident. It is not hard to imagine that a taxonomist unfamiliar with the genus, given two extremes from the ‘‘population,’’ might consider the specimens as different species. Yet all intergrading degrees of variation exist, sometimes even on the same plant. The only con- clusion that seems tenable is that individuals of Malvaviscus ex- hibit great variability, whether due to ploidy, hybridization, or other causes. Photographs of some specimens from this “рорша- [Vor. 29 194 ANNALS OF THE MISSOURI BOTANICAL GARDEN tion’’ are presented in plate 15. Similar examination of a “рорша- tion area’’ in Vera Cruz, Mexico, though not as extensive, sup- ported these findings on individual variabilty. Examination of greenhouse plants: That certain variations in leaf shape are of no taxonomie value (although, as will be shown di- reetly, the manner of variation may be significant) ean be shown by examination of living Malvaviscus plants. On the same plant from comparable stems can be found both semi-lobed and unlobed leaves. Figure 1 of pl. 16 shows two leaves from a plant which had been growing in the greenhouse of the Missouri Botanieal Garden for a number of years. That these greenhouse plants, evidently brought from southern Mexico, had the potentiality for ‘‘acting queer’’ as far as leaf shape is concerned, was demonstrated when cuttings were made and rooted. The first-formed leaves of the cut- tings were unlike any leaf ever seen on the parent plant, although genetieally parent and seion were the same. However, after about eight or nine nodes, the cuttings developed leaves like the parents. A photograph of one such cutting appears in pl. 14. Examination of herbarium material: From limited examination of living plants of Malvaviscus, it was evident that great variability was to be expected in herbarium material of the genus, and that many of the characters which normally (i.e., in other genera in other families) might be constant and of definite systematic value were worthless here. Examination of all available herbarium material of Malvaviscus shows complete intergradation of all forms in the Malvaviscus ar- boreus complex. Segregation of specimens of different appearance can be accomplished only with overlappings and intergradations. Intense segregation leads to a multitude of intergrading forms all of slightly different appearance (over 100 such segregated forms, all marginally indistinct, are possible in sorting Malvaviscus speci- mens). Obviously such intense segregation is worthless in any genus as variable as Malvaviscus. It leads to nothing more than artificial forms to which must continually be added others as more specimens are collected. Keying out of such forms is impossible, even on a distributional basis alone. Thus in order to find some clear-cut specific characters of taxo- nomic value in the Malvaviscus arboreus complex, individual char- acteristics were scrutinized separately in a series of specimens ranging from South America to Texas. In approximately 400 speci- mens examined for distinctive characters in leaf shape, pubescence 1942] TABLE I NUMBER OF SPECIMENS FROM DIFFERENT LOCALITIES SHOWING CONTRASTING CHARACTE RS. A J SCHERY— MONOGRAPH OF MALVAVISCUS 195 jagged Y*Hundnoooodooddodooo yg serrate Ao SUE CHI CP OE E ФА за dentate Ee E OG UIDI I IO WO в ANNA sinuate когоо жу ж-н ож жою чое å long JOVDDOADONNDAANO HOO «D вы interm. TN oO а ID y EX E Nod не і short O O O O Q t0 H O O t9 оо sf QU tQ Or t9 t0 rH „а | 2 4.5 om. ою а юкогооюонянооо Op Ew 5-4,5 om, + OUO QOO Ogoor odi] сі Ф 4 c3 <3 cm, e Ene Macs Gr ESO UU D P E Н thin ооой SOS 10 QD GOD «o зо nao in term. -аповоооопаццозцово ра i ANS anc E wide Фо + Оччооо0обснчоо NOOO Ес | continuous Sp ee eo о eS Ps ог. decurrent DO NDS DONDE со онова КЕ; pubescent гчоюююономюючномноо 0 O ра оч ha qoo tQ t2 CO c cl pilose © 75 о Od сы u е NONNA 85 simple ла ce рок р c 4 O stellate DE Eg Qu T QE TCI T$ OOOOONHOH^NOHON ON O O е. i e d d «C Ф Оч чс 0 00 ЧО О 0 «9 00 СО 0) ку + < з 3 «<> OOOOH HO HAH MOCO a Ф ~> бз ю 10:15 (0 ON HN N O O AOFM OO a - с) сч г r4 o900000000000020000 = n ооо $ $ ооооо e $ od че O *r *d «4 OND rd Е: н AK MN Xd см Ф +4 nt ос LE E RE E E e Bot SHUGE OS SH O OF парренсаосъ е аве purs 49 d'O اض‎ + Ppp > вс ANOR цао en gn ве ос Фнообонловасзсооов ооо п. Д > О д о 6 щ 0 0 Ф рч 0 осор дало 196 Cent. Texas S. Texas .6 5 1.4 .3 о о о r- в... Com мч Pan Colom pre соте Colombia Braz.& Venez. Peru ооо ою чес но Ah bebe Nee Cent. Texas Colombia Braz.& Venez. eru 4, Presence of decurrent ridges of hairs, Fig. 3. Type of pubescence Fig. of petiole, б Fig. 2. Pubescence of leaf Fig. 1. Leaf shape, [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN иест Panama Colombia Nicaragua Costa Rica Panama Colombia Braz.k Venez. eru Cent «Texas - т" 1 & Braz.& Venez. Peru Cent.Texas Te lom EI & p^ Fig. 8. Type of leaf margin, Fig. 7. Leaf length, Fig. 6. Flower length, lucré Fig. 5, Width of invo lobes, 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 197 of leaf, of petiole, and of stem, shape of involuere lobes, length of flower and of leaf, and type of leaf margin, not a single clear-cut differenee was found. Later examination dealing with the number of flowers per inflorescence, presence or absence of petiole callus, relative length of the staminal tube, type of venation, pubescence of veins, and concentration of stamens, showed a similar lack of clear-cut characteristics in Panamanian specimens alone. However, one point of significance was noted about several of these characters: they differed, on the average, for specimens from different geographical areas. Thus in Colombia leaves of speci- mens, by and large, are broad and cordate at the base, whereas in British Honduras ‘‘average’’ leaves are narrow and rounded. Again, flowers in South American specimens, although often the same length as those in Texas specimens, show an average length significantly greater than the average for Texas flowers. Table I gives the tabulated results of these examinations, and figs. 1-8 show these results put into graph form for individual characters. Of course the intrinsic value as read from the ordinants of these graphs (geographical distribution read from abscissas) is not significant. These values were obtained in mathematical form by arbitrarily assigning one extreme of a given character a high number, the other extreme a low number. Then an average for all ene of the region was taken and tabulated. Thus if half the specimens showed the high number character and half the low, the average would be midway between these two. Significance for the character can only be assumed when there is a pronounced rise or dip in the curve at a certain geographical region. For example, reference to figs. 1-8 will show a steep dip or rise for many char- acters in the region of British Honduras. From this it can be as- sumed that the available British Honduran specimens are more or less constantly different in many ways from those of neighboring regions and probably constitute a taxonomic unit. Yet examination of these graphs as a whole shows a remarkable lack of ‘‘significant characters’’ for any one region. Rather, the dips occur indiscriminately for various characters, now in one geo- graphical region now in another. With one or two exceptions, there is no continuous gradation of characters from South America to North America. These results were assumed to indicate further that in the Malva- viscus arboreus complex, there are no clear-cut, distinctly different taxonomic units, at least as far as geographical distribution is con- [Vor. 29 198 ANNALS OF THE MISSOURI BOTANICAL GARDEN cerned. In other words, that although no distributional differences of specifie degree existed, perhaps numerous overlapping groups, viz. from British Honduras, Texas, etc., were deserving of varietal separation. Also it seemed that of the characters examined, basic leaf shape and length of flower were perhaps fundamental and would be of value as key characters. The other characters, seem- ingly, were useless or of limited value. Of course it must be remem- bered that tabulations and investigations of this kind are apt to show up only degrees of differentiation due to geographical isola- tion, and may or more probably may not show speciation due to biological isolation. However, in a genus such as Malvaviscus, where scarcely a constant character is to be found, any approach that may give a hint as to natural groupings is worth investigation. Hybridization evidence: In another line of attack on the prob- lem of taxonomy in the Malvaviscus arboreus complex, hybridiza- tion was attempted between var. Drummond (Texas) and var. pen- duliflorus (southern Mexico?) growing in the Missouri Botanical Garden greenhouses. Although both varieties flower abundantly in the greenhouse, their blooming seasons do not coincide but do partly overlap. Variety penduliflorus blooms through the winter, spring and early summer; var. Drummondii in the autumn and sporad- ically at other times. This discrepancy between blooming seasons can be accounted for by the latitudinal season and day-length dif- ferences of the regions to which these varieties are indigenous. Flowers of both varieties were early castrated, hand-pollinated a day or two later with pollen from the other variety, and then bagged (both opaque and translucent paper bags were tried). However, in no case would var. penduliflorus flowers set seed, even when self-pollinated. This was evidently due to greenhouse conditions, for greenhousemen cannot recall ever having seen this variety set seed at the Missouri Botanical Garden. Neither did any Drum- mondii flowers pollinated with penduliflorus pollen set seed; but so few Drummondii flowers were available at the proper time that this result is not signifieant. Normally, only a very small percentage of Drummondii flowers set seed in the greenhouse, even when self- pollinated. Thus from attempted greenhouse hybridizations, no conclusions could be reached as to the closeness of relationship in these two varieties. However, an incidental observation noted in herbarium material seemed to be supported: that there is an inverse relationship between the flower length in Malvaviscus and quantity of seed set. Smaller-flowered plants (as var. Drummondu) seem 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 199 to set more seed than larger-flowered ones (as var. penduliflorus ), perhaps due to the fact that the pollen tube must push farther through the style in long-flowered types with a consequent lesser chance of reaching the ovary. Even though no evidence that hybridization is possible in Mal- vaviscus was obtained from the greenhouse experiments, much cir- eumstantial evidence exists that hybridization does occur in the genus. Firstly, many plants are intermediate in charaeter and have never been exactly duplicated in later collections (specimens of Malvaviscus arboreus, Pavonia firmiflora, ete.). Secondly, geo- graphie distribution of varieties makes contact possible with many other varieties (extended and overlapping range of Malvaviscus arboreus and such varieties as mexicanus, penduliflorus, etc.). Thirdly, there is usually a low percentage of flower fertility in liv- ing plants (plants examined in the field in Mexico, Costa Rica, and Panama showed that very few fruits matured in proportion to the number of flowers produced). Fourthly, hybridization is evidently not uncommon in the family Malvaceae (Kearney ' on Sphaeralcea, Webber!* on relationship in Gossypium, Skovsted!5 on chromosome numbers in Ше Malvaceae, еѓе.). In all probability appropriate ex- periments by competent geneticists and cytologists would show varietal hybridization to be exceedingly common in Malvaviscus, and even intergenerie hybridization with Pavonia possible. Chromosome counts: Lack of proper technique by the author in making microspore chromosome counts, probably more than natural diffieulties in the genus, made investigations along this line of no taxonomic value. Aceto-carmine smears of young pollen showed oc- casional evidence of a great many chromosomes, but these were so abundant, indistinct, and obscured by debris that counts were im- possible. However, Skovsted?? reports the occurrence of a diploid count of 84 chromosomes in Malvaviscus. This number fits in with the basic chromosome number of 7 found in the tribe Ureneae, a number Davie!® would like to consider as basic for the family as a whole but which others (viz., von Kesseler,!? Webber,?? Skovsted,!* “Univ. Calif. Pub. Bot. 19: 1-128. pls. 1-12. 1 m ee = мыры M., in Jour. Agr. Res. 58: 237—261. 1 Skovsted, A., in Jour. Genet. 31: 263-296. Ms vide p. 285); in Compt. Rend. Lab, Carlsberg 23: 195-242. 1941. i “ш cit. 1935 and 1941. avie, J. H., in Jour. Genet. 28: 33-67, 2 pls. 193 D н E. von, іп Am. Jour. Bot. 19: 128—130, и 9. 1932. 2 Webber, J. M., in Science, rai 81: 639-640. 1935. [Vor. 29 200 ANNALS OF THE MISSOURI BOTANICAL GARDEN Harland,”! ete.) find unacceptable as ancestral for the entire Mal- vaceae. Although chromosome counts have scarcely been attempted in Malvaviscus, it is to be expected that when such are made a high degree of ploidy, perhaps to the octoploid or decaploid, will be found as in other Malvaceous genera. Conclusions: From these several lines of investigation a con- servative taxonomie treatment for the genus seems necessary. Since clear-cut, stable morphological characters of taxonomic value are lacking, lines of demarcation must necessarily be rather arbitrary and inclusive. Evidently the taxonomy should be made practical as well as following the indistinct varietal groupings as much as possible. Since the pattern of relationships is evidently reticulate through hybridizations rather than linear by descent, a key can only be constructed which may include closely related specimens in all of its main subdivisions. There is bound to be great varietal vari- ability, and often the delimitation of a variety will need be arbi- trary. Few groupings will deserve specific rank, since all inter- grade and few show any evidence of either biologic or geographic isolation. In short, Malvaviscus (i.e. the Malvaviscus arboreus com- plex) seems best pictured as a reticulate background of several intermixing varieties (Malvaviscus arboreus, vars. penduliflorus, mexicanus, ete.) on which several newer varieties (vars. palmanus, brihondus, sepium, ete.) have superimposed themselves locally, per- haps due to a new favorable but non-isolating ‘‘mutation.’’ PHYLOGENETIC INFERENCES Although it is extremely hazardous to venture phylogenetic opin- ions, certain hypotheses can be presented regarding relationship of taxonomic groups and their derivation. From the previously mentioned circumstantial evidence concerning Malvaviscus and the Malvaceae as a whole, it seems reasonable to suppose that Malva- viscus is made up of a number of types, probably polyploid, which have not as yet become genetically isolated (except perhaps, Malva- viscus candidus) and are only partially geographically isolated. These types (Malvaviscus arboreus and its varieties) have evi- dently crossed back and forth with one another extremely often. As a result there is a reticulate relationship between Malvaviscus arboreus and its varieties, the biggest mixup occurring in southern Mexico. Marginal groups as vars. Drummondii in Texas, Wil- “ Harland, S. C. The Geneties of Cotton. London, Toronto. 1939. 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 201 liamsii in Peru, and perhaps brihondus of British Honduras and palmanus of Costa Rica, are probably the ‘‘purest’’ but neverthe- less intergrading types. The promiscuous exchange of characters, possibly coupled with degrees of ploidy, could, at least in part, ac- count for the great variability and intensification of phenotypes even in small populations. TABLE II var. cubensis var. mexicanus | var. penduliflorus bribondu: A 77 ИО var. palmanus var. longifolius var. Williamsii —— М. candidus —D, firmiflora Ф —— Pavonia М. palmatus URENEAE The genus as a whole is closely related to Pavonia and may have split off from that genus rather recently. Assuming this to be the case, Malvaviscus is in an early stage of speciation within itself, the numerous varieties as yet lacking barriers whieh would lead to complete speciation. Malvaviscus candidus must have been an early segregation from the Malvaviscus arboreus complex, perhaps oc- [Vor. 29 202 ANNALS OF THE MISSOURI BOTANICAL GARDEN curring soon after the split with Pavonia. The same may be true for Malvaviscus palmatus, or this species may be a broadeross hybrid or a monster. Table II represents a possible phylogenetie “tree” for Malvaviscus. Actually the figure should be three dimen- sional with all the Malvaviscus arboreus varieties at about the same level. Also it is not intended to show ''pure"' lineage for the Ureneae, for this tribe in its early stages may have had interconnec- tions with Hibiscus-like forms, ete. Common NAMES AND USES Malvaviscus has been recognized by many peoples who have given vernacular names to the plant. As mentioned previously, an early Aztec name for Malvaviscus was ‘‘ Аа Zopillin.''?? A supposed Maya name is ‘‘taman chich." The Maya-Spanish name for the plant was ‘‘manzanita,’’ a name still used in many parts of Central America. Other names reportedly used today in Latin America for M. arboreus and its varieties are: ‘‘algodoncillo,’’ ‘‘amapola,”’ **amapolilla,"" ‘‘arete,’’ ‘‘arito,’’ ‘‘candelillo,’’ ‘еро colorado,” ‘‘chilmecate,’’ ‘‘chocho,’’ ‘‘claveloncillo,’’? **esbequen blaneo,"' “Лог de arito,’’ ‘ог de avispa,’’ ‘‘fucsia silvestre,’’ ‘‘manzana,’’ ‘‘manzanilla,’’? ‘‘manzanita guesillo,’’ ‘‘mapola,’’ **monaguillo,"' ‘‘monacillo,’’ ‘‘monecillo,’’ ‘‘obelisco,’’ ‘‘panelita,’’ ‘‘papito de monte,’’ **para tisano’’ (for root), **pico de gorrion,’’ ‘‘quesillo,’’ *quesito," ‘‘resuscitado de monte [mente?],’’ ‘‘sobon,’’ ‘‘tuli- pancia,’’ ‘‘uba,’’ and doubtless several others. М. candidus is known as ‘‘monacillo blanco." ‘‘Waxmallow”’ and ‘‘Turk’s-cap Hibiseus"' are two English names for M. arboreus var. penduliflorus, which is frequently cultivated as an ornamental in our southern states where it has been introduced, evidently from Mexico. ‘‘May- apple"' is the name given to Malvaviscus arboreus var. Drummondu in Texas where the fruits are reported to be eaten both raw and cooked. Malvaviscus arboreus is also reportedly called ''sugar bark’’ in Jamaica. Perhaps the reason why Malvaviscus has acquired so many local names in Central America is that it is of some economic importance to the natives, and thus attention has been focused on it. Rubbing the head with leaves of the plant supposedly cures scaly head, and ? The early Spanish explorer, Francisco Hernandez, who was sent to Mexico by Philip II from 1570 to 1576 as expedition naturalist, records this name in his ‘ Nov Wisi Animalium, Mineralium Mexieanorum Historia,’ which was published at Rome in 1651 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 203 a decoction of the leaves is said to be used as a diuretic and in treat- ment of stomach ailments. Also the fruit of Malvaviscus 1s edible, though scarcely appetizing. TAXONOMY ABBREVIATIONS USED IN THIS SECTION A—Arnold Arboretum Harvard University, Jamaica Plain, Massa- chusetts. F—Field Museum of Natural History, Chicago, Illinois. G—Gray Herbarium of Harvard University, Cambridge, Massa- chusetts. L—Museo de Historia Natural de la Universidad de Lima, Lima, Peru. M—Missouri Botanical Garden, St. Louis, Missouri. NY—New York Botanical Garden, New York City. US—United States National Herbarium, Smithsonian Institution, Washington, D. С. MALVAVISCUS [Dill.] Adans. Malvaviscus [Dill. Hort. Elth. 2: 210. pl. 170. fig. 208. 1732] Adans. Fam. 2: 399. 1763; Cav. Tert. Diss. Bot. 131. pl. 48, fig. 1. 1787 ; Medie. Ма. 49. 1787; Juss. Gen. Plant. 304. 1791; Moench, Meth. Suppl. 208. 1802; HBK. Nov. Gen. & Sp. 5: 285. 1821 [1822] ; DC. Prodr. 1: 445. 1824; Descourt. Fl. Ant. 6: 11. pl. 383. 1828; G. Don, Gen. Hist. Dichl. Pl. 1: 475. 1831; Presl, Reliq. Haenk. 2: 135. 1835; Endl. Gen. Pl. 982. 1836-40; A. Rich. Bot.—PI. Vasc., in Sagra, Hist. Nat. Cuba, 131. 1845(?) ; Gray, Gen. РІ. U. S. 77. pl. 131. 1849; Benth. & Hook. Gen. Plant. 1: 206. 1862; Baill. Hist. Pl. 4: 148. 1873; Hemsl. Biol. Cent.-Am. Bot. 1: 118. 1879; K. Schum. in Engl. & Prantl, Nat. Pflanzenfam. 3°: 46. 1890; Mart. Fl. Bras. 193: 535. 1892; Gray, Syn. Fl. N. Am. 1: 297, 332. 1897; Small, ЕІ. S.E. U. S. 733. 1913; Standl. in Contr. U. S. Nat. Herb. 23: 773. 1923. Hibiscus L. Sp. Pl. 2: 694. 1753, in part; Browne, Hist. Jam. 284. 1756. Achania Sw. Prodr. Fl. Ind. Oce. 102. 1788; Ait. Hort. Kew. 2: 459. 1789; Schreb. in L. Gen. Plant. ed. 8, 2: 469. 1791; Sw. Fl. Ind. Oce. 2: 1221. 1800; Willd. in L. Sp. Plant. 3: 839. 1801; Spreng. in І. Syst. Veg. ed. 16, 3: 100. 1826. [Vor. 29 204 ANNALS OF THE MISSOURI BOTANICAL GARDEN Pavonia Cav. Ie. 5: 20. pl. 434. 1799, in part; DC. Prodr. 1: 445. 1824, as syn. Anotea (DC.) Kth. ex Ulbrich in Fedde's Rep. Spec. Nov. 14: 108. 1915, as doubtful section in DC. Prodr. 1: 445. 1824. Perennial vine-like or shrub-like plants with many-branched gray-brown terete stems. Leaves alternate, stipulate (stipules caducous), petiolate, variously lobed, usually stellate pubescent, especially on lower surface. Inflorescence leafy, with single axil- lary flowers or with terminal or subterminal cyme-like clusters of flowers. Involucre of 6-16 entire, linear, lanceolate, or spatulate lobes. Calyx campanulate, usually with 5 lanceolate or deltoid lobes, variously pubescent. Corolla contorted, tube-like, never expanded. etals 5, obovate-cuneate, usually emarginate, unguiculate and auriculate basally. Staminal column exserted, 5-parted apically, bearing many stamens toward the apex. Style branches 10, each with capitate stigma. Ovary 5-carpellate. Fruit with outer fleshy covering (dry at maturity) surrounding five stony earpels. Carpels 1-seeded, indehiscent but often separating at maturity when outer covering dries. Type species: Malvaviscus arboreus Cav. Tert. Diss. Bot. 131, pl. 48, fig. 1. 1787 KEY TO THE SPECIES AND VARIETIES > . Flower large, petals about 8 em. long; staminal column about 12-15 em. long, ore or less curved-ascending ; A about 1 em. long.......... 1. M. candidus . Flower smaller, petals less than 6 em. long; staminal column less than 7 em. long, not eurved-aseending ; filaments s than 3 n ong B. Leaves palmately or digitately lobed (known € from type MR of о SMELL ааа РС 1. palmatus BB. Leaves less deeply lobed, sublobate, or unlobed, never palmately or Pe obed. C. Leaves lobed or sublobate. D. Lateral lobes of leaves usually manifest and more or less acute; branch- lets, petioles or lower leaf-surface very densely hairy, often velvety (oe- easionally lightly pubescent or hirsute, or almost glabrous, but then lobes > > s usually large and very prominent); upper leaf-surface usually predominantly stellate- Силен emery lobes never foliaceous; flowers FR eo s sr КОЛ OT EDE 3. M. arboreus DD. Lateral lobes of leaves obtuse or small; branehlets, petioles and lower leaf-surface variously pubescent but нду not with long dense hairs; plants either with upper leaf-surface more or less predominantly straight-haired, with leaves lightly iiti and се lobate, with involueral lobes foliaceous, or with whitish ers.... (vars. of M. terii) E. Leaves slender (decidedly longer than ма. with allow jagged lobes, rounded or but slightly cordate at base; involueral lobes never foliaceous. 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 205 F. Flowers white; central Mexico............ 3d. M. arboreus var. Hintoni FF. Flowers red; Mexico to South America, West Indies. Ge Flowers more than 4.2 cm. long, торов ree Servi ее 3h. M. arboreus var. penduliflorus GG. Flowers 2.3-4.2 cm. long........... 3f. M. arboreus var. mezicanus GGG. Flowers less than 2.3 em. long; chiefly West Indian............ 3b. M. arboreus var. cubensis EE. Leaves broad (as broad or nearly as broad as long), usually conspicu- ously cordate at base. H. Involueral lobes ee and more or less В ol une lanceolate; South America............. 87. M. arboreus . Wiliamsii HH. Involuer а lobes E QM oblaneeolate, or linear- ud ЧИШУ broadest at or above the middle. I. Leaves uniformly and dein cd obtuse-lobate ; pubescence of the upper leaf-surface almost always predominantly simple; branch- lets and petioles usually а short-pubescent ; Gulf States of the United States and Mexico........ 8с. M. arboreus var. Drummondii II. Lobed and o leaves on same plant or lobes irregular an varying in siz of upper leaf- ла usually predomi- nantly Med Pers ре petioles usually glabrous, scabrous, long-haired, or with longitudinal ridges of hairs; Central Mexico to South America; West Indies J. ee slightly (less than trite) longer than broad, subturbinate, short-eylindrie with more or less flaring or spreading lobes, иы searcely longer than fruit. К. Flowers more than 4.2 em. long, TOOUSL. eese oM E EE I P сте . M. arboreus var. penduliflorus K. Flowers 2.3-4.2 em. long........ 3f. M. arboreus var. mexicanus KK. Flow KKK. Flowers less than 2.3 em. long; chiefly West Indies...... var. cubensis о. 3b. M. arborew: JJ. E. about twiee as long as broad, long-cylindrie, with tube estly longer than mo conte dated above fruit and en- p it; leaves almost always unlobed; South Ameriea...... 3e. M. arboreus var. longifolius CC. Leaves unlobed, essentially ovate- ne te. ranchlets, petioles or lower leaf-surface very heavily haired, with a dense felt-like or velvet-like oe or with long semi-rigid yellowish hairs A ee сы OIA mM Mn ue 3. M. arboreus LL. Branchlets, petioles and lower leaf-surface pubescent or glabrous, but pubescence never continuously so dense as to be velvet-like............ dnm O UM ECL dur dudo DUE (vars. of M. arboreus) M. Involueral lobes broadly laneeolate (at least 3 mm. broad at the ase), more or less foliaceous; South Атегїеа.................... No TR A CURE tio HE Aer ue 3j. M. arboreus var. Williamsii MM. Involueral lobes linear or linear-laneeolate to obovate or spatulate (always less than 3 mm. broad at base), never foliaceo М. Calyx about twice as long as broad, long-cylindric, with tube mani- festly о than fruit, eontraeted above fruit and enelosing it; ОПА ета o S OE 3e. M. arboreus var. loads NN. cal E (less than twice) longer than broad, subturbinate, or indrie with more or less flaring or spreading lobes, tube spa denos than fruit. | [Vor. 29 206 ANNALS OF THE MISSOURI BOTANICAL GARDEN O. Leaves almost perfectly elliptie, broadest at the middle, with 3 large palmate veins from which lateral veins emerge almost at right angles; Costa Riea.............. 39. arboreus var. palmanus OO. Leaves from linear-lanceolate to ovate- отд. broadest below the middle, with 3-7 large veins and variable reticulation patterns. P. Flowers white; central Mexico........ 34. M. arboreus var. Hintoni PP. Flowers red. ©. Leaves narrow, oblong-lanceolate, 2.5-3.5 times as long broad; pubeseenee of upper leaf-surface — stel- late; рена short (0.5-3.0 ет.) ; flowers small (2.0- A (24 QQ. Leaves broader than in var. brihondus, lanceolate to ovate- cordate; pubescence of upper leaf-surface often predominantly straight-haired; petioles and flowers variable in length, usually longer and larger than in var. brihondus R. Involucral lobes тен I above the middle; flowers small (2.0-3.2 em. long); -— leaf-surface pre- m straight-haired ; ps Cru and Nuevo Leon, онаа онат Келес ще а ind var. sepium RR. ct e lobes usually linear or linear-laneeolate (if spatulate, upper leaf-surface without straight hairs) ; flowers large or small; pubescence of upper leaf-surface generally predominantly stellate. S owers more than 4.2 em. long; robust.............. аа qao a wa Ra AE sachin 3h. M. arboreus var. penduliflorus SS. Flowers 2.3-4.2 em. long....3f. M. arboreus var. mericanus SSS. Flowers less than 2.3 em. long; chiefly West Indies...... 3b. M. arboreus var. cubensis *o55»5499»9*99$2599«79922523€*960949299 1 Malvaviscus candidus DC. Prodr. 1: 445. 1824; Мос. & Sessé ex A. DC. Caleq. des Dess. pl. 90. 1874; Standl. in Contr. U. S. Nat. Herb. 23: 774. 1923. Malvaviscus Pringlei E. G. Baker ex Robins. & Greenm. in Am. Jour. Sei. IIT, 50: 175. 1895; Standl. loc. cit. 774. 1923, in syn. Shrubs, with upper branches, petioles and pedicels densely stel- late-pubescent, the hairs often in clusters. Leaves large, up to 18 em. broad, densely or thinly stellate-pubescent on lower surface, upper surface predominantly straight-haired; lobes 5, triangular-acute, dentate, often irregularly so, terminal lobe еек: basal pair small- est; veins (5 large palmate veins and many small reticulate ones) conspicuous, especially elevated on lower surface, usually heavily haired; petioles 2-20 em. long; stipules linear, 5-7 em. long, cadu- cous. Inflorescence terminal or subterminal, or occasionally flowers solitary towards tip of the leafy branch; flowers large and showy, about 8 em. long; involucral lobes generally 12, linear or narrowly linear-lanceolate, somewhat shorter than the calyx, pubescent ; calyx about 2 em. long, often ineonspicuously ridged longitudinally, lightly 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 207 pubescent or subglabrous on the outside, glabrous inside except toward the pannose margins of the ovate-lanceolate calyx-lobes and the lower portion of the calyx-tube which is encrusted with fleshy hairs; corolla white, petals subretuse, lightly pubescent on the outside below the middle, with curled, simple, coarse hairs; re- ceptacle convex; ovary depressed-globose; style usually 12-14 em. long, curved upward, Ше style-branches 7-10 mm. long, pubescent ; staminal column arcuate, usually 11-13 em. long, 5-dentate, stami- niferous in its upper half; filaments linear, about 1 em. long, sub- glabrous; anthers about 2 mm. long. Fruit 1.5 em. in diameter, B Ty at M. candidus, x approx. 15. consisting of 5 brown-black, 1-seeded, ridged carpels which sep- arate at maturity; seed reniform, about 6 mm. long. Distribution: north-central Mexico (see Map 2). Specimens examined: MEXICO: COAHUILA—Saltillo, Palmer 686 (A, F, G, M, NY, US). GUANAJUATO— Guanajuato, Duges 281 (С). JALisco—Colotlan, Rose 2670 (С, US). QUERETARO—with- out exact locality, Agniel 10621 (US). MExIco—Mexico, Rose § Hay 6354 (US); Molino, MacDaniels 526 (К). MicHoacaN—Lake Cuitzeo, Pringle 4132 (А, Е, G, M, NY, US). This species is very distinct from Malvaviscus arboreus and its varieties, being easily recognized by the large white flowers with up-curving staminal column and long filaments. Also the nearly star-shaped leaf 18 seldom found in M. arboreus. It cannot be said that plate 90 of deCandolle’s ‘Caleques des Des- sins de Mociño & Sessé’s Flore des Mexique’ is undeniably Malwa- [Vor. 29 208 ANNALS OF THE MISSOURI BOTANICAL GARDEN viscus. The illustration, in some respects, rather indicates another genus, possibly Hibiscus. For example, only 5 style-branches are shown, the involuere or subcalyx appears almost foliaceous, the fruit somewhat capsule-like, and the seed is not reniform. Yet the whole set of Mociño & Sessé illustrations seems to a degree inaccu- rate as regards details, and it appears better to regard the above- mentioned characters as craftmanship inaccuracies rather than to consider the species name candidus as a ‘‘nomen dubium." This seems desirable, as in general appearance the Mociño & Sessé il- lustration could scarcely fit any other plant, having convolute corol- las, auriculate petals, curved staminal column, ete., characters mani- fest in Baker’s later-proposed type, Malvaviscus Pringlet. 2. Malvaviscus palmatus Ulbrich, in Verhandl. Bot. Ver. Prov. Brandenburg 50: 89, fig. 1. 1908. Since a specimen of this species (known only from the type) is not at present available for examination, the following free trans- lation of Ulbrich’s original description is given: Glabrous shrub 3 m. high, with terete, striate branches and long- petiolate palmate leaves; stipules caducous; leaves 7-nerved, pal- mately dissected, deeply cordate, glabrous, paler beneath and very sparsely pubescent with simple and 3-rayed stellate hairs, 7-lobed, lobes lanceolate, usually caudate-subacuminate, repand, 5-10 cm. long, 1.5-2.0 em. wide, cinereous-green, also puncticulate and pubes- cent with small adpressed hairs; petiole up to 15 em. long; in- florescence axillary, subumbellate-racemose, with several short- petiolate palmately disseeted leaves; flowers very large, with ped- icels 1-2 em. long; involuere 15-20 mm. long, 9-lobed, lobes linear- lanceolate, 12-15 mm. long, 1.5 mm. wide, acute, fringed with sub- rigid yellow hairs, after anthesis somewhat accrescent ; calyx cylin- dric-campanulate, 18-20 mm. long, 5-lobed, nerves and margins fringed with small, crisp, rigid hairs ; calyx-lobes deltoid-subacumi- nate, 5 mm. long; petals dark red, obovate-cuneate, obtuse, about 4 em. long, about 22 mm. broad at the broadest part, unguiculate, ciliate with minute, erisp hairs, above sparingly and below densely pubescent with coarse, 3-rayed and fine simple hairs; staminal tube 6 em. long or longer, tenuous, barely 1 mm. in diameter, apically with 5 acute lobes each 1 mm. long, 2 mm. from the top bearing sta- mens for 8 mm. of its length; filaments delicate, 1.5 mm. long ; style exceeding the staminal tube only about 2 mm., 9-parted into sub- liguliform style branches each about 3 mm. long; stigmas sub- 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 209 eapitellate; stigma, style and staminal tube dark violet; fruit sub- globose, black, baccate, glabrous, about 10 mm. in diameter; carpels prominently nerved on the outside, subangulate, 1-seeded ; seed sub- reniform, 6 mm. long, 2-3 mm. wide, glabrous, fuscous. Distribution: Brazil (see Map 3). Specimens examined: none available. If the description and figure presented by Ulbrich are entirely accurate, Malvaviscus palmatus is indeed unlike any other Malva- viscus known. So distinctive are the deeply lobed palmate leaves that this species could not be confused with any other in the genus. From description only, it is difficult to say with surety that this spe- cies may not belong in the genus Pavonia; yet Ulbrich’s description of the fruit and his excellent illustration of the type specimen indi- eate that this plant probably is a Malvaviscus. On the other hand, the original description does not state that the petals are auriculate, from which the inference is that the plant may be Pavonia. Cer- tainty of identification will have to await examination of the type or some authentic specimen ; meanwhile the species had best remain as Malvaviscus palmatus. « 3. Malvaviscus arboreus Cav. Tert. Diss. Bot. 131, pl. 48, fig. 1. 1787; HBK. Nov. Gen. & Sp. 5: 287. 1821 [1822]; DC. Prodr. 1: 445. 1824; G. Don, Gen. Hist. Dichlamyd. Pl. 1: 475. 1831; Spach, Hist. Nat. Veg. Phan. 3: 369. 1834; Schlecht. in Linnaea 11: 559. 1837; E. G. Baker in Jour. Bot. 37: 344. 1899; Standl. in Contr. U.S. Nat. Herb. 23: (175. 1923. Hibiscus Malvaviscus L. Sp. Pl. 2: 694. 1753. Achania Malvaviscus (L.) Sw. Prodr. Veg. Ind. Oce. 102. 1788; Ait. Hort. Kew. 2: 459. 1789 (possibly var. mexicanus). Achania pilosa Sw. loc. cit. 102. 1788; ?Ait. loc. cit. 459. 1789; Sw. Fl. Ind. Осе. 2: 1224. 1800; Lodd. Bot. Cab. pl. 829. 1817? (pos- sibly var. mexicanus or cubensis, but better not to be used as name- bringing synonym for either variety because of uncertainty of synonymy). Achania mollis Ait., loc. cit. 459. 1789. Achania coccinea Salisb. Prodr. 385. 1796. Malvaviscus cordifolius Moench, Meth. Suppl. 208. 1802. Malvaviscus acapulcensis HBK. Nov. Gen. & Sp. 5: 286. 1821 [1822]. Malvaviscus concinnus HBK. loc. cit. 286. 1822. [Vor. 29 210 ANNALS OF THE MISSOURI BOTANICAL GARDEN Malvaviscus Вази DC. Prodr. 1: 445. 1824. Malvaviscus cordatus Balb. ex DC. loc. cit. 445. 1824, nomen nudum, in synon. Malvaviscus mollis DC. loc. cit. 445. 1824; Mart. Fl. Bras. 123: 538. pl. 106. 1892. Malvaviscus pilosus DC. loc. cit. 445. 1824; Macfad. Fl. Jam. 1: 64. 1837. Achania concinna (Kth.) Spreng. in L. Syst. Veg. ed. 16, 3: 100. 1826. Malvaviscus acerifolius Presl, Reliq. Haenk. 2: 135. 1835; Standl. in Contr. U. S. Nat. Herb. 23: 775. 1923. Hibiscus racemosus Willd. ex Steud. Nom. Bot. ed. 2, 1: 760. 1841. Malvaviscus spathulatus Garcke in Otto & Dietr. Allg. Gartenz. 21: 321. 1853. Malvaviscus velutinus 'Triana & Planch. in Ann. Sci. Nat. Bot. IV, 17: 168. 1862; Mart. loc. cit. 1892, in synon. Malvaviscus speciosus Lind. & Planch. ex Mart. loc. cit. 1892. Malvaviscus arboreus var. pilosus Hitche. in Ann. Rept. Mo. Bot. Gard. 4: 64. 1893. Malvaviscus arboreus var. parviflorus E. G. Baker in Jour. Bot. 37: 345. 1899. Malvaviscus Malvaviscus (L.) Millsp. in Pub. Field Mus. Bot. 2: 13. 1900. Malvaviscus Cutteri Standl. in Pub. Field Mus. Bot. 4: 315. 1929. Shrub or vine-like plants with branchlets, petioles and pedicels densely velvety-pubescent (rarely almost glabrous), pubescence shorter and rougher in plants of certain subxerophytie regions. Leaves petiolate, variously lobed to entire, densely stellate-pubes- cent on lower surface, lightly to densely pubescent on upper sur- face, coarsely serrate to sinuate on the margins. Flowers 2.8—5.5 em. long. Involucre usually densely pubescent, with 6 or more linear, linear-lanceolate or spatulate lobes approaching or exceed- ing the calyx in length. Calyx variously pubescent, generally with 5 subdeltoid lobes. Corolla red, of 5 generally deeply retuse petals. Mature staminal column usually exserted 14—14 of its length. Distribution: Mexico to Peru and Brazil; occasional in West Indies (see Map 1). Specimens examined: ICO: CAMPECHE—Carasayal, Goldman 461 (US); Champoton, Steere 1924 (Е); Tuxpena, Lundell 970 (А, Е, M, NY, US). cHiAPAS—Comitan, Goldman 904 (US); 1942] SCHERY—MONOGRAPH OF MALVAVISCUS Fit Escuintla, Matuda 95 (US); Huehuetan, Nelson 3885 (US); а. Nelson 3005 (G, US); San Vicente, Goldman 864 (US); Teopisca, Nelson 3454 (G, US), Goldman 940, 988 (US); without E loeality, Ghiesbreght 642 (G, "Ab DURANGO— Tamazula, Gentry e (ера) ALGO—Dublan, Rose $ Hay 5302 (US); Jacala, Chase 7073 СЕЛА: NY), Вари dus (F), Lyonnet 1298, 1823 (US) ; Tula ined Pringle 8232 (А Е, d M, NY, US), 9455 (G, US), 9688 (Е G, M, NY, US). JALISCO—Barranca de Oblatos, Barnes $ Land 208 (Е, US); Etzatlan, Rose $ Painter 7541 d 08); Guada- M. arboreus, x approx. 18. lajara, Pringle 8498 (F, G, M, NY, US); Lake Chapala, Pringle bos (US); La Palma, R.E.J. 111 (US). MEx1co—Churubusco, Federal District, Orcutt 4 (F); Mexieo City, Rose ф т 4237 (US); Temascaltepec, Hinton 3861 (A, ы po 5145 (M), ?184 (A, F, NY); Tlalpam, Rose, Painter & Rose 8497 (G, NY, US). MICHOACAN— Morelia, dine $4 (Е), 2729 (US), 5494 (A, G, M, NY, US), Dugés 173 (С). мо RELOS—Cuer- navaca, Pringle 9275 (US), Rose $ Hough 4363 (US). NAYARIT—Acaponeta, Rose, Standley $ Russell 14210 (G, NY, US); between Tepie and Mazatlan, Gregg 1110 (M); San Blas, Maltby 1 (US). oaxaca—Cerro San Felipe, Conzatti ф Camino 2447 (US); [Vor. 29 212 ANNALS OF THE MISSOURI BOTANICAL GARDEN Oaxaca, Conzatti ф Gonzalez 27 (M, US); Sierra de San Felipe, Pringle 5609 (G, US). SAN LUIS POTOSI—Tamazunchale, Kenoyer A373 (Е). sinaLoa—Coacoyolitos, Ortega 6444 (G, US); Eseuinapa, Ortega 5183 (US), Il (A, G); Mazatlan, Rose, Standley ey $ Russell 13939 (NY, US). ноар Hda., Wooton (US); Tampico, Kenoyer 772a (F). VERA wii Pace Арне 9977 (Е); Juana Ramirez, Palmer 470 (US); Rineonada, Schery 204, 206 (M). TAN—Chiehankanab, Gawmer ф sons 23686 noise G, NY, US); Uxmal, руда 643 ua Pam re 2020 (F). GUA ee Pate id Standley 79935 (F, US); Patzum, Piden Nes (А, Е); Teepam, Skutch 541 (А, Е, NY, US). CHIQUIMULA—Amatillo, Steyermark 30505 (Е); усте Steyermark 30615 (Е) ; Ipala, Steyermark 30360 (Е). HUEHUETENANGO—Huehuetenango, Holway 766 (US), Standley 81929 (Е, M); Quen Santo, С. ф E. Seler 2681 (А, С, US); San Miguel Aeatan, Skutch 1021 (A, F, NY). Ja- LAPA—Jalapa, Steyermark 32851 (Е). JuTIAPA—Asuneion Mita, Steyermark 31959 (Е); Jutiapa, Standley 75248 (Е, M). PETEN—Tikal, Cook $ Martin 63 (G, US), 197 (US). QUEZALTENANGO—Zunil, Steyermark 34989 (Е). QuicHE—San Miguel Uspantan, Heyde $ Lux 2920 (G, M, NY, US); WITHOUT DEFINITE LOCALITY—Heyde 193a (US). SACATEPE- QUEZ—Antigua, Standley 61127 (Е); Duenas, Standley 63263 (Е); Santa Lucia, Popenoe 690 (US); Santiago, Gomez 822 (US); Volcan Acatenango, Kellerman 4819 9 (US). SAN MARCOS—Chamae, Standley 66190 (F); шем Steyermark 36553 (Е). SANTA ROSA —Malpais, Heyde $ Lux 6071 (С, US). soLoLA—Primavera, Shannon 419 (US). ZACAPA —Zacapa, Pittier 1749 (NY, US). DEPT. IN DoUBT—near Jaealtenanjo, Nelson 3563 (US). HONDURAS: ATLANTIDA— Tela, Standley 52756, 54127 (А, Е, US). COMAYAGUA—EI Achote, Yuncker, Dawson $ Youse 5879 (Е, С, M), 5880 (Е), 6239 (Е, С, M, NY). SANTA BARBARA—San Pedro Sula, Thieme 5168 (US). SALVADOR: SAN MIGUEL—Laguna de Olomega, Standley 21021 (G, US). SAN SALVADOR —San Salvador, Calderon 121 (Е, С, M, NY, US), Standley 22690 (С, NY, US); Volean de San Salvador, Standley 22975 (С, NY, US). SAN УСЕМТЕ—Зап Vicente, Standley 21402 (G, US). WITHOUT DEFINITE LOCALITY—Renson 8 (US) NICARAGUA: MANA ae Artemio 72 (US), Chaves 75 (US), Garnier es (US), (US), Greenman 5665, 5712 (M), Maxon, Harvey $ Valentine 7275, 7353 (US 7450 (NY, US); ель Smith 118 (G, M, US). RivAs—San Juan del Sur. "d 3552 (С, M). Costa RICA: ALAJUELA—Carrillos de Poas, Brenes 19308 (Е); Вап Ramon, Brenes 17050, pr 21917 (Е), Tonduz 17654 (US). caRTAGO—Cartago, Biolley 8977 (US), Cooper 54 (F), Standley 33368 (US); Copey forest, m 11693 (F, US). SAN JOSE —Cerro de Eseaso, Solis 266 (F); San Jose, Valeri 0 (Е); Santa Maria de Dota, I 42268 (US), Tonduz 11631 (US). PROV. IN Аз Santiago and Ріса- ondongo, Brenes 16967 (F); exact locality unknown, Worthen (M), Valerio 67 Waar INDIES: BARBADOS—Lodge Hill, St. Michael, Botanic Station 508 (NY, US). CUBA— Havana Botanical appe Britton § Wilson 511 (NY), Curtiss 726 (A, F, G, M, NY, 05); Havana, Leon 695 (NY); Olimpo Finea, Hioram 3976 (NY); Act Ekman 10069 (NY); Santiago faj las A i Baker ? (A, G, M, NY, US). MAICA—Mount Pleasant, Harris 11143 (F, NY, US); St. Andrew, Harris 11835 (F, d. M, NY, US). VIRGIN ISLANDS—St. Thomas, аи 8 (NY). COLOMBIA: статуама Bogotá, Triana 388, 3132 (US); El Colegio, Aristé-Joseph 1060 (A, F, US); El Paso, Arbelaez § Cuatrecasas 6565 (US); Girardot, Rusby ¢ Pennell 143 (NY), 170 (NY, US); Ubague, Aristé-Joseph (US). SAN даа Vegas, Killip ф Smith 16087 (С, US). SANTANDER DEL NORTE—Tapata, Xdho ф Smith 20179 (С, US). Torima—Honda, Aristé-Joseph A975, s.n. (US), Holton 748 (С, NY). VENEZUELA: DISTRITO FEDERAL—Caracas, Pittier 7121 (G, NY, US), 11116 (A, G, 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 213 NY, US). MÉRIDA—Tabay, Gehriger 565 (Е, US). MIRANDA—San Diego de los Altos, Pittier 13014 (A, F, NY, US ECUADOR: CHIMBORAZO—Huigra, Rose 22595 (US). PERU: JUNIN—Chanchamayo Valley, Schunke 21, 1492 (F); Juaja, Univ. of Lima 13 (L); San Ramon, Killip ф Smith 24714 (Е, NY, US), Schunke ney (Е, US). ымА— Univ. of Lima 75 (L). LoRETO—Balsapuerto, Klug 3015 (А, Е, G, M, US). BRAZIL: AMAZONAS—mouth of Rio Embira, Krukoff 5150 (A, НЕ МҮ, 08). This species shows an ‘‘influence’’ possibly derived from a cross with M. candidus in west-central Mexico some time ago and other later crosses with many varieties from Texas to South America. Thus it is a catch-all species for specimens approaching all the M. arboreus var. brihondus M. arboreus var. cubensis 14. x approx. 14. varieties but differing in some way due to what may be termed the ‘‘candidus influence." This **eandidus influence" may be manifest in either or both of two ways: (1) dense velvety pubescence of petioles, pedicels, upper branches, and lower leaf surface, or (2) distinctive lobation of the leaf towards the M. candidus type. 3a. Malvaviscus arboreus var. brihondus?? Schery, n. var. Shrub, upper branches pubescent, often scantily so, with x pressed-stellate or straight hairs. Leaves oblong-lanceolate, 2.5— 3.5 times as long as broad, unlobed, bluntly dentate to sinuate, sparsely pubescent on lower surface with large stellate hairs, upper surface with small appressed stellate hairs predominating over 2 Frutex foliis oblongo-lanceolatis stellato-pubescentibus 2.5-3.5 plo longioribus quam latioribus; petiolis brevibus, 0.5—3.0 em. longis; floribus parvis, 2.0-2.5 em. longis. [Vor. 29 214 ANNALS OF THE MISSOURI BOTANICAL GARDEN straight hairs; petioles short, 0.5-3.0 cm., usually finely pubescent. Flowers very small, 2.0-2.5 em. long. Involucral lobes usually wider toward the apex than toward the base. Mature staminal column ex- serted about half its length. Distribution: British Honduras (see Map 2). Specimens examined: British HONDURAS: All Pines, Schipp 708 (А, Е, С, M, NY); Belize, Lundell 4246 (F); Belize Е Record (С, US); El Cayo, Gentle 2373 (A); Honey Camp, Lundell 13, 58 (Е), 480 (Е, M түре, US), Meyer 163 (Е); Prospecto, Gentle 870 (A, Е, M, NY); Tower Hill К Karling 43 (Е, US). Distinguishing features of this variety are the small flower and the long, relatively narrow leaf which is generally short-petiolate and rounded at the base. 3b. Malvaviscus arboreus var. cubensis Schlecht. in Linnaea 11: 360. 1837. Malvaviscus Jordan-Mottii Millsp. in Field Col. Mus. Bot. 2: 73. 900. Malvaviscus Cokeri Britton ex Coker in Shattuck, Bahama Isl. 259. 1905. Woody shrub-like plants with upper branches, pedicels and pet- ioles subglabrous or pubescent, usually with long (1.0-1.5 mm.) airs. Leaves lanceolate to ovate-lanceolate, cordate or subcordate at the base, bluntly dentate to sinuate-margined, hairs preponder- antly stellate on both upper and lower leaf-surfaces. Flowers very small, less than 2.3 em. long. Involueral lobes linear or sublinear, densely or lightly pubescent. Mature staminal column exserted Yê to 14 its length. Distribution: West Indies and occasional near the Yucatan Peninsula (see Map 2). Specimens examined: MEXICO: oAxAcA—Tuxtepec, Nelson 363 (US). GUATEMALA: SACATEPEQUEZ—Acatenango, Kellerman 4990 (US). aa рим ВАНАМА ISLANDS—Abaco: Brace 1527 (А, Е, US). Great Bahama: Brac 1 (F), Britton d Pci TE Eds US). Cayman Brac: The Creek, mild 1166 (F). —Camarioca, Britton d Wilson 14082 (NY, US); Cerro de Esperon, Killip 13511 e Porn Bay, d ф Cowell 9902 (С, US); Guanajay, Baker $ Van Hermann 4253 (NY), Palmer $ Riley 697 (NY); San Juan Valley, Roig 3160 (NY); without exact locality, Wright 2064 (С, M). GRAND CAYMAN—Spot Bay, Millspaugh 1313 (Е); without иа fpe Hitchcock (F, M), Rothrock 180, 287 ISLE OF PINES— Boque , Wilson $ Selby 14498 (US). JAMAICA—Porus, Lloyd 1108 (F, M); са a sie Pel Hart (F). 1942] . SCHERY—MONOGRAPH OF MALVAVISCUS 215 This variety is close to var. brihondus from which it can be dis- tinguished most easily by the broader, generally cordate leaf. It is distinguished from var. mexicanus and other varieties by the very small flowers. 3c. Malvaviscus arboreus var. Drummondii (Torr. & Gray) Schery, n. comb. М. arboreus var. Drummondii, x approx. 14. Malvaviscus Drummond Torr. & Gray, Fl. N. Am. 1: 230. 1838; Eaton & Wright, N. Am. Bot. 314. 1840; Engelm. & Gray in Boston Jour. Nat. Hist. (Pl. Lindh.) 5: 6. 1845; Gray, Gen. Pl. U. 8. 2: 76, pl. 131. 1849; Gray in Smithson. Contr. to Knowledge (Pl. Wright.) 3: 22. 1852; Torrey, Bot. Mex. Bound. Surv. 40 1856; Wood, Class- book Bot. 269. 1865; Coulter in Contr. U. S. Nat. Herb. (Bot. W. Texas) 2: 43. 1891; Standl. in Contr. U. S. Nat. Herb. 23: 774. [Vor. 29 216 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1923; Small in Addisonia 15: 19, pl. 490. 1930; Man. Southeast. ЕІ. 854. 1933 Pavonia Drummondi Torr. & Gray, Fl. М. Am. 1: 682. 1840, fide КУ nae cit. 139. 1878; Walp. Rep. Ei = 1842; Dietr. Syn. Pl. 4: . 1847; Gray, Gen. Pl. U.S. 2: 1849. яни Drummondii Young, FI. PUE 186. 1873, fide S. Wats. Bibl. Ind. 139. 1878 Procumbent, clambering or erect shrub; stems glabrous toward the base, densely and minutely tomentose apically, with tomentum uniform, of stellate or substellate hairs. Leaves eordate at base, as broad as long, shortly 3-lobed, bluntly dentate, usually with 5 promi- nent palmate veins, lower surface with many small stellate hairs and less abundant larger ones, upper surface predominantly straight-haired. Flowers small, usually 2.5-3.0 em. long. Involueral lobes oblanceolate, widest above the middle, obtuse or acute apically. Mature staminal column exserted 15—14 its length. Distribution: Gulf states of the United States; eastern Mexico; evidently introduced into Cuba (see Map 2). Specimens examined: UNITED STATES: FLORIDA—Duval Co.: Jacksonville, Curtiss (С); Escambia Co.: Pensacola, MeCormick (G); Hillsborough Co.: without exact locality, Fredholm 6413 (G); Leon Co.: Tallahassee, ^id (NY). — в exaet dint Chapman (NY). MISSIBSIPPI—À dam Natchez, Lm Y) —Bexar Co.: Bracken, Groth 209 (F, G, US); San ak Bush 1211 (M, NY, бй}, poem вш (М), ies (US), Eggert (M), Havard (US), Heller 1833 (G, M, NY, US), Jer 91 (M), (US), (M), Metz (NY), — 115 (G, M, US), Reden 1197 (n ч 08), om 255518 (M), Thurber (G), Ward (US), Wilkinson 47 (M), 102 (M, NY); Brazoria Co.: Brazoria, Fisher 125 (08); irê Point, Fisher 191 (US); Brazos Co.: College Station, Palmer 10756 (M, US); Burnet Co.: Marble Falls, Biltmore Herb. 11079a, 11079b (US) ; Caldwell Co.: Columbia, Bush 312 (M, NY); Cameron Co.: Brownsville, Barber 23 (US), Ferris ф Duncan 3150 (M), Pringle 1959 (Е, С, M, NY, US), Runyon (M), Townsend 49 (F, US), Sargent (A); Chambers Co.: Anahuae, Hanson 2 (US); Comal Со: New рое Biltmore Herb. 11079e (US), Lindheimer 25 (Е, G, M, US), 685 (F, G, M, NY, US), (M); Dallas Co.: Dallas, Reverchon 1195 (F, US); Falls Co.: Gurley, Howell 247 (US); Gillespie Be: Fredericks! urg, Jermy 707 (M); Goliad Co.: ae eye 70 (F, M); Harris Co.: Hockley, dejes (F); Houston, Buckley (M), 3 (Е, G), Mohr (US), Palmer 8584 (M г, US), Ward (US); Sheldon, Ания E (M); Hays Co.: San Mareos, ы (M), Tharp (NY); Jaekson Co: Horseshoe Lake, without collector's name (M); Navidad River, Drushel 2844 (M); Kendall Co.: Spanish Pass, Clemens 609 (M, NY). Madison Co.: Trinity River, Dixon 441 (F, G, NY); Montgomery Co.: Willis, None (M); Travis Co.: Austin, Biltmore Herb. 110?9c, ^ iie Ages. нога 66 (M, NY), 102 (М, US), срце 115 (A); солды Co.: Vic , Clark 3977 (M); Wharton Co.: Pierce, Tracy 7476 (F, G, M, E, UB); с. Poner 6609 (M, US); Williamson Co.: Round БА Bodin 129 US). MEXICO: SAN LUIS POTOSI—Canoas, Rose g Hough 4883 (US); Taneanhuitz, Nelson 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 217 4392 (US). TAMAULIPAS—EI Milagro, Bartlett 11069 (Е, US); San Jose, LeSueur 283 (Е); Tampico, Palmer 94 (G, M, NY, US), 387 (US), 525 (С, US). West INDIES: CUBA—without exact locality, Wright 2065 (M), 2068 (С, M, NY, US). This is one of the most distinet varieties of M. arboreus. It is best distinguished by a combination of characters including the broad, cordate, obtusely lobed leaf; predominance of straight hairs on the upper leaf surface; and involucral lobes which are broadest above the middle. 3d. Malvaviscus arboreus var. Hintoni (Bullock) Schery, n. comb. Malvaviscus Hintoni Bullock in Kew Bull. 291. 1937. Shrub with upper branches, pedicels and petioles pubescent with long hairs (usually 1.0-1.5 mm. long). Leaves lanceolate, 2-3 times M. arboreus var. Hintoni M. arboreus var. longifolius x approx. 18. x approx. 14 as long as broad, sometimes obscurely 3-lobed, rounded at base, bluntly serrate, lightly pubescent, lower surface with large stellate hairs, upper surface with smaller stellate and straight hairs. Flowers 3.5—5.0 em. long, white or whitish. Involueral lobes linear or linear-spatulate. Mature staminal column exserted 1—1, its length. Distribution: east-central states of Mexico (see Map 2). Speeimens examined: МЕХхтсо: MEXICO—Temascaltepec, Hinton 690 (Е), 3928 (A, NY), 4289 (A, Е, NY), 5057 (А, M, NY), 5371 (A, NY), 7912 (US). VERA CRUZ—Chontla, Cardenas 375 (A, Е). This variety is very close to marginal types of M. arboreus and M. arboreus var. mexicanus but ean be distinguished by the white я [Vor. 29 218 ANNALS OF THE MISSOURI BOTANICAL GARDEN flowers (all other varieties have red flowers) and comparatively narrow lanceolate leaves. 3e. Malvaviscus arboreus var. longifolius (Garcke) Schery, n. comb. Malvaviscus longifolius Garcke in Otto & Dietr. Allg. Gartenz. 22: 321. 1854. Malvaviscus cuspidatus Turez. in Bull. бос. Nat. Мове. 31: 190. 1858. Malvaviscus leucocarpus Planch. & Linden ex Triana & Planch. in Ann. Sci. Nat. Bot., IV, 17: 169. 1862; Mart. Fl. Bras. 128: 536. 1892. Malvaviscus Funkeanus Linden & Planch. Trois Voy. Linden in Pl. Columb. 1: 41. 1863. Malvaviscus elegans Linden & Planch. ex Mart. loc. cit. 537. 1892. Malvaviscus maynensis Huber in Bol. Mus. Goeldi 4: 583. 1906. Malvaviscus integrifolius Ulbrich in Verhandl. Bot. Ver. Bran- denburg 50: 88. fig. 2. 1908. Malvaviscus Ulei Ulbrich in Notizblatt 6: 328. 1915. Shrub-like plants with upper branches, pedicels and petioles usu- ally with long (1.0-1.5 mm.) hairs, or glabrous. Mature leaves gen- erally large (seldom less than 8 cm. long), ovate-lanceolate to tri- angular, usually cordate at the base, bluntly dentate or sinuate mar- gined, entire or slightly lobed, upper surface variously pubescent, lower surface stellate-pubescent. Flowers large, usually longer than 4 em. Involueral lobes linear or linear-lanceolate, usually nar- row to acuminate-attenuate. Calyx long-cylindrie, usually twice as long as broad or longer, often yellow-setose, contracted above and enclosing the fruit. Mature staminal column exserted 14-14 its length. Distribution: northern South America (see Map 3). Specimens examined: COLOMBIA: ANTIOQUIA—Angostura?, André K862 (Е, С, NY). ATLANTICO—Barran- quilla, Elias 428 (US); Usiacuri, Dugand 887 (Е). BoLivAR—Cartagena, Heriberto 244 US). WITHOUT EXACT LOCALITY—Mutis 2262 (US). VENEZUELA: DISTRITO FEDERAL—Caracas, Allart 76 (NY, US); Cerro Avila, Vogl. 76 (Е); Los Flores a Papelon, Delgado 271 (Е, US), Tamayo 385 (US). LARA—Bar- quisimeta, Saer 65 (US). MÉRIDA—Tovar, Fendler 101 (С, M). ECUADOR: oro—Santa Rosa, Hitchcock 21141 (NY, US). PERU: AYACUCHO—Estrella, Killip ф Smith 23065 (Е, G, NY, US). JUNIN—Puerto Bermudez, Killip & Smith 26648 (NY, US). LORETO—Iquitos, Williams 8068 (Е); La Victoria, Williams 2675 (F); Rio Nanay, Williams 508 (F). SAN MARTIN—Juan Jui, Klug 3919 (Е, С, M, NY, US), 4382 (A, Е, M, NY, US). 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 219 BRAZIL: AMAZONAS—Manariao, Krukoff 4589 (A, F, M, NY, US); Seringal (Rio Acre), Ule 9591 (US). This variety approaches M. arboreus and M. arboreus var. pen- duliflorus, on the one hand, and M. arboreus var. Williams, on the other. It can best be distinguished by the very long-cylindrie calyx enclosing the fruit, the linear or linear-lanceolate involueral lobes, and the relatively large, broad leaves. 3f. Malvaviscus arboreus var. mexicanus Schlecht. in Linnaea 11: 359. 1837,as Mexicana; E. G. Baker in Jour. Bot. 37: 346. 1899. Pavonia spiralis Cav. Ie. 5: 20, pl. 434. 1799. Malvaviscus grandiflorus HBK. Nov. Gen. & Sp. 5: 286. 1821 [1822] ; Standl. in Contr. U. S. Nat. Herb. 23: 775. 1923. Malvaviscus ciliatus DC. Prodr. 1: 445. 1824. Malvaviscus pentacarpus Moc. & Sessé ex. DC. loc. cit. 1824; ex. A. DC. Calq. des Dess. pl. 88. 1874. Achania ciliata Spreng. Syst. Veg. 3: 100. 1826. Malvaviscus brevipes Benth. Bot. Voy. Sulph. 68. 1844. Malvaviscus pulvinatus A. Rich. Bot.—Pl. Vasc., in Sagra, Hist. Nat. Cuba, 133. 1845 (?). Malvaviscus Sagraeanus, A. Rich. loc. cit. 131, pl. 14. 1845(1). Malvaviscus Guerkeanus Hieron. in Engl. Bot. Jahrb. 21: 320. 1895. Malvaviscus arboreus var. Grisebachu E. G. Baker in Jour. Bot. 37: 345. 1899. Malvaviscus arboreus var. Sagraeanus (Rich.) Baker, loc. cit. 1899. Malvaviscus arboreus var. Sloane Baker, loc. cit. 1899. Malvaviscus brevibracteatus Baker, loc. cit. 347. 1899. Malvaviscus Polakowskyi Baker, loc. cit. 346. 1899. Malvaviscus rivularis Brandg. in Zoe 5: 211. 1905; Standl. in Contr. U. S. Nat. Herb. 23: 774. 1923. Malvaviscus oaxacanus Standl. loc. cit. 775. 1923. Bushy or vine-like shrubs, with upper branches, pedicels, and petioles variously pubescent or subglabrous. Leaves lanceolate or ovate-lanceolate, rounded or cordate at base, acute or attenuate at apex, serrate to sinuate-margined, unlobed or occasionally with slight marginal projections, variously pubescent. Flower of med- ium size, 2.3-4.2 em. long. Involucral lobes linear or sublinear. Calyx campanulate-cylindric, somewhat longer than broad, gla- brous or pubescent. Mature staminal column exserted 14-14 its length. [Vor. 29 220 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: southernmost United States to Panama; West Indies (see Map 5). Specimens examined: UNITED STATES: TEXAS—Cameron Co.: Brownsville, Hanson (M), Rose 4 Russell 24292 (US). Mexico: cHIAPAS—Escuintla, Matuda 2153 (Е, NY); Tumbala, Nelson 3346 (NY, US). солма--МапгапШо, Ferris 6078, 6231 (US), Palmer 963 (С, US); Paso del Rio, Emrick 167 (Е). GUERRERO—Acapuleo, MacDaniels 191 (Е), Palmer 536 (С, US); M. arboreus var. mexicanus, x approx. 14. Galeana, Hinton 10897 (M, NY, US); Mina, Hinton 959? (A, F, M, US); San Luis, Langlasse 924 (С, US). JALISCO—Tuxpan, Mexia 1023 (А, Е, M, NY, US). MEXICO— ге Hinton 4568 (A, NY), 5254 (A, их. US), Hinton 6719 (А, Е, NY, US). MICHOAC elia, Arsène 2729 (A, С, M, US). wonELos—Cuernavaea, Pringle 9662 (G, "US, Pha Painter $ Rose 10220 (US). ммен E Blas, Ferris 5436 (А US), Wright 1348 (Е, М); Tepie, Palmer 1955 (Е, С, NY, US). oaxaca—Ejutla, дов. zatti 8948 (US); Jamiltepec, Conzatti 4430 (US); Oaxaca City, pe ose d Hough 4587 (US) ; Oaxaca peri Nelson 1223, 1256 (G, US), Pringle ene (A, F, G, M, NY, US), Smith 296 (G), 688 (M); рен Nelson 348 (US) ; Ubero, Williams 9189, 9219 (Е). es LLAMAR, Goldm 2 (G, US); Puebla, Arséne 128, 389, 1957 (US). SAN LUIS POTOSI—Espinazo del Dan Pennell 17975 (US); Rio Tampaon, Chase 7481 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 221 (F, G); Tamazunchale, Edwards 482 (F, M). ee Ortega 5116 (08); Culiacan, rd (С, US). mTAMAULIPAS— Tampico, Kenoyer 772 (Е), s.m. (M). VERA CRUZ—Catemaco, Non 405 (08); Coatz sone эп ith 1023, 1030 (С, US), 1030 (M); pipe Wits ms 8336 (Е); Jalapa, Plunket 0 (F); Vera Cruz, Greenman 49 (F, G, NY); ubi exact locality, Orcutt is p^ YUCATAN—Chichankanab, Gaumer 1858 (Е, С, M, US), Gaumer $ sons 23686 (M); Chichen Itza, Seler 4913 (С, US), Steere 1127, 1618, 1642 (Е); Izamal, Gaumer (Е, US), Greenman 443 (Е, С, NY); Kaneabonot, Gawmer & sons 23523 (Е, С, M, US); Merida, Schott 177 (Е), s.n. (Е, US); Progresso, Schott tie KE, ШЕ); Опера 1728 (Е); Silam, Gawmer (Е); Suitun, Gawmer $ sons 23361 (А, Е, С, M, NY, US); exact locality unknown, Gawmer 580 (А, Е, С, M, NY, US), Goldman 579 (US); Millspaugh 42 (Е, bes 60 (F). STATE IN DOUBT —Miramar, Matuda 93 (US); without locality, oin на 6 (U BRITISH HONDURAS: Corozal, Gentle 208 (Е, US), [Zanden 4890] (Е, M EMALA: A pAz—Coban, Puerobhein 7 (08); Pamsnnls, Tuerck- ne Ag US); ng Desa 90542 (M). BAJA VERA PAZ—Santa 2 Standley Aes Tuerckheim 112812 (С, US). cHiquimMuLa—Jocotan, Steyermark 31634, we (Е). ESCUINTLA—El a ul, Rojas 4 Tonduz 56 (US); Pact, Smith 1991 (G, US), Standley 63420, 63919 (Е), 89189, 89488, 89571 (Е, М); Peinha, Pittier 1793 (US) ; Rio Guacalate, Standley 58277, 60179 (Е); San Jose, Standley бос (Е). GUATE- MALA—Chilloui, Rojos 67 (G, US); Guatemala, Ruano 405 (US), Tonduz 814 (US). HUEHUETENANGO—Canibal, Shannon 307 (US). ero ee nd Standley 54807 (G, NY, Е e Vargas, siii sd 790 (Е). PETEN—La ertad, Адш 7 (А, M). QUEZALTENANGO— Pal E bos 5811 (US); Los Eod. Stn Pu (F); vein Santa Maria, и ark 83581 (Е). qo eu 8824 j ; Las Delieias, Standley 88014, 88119 (F, Nueva Linda, ns 66586, 66539 (F); Retalhuleu, Standley 66702, 66719, pe sj e. 88262 (F, М); San Felipe, Holway 694 (US). SACATEPEQUEZ—Antigua, Standley 60322 (Е); Volean E Kellerman 4806 (US); Volean de Agua, Standley 59465 "res SAN MARCOS —Ocos, Steyermark 87878 (F); Tajumuleo, Steyermark 36665 (F). SANTA сср аа Standley 77752 (F, M); Cuazaeapan, Standley 78616, 78644 (Е, М); Sepultura, Standley 79410 (F, M); Taxisco, Standley 79019 (F, M). SUCHITEPEQUEZ —Las Animas, gray 884 (US); Mazatenango, siii 529 (US), Kellerman 4962 (US), Mazon $ Hay 3470 (US); Patulul, Standley 62146 (A, Е). zACAPA—Sierra de las Min pedis 29878 (F); Zacapa, Deam Ди ве NY, US), Kellerman 9019 (F), Bandi 72026 (F), 74212 (F, M). HONDURAS: ATLANTIDA— Tela, pines 55780 (F, US). pre pin ЖИ. Yuncker, Dawson 4 Youse 5680 (Е, G, M). voRo—fPijol n Hagen 1 (Е, N WITHOUT EXACT момите та Lima, Johansen 29 Js SALVADOR: AHUACHAPAN—— TE Standley 19718a, 19855 m NY, US); with- out exact locality, CEN ES Un 7 (A, M, US). LA UNION—La Union, Standley 20809 (С, US). soNsoNATE—Armenia, ptc 23520 (G, б, ге чай Chilata, Stand- ley 19339 (С, NY, US) ; Izalco, Standley 21837 (С, NY, US); Sonsonate, Standley 22300 (G, NY, US) NICARAGUA: MASAYA—Masaya, Baker 163 (G, M, NY), 618 (US). DEPT. IN DOUBT— Braggman's Bluff, E 75, 94 (F); without exact т Wright (G, M, 08). Costa Rica o—Carpintera Mt., Stork 333 (U M^ rtago, oe 5719 (Е, US); Вап atari, gu 9050 (US). ava ANACASTE—Cruz E nacaste, Pittier 2770 (US); Hacienda Santa Lo Dodge & Thomas 6324 a ; Los wd ade Tonduz 2884 (US); Nicoya, Mats 367 p. Tonduz 18485 (G, M). NTARENAS— Santo Domingo del Golfo Dule ge 2 (US). SAN ic aec at a ride dia 5433 (M); La УегЬепа, 038 32280 CoB Tonduz 8946 (US); San Jose, Tonduz 1092 (US). PROVINCE IN DOUBT: Ochonogo, Pittier 59 (US). [Vor. 29 222 ANNALS OF THE MISSOURI BOTANICAL GARDEN PANAMA: CHIRIQUÍ— Bajo — Woodson $ Schery 531 (M); Boquete, Pittier 2925, 3138 (U 4 Pena Blanca, Woodson $ Schery 302 (M); Quebrada Velo, Woodson $ Schery 276 (M); New „седани Allen 1364 (Е, С, M, NY); Peninsula de Buriea, Woodson ф Schery 932 (M). cocLÉ—EI Valle, Allen 91 (A, ге hi 1176 (Е, M) P. $ С. White 70 (G, M). slc a mio M. E. $ Е. A. Terry 1386 (M). PANAMA—Taboga Island, Standley 27924 (US). VERAGUAS—Sona, Allen 1045 a G, M EST INDIES: BAHAMAS—Watlings Island, Britton $ Millspaugh 6145 (US). cuBA —Havana, Britton ф Wilson 4535 (Е); Oriente, Ekman 2976 (Е); Pinar del Rio, Palmer $ Riley 597 (US), van Hermann 253 (Е, NY), Wilson $ Leon 1129? (US); Sierra de ть Leon 4745 (NY). JAMAICA—Blue Mountain Peak, Hitchcock (М); Cinchona, Harris $ Lawrence C15297 (US), Marble 193 (NY), Rehder (A); Diabolo Mt., Maxon $ кану 398 (US); Gordon Tan Hart 571 (US); John Crow Mts., Britton 3995 (NY); Latimer River, Nichols 6. EB US); Lucea, Britton 2917 (NY); Mandeville, Britton 999 (NY); Brown 78 (NY); Negril, Britton $ Hollick 2093 (NY); Port An- tonio, Hitchcock (F, M), Millepaugh 931 (F, NY), Wight 22 (F, NY); Rio Grande, Millspaugh 1920 (F); Walderston, Harris 12863 (F, G, M, NY, US); Wavels Rock, Fawcett 8022 (NY); without exact locality, Alexander (M). TRINIDAD—Queen’s Royal College grounds, LI 7690 (F, M); St. Ann's, Broadway 7324 (NY COLOMBIA: ANTIOQUIA—San Geronimo, Tomas 622 (US). aTLANTICO—Barranquilla, Elias 655 (US). ровнее Cuatrecasas 1850 (US). CUNDINAMARCA—San Javier, Aristé-Joseph (US); Tabio, Antonio 16E (US). MAGDALENA—Araeataea, Dugand Barriga 2476 (US). SANTANDER—Badillo, Pennell 3910 (NY, US). The only character which separates this variety from varieties penduliflorus and cubensis is the flower size. Admittedly arbitrary limits are given for flower size, but this is necessary in view of the fact that complete intergradation occurs between these varieties. Here the flower is smaller and less robust than in var. penduliflorus, with the mature staminal column, as a rule, more exserted. On the other hand, the flowers are not as dwarfed as in var. cubensis and the staminal column is usually less exserted. 3g. Malvaviscus arboreus var. palmanus (Pittier & D. Smith) Schery, n. comb. Malvaviscus palmanus Pittier & D. Smith in Bot. Gaz. 23: 238. 1897. Shrub or vine-like plants with upper branches, pedicels, and petioles generally thickly pubescent with short stellate hairs, some- times also with longer straight hairs, or almost glabrous. Leaves entire, almost symmetrically elliptie, broadest at the middle, usually large when mature (as long as 20 em. in extreme cases), 2 to 4 times as long as broad, with 3 prominent palmate veins from which smaller retieulate veins emerge almost at right-angles, stellate-pubescent on both surfaces; margins shallowly dentate or sinuate. Flowers large, usually 4-5 em. long. Involucral lobes essentially linear. Mature staminal column exserted 14—1% its length. Distribution: Costa Rica (see Map 2 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 223 Specimens examined: Costa RICA: ALAJUELA—La Ventolera, Standley 34685 (US); San Carlos, Smith H1670 (F, M); San Ramon, Brenes 5954, 13418, 21969, 21982, 22623 (F); Viento Fresco, Standley d Torres 47926 (US); cARTAGO: Orosi, Standley 39640, 39837 (US). GUANACASTE—El Silencio, Valerio 66 (US); Tilaran, Brenes 15630 (Е), Standley $ Valerio 44605, 46238 (US). HEREDIA—San Frideo, Pittier 14015 (US); Vara Blanca de Sarapiqui, Skutch 3266 (A, M). SAN JosE—La Hondura, Standley 36585 (US); La Palma, Maxon $ Harvey 8004 (US), Standley 33209 (US), Stevens 301 (US), Топдиг 7393 (Е, а, M, US), 8089, 12465 (NY, US); PROVINCE IN DOUBT: Zarcero, Smith H20 (F, M), 4316 (F); without exaet loeality, Smith? 7393 (M). (е D Ди “ас pe M. arboreus var. palmanus, x approx. 14. This variety approaches vars. penduliflorus and mexicanus but ean be distinguished by its almost perfectly elliptie leaves. 3h. Malvaviscus arboreus var. penduliflorus (DC.) Schery, n. comb. Malvaviscus penduliflorus Moc. & Sessé ex DC. Prodr. 1: 445. 1824; ex A. DC. Calq. des Dess. pl. 90. 1874; Standl. in Contr. U. В. Nat. Herb. 23: 774. 1923. Malvaviscus oligotrichus Тител. in Bull. Soc. Nat. Mose. 31: 190. 1858. Malvaviscus glabrescens Planch. & Lind. ex Triana & Planch. in Ann. Sei. Nat. Bot. IV, 17: 168. 1862, nomen nudum in synon. (= M. oligotrichus fide Triana & Planch.). Malvaviscus lanceolata Rose in Contr. U. S. Nat. Herb. 5: 175. 1899. Malvaviscus Conzattii Greenm. in Field Col. Mus. Bot. 2: 333. 1912; Standl. in Contr. U. S. Nat. Herb. 23: 774. 1923. [Vor. 29 224 ANNALS OF THE MISSOURI BOTANICAL GARDEN Bushy or vine-like shrubs ; upper branches, pedicels, and petioles with long (1.0-1.5 mm.) hairs, or glabrous. Leaves lanceolate or ovate-lanceolate, rounded or cordate at base, acute or attenuate at apex, serrate to sinuate-margined, variously pubescent. Flower robust, longer than 4.2 em., often appearing abruptly truncate at M. arboreus var. penduliformis, x approx. 16. the tip in herbarium specimens. Involueral lobes essentially linear in southern part of plant range, often broader and more or less spatulate in northern part. Calyx more or less cylindric, somewhat longer than broad, glabrous or lightly haired. Mature staminal column exserted 14—14 its length. Distribution: widespread, from central Mexico to Colombia (see Map 4). Specimens examined: ЕХІОО: CHIAPAS—Chicharras, Nelson 3807 (G, US); La Illusion, Mell 2019 (US); San Cristobal, Nelson 3170 (US). auERRERO—Chilacayote, Hinton 14181 (С); Petlacala, 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 220 Мега 9094 (С, M, NY); Pilas Pasion Filo Mayor, Hinton 10753 (Е, M, US); Plan de Carrizo, Hinton 11030 (G); San Marcos, Nelson 2264 (US); Sierra Madre, Langlasse 793 (G, US). JALISCOo—Puerto Vallarta, Mexia 1135 (А, Е, M, NY, US); Santa Cruz de Vallarta, Mexia 1267 (US); San Sebastian, Mexia 1442 (A, F, G, M, MEXICO— Temascaltepee, Hinton 4014 (A, NY, U MICHOACAN—Zitacuaro, tn 13529 (С). MORELOS—Cuernavaca, MacDaniels 305 (Е), Rose $ Hough 4351 (US). oaxaca—Cafetal Concordia, DAE n $ Makrinius 2891 (Е, US), Reko 3348 (US); Comaltepec, Nelson 926 (US); Plunia, Nelson 2500 p US); San Pablo Huitzo, Con- zatti 1981 (Е); Santo Domingo, Conzatti 1683 2 8); Yaveo, Mexia 9204 (G, M, NYJ: rM Nelson 2016 (US). ARO—Cerro de s las Campañas, Arsène 10058 (F, M, US); without exact locality, ho jn (A, F, G, M, US). SINALOA— without |. locality, Ortega 7327 (Е). VERA CRUZ— Cordoba, ии 188 (M); Tam- pico, Palmer 391 (M, NY, US). STATE IN poust—from Mexican ae Rose 4027 (US). GUATEMALA: ALTA VERA PAZ—Panzos, Mason $ Hay 3080 (US); Saxoe, Tuerck- heim 8185 (US). VERA PAZ—Paujal, Tuerckheim ny m CHIMALTENANGO —dQuisache, И Pa 24 Е), 62048 (Е). 1ZABAL—Boca del Polochie, Smith 1658a (US); Quirigua, Standley 24587 (G, NY, US), 72311 (F). QUEZALTENANGO—Aguas Amargas, Standley 65416 NS Chiquihuite, Standley 68102, 68112 (F); Santa Maria de Jesus, садо 68236 (Е, NY), Steyermark 33386 (Е); Volcan Zunil, Skutch 878 „ B») HULEU—San Felipe, И РЕЯ 34516 (Е). soLoLA—Volean Atitlan, Hateh ф UE 360 (F) HONDURAS: ATLANTIDA—La Ceiba, Yuncker, Koepper $ Wagner 8523 (Е, M, NY); Sus Mitchell 181 (F, A Standley 53743 n Е, US), 54021 (Е), 54786 (А, Е, US). А BARBARA— San ro Sula, Bangham 341 (A), Thieme 5153 (С, NY, US). YORO EN distriet, som 55498 (А, Е, aay, Quebrada Seca, Standley 53926 (A, Е, US); Progresso, Standley 54987 (А, Е, US). DEPT. IN DoUBT—Puerto Sierra, Wilson Ф М2 NDS NICARAGUA: ATAGALPA—Jinotega, Grant 7298 (A). Costa RICA: ALAJUELA—Capulin, Standley 40163 (US); Naranjo, Stork 1828 (Е); Palmira, Smith 4215 (Е); San Pedro, Brenes 16998 (Е). cARTAGO—Cartago, Holway 276 Ж Standley & Valerio 49600 (US), Stork 2834 (Е); Cerro Carpintera, Dodge $ Thomas 4784 (M), Standley 35756 (US); Irazu Voleano?, Pittier 13063 (US). GUANA- CAS im ш. Bay, Pittier 12020 (US). HEREDIA—Puerto Viejo & Sarapiqui River, Biolley 7403 (US); Santo at ge del Roble, Dodge 4 Goerger 9582 (F, M). LIMON —La Colombiana farm of United Fruit Co., Standley 36978 (US). SAN JosE—El Gen- eral, Skutch 2368 (A, G, M, rum US); San Francisco de Guadalupe, ioni 18081 (US), Tonduz 6973 (С, US); San Jose, Biolley 43 (Е), Holway 402 (US), Tonduz 726 (US); Santa Maria de Dota, йл Ж 41845 (F, US), Standley & Valerio yo (PUS) Stork 2975 (F); Tucurrique on Las Vueltas River, Tonduz 13149 (US); Ciruelas River, Tonduz 2218 (US); San Isidro Coronado, Alfaro 32378 (F, US); Llanuras de Santa ur Smith 6450 (US). AMA: BOCAS DEL ape cea Cooper & Slater 26 (US), Cooper 103 (F, NY); 2 del AN Carleton 81 (NY, US); Chan ime River, Dunlap 101 (F), 349 (F, G, US), 440 (Е). CHIRIQUΗBajo Chorro, Davidson 68 (А, Е, M); Bajo Mona, Wood- on $ Fees 550 (M); Boquete, Mazon 5000 (US); d Punta, Seibert 263 (A, M, NY); San Felix, Pittier 5210 (NY, US). cocLÉ—El Valle, Allen 1906 (M), Wood- son А Schery 188 (M). PAN XH la Maestra, Allen 52 (A, G). PROVINCE IN DOUBT — western жшн: Stork p (US). мш ом BO о A кш: $ пей 14052 (А, с, NY, US); Turbaco, Killip & PS pops (а, US). caqueTa—Cordillera Oriental, бшм аз 9158 (US). aedi UA Магы, Smith p (NY), bg (А, Е С, М, МУ 05), 235 (Е, d , NY, US), 2727 (G, NY), 2816 (NY). SANTANDER—Barranca Bermeja, Haugh [Vor. 29 226 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1532 (Е, US); California, Killip $ Smith 16890 (С, US). DEPT. IN DOUBT—Fouinnef, Arbelaez 2456 (US). VENEZUELA: DISTRITO FEDERAL—Cerros del Avila, Pittier 49 (NY, US); Galipan, Pittier 83 (NY, US ECUADOR: PROVINCE IN DOUBT—La Chonta, Rose, Pachano $ Rose 23478 (NY, US). This variety shows complete intergradation with var. mexicanus from which it must be separated rather arbitrarily. The difference between these varieties is in the flower, var. penduliflorus having a more robust flower, almost always greater in length than the ar- bitrary limit of 4.2 em. 3i. Malvaviscus arboreus var. sepium (Schlecht.) Schery, n comb. Malvaviscus sepium Schlecht. in Linnaea 11: 361. 1837; Standl. in Contr. U. S. Nat. Herb. 23: 775. 1923, as syn. M. arboreus var. sepiwm M. arboreus var. Williamsii X approx. 4. approx. 14. Shrub with upper branches, pedicels, and petioles lightly or thickly pubescent with long (1.0-1.5 mm.) hairs, pubescence usually decurrent in ridges on petiole and stem. Leaves small (less than 10 em. long, usually about 6 сш.), 1.5-3.5 times as long as broad, lanceolate to ovate-lanceolate, bluntly dentate, unlobed, upper leaf- surface predominantly straight-haired, lower surface with few to many straight hairs interspersed among stellate hairs. Flower small, 2.0-3.2 em. long. Involuerallobes spatulate. Mature staminal column exserted 1—14 its length. Distribution: Vera Cruz and occasional elsewhere in northeastern Mexico (see Map 2). Speeimens examined: MEXICO: NUEVO LEON— Monterrey, collectors unkonwn (Е, US). VERA CRUZ— Cordoba, Bourgeau 1512, 1669 (G, US), Greenman 191 (F, G); Coscomatepee, Matuda 1310 1942] SCHERY—MONOGRAPH OF MALVAVISCUS ОЕТ (Е); Fortin, Fisher 85506 (Е); VU bes. Chamberlain $ Land 34 (Е), MacDaniels 944 (F), Orcutt 2811 (F, M), Plunkett 45 (F), Pringle 7833 (G, US), 8202 (A, F, G, M, NY, US), Rose $ Hay 6184 an US), Rose $ Hough 4245 (US); Mirador, Mohr (US); Nogales, Seaton 800 (F, G, US); Orizaba, Fisher 152 (US, F, M), Mohr (US), Pringle 5914 (US); Azcuapan, Pures | 7430 (G, M, US), 10729 (US) This variety is close to var. mexicanus and is best distinguished by the spatulate involucral lobes, the small flowers and leaves, and the predominantly straight-haired pubescence of the upper leaf surface. 3j. Malvaviscus arboreus var. Williamsii (Ulbrich) Schery, n comb. Malvaviscus Williamsui Ulbrich in Notizblatt 11: 545. 1932. Shrub-like plant with upper branches, pedicels and petioles stel- late-pubescent, often densely so. Leaves broadly lanceolate to tri- angular, cordate at base, unlobed or occasionally with small mar- ginal projections, shallowly or deeply dentate, usually predomi- nantly stellate-pubescent on both surfaces. Blowers large, 4-6 cm. long. Involucral lobes large, more or less foliaceous, lanceolate or ovate-lanceolate, 4-8 mm. broad near the base, more or less cover- ing and concealing the long-cylindric, yellow-setose calyx. Mature staminal column exserted 14—14; its length. Distribution: Peru and Colombia (see Map 3). Specimens examined: COLOMBIA: PUTUMAYO—Umbria, Klug 1712 (Е). DEPT. IN pouBT—Tocamet, Schott 3 F | PERU: LoRETO—Florida, Klug 2077 (А, Е, С, M, NY, US); Gamitanacocha, Schunke 298 (A, F, NY, US). This variety is close to var. longifolius which it resembles espe- cially in calyx and leaf characters. However, it is the only variety with large, broad involucral lobes. PAVONIA Cav. 4. Pavonia firmiflora Schery, n. sp. Frutex undique stellato-pubescens; foliis maturis subrotundis, magnis, са. 14 ст. longis, 12 em. latis, aliquid 3—5-lobatis lobis mediis prominentibus, profunde cordatis, subtus pallidioribus et pubes- centioribus, 7-9 venis prominentibus palmatis; petiolis folia aequantibus vel longioribus; floribus terminalibus brevi-racemi- formibus ; pedicellis 0.5-6.0 ст. longis; involucris subpatelliformi- bus ca. 8-9-lobatis, stellato-pubescentissimis, lobis linearibus, acutis ; calycibus 1.0-1.5 em. diametro, extus stellato-pubescentibus, [Vor. 29 228 ANNALS OF THE MISSOURI BOTANICAL GARDEN intus brevi-mollissimis, 5-lobatis, lobis lanceolatis ; petalis obovatis, inaequilateralibus, retusis, non-aurieulatis, 3.5—4.0 em. longis ; tubis staminalibus non exsertis, columniformibus, supra medium in fila- menta congesta monanthera divisis, apice nudis, dentatis; ovariis subglobosis, rigide pubescentissimis, 5-carpellatis, carpellis uni- ovulatis et monoseminatis ; stylis 10; fructibus juventate rigide pu- bescentissimis; seminibus subreniformibus, glabris. Distribution: Jalisco, Mexico (see Map 6). Specimens examined: MEXICO: JALISCO— Tequila, Pringle 5447 (Е, С, M түре, NY, US). DN MA y NE 2.2 Pavonia firmiflora, x approx. 16. The Pringle specimens have for almost 50 years been labeled in herbaria аз“ Malvaviscus acerifolius Presl. (ex. char.)."" Only this collection ( Pringle 5447) has ever been made of specimens referable to the new species, P. firmiflora. 'The specimens differ from those of the original deseription of Malvaviscus acerifolius especially in having stiffly pubescent fruit and a shortened staminal column. A comparison of Pavonia firmiflora with a photograph of the M. aceri- folius type specimen shows P. firmiflora to have a coarser, stouter appearance and different leaf texture. Pavonia firmiflora is close to Pavonia Palmeri (Baker) Schery (formerly Malvaviscus Palmeri), but has much larger flowers, a condensed inflorescence, and only slightly lobed leaves. It is pos- sible that P. firmiflora is a hybrid between Pavonia Palmeri and 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 229 perhaps Malvaviscus candidus. The fact that only one collection has ever been made would tend to support this view. The border-line between Pavonia and Malvaviscus is vague. It is difficult to tell where a fleshy fruit stops and a dry fruit begins, to use this classical character of demarcation between Malvaviscus and Pavonia. This is true of Pavonia firmiflora; yet P. firmiflora is so similar to P. Palmeri that inclusion in the genus Pavonia seems entirely warranted, especially since P. firmiflora does not have auriculate petals, an additional definite character of Pavonia, sug- Pavonia Palmeri, x approx. 16. gested in the introduction to this monograph for use in distinguish- ing between Malvaviscus and Pavonia. The whole P. firmiflora plant has a coarse, dense, stellate pubes- cence. The cordate leaves are slightly lobate, a lighter gray color and more densely pubescent on the under side than above. The in- florescence is almost a condensed raceme, although occasional flowers are axillary in the upper leaves. The involuere is broad, with narrow linear-lanceolate lobes, shorter than the robust calyx. The calyx is minutely pannose within, with a circular bare area at the base surrounding the 5-carpellate densely haired ovary. There is one subreniform seed almost completely filling each сагре!. [Vor. 29 230 ANNALS OF THE MISSOURI BOTANICAL GARDEN 5. Pavonia Palmeri (Baker) Schery, n. comb. Malvaviscus Palmeri Baker f. ex Rose in Contr. U. S. Nat. Herb. 3: 313. 1895. Malvaviscus cinereus Baker f. ex Robins. & Greenm. in Am. Jour. Sci. III, 50: 176. 1895, nomen; E. G. Baker in Jour. Bot. 37: 347. 1899. Pavonia amplifolia Standl. in Pub. Field Mus. Bot. 4: 230. 1929. Distribution: western Mexico (see Map 6). ee examined: XICO: JALISCO—San Sebastian, Mexia 1480 (Е, С, M), Nelson 4061 (С, US). essai inq Palmer 1835, in part (F, G), 1990 (F, G, NY, US). The transfer of this species from Malvaviscus to Pavonia is made on the basis of the general structure of flower and fruit. The petals are not auriculate and the fruit can scarcely be termed ‘‘fleshy.’’ EXCLUDED SPECIES Malvaviscus cinereus Baker f. ex Rob. & Greenm. in Am. Jour. Sci. III, 50: 176. 1895, nomen subnudum ; E. G. Baker in Jour. Bot. 37: 347. 1899 = Pavonia Palmeri (Baker) Schery. Malvaviscus coccineus Medic. Malv. 49. 1787 - Hibiscus coccineus Walt., fide Ind. Kew., probably Pavonia sp. Malvaviscus floridanus Nutt. in Jour. Acad. Phila. 7: 89. 1834 - Hibiscus Bancroftianus Macfad., fide Ind. Kew. Malvaviscus fragilis Bory ex DC. Prodr. 1: 446. 1824, as syn. - Hibiscus liliiflorus Cav., fide Ind. Kew. (H. fragilis DC. loc. cit). Malvaviscus longifolius Spach, Hist. Veg. Phan. 3: 370. 1834 - Pavonia longifolia A. St. Hil., fide Ind. Kew. Malvaviscus montanus Mart. ex Garcke, Jahrb. Bot. Gart. Berlin 1: 222. 1881 - Pavonia montana Garcke, fide Ind. Kew. Malvaviscus multiflorus Spach, loc. cit. 1834 - Pavonia multiflora A. St. Hil., fide Ind. Kew. Malvaviscus Palmeri Baker f. ex Rose in Contr. U. S. Nat. Herb. 3: 313. 1895 - Pavonia Palmeri (Baker) Schery (see above). Malvaviscus populifolius Presl, Reliq. Haenk. 2: 135. 1853 = Hibiscus sp., as judged from the description. Malvaviscus populneus Gaertn. Fruct. 2: 253, pl. 135, fig. 3. 1791 - Thespesia populnea Soland., fide Ind. Kew. Malvaviscus puniceus Bory ex DC. loc. cit. 446. 1824, as syn. = Hibiscus liliiflorus Cav., fide Ind. Kew. Malvaviscus rosa-sinensis Moench ex Steud. Nom. Bot. ed. 2. 1: 760. 1841 = Hibiscus rosa-sinensis L., fide Steud. 1942] SCHERY—MONOGRAPH OF MALVAVISCUS 231 SPECIES OF DOUBTFUL STATUS Achania cordata Nees & Mart. in Nov. Act. Nat. Cur. 11: 99. 1823, not Malvaviscus, probably Pavonia. Same as Pavonia coccinea Willd., fide Nees & Mart. Achania floridana Raf. New Fl. N. Am. 1: 4. 1836? = Hibiscus Bancroftianus Macfad. (H. floridanus Shuttlew.), fide Ind. Kew. Achania Poeppigii Spreng. Syst. 3: 100. 1826 - Hibiscus Poep- pigii Garcke, fide Ind. Kew. (see Malvaviscus Роерріди). Achania stylosa Schrank in Flora 2: 449. 1819, nomen dubium. Achania tomentosa Sterler ex Steud. Nom. Bot. ed. 2,1: 12. 1841, nomen nudum. Anotea chlorantha (Malvaviscus chloranthus) Kth. ex Ulbrich in Fedde’s Rep. Spec. Nov. 14: 108. 1915, probably Pavonia sp. as judged from photograph of type and from the description. Anotea flavida (DC.) Ulbrich, loc. cit. 109. 1915, genus and spe- cies doubtful, probably Pavonia sp. Malvaviscus flavidus Moc. & Sessé ex DC. Prodr. 1: 446. 1824, genus and species doubtful, probably Pavonia sp. Malvaviscus pleurantherus Moc. & Sessé ex DC. loc. cit. 1824, genus and species doubtful. Malvaviscus pleurogonus Moc. & Sessé ex DC. loc. cit. 1824, genus and species doubtful. Malvaviscus Poeppigii (Spreng.) Q. Don, Gen. Syst. Dichl. РІ. 1: 475. 1831, nomen dubium - Hibiscus Poeppigii Garcke, fide Ind. Kew. Perhaps, were the deseription more complete and definitive, this name should be included as a synonym of Malvaviscus arboreus var. mexicanus. INDEX ro SPECIMENS CITED The number in italies refers to the collection number, the number in parentheses to the species or variety under which the specimen is cited: (1)—M. candidus; (3)—M. ar- boreus; (3a)—var. brihondus; (3b)—var. cubensis; (3e)—var. Птиттопай; (3d)— var. Hintoni; (3e)—var. longifolius; (3f)—var. mexicanus; (3g)—var. palmanus j (3h)—var. penduliflorus; (3i)—var. sepium; (3j)—vwar. Williamsii; (4)—P. firmi- Лота; (5)—P. Palmeri. Agniel, Bro. 10525 (3h); 10621 (1). pane Bro. Sig (3f). Aguilar, M. 87 (3f). Arbel КУР; J. Cuatreeasas. 2456 Alexander, R. C. — (3f). (ау, xo M Alfaro, A. 32378 (3h). Aristé-Joseph, Bro. 4975, 1060 (3); — Allart, A. ?6 (3 f) ME Е Allen, P. Н. 52 (3c); 91, 1045, 1176, 1864 Arsène, Bro. С. 34 (3); 128, 389, 1957 (32); 1906 (3h). (3f); 2729 (3, ЗЕ); 5494 (3); 10058 André, E. К. K862 (Зе). (3h). 232 ANNALS OF THE Artemio, Bro. 72 (3). Baker, C. Е. 7 (3); 163, 618 (3f); 824 (3) . A. Van Hermann, 4253 (3b). Bangham, C. M. 341 (3h). Barber, H. S. 23 (3e). Barnes, C. R., C. J. Chamberlain & W. J. G. Land. s Gi i). & W. J. G. Land. 203 (3). Bartlett, H. H. vid (3e). Berg, a Betlandie er, J. L. pi (3f). Biltmore Herb. 11079, a, b, c, d, & e (3e). Biolley, P. 43, 7403 (3h) ; 8977 (3). Bodin, J. E. 129 (3e). Boguseh, E. R. 115 (3e). Botanie ее 508 (3). Bourgeau, E. 1512, 1669 (31). Braee, L. J. К. 1527, 3541 (3b). Brandegee, Е 8. — (3#). Brenes, A. M. 5954, 13418, 15630 (3g); 16967 (3); 16998 (3h); 17050, 19308, 21480, 21917 (3); 21969, 21982, 22623 (3g). Britton, М. L. 999, 2917, 3995 (3f). —— — & J. Е. Cowell. 9902 (3b). —— — & А, Hollick. 2093 (3f). —— & C. F. Millspaugh. 2442 (3b); 6145 (3f). —— — & P. Wilson. 511 (3); 4535 (3f); 14032 st (3b). A. D. Selby. 14498 Um Broadway, W. E. 7324, 7690 (3f). Brown, S. 78 (3f). Buckley, S. B. — (3e). Bush, B. F. 312, 1211 (3e). Calderon, S. 121 (3). Cardenas, R. 375 (3d). Carleton, M. A. 81 (3h). Chapman, А. W. — (Зе). Chase, V. Н. 7073 (3) ; 7481 (3f). Chaves, D. ?5 (3). Clark, O. M. 3977 (3e). Clemens, J. 609, 610 Conzatti, C. 1683, 1981 (3h); 3948, 4430 ) -—— & Camino, 2447 (3). —— — & V. Gonzalez. 27 (3). Cook, O. F. & R. D. Martin. 63, 197 (3). Cooper, G. P. 54 (3); 103 (3h). —— — & G. М. Slater. 26 (3h). (VoL. 29 MISSOURI BOTANICAL GARDEN Cooper, J. J. 5719, 10367 (3f). Cuatrecasas, J. 1850 (3f); 9153 (3h). Curtiss, A. H. 726 (3); — (3e). Davidson, M. E. 68 (3h). Deam, C. C. 161 (3f). Delgado, E. 271 (3e). Dewey, L. F. — (3b). Dixon, R. A. 441 (3b). Dodge, C. W. & V. F. Goerger. 9582 (3h). —— — & W. S. Thomas. 4784 (3h); 6324 (3f). Drushel, J. A. #844 (3e). Dugand, A. 887 (3e). & Barriga. 2476 (3f). Duges, A. 173 (3); 281 (1). Dunlap, V. C. 101, 349, 440 (3h). Edwards, M. T. 482 (3f); 818 (3). t Hn. No J: man, E. L. 2976 sed 10069 (3). a Bro. 428 (3e); 655 (3f). Emriek, C. M. 167 Ag Englesing, F. C. ?5, 94 (3f). Faweett, W. 8022 (3f). Fendler, A. 101 (3e Ferris, R. S. 5436, 6073, n (3f). & С. D. Dunean, e). Fisher, G. L. | $e (31); 1 Fredholm, A. 6413 (3e). Gale. — (3e). Garnier, Bro, A. 291 (3). Gaumer, G. F. 580, 1858 — (3f). & sons. 23361, 23523 (3f); 23686 (3, Gehriger, W. 565 eae Р.Н. 208 (3f) ; 870, 2873 (8a). Gentry, H. S. 5258 (3). eri Hoe A. 642 (3). Goldman A. 22 (3f); 461 (3); 578 (3f); 864, 904, 940, 983 (3). Greenman, J. M. 49 (32); 191 (31); 443, 5433 (3f); 5665, 5712 (3). Gregg, J. 1110 (3 Groth, H. A. 209 (3e). Hagen, С. & W. von. 1104 (3f). Hall, E. 53 (3e). 1942] SCHERY——MONOGRAPH OF MALVAVISCUS Hanson, Н. С. 2 (3с); — (3f). Harris, J. за 11148, 11885 (3). . V. Lawrence. C15297 (3f). Harris, W. 12863 (3f). Hart, J. ВОТ (OL); & C — (3b). Hateh, W. R., & C. L. Wilson, 360 (3h). Haught, O. 1532 (3h). Havard, V. — (3e). Heller, A. A. 1833 (3e). Heriberto, EH Ки (Зе). Heyde, E. T a (3). & AAT $90 05 (3). Hinton, G. B. 690 (3d); 3861 (3); 3928 (3d) ; 4014 (3h) ; 4289 (34); 4563 (3£); 5057 (3d); 5145 (3); 5254 (3f); 5871 (34); 6719 (3#); 7184 (3); 7912 (34); 9597 (8f); 10758 (8h); 10897 (8f); 11030, 13529, 14181 (3h . 8976 (3). Hiteheock, A. L. 21141 (3e); — (3b); — (3f). Holton. 748 (3). Holway, E. W. D. (32); 766 (3). Howell, A. H. 247 (3c). 276, 402 (3h) ; 529, 694 Jermy, G. 191, 271, 707, — (3e). Johansen, H. 29 (3f). Karling, J. S. 43 (3a). 3f); 4819 (3); ). 49 Kenoyer, 3, 772a (3); 772 (3+). Killip, E. P. 13511 (1). C. Smith. 14052, 14187 (3h); 16087 (3); 16890 (3h); 20179 (3); 23065 (3e); Klug, G. 1712, 24714 (3); 26648 (3e). 4382 (3e). Krukoff, B. A. 4589 (Зе); 5150 (3). ,F. H. 399 (3). quus E. 798 (3h); 924 (3f). Leon, Bro. 695 o 4745 (3f). ) n 84, 685, — (3e). loyd, C. A. 1108 (3b ell, C. P p 52, 480 (За); 970 (3); 4246 (3a); 4890 (3f). Lyonnet, E. 1298, 1323 (3). McCormick, P. — (3e). 2077 (3j); 3015 (3); 3919, 233 MacDaniels, D. H. 191 (3f); 305 (3h); 626 (1); 944 (31). ) Matuda, Е. 83 (3f); 95 (3); 1310 (31); 2158 (8f). Maxon, W. R. 5000 (3f). — — & A.D. Harvey. 8004 (35). —— — & A. T. Valentine. 7275, 7353, 7450 (3). —— & В. Ha 3080 RUE 3470 (3f). 8 (3f). e): 1023 (3f); 1135, 1267, 1442 (3h) ; 1480 eu 9094, 9204 (3h). eyer, W. C. Millspaugh, C. P 42, 981 (32); 1166, 1313 (3b); 1728, 1920 (3f). om E. R. 131 (3h). Mohr, C. H. — (3c) ; — (31). mur C. V. & E. Makrinius. 2891 (3h). Mutis, J. C. 2262 (3e) Nelson, P W. 848 (3f); 363 ра 405 (3f); 926 (3h); 1228, 1256 (3f b ps (3h); 3005 (3); 3170 E Зи (3f); 3454, 3563 (3); 3807 (3h); 835 (3); 4061 (5), 4392 (3c). ед G. Е. 63 (3f). Northrop, J. I. 3 (3). Oreutt, C. В. 2811 (31); 2899 (3f); 4294 3). Ortega, J. С. 5116 (34); 5183, 6114, 6444 (3); 7327 (3h). Padilla, S. A. 193, 197 (3f). Palmer, Edw. 94, 115, 387 (3c); 891 (3h) ; Ди (3); 5 E. ee nie (8f); 686 (1); 3 (3f); 0 (5). iM E. J. pe pho 10756 (3c). Palmer, W. & J. H. Riley. 597 (3f); 69? (3b) Pennell, Е. W. 3910, 17975 (3+). Plunkett, O. A. 45 c 140 (3f). Pittier, E. 49, * 85; па (3); 15 2770, 0 (3h); 7121 (3); 10575 (3h) ; A (3); 12020 (3h); 13014 (3); 13031, 13063 (3h); 14015 (3g). 234 Popenoe, W. 690 (3). Pringle, С. б. 1959 (Зе); 4132 (1); 4923 (3f); 5447 (4); 5609 (3); 5914 (3i); 5975 (3); 7833, 8202 (31); 8232, 8498, 9275, 9455 (3); 9662 (3f); 9688 (3). Purpus, C. A. 7450, 10729 (3i). Record, 8. J. — (За). Rehder, A. — (3f). Reko, B. P. 3348 (3h). Renson, C. 8 (3). Reverchon, J. 1195, 1197, 8827, — (3c). ). 3f) Rose, J. N. 2670 (1) ; 4027 (3h — & В. Hay. 5302 (3); 6154 (31); 6354 (1). W. Hough. 423? (3); 4245 (31); 4351 (3h); 4363 (3); 4587 (3f); 4883 (3c). — —, A. Pachano & С. Rose. 23478 (3h). & J. H. Painter. ?541 (3). . Rose. 8497 (3); 10220 H (3f). — — & G. Rose. 22595 (3). —— & P. G. Russell. 24292 (3f). — —, Р. C. Standley & Р. С. Russell. 18989, 14096, 14210 (8). Rothrock, J. T. 180, 237 (3b). (3f). Ruano, M. 405 (3f Runyon, В. — (3c). Rusby, H. H. & F. W. Pennell. 143, 170 (3) Saer, J. 65 (3e). Sargent, C. S. — (3c). Schery, R. W. 188 (3h); 204, 206 (3). Schipp, W. A. ?08 (3a Schott, 8 (31); 177, 271 (3f); 643 (3); — (3f). Schunke, C. 21, 4116 (3); 293 (3j) ; 1492 (3) Seaton, H. E. 800 (3i). Seibert, R. J. 263 (3h). Seler, G. & E. 2681 (3) ; 4913 (3f). Shannon, W. E x 384 (3f); 419 (3). Skutch, A. F. 541 (3); 878 (3h); 1021 (3); 2368 (3h); 3266 (3g). Smith, A. H20, A316, H1670 (3g); 4215 (3h Smith, C. L. 118 (3); 296, 638, 1023 ¢ 1030, 1030 (3f). [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN Smith, H. H. 492, ?34, 735, 2727, 2816 3h Smith, J. D. 16584 (3h); 1991 (3f) ; 6450 (3h); 7393 (3g). Solis, F. 266 (3) Standley, P. С. 19339, 197 18a, 19855, 20809 (3f); 21021, 21402 (3); pool 22300, (3f) ; hi 22975 (3); , 23857 (3f); 24587 (3h); 27924, pes (3f) ; 33209 ^us 33368 (3); 34685 (3g); 35756 (3h); 86585 (3g); 36987 (3h); 39640, 39837 (3g); 40163, 41854 (3h); 53743, 58926, 54021 86 66775, 67894 (3f); 68102, 68112, 68236 (3h); 69853, 72026 (3f); 72311 (3h); 74212 (3); 7524 78644, 79019, 79410 (3f 35, 81929 (3); 88014, 88119, 88240, 89189, 89488, 89571, 90542 (3f). —— & E posce 47926 (3 —— & Valerio. 43278 (3h); 44605, 46238 op; 49600 teere, W. С. 1127, 1618, 1642 (3f) ; 1924, 2020 ( Stevens, F. L Steyermark, ; 30360, 43581 (3f); 34516 (3h); 34989, 36553 (3); nae ти Stork, Н. Е. Miei 333 (3f); 1828, ui 2975 m» Tamayo, F. 385 (3e). Terry, R. A. 1586 ea Tharp, B. € (Зе Thieme, C. 5153 (3h); 5168 (3). Thurrow, W. F. — (3c). Tomas, Bro. 622 (3f). Tonduz, A. 814, 2 (3f); 2818 (3h); 2884 (3f); 6973 (3h); 6982 (3f); 7260 (3h); 7291 (3f); 7393, 8089 (3g) ; 8036 (3h); $946 (3f); 11631, 11693 (3); 12485 (3g); 13149 (3h); 13485 (3f); 17554 (3). Townsend, C. H. T. 49 (3c). Tracy, S. M. 7476 (3c). 1942] SCHERY—MONOGRAPH OF rni W. — (3e Triana, J. J. 388, 3132 Ro uerckheim von, 482 (3f); 111721 (3h); E 2812 pos 8185 (3h). Ule, E. 9591 (3e). University of Lima. 13, 75 (3). Valerio, J. 66 (3g); 67 (3). Е гіо, М. 220 (3). п Негтап, Н. А. 253 (3#). Vogl, Fr. C. ?6 (3e). d, L. Е. — (3с). — (3e). White, P. & G. MALVAVISCUS Wilkinson, E. H. 47, 102, — (Зе). 70 ms, L. 506, 2675, 8068 (3e); 8336 (3f) ; 8977 (3); 9189, 9219 (3f). Wilson, P. 42 (3h). & Bro. Leon. 11297 (3f). Woodson, В. Е. & В. W. Берегу. 188 (3h); 276, 302, 531 (3f); 550 (3h) ; 932 (3+). Wooton, E. O. — (3). Worthen, G. C. — : Wright, C. 2064, 2065, 2068 (3e). Wright, W. а. 1348 (3f). Yuncker, T. G, В. Е. Dawson & Н. Е. Youse. 5680 (3f); 5879, 5880, 6239 (3). Yuneker, T. G., J. M. Koepper & К. A. Wagner. 8523 (3h) INDEX ro SPECIFIC AND VARIETAL NAMES New names, varieties and combinations are printed in bold-face type, synonyms in italics, and valid names in Roman type. Page y ni E Ae Е... 203 ошата шелш ее 219 сосстеа ы-ы сут Ул Ее 209 CONCINNA Е. 210 Сота. 231 YLOT Ana коопериране ы-ы» 231 Маа 209 ОН RA M MUERE 209 DHosdq рио MCN XD LE ee bee tees 209 IDOCDDVUT o a hae oa tesa cst E E 231 SEL OSes sos A ie ie A CREAN MS 231 COMENLOSG а ме. 231 ANOT EOM э ct AGRE tee career eae 204 ОО а. 231 Соат 231 Hibiscus «BancroftjAnus scm uv sous 230, 231 COCGINGUS: о. dft es тел Drummond Sra entities muris 216 Tlortdanus Eo trices scaled tiene 231 аба о оо RES 230 ПО н 230 А И ов 209 FOGPPIRIT 2a a S 231 r BUR PM EE ео 210 това Вен ео: 230 ао 203 Page MAaÍVRVISCUS CCR To tel о 20 СОСЕТ. 209 асас! ойе аа. 210 ЗО Муш ух шл ir 209 var Ъгщопанз, А 213 var. CODON 0 ees 214 var. D ONG а поради 215 таг; GIISCOGCIIT ааа ааа IS AES 219 var. Hintonb 0270s 217 var-longifolius E ЕО 218 о е0 АГЕН С 219 уат. р ВОЯ ОЕТ 222 VAY: пао отива 210 var. enduliforus ОС 223 var QUIM LI СИЕ 210 таг, чаргагатиз 2...2 ad 219 var piae a ee Sb ld as EIS 226 VAT: IGOR по ана UTE 219 уаг. и TCU NC 226 В GLD ASUS rere telecine ее 210 brevibracteatus о а. 219 DIODES Eo но е 219 candidus а е 206 ohloranithus- о 281 к Уз ТАА 219 CURET CUS Mien T EE ле узе ат eae 230 [Уог. 29, 1942] 236 ANNALS OF THE MISSOURI BOTANICAL GARDEN P Page DONE а vi 6 za s Paw een eee EES 214 penduliflorus celo ss ebe 223 о kc Дзе ана сър 209 pentacarpus «cies 219 CONO nesses eek 223 pilosus seen 210 TT 210 Pleurantherus ................... 231 17 ТОРО OTI 209 pleurogot ROO ОА 281 lass и 218 Poeppigü |... 281 OP NCMO БАДАН gio ТЭНИН... a reri ioi Drummond ..................... 215 ВЕРНИ сз s pectine era T ПОРУМ E VERRE Hi pa ute 230 megane оао свора очень 218 ЖИЙИ ИО ИИ 206 favidus ......................... 281 pulvihatus . „аъ E но tenen 219 BERN э, эхлак аку» aa 230 ПОШО сеа Дус... ted 230 TOM кк 230 Mule ool баса. i. dit cx 219 Funkeamus ...................... 218 а с: екет атлет 230 glabrescena ...................... 223 барами ое асус 219 ИСНЕ EPET ei ee 219 ЗОМИН Ж УГ ГК ГАТ УЛО ГТ 226 COOTER ...................... 219 APOCRINE ЖИЙЕН АНЕЛ rr 210 Doo i О И ITE T 217 T AUI MI NAA ROS so ca IET 210 integrifolius ..................... 218 DM И ИЕ 218 Jordan- Мо... вае 214 а тт Lei EUREN IN DUI COREL EGO A 210 aa E EE EI E EE T E EE EE E TE DE CIT 223 [Dis CETT 226 ا‎ аавосрекофокъж веки Ку 21 d iso! АРТ 227 longifolius ууз скала у» а» Куз 218, 230 ROUT OMG о T 230 Malwavitotis on cc eorr UA 21 ONE ЕС 231 тауплепзїв ..................... ».. 218 DFMO iier ror aas 216 BOUM Cada E xar er owl ee we. 210 SHON копеле око 227 MONTANE .................... ... 230 longifolia cua es Pee носови 230 са, LUC es era X KR Yd 230 ПОПОВО ар роза ааа Веси E Race ik 230 ОДЮЩОШАШ а.о оо Пел, ve ne 219 pln. fo] РРО ЕК КЛ 230 ПИСОИЧОВИЕ 2.2.» ез5, „+++ Qa 223 DENT каска ООо сене сах 230 e ТИ 222 . 1 ers ee PPP 219 SEDED ааа o reru YS ES 0. DONNE ere ey Casadei 204 EXPLANATION OF PLATE PLATE 14 Cutting from a plant of Malvaviscus arboreus var. penduliflorus, showing unusual - : i i : З i > ANN. Mo. Вот. GARD., Vor. 29, 1942 PLATE 1 SCHERY—MONOGRAPH OF MALVAVISCUS (VoL, 29, 1942] 240 ANNALS OF THE MISSOURI BOTANICAL GARDEN PLATE — EXPLANATION О PLATE ‘igs. 1-6. Six different specimens of М. arboreus vars. mexicanus and penduliflorus collected in the same locality in Chiriquí, Panama, Note the variation in leaf and petal shape, number and form of calyx and involucral lobes, differences in leaf margin, ete. The specimen of fig. 3 would be classified as var. penduliflorus on the basis of flower size, while the other specimens would fall into the var. mexicanus group. The specimens of figs. Тапа 2 would be almost on the border-line between these two varieties, x №. ANN. Mo. Bor. Garp., Vor. 29, 1942 PLATE л SCHERY—MONOGRAPH OF MALVAVISCUS [Vor, 29, 1942] 242 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 16 Fig. 1. Two leaves from the same Jin int “ М. arboreus var. penduliflorus growing in the Missouri Botanical Garden greenhouse. > Fi ig. 2. Dense velvety pubescence of ем i aus of a leaf of M. arboreus. х 10. Fig. 3. Stellate pubescence on upper surface of a leaf of M. arboreus var. brihondus. Note occasional larger hairs. x 10. Fig. 4 Glabrous upper surface of a leaf of M. arboreus var. mexicanus. Note the longitudinal ridge of hairs on the upper side of the petiole in this specimen, x 1 ANN. Mo. Bor. GARD., Vor. 29, 1942 PLATE 16 SCHERY—MONOGRAPH OF MALVAVISCUS [VoL 244 ANNALS OF THE MISSOURL BOTANICAL GARDEN EXPLANATION or PLATE PLATE 17 Fig. 1. Calyx of М. arboreus var. mexicanus with other floral organs removed: left— external view; right—-calyx split open to show inner surfaces. x 4. Fig. 2. Single petal of M. arboreus var. mexicanus. х 2. Fig. 3. Upper portion of staminal column in three specimens of M. arboreus var. mericanus. Note the erect position of the style branches and the five-toothed tip of the staminal column in the younger (upper) specimen. Fig. 4. Involucre of M. arboreus var. mexicanus with all other floral organs removed: upper—involuere split open with inner surfaces exposed; lower—external view, х Fig. 5. Upper row, from left to right: empty opened carpel, seed, seed in position in opened carpel, single unopened carpel, and mature fruit of М. arboreus var. mexicanus. Lower row: same for M. candidus except mature fruit is omitted, x 2. ANN. Mo. Bor. GARD., Vor. 29, 1942 SCHERY—MONOGRAPH OF MALVAVISCUS PLATE d Annals of the Missouri Botanical Garden Vol. 29 NOVEMBER, 1942 No. 4 CONTRIBUTIONS TO OUR KNOWLEDGE OF AMERICAN CARBONIFEROUS FLORAS! IV. A New SPECIES or LEPIDODENDRON ELOISE PANNELL Formerly Graduate Assistant, Henry Shaw School of Botany of Washington University Previous investigations have established Lepidodendron as one of the most abundant and widely distributed genera during the Carboniferous period, and the results of the present study indicate that it was the dominant element of the southern Illinois flora in middle Pennsylvanian times. Lycopod remains are present in almost every specimen of the hundreds of coal-balls collected during the past two years from the Pyramid mine of the Binkley Coal Company in Perry County, Illinois. Over 100 stems referable to Lepidodendron have been found thus far. This study is based chiefly on a selection of about 25 of the best-preserved specimens exhibiting a representative pieture of the structural variations found in the various orders of branching. Associated with the stems are leaves, roots, and repro- ductive organs. Some of these have already been described (An- drews and Pannell, '42) while others will be considered here and in later studies. Lepidodendron scleroticum sp. nov. 'The designation of a new species of this supposedly well-known genus probably needs a few words of explanation. Since Sternberg’s 1A study aided by a grant from the Penrose Fund of the American Philosophical Society. ? A detailed description of the occurrence of coal-balls at this locality has been given in the first of this series of contributions (Ann. Mo. Bot. Gard, 29: 1—18. 1942). Issued December 18, 1942. (245) [Vor. 29 246 ANNALS OF THE MISSOURI BOTANICAL GARDEN account of the type species in 1823, more than 100 others have been deseribed from compression material and over 20 from petri- fied specimens. There is little doubt that a considerable amount of synonymy exists in this long list of names. It is equally obvious to any investigator who has studied the Carboniferous Lycopods at all carefully that we are only beginning to understand the morpho- logical boundaries of the group. Very notable contributions relative to their reproductive organs have been made during the past few years (Arnold, '40; Hoskins and Cross, '41; Schopf, '38; Andrews and Pannell, '42). It is hoped that future studies will shed more light on the relationships between the cones and isolated vegetative remains. Lepidodendron stems constitute the dominant element of our Pyramid Mine collections, and the genus is abundantly represented in the roof shales of certain southern Illinois mines. The large num- ber of specimens at hand makes possible a much more complete description of the stem structure of the plant as a whole than do most of the previously described species, many of which are based on a single specimen. It will also be shown that L. scleroticum pre- sents highly distinctive cortical eharaeters which in themselves ade- quately justify the new specific name. The stems range from 3 mm. to 9 em. in diameter, while larger stelar fragments or portions of bark were found indicating stems as large as 30 em. in diameter. Two specimens with abundant secondary growth have been selected to show the typical organiza- tion of the tissues and the distinetive features of the species (pl. 18, figs. 1 and 2). They are numbers WCB55 and WCB56 respectively, in the Washington University fossil plant collection. A description of these specimens is followed by a shorter consideration of certain others in order to present a composite pieture of the shoot system as a whole. The pith, which is present in all of the larger stems, is surrounded by the xylem cylinder of primary and secondary origin, which is in turn surrounded by the phloem. The cortex may be divided into three zones, although these are not always distinct. The inner cortex immediately exterior to the phloem is composed of delicate paren- chymatous and small sclerotic cells. The parenchyma cells are often poorly preserved and the sclerotic cells crushed. The highly dis- tinctive middle cortex consists of sclereids grouped into nests and surrounded by parenchyma. The outer cortex presents a dicty- 1942] PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV 247 oxylon arrangement of slightly elongate thickened cells and paren- chyma. The periderm forms a large portion of the stem. It is composed of compactly arranged cells that were intimately connected with the mechanical support of the tree. The cork cambium is well pre- served in some of the smaller stems. Development of the Primary Xylem.— In Lepidodendron scleroticum there is a marked relationship be- tween stem size and primary xylem development. In small twigs from 3 to 10 mm. in over-all diameter, the xylem cylinder is typically protostelic, measuring from .26 to .78 mm. (pl. 19, fig. 6). The proto- xylem is exarch and unevenly distributed around the periphery. Branches with steles as small as this never exhibit any secondary xylem. Stems from 10 to 22 mm. show a mixed protostele varying from 1.04 to 1.56 mm. in diameter (pl. 19, fig. 3). When the cylinder of the primary xylem attains a diameter of 1.5-2 mm. secondary wood begins to form. In older stems from 3 to 5 em. in diameter the TABLE I SHOWING THE RELATIONSHIPS BETWEEN DIMENSIONS OF PITH AND PRI- MARY AND SECONDARY XYLEM IN THE VARIOUS BRANCH ORDERS Radius Radius Radius Total of the of the of the Stem No. diameter primary primary secondary of stele body xylem (Х,) xylem (X;) mm mm. mm. WCBI16 (2) .26 .13 .13 None WCB55A (2) .39 .18 .18 None WCB54 (2) 45 .23 .20 None WCB265 (2) 52 .26 .24 None WCBI6B (3) 52 .26 .24 None WCB54 (3) 1.04 53 31 None WCB55A (2) 1.10 55 45 Мопе МСВ124А (1) 1.30 .65 .55 None WCB54C (7) 1.56 .78 52 None WCBI6B (3) 1.69 .64 58 one WCB54A (8) | 1.69 ‚84 ona X, beginning WCB58A (2) | 1.88 94 .78 X, beginning WCB106A (1) 2.08 1.04 .80 None WCBI124A (1) 2.08 .84 62 м 9С (1) 4.81 1.23 65 1.17 CB135A (2) 5.20 1.69 91 0 WCBI180A (2) | 5.72 1.30 84 1.56 WCB57A (2) | 5.98 1.69 91 43 WCB56B (7) | 6.24 97 .30 CB148A (2) 6.76 1.82 1.04 1.52 WCB55A (10) 6.86 1.56 1.04 1.82 WCBOIB (1) 7.80 1.82 Ei 2.08 WCB18A (2) 7.8 1.95 LET 2.21 WCB20X (5) | 8.12 2.04 1.04 2.20 [Vor. 29 248 ANNALS OF THE MISSOURI BOTANICAL GARDEN xylem cylinder averages 6.7 mm., of which 3.1 is primary and 3.6 is secondary. The pith is now well developed, forming about one-half the total diameter of the primary body. Figure 6 is a photograph of the largest stele found, the secondary xylem measuring 22 mm. in radius and the primary xylem 6-7 mm. "Table 1 presents a more detailed eompliation of the dimensions of 24 well-preserved stems. The sequence of development shown in the table is uniform, and since only stems with well-preserved cortical tissues were used there is no reason to believe that they are not all referable to L. scleroticum. A study of this series not only clearly illustrates the intra-stelar origin of the pith, but it presents evidence that the whole primary body inereases in size even after the initiation of secondary wood. The evidence is based on the three following points: (1) no see- ondary wood around a protostele or a small primary cylinder was observed; (2) all the larger stems showed a siphonostele which probably developed from the smaller mixed protostele during the formation of the first few mm. of secondary growth; (3) wedge- shaped gaps exist between the secondary tracheids along the con- tact zone between the primary and secondary xylem (fig. 8) which are constant in all stems of any appreciable secondary growth. 'The gaps may be due either to pressure exerted by the increased cir- eumference of the primary cylinder or to the decay of paren- chymatous tissue laid down during the early stages of cambial aetivity. The rather fine preservation of the immediately adjoining parenchyma cells in both primary and secondary wood leaves little doubt that the first explanation is the correct one. The increase in size of the metaxylem was probably accomplished both by the maturation of the individual tracheids (cf. figs. 3, 6) as well as by the division of parenchyma cells associated with the tracheids. Bower (730) has been the most active investigator of the develop- ment of the primary body of vascular eryptogams. His research attempted to determine the physiologieal relationships between tracheids and living parenchyma cells which seemingly govern the ontogeny of the primary stelar body. He dealt largely with the different types of steles exhibited in the various species of Lycopods and concluded that the primitive xylie column in the fossil forms may undergo one or more of four types of progressive changes in order to maintain a more or less constant tracheid-parenchyma relationship. These may be briefly summarized as: 1942] PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV 249 1. A ‘‘fluting’’ of the periphery of the primary body is noted in certain species (L. Harcourtii Witham, and L. selaginoides Binney). This fluting resulted in a series of concave crenula- tions which increased the surface area. 2. Medullation occurs in most Lepidodendrons, but some are reported as protostelic while others develop a mixed proto- stele or siphonostele. Bower described branches of Lepido- phloios Wunschianus Carruthers which show a general rela- tionship of stem size to pith development, but he made no ob- servations on the beginning of secondary growth in the various twigs. 3. In species which developed secondary xylem the wood rays apparently served to retain a more or less constant tracheid-parenchyma relationship. Secondary wood may occur in both protostelic and medullated species although in some it is reported as absent. 4. Segregation of the primary xylem into separate strands occurs in certain Lepidostrobus cones (L. Brownit). These four tendencies apparently helped maintain a nearly con- stant ratio of living to dead cells in the progressive evolution of the vascular tissue in the fossil Lycopods. The Secondary Xylem.— The fossil Lycopods produced little secondary wood in compari- son with the size of the trunk. In most of the Lepidodendron scleroticum stems studied, which were about 4 cm. in diameter, the ring of secondary xylem measured from 1 to 2 mm. in width. Iso- lated fragments in which the secondary wood reached a radial dimension of 5 em. were found, but it cannot be said definitely that they belong to L. scleroticum, since the outer parts of the stem were not preserved. However, the specimen previously discussed (pl. 19, fig. 4), which developed secondary wood 3 em. in thickness, possessed enough fragments of the characteristic cortical sclerotic nests to be assigned to this species. At first, secondary growth is slight and develops frequently only on one side, with gaps between the active cambial cells. In such specimens the cambial development is closely comparable to that found in Lepidodendron fuliginosum Williamson, L. intermedium Williamson, and L. obovatum Sternberg. This peculiar type of cambial activity is described by Scott (220, p. 137), for the above three species as follows: [Vor. 29 250 ANNALS OF THE MISSOURI BOTANICAL GARDEN ‘‘ The cambium was an anomalous one, arising in various parts of the phloem zone and perieyele. It produced a good deal of the secondary parenchyma, among which there are usually scattered groups of wood; the secondary tracheids have a very sinuous and irregular eourse. We may regard these species as exhibiting either a primitive and rudimentary or reduced form of secondary wood." Lepidodendron scleroticum shows this supposed primitive char- acter only at the beginning of the growth of secondary wood, for after 10 to 20 xylem cells have been formed in an irregular manner the tracheids around the entire primary stele develop actively. There is frequently unequal growth of secondary wood on one side of the cylinder, probably due to some environmental condition. The first few secondary tracheids formed average 52 y in width; but after 10 to 20 cells have developed, the tracheids range from 88 to 104 и. None of the secondary tracheids reached the robust size of the primary ones which average 130 и, some even 195 и. Although the length of the tracheids is difficult to follow, due to their frequent distortion by rays and traces, approximately 30 cells were fol- lowed in their entirety. These were found to average 12 mm., some reaching 15 mm. Near the ends, the tracheids gradually tapered to a point. Using the maceration technique, the pronounced scalariform seulpturing of the secondary walls of the primary and secondary tracheids was studied in detail (fig. 5). The transverse bars are about 5.3 и in width and are spaced at a slightly greater distance. They extend completely across the wall face and at the corners merge into a continuous layer. The bars frequently dichotomize but with no regularity. Extending aeross the openings between the bars is a series of numerous fine threads which have been de- seribed in several other species of Lepidodendron as well as in Lepidostrobus Bertrandi Zalesky, Sigillaria Boblayi Brongn. and Stigmaria ficoides Brongn. Three interpretations of their mor- phology have been recorded: 1. Seward and Hill (’00) describe them as post-mortem changes in the primary wall which made the pit-closing membrane torn and fragmentary. 2. Calder (732) considers them, in Lepidodendron Brownti Un- ger, as tertiary thickenings laid down in a thin layer around the bars of the scalariform thickenings, with the fine threads connecting the layers. 3. Arnold (240), in his description of Lepidodendron Johnsonii, regards them as part of the secondary thickenings which lay against the primary wall and bridged the pit cavity on the inside. 1942] PAN NELL—AMERICAN CARBONIFEROUS FLORAS. IV 1 1 dem 0 к=н i 9 р i 7) j | E 1f E t | 0: Sj l i ppm 4 Е V | D Ж | Pe: Ф D : V t ; G # $ j 4% E: ред Ф Я $ $ Eus Ф $ [—— =e IZAN: И In m ma АИ а = БЫ ла "RD 2i 11 Text-figs. 1-2. ит ни scleroticum: fig. 1 B section of sealariform AS x 100; , tangential section of wood, x 251 [ VoL. За 252 ANNALS OF THE MISSOURI BOTANICAL GARDEN A study of the macerated tracheids of L. scleroticum presents strong evidence supporting their secondary nature as described by Arnold. The threads show definite connection with the secondary thickenings (text-fig. 1). They are also too evenly spaced to be torn fragments of the primary thickenings as Seward and Hill regarded them. Their structure appears similar to the outgrowths of the secondary walls in the vestured pits of the Angiosperms described by Bailey (238). The sealariform thickenings and the threads seem to be of the same material with no outer layer of the bars or tertiary thickenings as reported by Calder. The fine strands are, as far as it is possible to determine, secondary thickenings of the same ma- terial as the scalariform bands lying next to the primary wall con- necting adjoining bars. The wood rays vary from a single row of cells to many cells high and several cells wide (text-fig. 2). The cells range from 15 to 30 u in diameter and are about three times as long (radially). The walls of all the ray cells are characterized by delicate scalariform thick- enings although these never become as pronouneed as in the tra- cheids or leaf trace elements. There is no reason to believe, how- ever, that they were not living cells, and they present the only possibility of maintaining the ratio of living to dead cells in the secondary xylem. On the basis of studies in other plant groups the rays of L. scleroticum are clearly of an exceedingly primitive nature. This is suggested both in the variability in size of the rays as a whole, as well as the wide range in individual cell size. A less specialized ray strueture would be difficult to conceive. The leaf trace originated from the protoxylem cells at the per- iphery of the primary xylem and followed a horizontal course through the entire extent of secondary wood in all the specimens studied. This feature is of considerable interest since the largest stems had certainly shed their leaves before any appreciable amount of secondary growth. Both the tangential sections of the leaf bases and the impressions of the Lepidodendron stems in the shale indi- cate that only small twigs actually bore leaves. This persistent growth of the leaf trace has been noted in certain species of Arau- caria, living and fossil. The character is probably present in other species of Lepidodendron, but it has not been studied in detail. Near the primary wood the trace is surrounded by only a few ray cells, but as it progressed through the secondary wood the surrounding ray tissue inereased in height and breadth. Text-fig. 2 1942] PANNELL—-AMERICAN CARBONIFEROUS FLORAS. IV 253 was drawn from а tangential peel taken about 3 mm. from Ше periphery of the primary wood. The amount of xylem cells remains relatively constant in all the traces, as well as within the same trace as it progresses through the secondary wood. The protoxylem, if such were differentiated, is not distinguishable. The leaf trace follows an upward course through the cortex and then resumes a horizontal course through the extensive periderm tissue. This will be considered in more detail in the discussion of the extra-stelar tissues. Branching.— An excellent specimen showing monopodial branching is included in our collections (No. WCB254). The block was sectioned into nine parts and the stele studied at the various stages of development. The progressive origin and departure of the branch trace, with the closing of the gap, are illustrated in pl. 21, figs. 13-16. In fig. 13 the stele is seen before the departure of the branch, while fig. 14, which is 5 mm. above, represents the beginning of the trace. In fig. 15 the branch trace has progressed in its upward course 10 mm., and in fig. 16 about 20 mm., from the point of origin. The new stele is at first crescent-shaped, and as it departs, some secondary wood is carried with the primary tracheids (fig. 15). The branch gap is then bridged and the stele of the branch becomes circular, losing the secondary wood. Shortly above the last stage figured, the speci- men becomes poorly preserved and the newly formed branch is completely lost. In fig. 16 at pa band of tissue, secondary in origin and presumably periderm, may be noted. In all probability this represents the junc- tion of the stem and branch cork but it was not possible to follow it up high enough to observe the actual separation of the branch. Most of the stem compressions observed in the roof shales in southern Illinois mines indicate that dichotomy was of more fre- quent occurrence than monopodial branching, and only a few poorly preserved specimens branching dichotomously have been found in the coal-balls. Although a reconstruction of L. scleroticum probably compares closer to Hirmer’s for L. obovatum (Hirmer, ’27) than it does to Scott’s for L. elegans (Scott, '20) the types of branching un- doubtedly varied considerably in different species. Cambium and Phloem.— The delicate nature of the cambium and phloem prevents good preservation in fossil plants. In L. scleroticum, these tissues are [Vor. 29 254 ANNALS OF THE MISSOURI BOTANICAL GARDEN partially preserved only in the better specimens. 'The cambium ap- pears to have been eomposed of several layers of cells, and the surrounding phloem area is very small when compared with the large amount of secondary wood. These cells are represented in fig. 11. There have been two interpretations of the structure of the phloem cells in Lepidodendron: one, supported by Weiss ('01), states that sieve cells are present; the other view, that of Seward (702), holds that the phloem cells were not morphologically sieve cells but a type of secretory cell. It is only possible to say that the phloem in L. scleroticum probably consisted of very primitive un- differentiated parenchymatous cells not great in extent and little of secondary origin. The Cortex.— The cortical tissue is composed of three zones which are distinet even in young twigs. The inner cortex directly adjoining the phloem consists of delicate parenchymatous cells and small heavily thick- ened sclerotic cells. The parenchymatous cells are usually poorly preserved and the sclerotic cells crushed. The disorganized tissue forms a band next to the xylem in most of the stems (fig. 21). The remaining cortical cells, separated from the inner cortex by a cavity caused by the decay of that tissue, are well preserved. The middle cortex is composed of heavily thickened sclerotic cells grouped into nests and surrounded by parenchymatous cells most abundantly associated with the leaf traces which follow a gradual upward course through the cortex. The position, arrangement, and size of these cells may be noted in figs. 17, 18, and 19. The outer cortex consists of parenchymatous and sclerenchym- atous cells which are less thickened and more elongated than the sclerotic ones of the middle cortex. At first there are only a few small elongate sclerenchymatous cells among the sclerotic nests, but in the outer portions of the cortical region a reticulate network is formed from the decayed parenchymatous cells surrounding the leaf trace and the fibrous sclerenchyma (fig. 20). The composition and arrangement of the cortical tissue present the most distinctive characters of Lepidodendron scleroticum. The sclerotic nests, which have never been reported in a species of Lepidodendron, make possible specific recognition when only frag- ments of the stems are preserved. The reticulate network'of the outer portions of the cortex has been described in L. esnostense and L. rhodumnense Renault (’79). These species differ from L. sclerot- 1942] PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV 255 icum not only in cortical tissue, for they have no sclerotic nests, but they were protostelie with no secondary xylem. The Periderm.— The periderm of the fossil Lycopods is one of their most char- acteristic features, since it was not primarily a protective tissue, but served as the main support of the stem. The early English paleobotanists regarded it as vascular tissue or ‘‘pseudo-wood’’ (Binney, 1862, Witham, 1833). Williamson (1872) was the first in- vestigator to dispel this erroneous concept. Since that date our knowledge of this tissue has rapidly increased, terminating in the thorough study by Kisch (713). In Lepidodendron scleroticum, the cork cambium arises early in the growth of the twig, often before the mixed stele has developed an active vascular cambium. The phellogen originates 5-10 cells within the epidermis and divides rapidly. As in other species of Lepidodendron, the division is the exact reverse of the phellogen activity in modern trees, for the cambium lays down the great bulk of the tissue centripetally with only a small portion of phellem. A young twig shows little cellular differentiation of the cork (fig. 23) until after 2 to 3 mm. of growth. Its cells near the leaf bases divide radially, which increases the ci ference of the stem and keeps the leaf bases intact. Most of the periderm cells are 6-10 times longer than broad and are sharply pointed at the ends. They vary from thick and heavy to fragmentary and thin (fig. 22). The transition occurs in frequent succession and the partial decay of the thin-walled cells produces a series of holes. Such a periderm structure has been described in Lepidodendron selaginoides and Lepidophloios Wunschianus, but the degree of thickening of the preserved cell walls in those species is more constant. These gaps in the periderm have been described by Hovelaeque (792), in Lepidodendron selaginoides, as less-resistant layers formed at periods of sluggish growth, but recently they have been interpreted in another species as secretory strands (Arnold, '40). Although some of the cells surrounding the cavities show horizontal septa in tangential section, there are no other indications that the gaps were anything but decayed cells of less-resistant structure. The leaf trace follows a horizontal course through the great mass of periderm. Small sclerotic cells of the cortex may follow the trace for a distance into the cork, but the parenchyma cells associated with the trace have greatly decreased in mass. [Vor. 29 256 ANNALS OF THE MISSOURI BOTANICAL GARDEN The Leaf Bases.— The structurally preserved leaf bases of L. scleroticum were studied by peels made tangential to both young and old stems and were found to have the characteristic shape shown in fig. 26. The general appearance of the leaf cushions is a spiral arrangement of rhomboid struetures separated by a narrow groove where the stem surface was exposed. A small ligule, sunken in a pit, ap- pears near the apex of the cushion. Beneath the ligule is the vas- cular bundle surrounded by a cavity of decayed parenchyma and phloem. In most species, parichnos strands are present on either side of the vein. Young twigs of L. scleroticum show parichnos composed of 10-15 parenchyma cells, but in older cushions the parenchyma cells and phloem have decayed, leaving a cavity sur- rounding the lower parts of the xylem cells of the trace. Several isolated fragments of leaf bases referable to Lepido- phloios were also found in the Pyramid Mine coal-balls (fig. 24). Judging from the relatively few specimens found, however, this genus constituted but a minor element in the flora. A considerable number of well-preserved supposed Lepidodendron stems have been checked by means of sections taken through the leaf cushions to insure their correct identification. It is hoped that future collections may shed additional light on this Lepidophloios species. An attempt has been made to correlate Lepidodendron scleroticum with certain of the large Lycopod trunk compressions occurring above the same coal seam in adjacent parts of Illinois. Approxi- mately 25 specimens of stem compressions of various sizes were collected from the shale above coal #6 at the Old Ben Mine #11, Franklin County, Illinois. Exact specific determinations of im- pressions and compressions is difficult due to the varying degrees of preservation. It was possible, however, to separate the collection with reasonable accuracy into two species of Lepidodendron and one of Lepidophloios. The impressions of the leaf bases of Lepido- phloios laricinus Sternberg closely resemble those of the petrified Lepidophloios figured. Of the two Lepidodendrons, L. Volkman- nianum Sternberg (’25) compares more favorably with leaf bases of L. scleroticum. The other Lepidodendron impressions are prob- ably L. rimosum Sternberg, for the leaf bases are slender. Lepido- drendon V olkmannianum (fig. 26) has been reported a number of times from the southern Illinois area, although it was described under different specific names (Lesquereux, '66, '70; Noé, 25). 1942] PANNELL——AMERICAN CARBONIFEROUS FLORAS. IV 257 The size of the impression specimens of L. Volkmannianum on the shale above the coal shaft makes a probable reconstruction pos- sible. Figure 26 is one of the smaller stems, about 115 inches in diameter. It was selected as an illustration because its size 1s com- parable to the majority of the Lepidodendron scleroticum specimens, and although the trace and parichnos cannot be clearly observed the leaf scar is easily recognized. The stem branches dichotomously at the top, but the preservation of one branch is poor. Other speci- mens show leaf bases from № to 34 inches wide. Judging from the size relationship of leaf bases and stem in the completely preserved impressions, these larger specimens would probably have measured 1-2 feet in diameter. Economic Importance.— In view of the fact that Lepidodendron stem remains are by far the most frequently encountered fossils in the Pyramid Mine coal- balls it is of interest, economically, to note that the coal itself is in all probability composed very largely of the stems and leaves of this plant. Since the periderm constitutes the greater part of the stems it is that tissue which is largely responsible for this rich deposit of coal. Diagnosis: Primary body protostelie to siphonostelie depending on stem size; secondary xylem present, abundantly so only around fully developed siphonosteles ; inner cortex characterized by prom- inent sclerotic nests, outer cortex a reticulate sclerotic net in tan- gential section; massive periderm irregularly zoned by decay of less-resistant cells; leaf cushions resembling closely those of L. Volkmannianum. Locality and Horizon: Pyramid Mine of the Binkley Coal Com- pany, Perry County, Illinois; coal #6, Carbondale formation, mid- dle Pennsylvanian. Type specimens: WCB55 and WCB56, Washington University (St. Louis). Lycopod Organs Associated with L. scleroticum.— Leaves: Many Lyeopod leaves were found preserved near the petrified Lepidodendrom stems. Although the evidence strongly suggests their affinity with L. scleroticum, no stems were found ac- tually connected with the Lycopod leaves. The leaves vary from 1 [Уог. 29 258 ANNALS OF THE MISSOURI BOTANICAL GARDEN to 5mm. in width. Although their shape is frequently distorted, all their internal tissues show the same general arrangement. The length of the leaves is impossible to determine accurately since they were curved and broken before petrifaction. Several could be measured from 2 to 3 em. in length without any appreciable change in thickness. The xylem occurs in a single horizontally elongated strand com- posed of 40—50 tracheids, the amount of xylem varying with the leaf size. The protoxylem cells are difficult to distinguish, but they prob- ably oceur along the abaxial margin of the bundle. The metaxylem consists of scalariform tracheids. The phloem is concentrated on the abaxial side of the xylem and in some cases may enclose the entire xylem strand. A bundle sheath of thick-walled sclerenchyma cells surrounds the vein but is especially concentrated near the phloem. Around the bundle sheath, but more pronounced on the lower side, is an additional sheath of transfusion cells from 4 to 6 layers thick. Their diameter is equal to or greater than the xylem tracheids. Parenchymatous cells of about the same size but with- out the thickenings are mingled with the transfusion cells. The hypodermal layers, from 4 to 6 cells in thickness, add another unique character to the leaf structure. The tissue is composed of elongate sclerenchyma with transverse end walls, and presents a uniform transition from small, heavily thickened cells immediately beneath the epidermal layer, to larger, less-thickened ones compos- ing the last few rows. The stomatal grooves appear on either side of the thickened por- tion of the blade surrounding the vein. The sub-stomatal area con- sists of modified hypodermal cells with air passages between the groups of small rectangular cells. The mesophyll is poorly-preserved but, as far as can be deter- mined, it consists of more or less isodiameteric cells closely packed together. There appear to be no air spaces between cells in the blade itself, but there are large ones in the stomatal chamber. No evidence of differentiation into palisade and spongy tissue exists. The meso- phyll, which was probably the only ehlorophyllous tissue, constitutes a relatively small portion of the leaf. Lycopod leaves have been described in this country by Graham (35) and Reed (241) under the generic name Lepidophyllum. These leaves, associated with Lepidodendron scleroticum, compare closely with Lepidophyllum Thomasi Graham. 1942] PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV 259 Associated Fructifications: Several Lycopod microsporangiate cones and many seeds, one containing a well-preserved gametophyte, were found in the coal-balls. A complete description of these fossils has been presented under the name Lepidocarpon magnificum (An- drews and Pannell, '42). Although the evidence is incomplete it seems likely, on the basis of association, that these fructifications belong to Lepidodendron scleroticum. Discussion.— An observer of the abundant Lycopod remains of the Carbonif- erous period cannot help postulating an explanation for the ex- tinction of that great flora. The climatic conditions of that era have been considered as producing a swampy habitat similar to the present-day great Dismal Swamp in Virginia. This belief is sub- stantiated by the associated fern and horsetail remains occurring with the Lycopod fossils. The great abundance of Lycopod organs and the excellent preservation of the external surface of the stems give considerable evidence for the ‘‘in situ’’ origin and fossilization of the material. A study of the petrified Lepidodendrons presents a striking varia- tion from the usual concept of hydrophytic characters. The evolu- tionary tendency in these Lycopods was towards extreme bulk of dead cortical tissue. This tendency expressed itself in the stem by the bulky periderm and the small xylem and phloem cylinder, and in the leaves the hypodermal layer decreased the photosynthetic area. The apparent xerophytic modification of the leaves may have come about to prevent evaporation, not because of the scarcity of water but because the small stele was unable to conduct rapidly enough. The tree would be efficient under moist conditions as the seeds were probably quick to germinate and growth was apparently regular and rapid. But with the increasing dryness of the Permian period, the tree with its excessive modification in bulk and poor conducting and photosynthetic abilities rapidly became extinct. Acknowledgement.— The writer wishes to express her appreciation of the many serv- ices rendered by the staff of the Missouri Botanical Garden during the course of this work and is especially grateful to Dr. Henry N. Andrews for his patient guidance and constructive criticism. Thanks are also due the Binkley Coal Company for their continued good will and cooperation in allowing coal-ball collecting in the Pyramid Mine. [ Vor. 29, 1942] 260 ANNALS OF THE MISSOURI BOTANICAL GARDEN BIBLIOGRAPHY Andrews, H. N., and E. Pannell (1942). — г) to our knowledge of Ameriean arboniferous floras. II. Lepidocarpon. Ann. Mo. : 19-34. Arnold, C. A. (1940). Lepidodendron Johnsonii, sp. nov., ges the Lowe Pennsylvanian of Central Colorado. Univ. Mich. Mus. Paleont. 6: 21-52. Bailey, I. W. (1933). The eambium and its derivative tissues. Jour. Arn. Arb. 14: 259- 273 Binney, E. W. (1862). On some plants showing structure from the lower coal measures m Quart. Journ, Geol, Soc. 18: 106 Bower, F. үч (1930). Size and form in plants. London Calder, M. G. (1932). Notes on the Kidston collection of fossil plant slides. Trans. Roy. Soe, Edinb. 57: 547-555. Graham, R. (1935). An anatomical study of the leaves of the earboniferous arborescent eopods. Ann. Bot, 49: 587-608. Hirmer, M. (1927). Handbuch der Palüobotanik. Munich und Berlin. Hoskins, J. H., and A. T. Cross (1941). A consideration of the strueture of —— Seott based on a new strobilus from Iowa. Am. Midl. Nat. 25: 523- Hovelaeque, M. (1892). Recherches sur le Lepidodendron selaginoides бика; Mém. Soe. Linn. Normandie, 17: 1—161. Kiseh, М. L. (1913). The physiologieal anatomy of the periderm of fossil Lycopadiales. Ann. Bot. 27: 281-345. Lesquereux, L. (1866). Palaeontology: Deseriptions of plants. Geol. Surv. Ill. 2: "P — — — ——— (1870). Paleontology of Illinois: Descriptions of plants, ete. Ibid. Noé, A. C. (1025), Pennsylvanian flora of northern Illinois. Ill. State. Geol. бен: sd 52: 9-1 Reed, F. D. «ius Coal flora studies: Lepidodendrales. Bot. Gaz. 102: 663-682. Renault, B. ^ il Strueture comparée de quelques tiges de la flore Carbonifére. Nouv. Areh. Mus. Paris. Schopf, J. 7 (1938). Two new e seeds from the Illinois Pennsylvanian. Ill. State Geol. Surv., Cire. 28: 139-1 Scott, a H. (1920). Studies in Ruf Botany. I. Seward, A. C. m On the so-called phloem of Lepidodendron. New. Phyt. 1: 38-47. — — — ——, and А. W. Hill (1900). On the structure and affinities of а Lepidodendron ед tion "m calciferous sandstone of Dalmany, Scotland. Trans. Roy. Soc. Edinb. 39: 907. Weiss, F. E. (1901). On the phloem of Lepidophloios and Lepidodendron. Mem. Proc. Manchester Lit. and Phil. Soc. 14: 152. Williamson, W. С. (1872). On the organization of the fossil plants of the coal measures. Phil. Trans. Roy. Soe. Lond. 180: 197-235. Witham, H. (1833). The internal strueture of fossil vegetables found in the Carbonifer- ous and Oolitie deposits of Great Britain, Edinburgh. [Vor. 29, 1942] 262 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 18 Lepidodendron scleroticwm Fig. 1. Transverse section of stem: oc, outer cortex; me, middle cortex; Lt, leaf trace. WCB55A.7, x 3. Fig. 2. Longitudinal section of stem. WCB56I.20, x 3. ANN. Mo. Вот. Garb., Vor. 29, 1942 PLATE 18 PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV [ VoL. 29, 1942] 264 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 19 Lepidodendron scleroticum Fig. 3. Transverse section of mixed protostele. MCB54.3, x 57. Fig. 4. Transverse section through a portion of a large stele. WCB42.2, x 3.6. Fig. 5. Macerated tracheids showing ‘‘vestured ue iic е 1213, х 290. Fig. 6. Transverse section of protostele. WCB54. Fig. 7. Transverse section of siphonostele. NONI de x 14. Fig. 8. Transverse section through contact zone of primary and secondary xylem. WCB56B.7, x 57. Ann. Mo. Bor. Garb., Vor. 29, 1942 PLATE 19 PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV [ Vor. 29, 1942] 266 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 20 Lepidodendron scleroticum Fig. 10. Wood-ray cells in radial section, WCB90B.1 200. Fig. 11. Transverse section through outer part of oun xylem and phloem. WCB55A.10, x 75. Ee 19. Showing course of leaf trace from primary into secondary xylem. WCB561.2, Fig. 9. Leaf trace passing through secondary xylem, in radial section, WCB82B.83, 00 u—— EE IC H—————H—— C -— PLATE 20 ^a Sr р Ta на ) h mixes. TRIAN , Т ы mme ВИ MES AST „др * ы. Sg c "à е. ра > < (E вв er Е в: ма уз Ann. Mo. Вот. Garp., Vor. 29, 1942 IW М № | +. : ае": ` 4 - + ` ва , A e - КА Ti PAAA PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV [ Vor. 29, 1942] 268 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 21 Lepidodendron scleroticum Figs. 13-16. Part of a series of peels tracing the origin and departure of a branch stele: p, periderm. All figures x 8. ig. 13. WCB253A.1 Fig. 14. WCB253C.T2. Fig. 15. WCB253D.TI. Fig. 16. WCB253F.T?2. ANN. Mo. Bor. GARD., Vor. 29, 1942 PLATE 21 PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV [Vor. 29, 1942] 270 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 22 Lepidodendron scleroticum Fig. 17. Transverse section of cortex. WCB554A.7, x 8. Fig. 18. Radial section of cortex. WCB506I.20, x 8. Fig. 19. Tangential section бод central sapiat of cortex. WCB55III.515, х 7. Fi g. 20. Tangential section through outer region of cortex. WCB56I.B2, x 5. See text for detailed explanation, PLATE 22 Ann. Mo. Bor. Слко., Vor. 29, 1942 -$ he s? a AR, ; B ф» $ % ET я! | Ё EY T PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV [ Vor. 29, 1942] 272 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION or PLATE PLATE 23 Lepidodendron scleroticum Fig. 21. Transverse section through young stem, WCB106C, x 10. Fig. 22. E seetion through peride stem. WCB55A.7, 0 Fi ig. 23. Transverse section through periderm and leaf bases of a young twig. WCB54A.8, x 40. rm and leaf bases of a comparatively old ANN. Mo. Вот. GARD., Vor. 29, 1942 PLATE 23 Е 41! ut р ит "f і Pr 1 4 "ar 1 , p^ MD А ро ENDO) PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV [Vor. 29, 1942] 274 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION ОЕ PLATE PLATE 24 Fig. 24. Lepidophloios sp. from the Pyramid mine, Perry County, Illinois. WCB58B.9, x 5. ы 25. Lepidodendron scleroticum, tangential section through leaf eushions. WCB91.6, "Te. 26. Lepidodendron Volkmannianum (?) from Franklin County, Illinois. No. 1447, x 1. ANN. Mo. Bor. Garb., Vor. 29, 1942 PLATE 24 b axe > PANNELL—AMERICAN CARBONIFEROUS FLORAS. IV CONTRIBUTIONS TO OUR KNOWLEDGE OF AMERICAN CARBONIFEROUS FLORAS! V. HETERANGIUM HENRY N. ANDREWS Paleobotanist to the Missouri Botanical Garden Assistant Professor, Henry Shaw School of Botany of Washington University Our knowledge of Pteridosperm stem remains referred to the genus Heterangium is based chiefly on the earlier works of William- son and Renault and the later detailed studies of Scott and Hirmer. Judging from its rather frequent occurrence in England, Scotland and on the Continent, the genus was well established throughout the greater part of the Carboniferous period. Of the various plant organs that are assigned with some degree of certainty to the Pteridospermeae, Heterangium is of special sig- nificance because of its comparatively primitive structure as well as its occurrence in the early Calciferous Sandstone Series. It seems very likely that when its reproductive organs become better known we shall have a much clearer concept of how this early group of seed plants originated. An exhaustive review of the genus is unnecessary here inasmuch as the better-known species are adequately treated in certain of the standard texts and more detailed accounts may be found in the contributions of Scott (217) and of Hirmer (733). The latter work contains a useful key which outlines the major varia- tions within the genus. Although it has been known for some few years that Heterangium occurs in American Pennsylvanian coal-balls, no descriptions have been published. In 1935 Graham listed H. tiliaeoides as present in eoal-balls from the MeLeansboro horizon in Illinois, but his only comment relative to its occurrence was that, ‘‘Several stems were identified." In 1938 Fisher and Noé also reported Heterangium species from Calhoun coal-balls, but there were no accompanying descriptions. They also listed H. Grieviit as having been found, which is rather surprising in view of its much earlier occurrence in the Calciferous Sandstone Series in Scotland. These brief refer- ences constitute, so far as I am aware, the only published accounts of the occurrence of the genus in America. 1 Issued December 18, 1942. (275) [Vor. 29 276 ANNALS OF THE MISSOURI BOTANICAL GARDEN Heterangium americanum, sp. nov.— The following description is based on a number of stem and petiole remains found in coal-balls from the Calhoun coal, Richland County, Illinois. The locality is in the upper part of the McLeans- boro formation and is of upper Pennsylvanian age. A single speci- men of /Teterangiwm has been collected by the author from the Herrin (No. 6) coal at the Pyramid Mine three miles south of Pinckneyville, Illinois. This constitutes the top of the Carbondale formation, and although it lies somewhat below the Calhoun coal (Schopf, '41, chart p. 9) this particular specimen is referable to H. americanum. The specimens from Richland County consist of stem fragments bearing petioles, as well as isolated remains of both, and although the size and comparative development of secondary wood vary somewhat there are no distinctive features that necessitate segrega- tion of the specimens into more than one species. Primary Wood.— The diameter of the primary xylem varies in different specimens from 1.75 mm. to 4 mm. As in the European species of Heterangium, the protoxylem groups occupy a position very close to the periphery. Where it is possible to distinguish the protoxylem with certainty a few primary tracheids can be observed external to it although these pass almost imperceptibly into the secondary xylem. The proto- xylem thus occupies an excentrically mesarch position though it is very close to being exarch. The large metaxylem cells (pl. 25, fig. 3), which average about 260 ит diameter, are uniformly distributed throughout the parenchyma of the central cylinder. They may occur singly or in small groups of 2, 3 or 4 cells, but usually not more than that number except in the peripheral region where they pass into the protoxylem. The tracheids of the latter may be as small as 15 y in diameter, while the secondary tracheids average about 65 y. The pitting of the primary tracheids differs in no way from that deseribed for other species. Secondary Wood.— Some of the stems exhibit no secondary growth at all, while others show as much as 5 mm. (figs. 1, 4). The wood rays vary greatly both in height and breadth as well as in the size of the component cells. Text-fig. 2, drawn with the aid of a camera-lucida, shows a representative tangential section. The rays may consist of but a 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. V hii Text-fig. 1. Heterangium americanum: A drawing prepared with the aid of a eamera lucida showing division of the lateral and lobing of the central trace branches in a petiole. CB386.C. T11, x 17.5. [Vor. 29 278 ANNALS OF THE MISSOURI BOTANICAL GARDEN single row of cells while others are uniseriate but attain a height of 3mm. or more. One such slender ray may be observed near the right-hand side extending the entire height of the figure. Elongate fusiform bi- or triseriate rays are common and may even reach a width of 6 or 7 cells although all of these broad rays that were ob- served were undergoing division. The radial walls of the tracheids are characterized by closely compacted angular pits (text-fig. 4) like those found in other species of Heterangium, as well as certain other probably closely related Pteridosperms such as Lyginopteris, Rhetinangium and Steno- myelon. The pit borders are not well preserved in the radial walls, but where they can be observed the orifice appears to be rather broadly oval-shaped and horizontally elongated. The pits in the tangential walls are quite distinct from those in the radial walls, being nearly circular and loosely arranged in 1, 2 or 3 rows (text-fig. 3). The narrow orifice extends almost the entire diameter of the pit, and forms a cross with the pit in the wall of the adjacent tracheid. Leaf Trace and Petiole.— Owing to the fact that all the stems and petioles are short frag- ments, it has not been possible to follow any individual leaf trace from its origin in the stele out into the petiole. However, a sufficient number of specimens showed various stages in the course of the trace to give a reasonably complete picture of its anatomy. The leaf trace originates from the periphery of the primary xylem as two distinct bundles about two mm. apart (text-fig. 5). They start to divide almost immediately and are two-lobed at the time of their departure from the stele. In their passage through the cor- tex the division is completed and four separate bundles enter the base of the petiole (fig. 2). The outer or lateral branches (fig. 2, Та, Ls) resulting from the first division then divide into two small bundles which pass out toward the wing of the petiole (fig. 9, Lis, Lin; text-fig. 1). The two central bundles (fig. 2, cı, се) re- sulting from the first division then start to divide into three bundles each (fig. 5; text-fig. 1). That this division began prior to the sepa- ration of the petiole from the stem is evidenced in text-fig. 1 which represents a petiole still connected to its stem. One of the large cen- tral bundles is clearly three-lobed while the other has not quite reached this stage. Another petiole found isolated (fig. 5) exhibits the three-lobed condition in both bundles. No specimens have been 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. V 279 SEE P1304] за CASS өе SHORES 6200020 С а n I . oe ¢ 5 Text-figs. 2-5. Heterangium americanum: fig. 2, tangential section through the secon- dary wood. CB542.B1.F48, x 62; fig. 3, pitting in the tangential walls of the secondary tracheids. CB542.B1.F48, x 265; fig. 4, pitting in the radial walls of the secondary tra- cheids. CB542.B1.F48, x 265; fig. 5, portion of a stele in transverse section showing the two traces that will enter a petiole. CB386.D2.T10, x 17.5. [Vor. 29 280 ANNALS OF THE MISSOURI BOTANICAL GARDEN found showing a higher level of the petiole than that illustrated in this figure. It seems very likely, however, that the two central strands soon divided, resulting in a total of ten at a slightly higher level. No evidence is available pertaining to the further division of the traces or the petiole as a whole. Outside the bundles of the petiole shown in text-fig. 1 there may be noted a series of sclerotic nests embedded in large, thin-walled parenchymatous cells. The outer cortex is not well preserved, but judging from the more or less regular sequence of cavities (text- fig. 1) it seems likely that this zone consisted of alternate bands of fibrous and parenchymatous cells, the latter having decayed. The extra-stelar tissues of the stems are not well preserved al- though specimen CB542 displayed the horizontal sclerotic plates characteristic of the outer cortex of the genus. Diagnosis.— Primary xylem from 1.75 to 4. mm. in diameter, metaxylem tracheids about 260 и in diameter and uniformly distributed singly or in groups of 2, 3 or 4 cells; secondary wood well developed in some specimens, tracheids pitted on radial and tangential walls, rays variable in vertical and tangential dimensions; leaf trace double from the time of its origin, each trace dividing in its course through the cortex ; of the four bundles entering the petiole the two central ones become three-lobed while the two marginal ones each divide to form two distinct bundles. Locality and horizon: Richland County, Illinois; Calhoun coal, McLeansboro formation, upper Pennsylvanian and Herrin coal, Pyramid Mine, Perry County, Illinois; upper Carbondale forma- tion. All figured preparations are preserved in the Washington Uni- versity collections, St. Louis. The original blocks and a representa- tive set of preparations are deposited with the Illinois State Geo- logical Survey. Discussion.— In his consideration of the British Coal-Measure Heterangiums Scott ('17) proposed the sub-genera Eu-heterangium and Poly- angium to include those species characterized by having the vascular system of the petiole originate as a single or double bundle re- spectively. Hirmer (733) recognizes six species in the Polyangium group, in whieh H. americanum belongs. The two German species, 1942] ANDREWS—AMERICAN CARBONIFEROUS FLORAS. V 281 H. Kukuki and H. Hoppsteadteri, differ rather strikingly from the Illinois Heterangium in the arrangement of the metaxylem tracheids. In the last they usually occur in 2- or 3-celled groups which are uniformly distributed. The contrast is especially true in H. Kukuki, where the tracheid cells are in large groups of 10 or more and the groups are separated by prominent parenchymatous ‘‘ravs’’ (сЕ. Hirmer, 733, pl. 8, fig. 2). The affinities of the Illinois species seem to lie close to the English H. tiliaeoides and H. shorense. Graham recognized this relation- ship when he assigned his specimen to the former. The small size of the metaxylem tracheid groups in H. americanum offers a char- acter which likewise sets it apart from these English Heterangiums and which I believe is worthy in itself of specific distinction. Another feature, however, that has not been described for the other species is the apparent tri-partite branching of the central traces in the petiole (fig. 5). It may be, of course, that comparable portions of the petiole are present among the English specimens but until such is shown to be the case this may be considered as characteristic of H. americanum. In view of these differences it seems advisable to assign this new specific name to our specimens. Acknowledgement.— I wish to express my appreciation to the Illinois State Geological Survey, and especially to Dr. James M. Schopf of that organization, for the opportunity of studying and describing these American specimens of Heterangium. Literature Cited.— Fischer, M. C. and A. C. Noé ee A list of coal ball plants from Calhoun, Richland nty. Trans. Ill. Acad. . 31: 178-181 Bu R. (1935). тийе Үн in flora of Illinois as revealed in coal balls. IT. Bot. Gaz. 97: 156-168. Hirmer, M. (1933). Zur Kenntnis der strukturbietenden Pflanzenreste des jungeren Palaeozoikums. Palaeontographiea. B. 78: 57-113. Sehopf, J. M. (1941). Contributions to Pennsylvanian paleobotany. Mazocarpon oedip- PME sp. nov., and Sigillarian relationships. Ill. State. Geol. Surv., Rept. of In- t. 75: 1—53. Pls. 1—6. Scott, D. H. (1917). The ред сори of the British eoal measures. Jour. Linn. бос. Bot. 44: 59-105. Pls. 1-4. [Vor. 29, 1942] 282 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 25 Heterangiwm americanum Fig. 1. IM pU transverse section showing well-developed secondary wood. CB542.A1.813, x Fig. 2. "deli erse section of a portion of a stem and attached petiole. CB386.D.T21; Ly, e; and Lg, e, are the branch traces resulting respectively from the two traces enter- ing the petiole, x 8.5. Fig. 3. Stele of same, x 12. st 4. A smaller stem showing secondary wood. CB323.A2.89, x 8.5. Fig. 5. An isolated petiole showing partial division of the central trace branches, CB323.A2.89, x 8.5. Photographs by Eloise Pannell. Ann. Mo. Вот. Garp., Vor. 29, 1942 PLATE 25 ANDREWS—AMERICAN CARBONIFEROUS FLORAS, V A FOSSIL ARAUCARIAN WOOD FROM WESTERN WYOMING! HENRY N. ANDREWS Paleobotanist to the Missouri Botanical Garden Assistant Professor, Henry Shaw School of Botany of Washington University AND ELOISE PANNELL Formerly Graduate Assistant, Henry Shaw School of Botany of Washington University A few years ago the senior author made a collection of silicified woods from the Gros Ventre Canyon in western Wyoming. We were attracted to the region partly by some petrified log fragments on display outside the Jenny Lake Post Office, which were said to have been collected in the canyon, and partly by a landslide that presents a conspicuous scar on the lower end of the canyon, a land- mark readily seen from the vicinity of Teton Park. Although two visits to the slide area failed to reveal anything of paleobotanical interest, certain regions farther up the canyon proved more pro- duetive. Approximately twelve miles up the road there is a second and considerably older slide now partially concealed by a fairly heavy vegetation. When this older landslide occurred the Gros Ventre Canyon was blocked and a lake formed in Cole Hollow. At the time of our 1936 visit a petrified trunk some three feet in diameter was exposed approximately 200 yards from the south bank of the river. The slide area was subsequently followed up to its apparent origin, a point about one mile from the river (Mt. Leidy quadrangle, В, 112 W, T 42 N). There, a small badlands area, some few acres in extent and light gray color, is quite prominent when viewed from the opposite side of the valley. Sections of silieified trunks and small twigs are comparatively abundant in the gullies, weathering out of rocks of Cretaceous age. A more precise determination of the horizon is not possible at present, since neither the stratigraphy nor paleontology of the region has been studied in detail. Very small-scale coal-mining operations have been carried out on the opposite side of the river, likewise in Cretaceous strata. A seam some four feet thick outcrops about one-half mile north of the river and has been exploited in the past. The seam includes a number of bands of sandy clay some of which contain fern and dicot 1 Issued December 18, 1942. (283) (Vou. 29, 1942] 284 ANNALS OF THE MISSOURI BOTANICAL GARDEN leaf impressions. However, the matrix is very friable near the surface, although it might be possible to obtain better specimens if deeper excavations were made. The silicified woods in our collection include both dicotyledons and conifers. No attempt has been made to identify the former. The coniferous specimens include at least two distinct species. One of these is referable to Cupressinoxylon but, because of its otherwise indeterminate natural affinities, will not be considered, and the other can with reasonable certainty be placed in the Araucariaceae. Araucarioxylon wyomingense sp. nov.— Annual rings.—Present but indistinct, defined only by 2-3 rows of slightly thicker-walled summer cells (pl. 26, fig. 5). Resin canals.—Absent. Wood parenchyma.—Absent. Wood rays.—Strictly uniseriate and mostly only 1-3 cells high (fig. 3) ; cross field pitting of the cupressoid type.? Tracheids.—Quite constant in size throughout the annual ring, average 18 x 21 и (tangential and radial dimensions re- speetively) in transverse section. Pitting uniseriate and biseriate, pits closely erowded and angular when biseriate (figs. 1, 2, 4, 6), crassulae absent, tangential pitting absent. T ype.—Slides 617, 623, and 631, radial, tangential, and trans- verse sections respectively, from a fragmentary specimen, Washington University, St. Louis. The description is based on small fragments as well as a section of a trunk 9 inches in diameter and 5 inches long (No. 1396). Al- though this is the largest specimen in our collection it is a decorti- eated one and was evidently somewhat larger in life. The most distinctive feature of the wood lies in the diminutive size of the rays. In a total of over 700 rays counted from four different slides 50 per cent were but one cell high, 28 per cent two cells high, 12 per cent three cells high, while only 10 per cent were more than three cells high. They present a striking appearance when viewed in tangential section as may be noted in fig. 3. This ray structure is quite distinct from that of any other described Araucarian wood and constitutes the chief basis of our specific designation. The supposed Araucarian relationships of the species are supported by the weakly defined annual rings, the eupressoid type cross field pitting and especially the tracheidal pitting. * Phillips, E. W. J. The identification of coniferous woods by their microscopic struc- ture. Jour. Linn. Soc. Bot. 52: 259-314. 1941. [Vor. 29, 1942] 286 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 26 Araucarioxylon wyomingense Figs. 1, 2, 4, 6. Pitting in the radial walls of the secondary tracheids, Slide 617, x 300. Fig. 3. Tangential section. Slide 623, x 56 Fig. 5. Transverse section, Slide 631, x 22. PLATE 26 Ann. Mo. Вот. Garp., Vor. 29, 1942 ч = = "TRO" ке‏ = سے س سے m — == = pm : piens " omy” B - т го 4 SF am "0 et "^ p —— е € i m on —— — "в. LAU eee ets DIT 42 агі ANDREWS & PANNELL— FOSSIL ARAUCARIAN WOOD MONOGRAPH OF SCHOENOCAULON? ROBERT ROLAND BRINKER Formerly Graduate Student, Henry Shaw School of Botany of Washington University INTRODUCTION Interest in a monographic treatment of the genus Schoenocaulon, a member of the Liliaceae, tribe Veratreae, was aroused by the state of taxonomic confusion into which it had fallen. Though the first plants of this genus were discovered more than a hundred years ago and more have been constantly collected since then, no monographic study has as yet been made of Schoenocaulon. From time to time new species were described which, in some cases, only increased the confusion. The problem resolved itself into a delimitation of the genus, a determination of the true generic name, the study of the morphological features which distinguish the species, and the geo- graphical distribution of the genus as a whole and of the several species. History oF THE GENUS The genus Schoenocaulon was first proposed by Asa Gray in 1837. It was based on Helonias dubia Michx. Michaux gave us the first description of the plant in 1803.? He had his doubts as to the genus when he designated some plants from Georgia and Florida as Helo- nias dubia. Willdenow,? Pursh,* Nuttall,* and others considered the species as doubtful Helonias. Later Gray made the segregation but ealled the plant Schoenocaulon gracile.® Schiede and Deppe collected some plants on the eastern side of the Mexican Sierras. These were recorded by Schlechtendal and Chamisso as Veratrum, and in 1831 they proposed Veratrum offici- nale as a new species." Hartweg found some more in Ше neighbor- 1A dissertation carried out in the Graduate Laboratories 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 master of science in the Henry Shaw School of Botany of Washington University. 2 Michx. Fl. Bor. Аш. 1: 213. 1803. з Willd. in Ges. Naturf. Freunde Berlin, Mag. 2: 29. 1808. * Pursh, Fl. Am. Sept. 1: 244. 1814. 5 Nuttall, Gen. М. Am. Pl. 1: 234. 1818. в Gray, Ann. Гус. М. Y. 4: 127. 1837. 1 Schlecht. & Cham. in Linnaea 6: 45. 1831. Issued December 18, 1942. (287) [Vor. 29 288 ANNALS OF THE MISSOURI BOTANICAL GARDEN hood of Vera Cruz and sent them to the Royal Horticultural Society of London. Lindley had no doubt that these were the same as Schiede and Deppe's plants. However, he saw they were not of the genus Veratrum and proposed the name Asagraea officinalis.5 No sooner had Dr. Gray seen the illustration and description of the plant than he realized that it belonged to his genus Schoenocaulon. Examination of Schiede and Deppe's specimen confirmed his opin- ion." When in 1836 [1837] Brandt and Ratzeburg treated Veratrum officinale’ they inserted a footnote calling this plant Sabadilla offici- narum, but they used Sabadilla as a subgenerie name in their main description. The next species appeared in literature in 1838 when Schlechten- dal described an Ehrenberg specimen as Veratrum caricifolium.™ Six years later Schlectendal tried to correct his generic designation and change the name to Sabadilla caricifolia.? But he was too late, for Gray in 1840 had examined the same specimens and rightfully called them Schoenocaulon caricif olium.!? Schoenocaulon Drummondii was described by Gray in 1840, from a plant in Drummond's Texas соПесііоп.!* The species Schoeno- caulon tenuifolium was first called Veratrum tenuifolium by Mar- tens and Galeotti in 1842,15 and Asagraca tenuifolia by Kunth a year later.'! Robinson and Greenman made the proper combination in 1895.11 Lindheimer collected Texas plants in the spring of 1846 and sent them to Scheele who called по. 543 Schoenocaulon texanum.!* From this description Gray mistook the plant for his S. Drummondài and reduced S. texanum to synonymy.!? These two species were contin- ually eonfused until 1916, when Pennell pointed out their several points of difference.?° Schoenocaulon Coulteri is the name given by Baker in 1879 to a * Lindl, in Edwards’ Bot. Reg. 2: 1.33. 1839. * Gray in Hook. & Arn. Bot. Beechey's Voy. 388. 1840. ? Brandt & Ratzeb. in Hayne, iiam v. Besehreib. d. Arzneig. 13: 1.27. 1836. [1837]. “ Schlecht. in Ind. Sem. Hort. Hal 1838. ? Schlecht. іп Linnaea 18: 444. Tel " Gray in Hook. & Arn. Bot. Beechey’s Voy. p. 388. 1840. “Gray in Hook. & Arn. Bot. Beechey’s Voy. 388. 1840. * Mart. & Gal. in Acad. Roy. Brux. Bul. 9°: 380. 1842. * Kunth, Enum. Pl. 4: 700. 1843. " Robins. & Greenm. in Am. Jour. Sci., ПТ, 50: 168. 1895. * Scheele in Linnaea 25: 262. 1852. ? Holographic notes of Dr. A. Gray, in Libr. Mo. Bot. Gard. * Pennell in Bull. Torrey Bot. Club 43: 408. 1916. 1942] BRINKER—MONOGRAPH OF SCHOENOCAULON 289 plant which Coulter found near Zimapan, Mexico.’ Hemsley”? and Kuntze?? have the only other references in literature to this plant. Greenman deseribed the next species, S. Pringlei, in 1897?*; and S. calcicola, S. Ghiesbreghtu and S. jaliscense in 1907.25 The most recent description is that of S. megarhiza by Marcus E. Jones in 1912.26 Schoenocaulon intermedium has had a career of confusion ever since its description by Baker in 1879.27 Various plants have been placed in and taken out of the species. The original description was too general and the type specimens so poor that no definite idea of the real species could be formed. Evidence seems to point to the correctness of Hemsley’s statement that **probably S. intermedium may not be a distinct species’’?* and should be reduced to synonymy under S. texanum. MORPHOLOGY The bulb of Schoenocaulon is generally egg-shaped and covered with either dark, thin, dry scales, as in S. Drummond and S. du- bium, or, as is more common, by a tangle of fibers. The fibers are the fibro-vascular bundles which persist after the mesophyll of the first leaf-bases has decomposed. In young plants these fibers are not thickly coated over the bulb and around the base of the leaves, but in plants several years old they show specific characters in the density, color and texture. Some species, for example S. jaliscense, have fine and rather soft fibers, while others, as S. comatum, are charac- terized by eoarse and stiff fibers forming mats which resemble hanks of horsehair. The grass-like leaves are all basal and afford specific characters as to length and breadth only when the average can be taken from a number of full-grown plants. The intergradation between a short leaf of a long-leaf species and an unusually long leaf of a short-leaf species makes leaf characters quite unsatisfactory for classifica- tion. As to number of veins in a leaf, it has been found impossible to use the character of venation with any degree of satisfaction. п Baker in Jour. Linn. Soc. Bot. 17: 477. 1879. 23 Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 383. 1885. а Kuntze, Rev. Gen. Pl. 2: 713. 1891. ^ Greenm. in Proc. Am. Acad. 32: 295. 1897. 25 Greenm. in Proc. Am. Acad. 43: 19-20. 1907. * Jones, Contr. West. Bot. 14: 29. 1912 п Baker in Jour. Linn. Soc. Bot. 17: 477. 1879. 3 Hemsley in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 382. 1885. [Vor. 29 290 ANNALS OF THE MISSOURI BOTANICAL GARDEN Extreme cases of narrowness or breadth, as in S. tenue and S. offi- cinale respeetively, coupled with other features, may be of value in determining some of the species. The зсаре is always naked, straight, and terete above. Usually the base is somewhat angled because of the pressure exerted on it in the bulb. Some of the most valuable diagnostic features are to be found in the inflorescence, particularly its length and its diameter at anthe- sis. Other points to be noted are the disposition and spacing of the flowers on the axis, whether evenly or irregularly placed, whether crowded or distant from each other. In some species, as S. regulare, the flowers are all uniform in position and mature simultaneously. Other species, as S. texanum or S. officinale, have the lowermost flowers fruiting before those at the tip of the spike shed their pollen. The six perianth segments are, with the exception of S. Conzattzi, S. Ghiesbreghtu, S. Pringlei, and S. tenuifolium, about 2-3 mm. in length. Schoenocaulon tenuifolium is the only species with broadly ovate perianth-segments. All the other species have either ligulate or narrowly ovate segments. It is the margins of the segment which affords the most valuable character for specific determination. Schoenocaulon Drummondu, S. tenuifolium and S. yucatanense are erose-margined; S. comatum, S. dubium, and S. officinale are entire- margined; the majority of the species have a single hyaline tooth projecting from the sides of the segment near the base; others, such as S. Ghiesbreghtu, have two teeth on each margin. Schoenocaulon Pringlei is subscarious-margined and 5. caricifolium is subdentate- margined. In all species, except S. officinale, the nectary gland is too inconstant and inconspicuous a character to be of much value. The hypogynous stamens are six in number with large, reniform, single-celled anthers. In most species the filaments are twice the length of the perianth, while in S. megarhiza, S. obtusum, and S. Pringlei they scarcely project at all, and S. yucatanense has fila- ments three times as long as the perianth. The mature fruit of most of the species is ovoid and somewhat inflated. Schoenocaulon macrocarpum has a capsule which is more linear than oval; that of S. caricifolium and S. tenuifolium shows marked inflation. The former has a regular oval shape, whereas the latter is obovate. The seeds, if mature and obtained in sufficient abundance, may perhaps furnish added traits for determination of the species. 1942] BRINKER—MONOGRAPH OF SCHOENOCAULON 291 Economic IMPORTANCE The entire plant is poisonous, particularly the seeds, which are without odor but have an acrid burning taste. The medicinal virtue of the seed is due to the alkaloid veratrine. Tinctures, extracts, and powders were once made from the seeds and used for rheumatism, paralysis, cardiac conditions, dropsy, and as a taenicide. Because it is such a drastic purgative and productive of such severe vomit- ing and internal bleeding it has been discarded as a remedy in inter- nal medicine. Several deaths from its employment have been recorded. Its use nowadays is restricted to veterinary medicine or for destroying body lice and other vermin. Specimens received recently from H. S. Gentry bear the following notation concerning the plants he collected in Sinaloa, Mexico: ‘‘Roots employed for killing maggots in wounds ; powdered and applied or decocted as a wash." RELATED GENERA Schoenocaulon has been confused with several genera, particu- larly Helonias, Veratrum, Tofieldia, and Zygadenus. The follow- ing tabulation will furnish some salient morphological features use- ful in distinguishing Schoenocaulon from related genera. Schoenocaulon: Base of plant a fibrous-tunieate bulb; leaves basal, narrow; scape simple; flowers disposed in a spike or dense raceme; perianth-segments glandular or eglanular. Helonias: Base of plant a short tuberous root-stock ; leaves basal, broad and fleshy; scape simple; flowers disposed in a raceme; perianth-segments not glandular. Veratrum: Base of plant a fibrous-tunicate bulb; leaves cauline, broad; inflorescence a dense terminal panicle; perianth-segments not clawed, egland- ular. Tofieldia: Base of plant a short creeping rhizome; leaves 2-ranked, equitant; flowers disposed in a dense raceme or spike; perianth-segments not clawed. Zygadenus: Base of plant a rhizome or a tunicated bulb; leaves basal, linear; scape branched; flowers disposed in an open panicle; perianth-segments glandular, distinctly clawed. [Vor. 29 292 ANNALS OF THE MISSOURI BOTANICAL GARDEN GEOGRAPHICAL DISTRIBUTION Schoenocaulon has been found in the southernmost parts of the United States, in Mexico, Central America, and in two states of South America. Schoenocaulon dubium is the only Floridan species. S. Drum- mondi and S. texanum are found in Texas and northern Mexico. The latter species grows as far north as the southeastern counties of New Mexico. The widest distributed species is S. officinale, hav- ing been collected in Mexico, throughout Central America, and in parts of Venezuela and Peru. The paucity of specimens makes an adequate distribution picture of the genus Schoenocaulon impossible. Those states of Mexico and Central America which have been more extensively botanized have contributed a greater number of species. Gaps between localities where such species as S. comatum and S. regulare are found indicate the need of more field work and collecting. Common NAMES Schoenocaulon has a number of common names, such as: Green Lily (Texas), Feather-shank (Florida), Sabadilla, Sabadille, Ceba- dilla, Cevadille, Cevadilha, Matu Curros, Semen Sabadillae, Saba- dillsamen, Capuziner Samen, Laeusekraut, Mexicanischer Laeuse- samen and several others. ABBREVIATIONS Sincere appreciation is expressed to all those institutions whose cooperation and facilities have made this work possible. The her- baria cited in this monograph are indicated by the following abbre- viations : CA—California Academy of Sciences. D—Dudley Herbarium of Stanford University. F—Field Museum of Natural History. ray Herbarium of Harvard University. M—Missouri Botanical Garden P—Pomona College. PA-—Academy of Natural Sciences of Philadelphia. UC—University of California, Berkeley US—United States National Herbarium. TAXONOMY Schoenocaulon Gray in Ann. Lyc. №. Y. 4: 127. 1837; Endl. Gen. Р]. 1357. 1840; Meisn. Pl. Vase. Gen. 1: 405. 1836-43; Kunth, Enum. Pl. 1942] BRINKER— MONOGRAPH ОЕ SCHOENOCAULON 293 4: 185. 1843; Chapm. Fl. South. U. S., 490. 1860; Baker in Jour. Linn. Soc. Bot. 17:476. 1879; Benth. & Hook. f. Gen. Pl. 3: 836. 1880; Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 383. 1885; Engl. & Prantl, Nat. Pflanzenfam. П. 5: 23. 1888; Small, Fl. Southeast. U. 8. 250. 1903, and ed. 2, 1913; Small, Man. Southeast. Е1., 277. 1933. Veratrum Schlecht. & Cham. in Linnaea 6: 45. 1831, not L. Helonias Don in Edinb. N. Phil. Jour. 234. 1832, not L. Sabadilla?® Brandt [Brandt & Ratzeburg] in Hayne, Arzneig. 13: t. 27. 1836 [1837], as subgenus; Schlecht. in Linnaea 18: 444. 1844, as genus; Kuntze, Rev. Gen. Pl. 2: 713. 1891; Dalla-Torre & Harms, Gen. Siph. 60. 1900-1907 ; Engl. & Prantl. Nat. Pflanzenfam. ed. 2, 15a: 261. 1930. Skoinolon Raf. Fl. Tellur. 4: 27. 1836, nomen prius. Asagraea Lindl. in Edwards’ Bot. Reg. 2: pl. 33. 1839; Hook. & Arn. Bot. Beechey's Voy. 388. 1840; Kunth, Enum. Pl. 4: 184. 1843; A. Rich. in Orb. Diet. 2: 199. 1845 ; Spach in Hist. Nat. Veg. Phan. 12: 245. 1846, not Asagraea Baill. in Adansonia 9: 233. 1870. DESCRIPTION OF THE GENUS Herbaceous perennials ; bulb oblong to ovoid ; basal portion of the plant surrounded by a cylindrical covering of brownish-black to black scales and fibers, the remnants or the first bulb-scales and outer leaves; leaves all basal, grass-like, long and narrow, attenu- ate into a hair-like tip, flat, coarsely and strongly nerved, firm in texture, glabrous on both sides, slightly or not at all glaucous, mar- gins serrulate; scape erect, simple, naked, glabrous, glaucous and terete above, purplish and angled below; spike virgate, many-flow- ered, bracteate, terminal portion bearing sterile flowers; bracts small, solitary, partly clasping the axis, broad-triangular or deltoid, rounded or acuminate-tipped, thin, scarious, dull hyaline, erose- margined; flowers perfect, at first crowded, later more or less scat- tered, regular, small, erect, sessile or on short stout pedicels, base ? Brandt and Ratzeburg, in 1836 [1837], in Hayne’s ‘Darstellung und Beschreibung der Arzneigewiichse,’ recognized Veratrum officinale Schlecht. as a valid name and de- seribed and illustrated this plant in detail. Sabadilla was definitely included as a sub- genus of Veratrum. In a footnote they suggested the possible binominal Sabadilla offici- narum. This name, however, is accompanied by the following statement. ““ Es schien uns daher besser, fiir jetzt ein Subgenus! unter dem Namen Sabadilla vorszuschlagen, um jene auffallenden Eigenthümlichkeiten anzudeuten.’’ Since Sabadilla was here placed in the rank of subgenus it cannot, according to the International Rules of Botanical Nomen- clature, supersede the generic name of Schoenocaulon Gray which was validly published in 1837. [Vor. 29 294 ANNALS OF THE MISSOURI BOTANICAL GARDEN broadly bell-shaped to hemispherical, pale green to yellowish-white ; perianth-segments 6, hypogynous, cyclic, subequal, sessile, free or scarcely united at the base, somewhat spreading, persistent, faintly 3—5-nerved, leathery to scarious, obtuse, oblong to narrowly linear, entire, subentire, 1-2-dentate on either margin, or erose, often hye: line-margined, glandular or eglandular; stamens 6, hypogynous, subequal, руните at Ше base of Ше perianth-segments, persistent, erect; filaments filiform, yellow to purplish-red, awl-shaped, at first short, later elongating, recurved; anthers large, versatile, yellow, reniform before dehiscence, later elypeolate or peltate, unilocular, extrorse; ovary ovoid, constricted at the base, superior, free from the perianth and stamen-cycle, triearpellary ; ovules 6-8 in each cell, biseriate, axillary, ascending, anatropous; styles 3, distinct, diver- gent, slightly recurved, short, tapering to the apex; stigma termi- nal, simple, minute; capsule ovoid to oblong, acuminate, charta- ceous, glaucous, persistent, 3-celled, septicidally dehiscent, pedicels when present stout, recurved or ascending; seed ovate to slender- oblong, nearly terete, often curved, slightly winged at the apex, coat corrugated, dark brown and shiny, 1—5 superposed and compressed in each cell, albumen firm, embryo minute, near the hilum. Type species: Schoenocaulon dubium (Michx.) Small, Fl. South- east. U.S., 250. 1903 (S. gracile Gray in Ann. Lyc. N. Y. 4:127. 1837). KEY TO THE SPECIES "ҮЧТҮК eee ee Ie eT а 14. S. officinale AA. Perianth- — i sera to ovate, margins with or without teeth; nectaries absent o № РОВ, a ооо свод неа оов оова ене езт ев ORO A anes B. rehen тича mun dme. ара ani о Oa Rai NE C Perianth-segments ovate, 3 mm. broad.................... 18. S. tenuifolium ec. речей -égmenta liguiate, 1 mm. bro... ere o ena ES RIA ea VIN D . Perianth-segments with erose margins................4-. 6. S. Drummondii "m Рен, erd -segments with entire or with few-toothed — оиа es © E E. Flowers crowded on the вріке...................... D Ghiesbreghtii EE. Flowers loosely disposed on the spike................... . S. Conzattit ВВ. Inflorescence 5—15 mm, in diameter.,......... oc eo очно ето оное ново C . Perianth-segments with erose margins...................2 20. S. yucatanense CC. Perianth-segments not егозе-тшагрїпей................................... 1 D. Perianth-segments obscurely or not at all dentate, not searious-margined.. E. Bulb covered with coarse stiff fibers; scape 15-30 em. long............. E: Е. Scape 15-17 em. long; capsules strongly inflated....2. S. caricifolium FF. Scape 18-30 cm. long; capsules not strongly inflated....3. S. comatum EE. Bulb covered with scales or fine fibers; seape 30 em. or more м оваа Е Е. Spike narrow, 5-8 mm. in diameter.................... . S. dubium Ке wider, 10-15 mm. in diameter.................. P S. texanum DD. Pelei phos distinetly diste or scarious-margined............... E 1942] BRINKER— MONOGRAPH OF SCHOENOCAULON 295 EL Пенев nafTUW,20.0-2.5 MM ИОВ or. e rade v cR ка ERSTE NE F F. Flowers crowded on the axis; capsules 10-12 mm. long..15. S. Pringlei FF. Flowers distantly disposed on the axis; capsules 8-10 mm. long... о ооо ЕАУ ПАРЕ 17. S. tenue G. Spikes 4—10 em. long, 10 mm. in diameter.............. 17. S. tenue GG. Spikes 1.5-6 em. long, 13-14 mm. in diameter............ 5. S. Coulteri E вати proader 29558 mim wIdeu. йт ыгы. di ноев Е пе а COL IONE E АМ ohn И de ah РР А G Spike 90-40 em Ong sy nee ost ооа 9. S. jaliscense OG Spike PE ZI Gm JOlp ovat cae РР 2. S. Mortoniüi DIT SCADOR 2510 CI TONG тын ы о For EM beet ers all G G. Flowers all uniform in size and disposition on the axis............. H H. Filaments seareely projecting beyond the perianth; capsules numerous, 5—7 mm. in diameter, closely imbricated and ap- ргевзей 60 UNG TAXIS. о e В. 18. S. obtusum HH. Filaments twice the length of the perianth-segments; capsules fe d 4—5 mm. in diameter, small and distantly placed on the axis, mbricated or appressed to the axis........... 16. S. regulare GG. Flowera s» different sizes and irregularly disposed on the axis...... н eaf 3-6 mm. broad; seape 40—60 em. long; flowers subsessile e EU TES Част E Бове 11. S. megarhiza II. Leaf 2-3 mm. broad; seape 30—45 em. long; flowers pedicelle ave coi ITO "= ат PME 10. 8. macrocarpum 1. Schoenocaulon calcicola Greenm. in Proc. Am. Acad. 43: 19. 1907. Bulb ovoid, 1.5—3 em. in diameter; basal portion of the plant cov- ered with coarse black fibers extending to a height of 5-14 em.; leaves thin, 3-10 dm. long, 2-4 mm. broad; scape 5.5-7.5 dm. long, slender, somewhat flexuous, purplish toward Ше base; spike 8-23 em. long, 8-10 mm. in diameter, the flowers loosely and irregularly disposed on the axis, small, sessile, short-pedicelled; perianth-seg- ments slender, 2.5-3 mm. long, bearing a single tooth on each mar- gin; filaments twice the length of the perianth; eapsule oblong-lan- eeolate, 8-10 mm. long, 5 mm. in diameter, reflexed; 2—3 seeds in each cell. MEXICO—OAXACA: Las Sedas, ealeareous banks, alt. 1830 m., 19 July 1897, Pringle 6740 (G, M, NY, PA, UC, US) ; hillsides, alt. 1830 m., 1 Aug. 1894, Pringle 5754 (G TYPE, US) 2. Schoenocaulon caricifolium (Schlecht.) Gray in Hook. & Arn. Bot. Beechey's Voy. 388. 1840; Wats. in Proe. Am. Aead. 14: 281. 1879, excluding сои 1568; Hemsl. in Salvin & Godman, Biol. Cent: -Am. Bot. 3: 383. 1885 in part; Greenm. in Proc. Am. Acad. 43 19. 1907, excluding Concent Ф Gonzalez 323. (Vou. 29 296 ANNALS UF THE MISSOURI BOTANICAL GARDEN Veratrum caricifolium Schlecht. in Ind. Sem. Hort. Hal. 8. 1838; in Linnaea Litt. Ber. 13: 100. 1839 Asagraea caricif olia Kunth, Sidus. Pl. 4: 666. 1843. Sabadilla caricifolia “ККК in Linnaea 18: 444. 1844. ulb 15-25 mm. in diameter; base of the plant covered for 7-17 em. with dark brown, stiff fibers; leaves narrow, flexuous, 3-3.5 dm. long, 1-3 mm. broad; scape stout, short, 15-17 cm. long; spike short and narrow, 4—6 em. long, 10 mm. in diameter; flowers small and erowded on the axis; perianth-segments obseurely subdentate, oblanceolate, 3 mm. long; capsule broadly ovate, inflated, 15 mm. long, 10 mm. in diameter, compacted into a fruiting spike 3.5 em. in diameter ; seed 6-7 mm. long, 1-3 in each cell. MEXICO—COAHUILA: ‘‘mountain borders,’’ near Saltillo, 25 June 1848, Gregg 214 ‚ M). HIDALGO: near los Baños de Atotonilco el Grande, calcareous mountains, Oct. and Dec., Ehrenberg (С TYPE, NY, P). OAXACA: near City of Oaxaca, 16-21 June 1899, Rose § Hough 4972 (US); Cerro Verde, Aug. 1908, Purpus 4389 (UC). 3. Schoenocaulon comatum Brinker, n. sp.?? Bulb ovoid, 1.5-3 em. in diameter; base of the plant covered for 8-26 em. with a thick cylinder of blackish, coarse, stiff fibers ; leaves longer than the scape, straight, as much as 6 dm. long and 2-4 mm. broad; scape 18-30 em. long, erect, purplish toward the base; spike loosely flowered, attenuate, 3.5-15 em. long, 7-12 mm. in diameter at anthesis; flowers small, subsessile; perianth-segments entire- margined, 2-3 mm. long, ligulate; filaments thin, twice as long as the perianth; mature capsule ovate-oblong, chartaceous, erect, inflated, 10-18 mm. long, 6-7 mm. in diameter; pedicel of the fruit 3—4 mm. long ; 2—3 seeds in each cell. MEXICO—SAN LUIS POTOSI: Край of San Louis Potosi, San Miguelito 7: coll. of 1876—1878, pet 5636 (PA, G, NY); without definite loeality, Schaffner 228 (CA, NY ТҮРЕ, Р, РА, US); inita of San Luis Potosi, alt. 1830-2440 m., coll. of 1878, Parry Е Palmer _ (а, М, PA, US); Chareas, July- ds 1934, Whiting 523 (D, NY, U UEBLA: Cerro de pet moist soil, Me 1907, pns 2731 (UC); vicinity of San iub баара June 1908, Purpus (NY, US). oaxaca: Canada de San Gabriel, Etla, alt. 300 m., 8 pio 1897, ка d Дацов 323 (9). *Schoenocaulon comatum sp. nov. Bulbus ovoideus, 1.5-3 em. diametro; caudice eylindro, 7-26 em. longo, fibris sy tT crassis rigidis dense obtecto; foliis веаро longioribus, rectis, usque ad 6 dm. longis, mm. latis; — 18-30 em ‚ар; р basin versus purpureo; spiea la sni ra, attenuata, 3.5-15 . longa, з -1 diametro; floribus parvis, subsessilibus; perianthii pins s integris, 2-3 mm. longis, ligulatis; filamentis tenuibus, perianthio duplo longioribus; capsulis maturis ovato- oblongis, chartaceis, erectis, inflatis, 10-13 mm. longis, 6-7 mm. diametro; pedicello fructus 3-4 mm. longo; seminibus in loculo 2-3 1942] BRINKER— MONOGRAPH OF SCHOENOCAULON 297 4. Schoenocaulon Conzattii Brinker, n. sp.?! Bulb ovoid, 1.5-2 em. in diameter; base of the plant covered for 7-23 em. with a dense sheath of dark brown, coarse fibers; leaves 6-11 dm. long, 4-7 mm. broad; scape 5-7 dm. long; spike 10-25 em. long, 15-17 mm. in diameter; mature flowers loosely disposed, erect, shortly pedicelled; perianth-segments with a single tooth on each margin, linear-lanceolate, 3.5-4 mm. long; filaments twice or three times as long as the perianth; immature capsule 8 mm. long, 4 mm. wide, imbricated, pedicel arched and 3-4 mm. long; 4-5 seeds in each cell MEXICO—HIDALGO: Sierra de Pachuca, alt. 2981 m., 1 Sept. 1906, Pringle 13841 (С, US). sTATE OF MEXICO: Temasealtepee, Rincon del Carmen, woods, alt. 1340 m., 23 Nov. 1932, Hinton 2690 (NY); Nanchititla, oak woods, 8 Oct. 1933, Hinton 4970 (G, M, NY). OAXACA: Cerro de San Felipe, alt. 2000 m., 29 Aug. 1897, Conzatti ф Gonzalez, 449 (С, US TYPE). PUEBLA: Boca del Monte, moist grassy soil, J 1907, Purpus 2490 (UC). 5. Schoenocaulon Coulteri Baker in Jour. Linn. Soc. Bot. 17: 477. 1879; Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 382. 1885; Coulter in Contr. U.S. Nat. Herb. 2: 440. 1894 (Bot. West. Texas), as to name only. Sabadilla Coulteri Kuntze, Rev. Gen. Pl. 2: 713. 1891. The only specimen available for examination comprises frag- ments of leaf tips and parts of flowering stalks. The spikes measure 1.5-6 em. in length and 13-14 mm. in diameter; the flowers are erect, somewhat loosely set on the axis of the inflorescence; perianth-seg- ments leathery, with a single tooth on each margin, 3 mm. long; filaments reflexed, twice the length of the perianth, anthers small, elobular. MEXICO—HIDALGO: near Zimapan, Coulter 1569 (G ISOTYPE). 6. Schoenocaulon Drummondii Gray in Hook. & Arn. Bot. Beechey’s Voy. 388. 1840; Torrey in Bot. U.S. & Mex. Bound. Surv. 2: 222. 1859, in part, excluding Pl. Lindh.; Wats. in Proc. Am. Acad. 14: 281. 1879; Baker in Jour. Linn. бос. Bot. 17: 477. 1879, in part, excluding Lindheimer 543, 711; Wats. in Proc. Am. Acad. 18: 166. 31 Schoenocaulon Conzattii sp. nov. Bulbus ovoideus, 1.5-2 em. diametro; eaule fibris densis atrobrunneis erassis ad 7—23 em. dense obtecto; foliis 6-11 dm. longis, 4-7 mm ш seapo 5—7 dm. longo; spiea 10—25 em. longa, 15—17 mm. diametro; floribus maturis axe dispositis, ereetis, breviter pedicellatis; perianthii segmentis utraque margine dente unico, lineari- ee 3.5-4 mm. longis; filamentis duplo vel triplo perianthio longiori- bus; capsulis immaturis 8 mm. longis, 4 mm. latis, imbrieatis, pedicello arcuato, 3-4 mm. longo; Зета три in cellulo 4-5. [Vor. 29 298 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1883, in part, as to Palmer 1322 only ; Hemsley in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 382. 1885 in part, as to name only ; Small, FI. Southeast. U.S. 250. 1903. Schoenocaulon aletroides Gray in Hook. & Arn. Bot. Beechey's Voy. 388. 1840 Sabadilla Drummondii Kuntze, Rev. Gen. Pl. 2: 713. 1891. Bulb ovoid, 2-3.5 em. in diameter; base of the plant covered with thin scales and few fibers for 5-12 ст. ; leaves 2-5 dm. long, 1.5-5 mm. broad; scape 2.5-6 dm. long; nike: dense, 3-20 em. long, 15-20 mm. in diameter; flowers subsessile; perianth-segments oblong- ovate, 3 mm. long, 1 mm. broad, nearly membranaceous, with broad scarious, erose margins; filaments stout, subclavate-filiform, much dilated above, 5-7 mm. long; flowering in autumn. Distribution: southwestern United States and Mexico. UNITED STATES: TEXAS—BEE CO.: % mi. s. of Tuleta, 21 Sept. 1936, Cory 20644 (С). BEXAR CO.: San Antonio, Oct. 1837, Riddell ny COLORADO CO.: 3 mi. e. of Alleyton, sandy soil, 22 Sept. 1913, Pennell 5557 (NY). WITT CO.: roadside near Yorktown, 6 Oct. 1857, Schott (NY). DUVAL CO.: 25.6 mi. п.м. of San Diego, 9 Oct. 1935, Cory 16146 (С). GOLIAD CO.: near Goliad, open prairie, 7 Oet. 1926, Williams 91 (P A). GUADALUPE CO.: Seguin, lime- stone soil, 22 Aug. 1903, Groth 202 (CA, G, NY, US). SAN PATRICIO CO.: s. of Mathis, 20 et. 1927, Rose ф Russell 24160 (US). WILSON co.: Sutherland Springs, Aug. 1879, Palmer 1322 (G). WITHOUT DEFINITE LOCALITY: coll. of 1880, Palmer 1322 (US); Valley of the Rio Grande, below ers Parry, Bigelow, Wright 4 Schott 1482 (US); Drummond I 284 (G) ; Drummond III 284 (NY TYPE) ; 1 Oct. 1935, Parks (G) ; 1840, Riddell 52 (NY); 188-, Schlottmann ( 8). MEXICO—SAN LUIS POTOSI: San Dieguito, 13-16 June 1904, Palmer 119 (G, M, NY, ES; шыбы, Cerro de 8. Francisco, 11 May 1891, Manry 6257 (С). 7T. Schoenocaulon dubium (Michx.) Small, Flora Southeast. U.S. 250. 1903, and ed. 2, 1913; Small, Man. Southeast. Fl. 277. 1933. Helonias dubia Michx., Fl. Bor. Am. 1: 213. 1803; Willd. in Ges. Naturf. Freunde Berlin Mag. 2: 29. 1808; Poiret in Lamarck, Encycl. Meth. Bot. Suppl. 3: 28. 1813; Pursh, Fl. Am. Sept. 1: 244. 1814; Nutt. Gen. N. Am. Pl. 1: 234. 1818; Roem. & Schult. Syst. 7: 1565. 1830. Schoenocaulon gracile Gray in Ann. Lye. N.Y. 4: 127. 1837, and in Hook. & Arn. Bot. Beechey's Voy. 388. 1840; Kunth, Enum. PI. 4: 185. 1843; Chapman, Fl. South. U.S. 490. 1860; Wats. in Proc. Am. Acad. 14: 280. 1879; Baker in Jour. Linn. Soc. Bot. 17: 477. 1879. Sabadilla dubia Kuntze, Rev. Gen. Pl. 2: 713. 1891. Bulb ovoid, small, 1.5-3 em. in diameter; base of the plant cov- ered for 4-13 em. with scarious brown scales, separating above into fibers; leaves linear, narrow, suberect, 3-9 dm. long, 1-3 mm. broad; 1942] BRINKER—MONOGRAPH OF SCHOENOCAULON 209 scape very thin, 3-7 dm. high; spike slender, 10-85 em. long, 5-8 mm. in diameter, virgate, flowers interruptedly and loosely ar- ranged on the axis, pale green, very small, sessile; perianth-seg- ments oblong, leathery, obscurely or not at all dentate, less than 2.5 mm. long ; filaments filiform, very slender, twice the length of the perianth ; eapsule conie, ovate, nearly sessile, 8-10 mm. long, 5 mm. in diameter, appressed to the axis; seeds 5 mm. or less long, 2—5 in each cell. d UNITED STATE A—ALACHUA CO.: CM 3 May 1897, Lighthipe 479 (M, NY, US), Mareh Ts S (NY), June 1876, Garber (US), open = land, 15 June 1910, Hood (M, 850094), 30 April 1936, and 10 July 1936, Murrill S : Indian River, 1897, Crawford (PA); Eau Gallie, dry pine barrens, July an T S Ner (С, МУ, М, US); Okeechobee Region, serub-oak land, 10 July 1903, К и. 6032 (G, IH ; bor La- goon, dry pine barrens, June 1879, Curtiss 2900 (G, M, NY, UC, I RNANDO CO.: Bro p p woods, 26 April 1930, Moldenke 5943 (NY). QOO: DERE 3 March 1888, Burk (PA); Eustis, June and July 1894, а E 210134), June and July ed B (M 210138), Marsh (M 760846), high pine land, 1-15 April 1898, Nash 297 (G, M, NY, UC, US), 16-31 July 1894, Nash 1447 (PA, US), and 19-30 June 1895, 2016 (NY, US). MARION co.: east of Flemington, dry sandy woods, 29 April 1930, Moldenke 1087 (M, NY, US). ORANGE CO.: take Brantley, 12 July 1890, Kline (PA 517892) ; Aug. 1894, Kline is 517891); 7 July Bee Lewton (NY), and 18 July 1894 (PA 517890) ; Aug. 1894, ex Herb. Williamson (PA 509795) ; Lake Conway, 11 April 1900, Huger 18 (M) ; dry pine ub. 12 July 1902, жем 5387 (G, NY); Lake Helen, pine land, 24 April 1911, Hood (M 1073119). OSCEOLA co.: Lake Gentry, 26 April 1925, Howell 1105 (US). PASCO co.: St. Joseph, dry pine lands, 24 April 1927, O'Neill (M 953079). POLK CO.: vicinity of Winter H: ey кы pine land east of Lake Marion, 12 May 1931, McFarlin 5261 (CA). VOLUSIA CO.: near Volusia, dry pine ridges, 24 March 1882, Mohr US), dry sandy ridges, 4 April 1882, i (US) ; Seville, dry pine barrens, 7 May 1 Curtiss 6606 (G, M, NY, PA, UC, US); De Land, Майен 1891, Hulst (NY), 25 March 1891, Hulst [Hurst] (UC), 27 Mareh 1891, Hulst (NY). FORT KING AND EASTERN FLORIDA : April and May, Leavenworth (NY). TAMPA BAY: Leavenworth (С); Alden (NY); 1834, Burrows (NY). LOCALITY INDEFINITE: 1839, Buckley (G, M, NY); Chapman (M 210141, US 968641) ; Wright (D 88718, NY). 8. pea ae Ghiesbreghtii Greenm. in Proc. Am. Acad. 43: 20. 1907. Bulb 1-1.5 em. in diameter; basal portion of the plant covered with brownish-black scales and fibers to a height of 10-12 em.; leaves narrow, 4-8 dm. long, 2-6 mm. broad; scape straight, 5-6.5 dm. long; spike 1-2.5 dm. long, 1.5-2 em. in diameter ; bract broadly ovate, 2.5 mm. long, obtuse, 5-nerved; flowers sessile or on short pedicels ; segments of the perianth oblong, glandular, with two teeth on each margin, 4—4.5 mm. long, 1 mm. broad; filaments more than twice the length of the segments; fruit unknown. MEXICO—CHIAPAS: Ghiesbreght 672 (G TYPE, M), and at alt. 2135 m., Berendt (G). [Vor. 29 300 ANNALS OF THE MISSOURI BOTANICAL GARDEN 9. Schoenocaulon jaliscense Greenm. in Proc. Amer. Acad. 43: 20. 1907. Bulb oblong-ovoid, 2.5-3.5 em. in diameter; basal portion of the plant eovered with densely massed, tangled, fine, brownish fibers to a height of 14-16 cm. ; leaves 6-10 dm. long, 2-7 mm. broad; scape 10-15 dm. long; spike 3-4 dm. long, 12 mm. in diameter; flowers small, shortly pedicelled, irregularly spaced on the axis; perianth- segments linear, with a single tooth on each margin; filaments red- dish-purple, twice the length of the perianth; capsule ovate, 6 mm. long, 3 mm. in diameter, on curved pedicel and appressed to the axis. MEXICO—JALISCO: near Guadalajara, cool grassy sides of canyons, Pringle 2938 (G TYPE); same locality, steep bluffs of ravines, Pringle 11853 (G, US) 10. Schoenocaulon macrocarpum Brinker, п. sp.?? Bulb ovoid, 2.5-3 em. in diameter; base of the plant covered for 12-14 em. with coarse black fibers; leaves about 5 dm. long, 2-3 mm. broad; scape 3—4.5 dm. long; spike 10-23 em. long, about 8 mm. in diameter ; flowers small, subsessile, erect, not very closely placed on the axis; perianth-segments ligulate, linear, with a single tooth on each margin; filaments reddish, twice the length of the perianth- segments ; capsule linear-oblong, imbricated, appressed to the axis of the inflorescence, 16 mm. long, 4 mm. in diameter; 1-3 seeds in each cell. MEXICO—NUEVO LEON: Sierra Madre Oriental, Canyon de los Charcos y Mesa de la Camisa, above Alamar, 15 mi. s.w. of Galeana, common in open wood in lower canyon, 4 June 1934, Mueller 724 (С TYPE, Е). 11. Schoenocaulon megarhiza Jones, Contr. West. Bot. 14: 99. 1912 (as megarrhiza). Bulb 2.5-3.5 em. in diameter; basal portion of the plant covered for 10-14 em. with coarse, brownish-black fibers; leaves 3-7 dm. long, 3-6 mm. wide; scape 4-6 dm. long; spike 2-3 dm. long, 10-12 mm. in diameter; flowers small, sessile to subsessile; perianth-seg- ments 3 mm. long, with a single tooth on each margin; filaments not much longer than the perianth ; immature capsule 6 mm. long, 3 mm. in diameter, on а pedicel 3 mm. long; 3—4 seeds in each cell. ? Schoenocaulon macrocarpun sp. nov. Bulbus ovoideus, 2.5-3 em. pes caule fibris erassis nigrisque ad 12-14 em. obtecto; foliis cirea 5 dm. longis, 2-3 m atis; зсаро 3—4.5 dm. longo; spiea 10-23 em. longa, cirea 8 mm. diametro; floribus parvis, sub- sessilibus, erectis, haud arcte dispositis; perianthii segmentis Нана, linearibus, utra- que margine dente uno; filamentis rubris, perianthii segmentis duplo longioribus; eap- sulis lineari-oblongis, ейн», axi appressis, 16 mm. longis, 4 mm. diametro; seminibus in loeulo 1-3. 1942] BRINKER— MONOGRAPH OF SCHOENOCAULON 301 Ep O—CHIHUAHUA: Sierra Madre Mts., Guayanopa Canyon, alt. 1525 m., 23 Sept. 1903, es (P TYPE); transition pine TN Sierra Charueo, Rio Fuerte, infrequently We Rh shady banks and under trees, 23 July 1936, Gentry 2315 (G, M); transi- tion, pine-oak country, solitary in shaded rock D Sierra Canelo, Rio М. 30 Aug. 1936, ek 2515 (G, M, UC); San Jose de Pinal, e transition, de slopes, in rocky terrain with harsh grass, 5 Sept. 1936, Pomp 2591 (M). ALOA: brado de мы, Sierra Surotato, open grassy slope, oak forest, ae Bo o m., 10-14 Sept. 1941, Gentry 6555 (M). 12. Schoenocaulon Mortonii Brinker, n. sp.?? Bulb unknown; leaves 6-7 dm. long, 3—5 mm. broad; scape erect, very long, 8.5-10 dm. long; spike many-flowered, 14—23 em. long, 10-12 mm. in diameter; flowers sessile or on short pedicels, loosely disposed on the axis of the inflorescence; perianth-segments with a single wide tooth on each margin, 3 mm. long; filaments reddish- purple, twice as long as the perianth ; capsule unknown. MEXICO—MICHOACAN: Zitaeuaro, Ypasote Hill, alt. 2175 m., steep rocky slope in sparse oak forest, 18 Nov. 1938, Hinton 13465 (M TYPE, G). 13. Schoenocaulon obtusum Brinker, n. sp.“ Bulb ovoid, 2—8 em. in diameter; base of the plant sheathed for 10-20 em. by a dense cylindrical growth of brown, coarse, rigid fibers; leaves about 5 dm. long, 3-8 mm. broad; scape 25-45 cm. long; spike dense, regular, 8-20 em. long, 8-10 mm. in diameter; all the flowers maturing simultaneously and of the same disposition on the axis, touching each other, sessile; perianth-segments short, 3 mm. long, obtuse, with a single "n on each margin; filaments scarcely longer than the perianth; capsules numerous, SERE TNI and appressed to the axis of the inflorescenee, 8-10 mm. long, 5-7 mm. in diameter, 1—3 seeds in each ce MEXICO—HIDALGO: Sierra de Pachucha, Sept. 1903, Rose $ Painter 6722 (С, US TYPE); El Chico, ‘‘ Entre Roca Lumate,’’ July 1929, Lyonnet 329 (M, NY, US). 14. Schoenocaulon officinale (Schlecht. & Cham.) Gray in Hook. & Arn. Bot. Beechey's Voy. 388. 1840; Benth. Pl. Hartw. 29. 1840; 3 Schoenocaulon Mortonii sp. nov. Bulbus SEA ETE 6-7 Үй longis, 3-5 mm. latis; die erecto, perlongo, 8.5-10 dm.; spica multiflora, 14—23 em. longa, anthesi 10—12 mm. diametro; floribus sessilibus vel breviter pediccllatis "m ee perianthii seg- шсш utraque margine dente lato uno, 3 шт. longis; filamentis rubro-purpureis peri- ae duplo longioribus; capsulis ignotis. choenocaulon obtusum sp. nov. Bulbus ovoideus, 2-3 em. diametro; caudice cy- lindro ad 10-20 em. fibris brunneis crassis rigidis dense obtecto; foliis circa 5 dm. longis, 3—8 mm. latis; seapo 25—45 em. longo; spica densa, regularis, 8-20 ет. longa, 8-10 mm. diametro; floribu us omnibus aequalibus maturitate et dispositione, AA tangentibus, sessilibus; deii nthii segmentis brevibus, 3 mm. longis, obtusis, utraque margine dente uno; filamentis perianthio haud longioribus; eapsulis numerosis, imbrieatis et inflores- eentiae axi appressis, 8-10 mm. longis, 5—7 mm. diametro; seminibus in loculo 1-3. [Vor. 29 302 ANNALS OF THE MISSOURI BOTANICAL GARDEN loe. eit. 96. 1842; Wats. in Proc. Am. Acad. 14: 281. 1879; Baker in Jour. Linn. Soc. Bot. 17: 476. 1879; Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 383. 1885, in part; Yuncker in Field Mus. Publ. Bot. 17: 323. 1938. Veratrum officinale Schlecht. & Cham. in Linnaea 6: 45. 1831; Nees, Pl. Offic. pl. 50. 1821-33. H elonias officinalis Don in Edinb. N. Phil. Jour. 234. 1832 ; Lindley, ЕІ. Med. 586. 1838. Sabadilla officinarum Brandt (Brandt & Ratzeburg) in Hayne, Arzneig. 13: pl. 27. 1836 [1837] ; Schlecht. in Linnaea 18: 444. 1844; Kuntze, Rev. Gen. РІ. 2: 713. 1891, as officinalis; Standl. & Cald. in List. Prelim. Pl. El Salvador, 49. 1925, as officinalis; Knuth in Fedde Rep. Beih. 43: 199. 1927, in part. Asagraea officinalis Lindl. in Edwards’ Bot. Reg. n.s. 2: pl. 33. 1839; Hook & Arn. Bot. Beechey's Voy. 388. 1840; Kunth, Enum. P1. 4: 184, 1843; Spach in Hist. Nat. Veg. Phan. 12: 245. 1846; Lindl. Med. & Oecon. Bot. 55, fig. 90. 1849; Hare, Caspari & Rusby, Nat. Stand. Dispens. 1336, fig. 354. 1905. Asagraea caracasana Ernst in Jour. Bot. 9: 91. 1871; Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 383. 1885, in synonomy Bulb large and thick, 2.5—4 em. in diameter ; base of the plant cov- ered for 8-20 em. with dark brown scales separating above into fibers; leaves coarse, 3-10 dm. long, 3-15 mm. broad ; scape angled below, terete above, 6-12 dm. long, 3-12 mm. in diameter; flowering spike dense, cylindrical, 1-5.5 dm. long, 12-18 mm. in diameter; perianth-segments ligulate, narrowly oblong lanceolate, 3-4 mm. long, margins entire, nectaries prominent at the base of the seg- ment; filament thick, stout, 6-7 mm. long; fruiting spike 2 ст. in diameter; capsules numerous, crowded, elliptic-oblong, 10-13 mm. long, 4—5 mm. in diameter, on pedicels 4 mm. long; bracts deltoid, 3 mm. long; seeds 1—4 in each cell. Distribution: Mexieo, Central America to Venezuela and Peru. NORTH AMERICA: MEXICO—CHIAPAS: Fenia, open pine and oak forest, June 1925, Purpus 405 (US); Siltepec, 9 Aug. 1937, Matuda 1600 (Е, NY). солма: 9 Jan.-6 Feb. 1891, Palmer 1410 US). MEXICO STATE: Temasealtepee, Ин rm alt. 1840 m., 17 Nov. 1932, Hinton 2420 (CA); Temascaltepec, Chorrera, alt. 1230 m., 9 Oct. 1932, Hinton 2039 (G, F). MICHOACAN: Zitacuaro, Piedra de Cal, alt. 1300 m., steep rocky slope, 10 Sept. 1938, Hinton 13226 (G, M, NY, US). MORELOS: Нобл, on sunny heights, 29 Sept. 1910, Seler 5379 (296) (С). GUERRERO: Mina, Santa Teresa, grassy hill, oak woods, alt. 1040 m., 12 Sept. 1936, Hinton 9397 (US); Mina, Puerto del Clarin Cayunche, grassy hill, alt. 750 m., 21 Sept. 1936, Hinton 9506 (US); Mina, Calavera, Puerto, oak woods, 19 Sept. 1936, Hinton 9539 (G, NY, US). oaxaca: Santa Efigenia, alt. 150 m., 18 July 1895, 1942] BRIN KER—MONOGRAPH OF SCHOENOCAULON 303 ieee ра P and 2849 (F); San Benito, near Apango, alt. 500 m., 11 Oct. 1917, (US). CRUZ: Orizaba, June 1857, Mohr ф Botteri (US), Botteri 1186 ven Es л» ii. ador 2981 (G, US), 7 Oct. 1853, Mueller 222 (G, NY), Uluapam, Oct. 1853, Mue ESTE ONS). о 1841-48, Liebmann 14625 (Е), Mirador, Oct. 184 pei 14627 (F), . 1841, 14628 (G, F, UC, US), and 14629 (ЕУ) S Malus May 1937, Matuda 1346 A E Zacuapan, Fortin, sunny slopes, open woods, Aug. 1906, Purpus 2023 (Е, С, М, NY, UC, US); Cerro de Borego, dry calcareous hills, collector unknown, 496 (PA). **On o. eastern slopes of the Mexican Andes," in the а de Tioselo near the Hacienda de la Laguna, Schiede $ Deppe 982 (М TYPE). A specimen in the Gray Herbarium labeled **Mexieo (Hooker dupl. 1839) "' without further data is referable to this species. CENTRAL AMERICA: GUATEMALA—DEPT. SANTA ROSA: Rio de Los Esclavos, alt. 750 m., Aug. 1892, Heyde $ Lux (Smith 3874) (С, NY, US); Alameda, 8 Aug. 1937, Johnston 961 (Е, NY). DEPT. HUEHUETENANGO: Uaxackanal, Quen Santo, open wooded limestone hills, 23 Aug. 1896, Seler 3220 (G); bir и wooded limestone hills, 7 Sept. 1896, Seler 3273 Gb ) PT. GUATEMALA s. of San Raimundo, damp wooded barranea, bushy slope, alt. xm m., 18 um TUR ри 62922 (К); 1939, Aguilar 128 (Е). DEPT. ZACAPA: lower slopes of Sierra de las Minas, p trail above Rio Hondo, 250-900 m alt., grassy area, 11 Oct. 1939, Steyermark 29548 (Е). DEPT. CHIQUIMULA: Моана Castilla, vicinity of Montana баа along M ae ia Saso, 3 mi. s.e. of Quezaltepeque, alt. 1200-1500 m., 6 Nov. 1939, Си di (F). JUTIAPA: Lago Retana, between Overo and Progresso, alt. 600 26 Nov. 1939, icd 2025 (F). wiTHOUT DEFINITE LOCALITY: in rocky places, a Basten 627 (NY). HONDURAS—DEPT. OF COMAYAGUA: 6 km. w. of Siguatepeque, moist soil, river bank, alt. 1250 m., 8 Aug. 1936, Yuncker, Dawson ф Youse 6358 (Е, С, M). Eu Satvapor—Vicinity of San np Renson 167 (NY, US); Cerro de San Ja- cinto, Е 1922, Calderon 1060 (G, M, A Riek ats JUELA: San Jose, um Um alt. 800 m., July 1912, Jimenez 637 D SOUTH AMERICA: sie 1929, Elias 71 (Е); grassy hills, Ernst (US 601444) ; 24 June 1917, Curran $ Ei 1196 (CA, F, G, NY, UC, US); os sides, alt. 600— 050 m., Dec. 1935, p 888 (NY); Middle Catouche wood, in forest, in savannas, alt. 1200-1400 m., 2 Sept. 1917, Pittier Md (G, US) ; La Guaira, Ы road, alt. 1100— 1300 m., 6 June 1921, Pittier 9556 (G, US); 31 Oct. През Rose 21887 (NY, US); Los eie alt. 950 m., Dee. 1939, Wiliams 13604 (F); between Caracas and La Guaira, rough Brett, дай 21887 (US). vAR: 1854—5, Fendler 1506 (G, M). MIRANDA: hills ko ve Los Teques, in open places, E Sept. 1924, Pittier 11611 (NY, US); Los Chorros, alt. 960 m., 20 Nov. 1939, Williams 12398 (F). LOCALITY UNSPECIFIED: 22 Sept. 1891, Eggers 13361 (F, US). PERU—San Miguel, Urubama valley, alt. 1800 m., 31 Мау 1915, Cook ф Gilbert 1013 (US); Santa Ana, alt. 900 m., 27 June 1915, Cook ф ‘Gilbert 1538 (US). CUZCO: POTRER Convencion, alt. 1300 m., 4 March 1940, Vasgare 1844 (G). 15. Schoenocaulon Pringlei Greenm. in Proc. Am. Acad. 32: 295. 1897. Bulb ovoid, 1.5-3 em. in diameter; base of the plant covered with coarse dark brown fibers to the height of 8-24 сш.; leaves narrow, 2.5-7 dm. long, 1-3 mm. broad; scape straight, 3-7.5 dm. long; flow- ering spike 2-10 em. long, 8-10 mm. in diameter; fruiting spike [Vor. 29 304 ANNALS OF THE MISSOURI BOTANICAL GARDEN dense, 15 mm. in diameter ; bract rounded at the apex; flowers closely appressed, sessile to subsessile; perianth-segments oblong-ovate, dark-brown, searious-margined, 4—5 mm. long; filaments scarcely projecting beyond the perianth; capsules imbricated, oblong, 10-12 mm. long, 5 mm. in diameter; seeds small, thin, 4-5 mm. long, 1-5 in each cell. MEXICO—HIDALGO: between Somoriel and Las Lajas, 5 Aug. 1905, Rose, Painter ф Rose 9243 ). FEDERAL с lava beds, Serrania de Ajusco, alt. 3050 m., 23 Aug. 1896, Pringle 6451 (CA, С, M, NY, PA, UC, US ТҮРЕ); Cima Station, lava fields, alt. 3050 m., 19 Sept. 1903, Deni 11716 (G, US), alt. 2981 m., Pringle 13621 (G, US); La Cima de A jusco, lava fields, alt. 2981 m., 2 Aug. 1906, Pringle 18778 (G, US). NAYARIT: near Santa Teresa, top of Sierra Madre, 13 бус, 1897, Rose 2227 (US). PUEBLA: Mount Orizaba, 25-26 July 1901, Rose $ Hay 5690 (US). 16. Schoenocaulon regulare Brinker, n. sp.?? Bulb ovoid, 2-4 em. in diameter; base of the plant covered for 6-14 em. by a collar of dark-brown, coarse fibers; leaves shorter than the seape and inflorescence, 2-6 dm. long, 2-5 mm. broad; scape 2.5-6 dm. long; spike loosely flowered, tapering, 10—24 em. long, 7-10 mm. in diameter; flowers very small, sessile, equal to each other in size and maturity, regularly disposed; perianth-segments very small, with a single tooth on each margin, 2.5 mm. or less long; filaments twice the length of the perianth; capsules subsessile, 8-12 mm. long, 4—5 mm. in diameter; 2—4 seeds in each cell. MExICO—JALISCO: Sierra Madre Occidental, trail from San Sebastian to Real Alto, Loma del Oregano, alt. 1500 m., pine forest on steep hill-slope, 18 Feb. 1927, Mexia 1702 (CA, D, M, NY, UC, US TYPE) ; w. of San oo Hacienda del bony Arroyo de los оне, alt. 1500 m., near stream, 6 March 1927, Mexia 1823a (CA, UC, US) ; Sierra Madre, w. of Polanos, 15-17 Sept. 1897, Rose p (G, US). DURANGO: E Papasquiaro, April-Aug. 1896, Palmer 419 (G, M, NY, UC, US). 17. Schoenocaulon tenue Brinker, n. sp. Bulb small, 1-1.5 em. in diameter; base of the plant covered for 3 Schoenocaulon regulare sp. nov. Bulbus ovoideus, 2—4 ст. diametro; caudice fibris atro-brunneis erassis ad 6—14 em. dense obteeto; foliis seapo et inflorescentia — 2—6 dm. longis, 2—5 mm. latis; seapo 2.5—6 dm. longo; spica laxiflora, attenuat em, longa, 7-10 mm. diametro + боғ ibus minimis, sessilibus, — et баба sta ibus, regulariter родни perianthii segmentis minimis, utraque margine uno, 2.5 mm. vel minus longis; filamentis perianthio duplo longioribus; capsulis paa: 8-12 mm. longis, 4—5 mm. oes on seminibus in loculo 2— *Schoenocaulon tenue sp. nov. Bul arvus, wr em. diametro; caule fibris brunneis filamentosis ad 6—13 em. alto сч foliis tenuissimis, Kes, prep ibus, 2.5—3.5 dm. longis, .5-2 mm. latis; seapo — erecto, graeili, 16-35 e spica 4—10 em. longa, 10 mm. tantum diametro; floribus paueis, minimis, пребит, Тм. dispositis; perianthii segmentis utraque margine dente uno, ligulatis, 2.5-3 mm. longis; filamentis duplo perianthio пес iab TP eapsulis maturis parvis, 8-10 mm. longis, 4-5 mm. ues seminibus in loeulo 1942] BRINKER— MONOGRAPH OF SCHOENOCAULON 305 6-13 em. with brown hair-like fibers; leaves very narrow, short, sub- filiform, 2.5-3.5 dm. long, 0.5-2 mm. broad; scape short, erect, slen- der, 16-35 em. high; spike 4-10 cm. long, only 10 mm. in diameter; flowers few, very small, pedicelled, distantly disposed on the axis of inflorescence; perianth-segments with a single tooth on each mar- ein, ligulate, 2.5-3 mm. long; filaments twice as long as the perianth; mature capsules small, 8-10 mm. long, 4-5 mm. in diameter, 1-2 seeds in each cell. MEXICO—MORELOS: cooler grassy slopes of the knobs of the Sierra de Tepoxtlan, alt. 2285 m., Sept.—Nov. 1900, Pringle 8356 (G, M түрк, NY, P, PA, UC, US). SAN LUIS POTOSI: Charcas, July-Aug. 1934, Lundell 5464 (US). Further collections of this plant may prove its identity with the obscure S. Coulter. 18. Schoenocaulon tenuifolium (Mart. & Gal.) Robins. & Greenm. in Am. Jour. Sci. ПТ, 50: 168. 1895. Veratrum tenuifolium Mart. & Gal. in Acad. Roy. Brux. Bul. 9: 380. 1842; Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 383. 1885. Asagraea tenuifolia Kunth, Enum. Pl. 4: 700. 1843; Hemsley, l.c. Bulb 2.5-4.5 em. in diameter; base of the plant covered with a sheath of very coarse fibers to the height of 15-30 cm.; leaves 5-9 dm. long, 3-10 mm. broad; scape stout, 10-36 em. long; flowering spike 7-16 ст. long, 16-20 mm. in diameter at anthesis; subtending bract large, 4 mm. long; fruiting spike fertile at the base only, 4—5 em. in diameter; flowers large, sessile, densely arranged on the axis; perianth-segments broadly ovate, green tipped with red, margin erose, 5 mm. long, 3 mm. broad; filaments thick, red, twice the length of the perianth; capsule large, inflated, obovate, 18-20 mm. long, 10-12 mm. in diameter; seeds large, oval, 7 mm. long, 5 mm. in diam- eter, 1-2 in each cell. MEXICO—OAXACA: Cerro San Felipe, alt. 3000 m., Conzatti 688 (G); mountains s. of Miahuatlan, alt. 300 m., Nelson 2530a (G, US); Sierra de San Felipe, alt. 3011 m., Pringle 5857 (G, US), alt. 2890 m., 6653 (F, G, NY, M, PA, UC, US), and alt. 3200 m., 10184 (F, G, M, NY, PA, UC, US), alt. 300 m., Smith 753 (M, US). 19. Schoenocaulon texanum Scheele in Linnaea 25: 262. 1852. Schoenocaulon Drummondii Torrey in Bot. U.S. & Mex. Bound. Surv. 2: 222. 1859, as to Pl. Lindh. only; Wats. in Proc. Am. Acad. 14: 281. 1879, as to synonym; Baker in Jour. Linn. Soc. Bot. 17: 477. 1879, as to synonym and Lindheimer 543 and 711; Wats. in Proc. Am. Acad. 18: 166. 1883, as to Coulter 1570 only ; Hemsl. in Salvin & Godman, Biol. Cent.-Am. Bot. 3: 382. 1885; Small, Fl. South- [Vor. 29 306 ANNALS OF THE MISSOURI BOTANICAL GARDEN east. U.S. 250. 1903; Wooton & Standl. in Contr. U.S. Nat. Herb. 19: 129. 1915, as to synonym. Schoenocaulon intermedium Baker in Jour. Linn. Soc. Bot. 17: 477. 1879, as to Coulter 1568 and 1570. Bulb 1.5-2.5 em. in diameter; basal portion of the plant covered with a dense cylinder of brownish-black fibers 7-15 em. long; leaves narrow, recurved, 3—6 dm. long, 2-5 mm. broad; scape slender, 3-4.5 dm. long; spike pointed at the apex, densely flowered, 7.5-18 em. long, 10-15 mm. in diameter; braets acutish to acute, 2 mm. long; flowers erect, crowded, subsessile, pedicel less than 1 mm. long; perianth-segments linear-oblong, thickened, entire or obscurely den- tate, 2 mm. long; filaments slender, slightly dilated toward the base only, 3—4 mm. long; capsules somewhat appressed to the axis of the inflorescence, 10-14 mm. long, 4-7 mm. in diameter; fruiting pedicel 2mm. long; seeds 5—6 mm. long, 1-3 in each cell; flowering in spring. Distribution: southwestern United States and northern Mexico. UNITED STATES TEXAS— : Deni 30 March 1932, Jones 29080 (M, P, UC). BEXAR CO.: Leon Springs, Needs life z zone, 17 May 1911, Clemens 483 (CA, M, P); nw. of San Antonio, hard limestone hillside bns scenie loop, 26 € 1936, Metz 2436 (NY); San Antonio, April 1922, Schulz 792 (US). BREWSTER со: 5 m of Nichols’ Ranch House, 4 April 937, Warnock 1625 (С). сомль 00.: New Braunfels, stony prairies, dry places, April 1846, Lindheimer 543 (G, M, МУ, PA, UC, US, all isorYPES), and April 1848, 711 (G, M, NY, veu a Spring, 1850, Lindheimer 1220 (M), and May 1851, 1221 (G, M, NY , UC, US). cULBERSON CO.: ridge above McKittrick Canyon, shaded rocky places, ^ въ 1931, Moore ¢$ Steyermark 3485 (CA, D, С, M, NY, PA, UC, US). GILLES- PIE CO.: Nibo (f) Mt., Jermy 329 (M, US). GOLIAD co.: Goliad, open prairie, March 1927, Williams 50 (PA). HAYS co.: San Marcos, rocky uplands, 8 April 1918, Palmer 15309 US); San Mareos and peri spring 1898, Stanfield (NY). JEFF DAVIS sw. of Toyahvale, 30 Oet 5, Cory 17533 (С). KENDALL CO.: 8% mi. nw. of Boerne, 24 May 1935, Cory 13988 re rocky bluffs, June 1885, Reverchon 1607 (D, G, M, NY, P, US) ; Boerne, dry calcareous her 22 May 1916, Palmer 9841 (D, M, PA, US) ; on lime- stone hill, e. of Comfort, 29 April 1940, von Schrenk (M). KERR со.: Turtle Creek, 2 May 1899, Bray 227 (US) ; about Kerrville, 480—600 m, alt., 12-19 June 1894, Heller 1626 (G, Y, PA, UC, US) ; Kerrville, sandy loam of stony hill, 9-10 May 192 20, Pennell 10373 (NY, PA). TRAvis co.: Bull Creek, 2 May 1926, Bogusch 546 (08); Austin, 13 May 1872, Hall 644 (G, M, NY, P, US); near Austin, 24 April 1914, Young (M); Austin, 18 April 1903, ex Herb. Biltmore 14799 (US). Between Kerrville, KERR со. and San Antonio, BEXAR СО.: 23 April 1931, Jones 78405 (M, P, UC). Between Mason, MASON co. and Fred- ericksburg, GILLESPIE со., 14 Мау 1932, McKelvey 2780 (G). WITHOUT DEFINITE LOCALITY : uadalupe Mts., w. Texas, 1882, Havard 29 (G, PA) ; 1881, Havard (US). ‘‘ Between w. Texas and El Paso, New Mexieo,"' May—Oct. 1849, Wright 697 (G, UC, US). Upper Gua- m stony prairie, April 1845, Lindheimer 416 (G, M); mountainous prairies, May- June 1884, Reverchon 1607 (M). Nm MEXICO—CHAVES CO.: 10 mi. w. of Roswell, 28 July 1905, Wooton (US). : Guadalupe Mts., e. of Queen, crevices of rocks, 19 May 1932, Wilkins 2026 (PA, 8). 1942] BRINKER—MONOGRAPH OF SCHOENOCAULON 307 MEXICO—CHIHUAHUA: Santa Eulalia Mts., May-June 1885, Pringle 40 (G, NY, PA, US); same gare 18 Aug. 1887, Pringle 40 (M 207467), and coll. of 1885, Wilkinson (NY). HIDALGO: near Zimapan, Coulter 1568 and 1570 (G). EVO LEON: Monterrey, Sierra Е. Mts., 27 July 1933, Mueller 12 (G); foothills ue Pablillo, 15 mi. sw of Pueblo Norma alt. 2256-2440 m., rare in oak scrub, 21 May 1934, MUN 518 (с). UEBLA: vieinity of Puebla, Hueyotlipan, alt. 2180 m., 15 June 1908, re 10227 (08); кил 20 July 1910, Nicolas 2302 (US); Arsóne (US 1031247). TAMAULIPAS: vicinity of Marmolejo, Pico del Diabolo, 12 Aug. 1930, Bartlett 10918 (US) ; kilo. n.w. of Е Progresso, which is 18 kilo. n.w. of Ocampo, on mountains with m Жкн м alt. 1450 m., 22 Aug. 1941, Stanford, Retherford $ Northcraft 1055 (M). 20. Schoenocaulon yucatanense Brinker, п. sp.?? Bulb unknown; leaves 6 dm. and more long, 6-7 mm. broad; scape slender, 44 ст. long; spike 11 em. long, 15 mm. in diameter at anthe- sis; all flowers sessile, erect; bracts small, scarious, very erose; peri- anth-segments erose-margined, ovate-oblong, about 2.5 mm. long; filaments three times as long as the perianth, slightly dilated above, reflexed, yellow; eapsule unknown. MExiCo—vUCATAN: Uxmal, on base of large pyramid, 20-21 July 1932, Steere 2093 (NY TYPE). * Schoenocaulon yucatanense sp. nov. Bulbus ignotus; foliis 6 dm. et ultra longis, 6-7 mm. latis; seapo gracili, 44 em. longo; spica 11 em. longa, anthesi 15 mm. diametro; floribus omnibus sessilibus, erectis; braeteis parvis, seariosis, valde erosis; perianthii seg- mentis margine erosis, ovato-oblongis, eirea 2.5 mm. longis; filamentis Бона triplo longioribus, superne sensim dilatatis, duxit luteis; capsulis ignoti EXCLUDED SPECIES Asagraea frigida (Schlecht. & Cham.) Lyons, Pl. Names Scientif. and Pop., ed. 508. 1907 = Stenanthium frigidum Kunth (Veratrum frigidum Schlecht. & Cham.), fide Index Kewensis. Asagraea longiflora Rusby in Bull. N. Y. Bot. Gard. 6: 491. 1910 = Tofieldia falcata Pers. Syn. 1: 399. 1805. List or EXSICCATAE The numbers in the parentheses indicate the numbers of the species as treated in the monograph. e collector’s numbers are in italics; the abbreviation s.n. signifies that the specimen is without a collector з number Aguilar, I. 728 (14). Bourgeau, E. 2981 (14). Alden, Lieut. s.n. (7). Bray, W. L. 227 (19). ruine ; Br. G. s.n., 10227 (19). Buckley, S. B. s.n. (7). Bartlett, H. H. 10918 (19). m I. s.n. dr Berendt, s.n. (8). rows, Dr., s.n. (7). Biltmore (ex Herb.) 14799 (19). одаи в. 1060 (14). Bogusch, Е. R. 546 (19). Chapman, A. W. (ex Herb.) s.a. (7). Botteri, M. 1186 (14). Clemens, Mr. and Mrs. J. 483 (19). 308 Conzatti, C. 688 (18). Conzatti, C. & Gonzalez, V. 323 (3); 449 4). Cook, O. Е. & Gilbert, С. В. 1018, 1538 (14). Cory, V. L. 16146, 20644 (6) ; 13988, 17533 Coulter, T. 1569 (5) ; 1568, 1570 (19). Crawford, J. s.n. (7). Curran, H. M. & Haman, М. 1196 (14). Curtiss, A. H. 2900, 6551, 6606 (7). Drummond, T. 284 (6). Eggers, H. F. A. von. 73361 (14). ее, К. г.п. (2). Elias, Вт. 71 (14). Ernst, А. s.n. (14 Fendler, A. 7506 (14). Fredholm, A. 5387, 6032 (7). я А. = s.n Gentry, H. £515, 2515, 2591, 6555 (11). na dod A. 672 (8). Gregg, J. 214 Ai Hartweg, T. 627 (14). Havard, V. s.n., 29 (19). Heller, A. A. 1629 (19). Hinton, G. B. 2690, 4970 (4); 13465 (12); 2039, 2420, 9397, 9506, 9539, 13226 (14). Hitchcock, А. В. s.n. (7). Hood, 8. C. s.n. (7). Howell, A. H. 1105 (7). Huger, A. M. 13 (7). Hulst, G. P. s.n. (7). Jermy, G. 329 (19). Jimenez, O. 637 (14). Jones, M. E. s.n. (11) ; 29080, 78405 (19). Johnston, T. R. 961 (14). Kline, s.n. (7). Lawranee, A. E. 888 (14). Leavenworth, М. C. s.n. (7). Lewton, F. L. s.n. (7). Liebmann, F. м, 14625, 14627, 14628, 14629 Lighthipe, І. н. 479 (7). Lindheimer, Е. 416, 543, 711, 1220, 1221 (19). Lundell, С. L. 5464 (17). Lyonnet, C. E. 329 (13). Manry, P. 625? (6). arsh, E. G. s.n. (7). Matuda, E. 1346, 1600 (14). McFarlin, J. B. 5261 (7). [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN McKelvey, 8. 2 2780 (19). Metz, M. C. 2436 (19). Mexia, Y. 1702, ea (16). Mohr, C. A. s.n. (7). Mohr, C. A. & Botteri, M. s.n. (14). Moldenke, H. N. 1087, 5943 (7). Moore, J. A. & Steyermark, J. A. 3485 (19). Mueller, C. H. & M. T. ?24 (10) ; 12, 513 19 (19). Mueller, F. 222, 231 (14). Murrill, W. A. s 15. Nash, G. V. 297, peur S Nelson, E. W. 2847, 2849 m. 2530a (18). nega. F. 5302 (19). O'Neill, H. s.n. PC E. 119, 1322 (6); 1410 (14); 419 (16). Palmer, E. 7. 9841, 13309 (19). Parks, H. B., 6). Parry, C. С., Bigelow J. M., Wright, C. & А «3 , E. 882 (3). Pennell, F. W. 555? (6); 10373 (19). Pittier, H. 7552, 9556, 11611 (14). ringle, C. G. 5754, 6740 (1); 13841 (4); 2938 (9); 6415, 11716, 13621, 13778 (15); juge (17); 5857, 6653, 10184 (18) ; 40 (19). Purpus, " А. 4387 (2); 2731 (3); (4) ; 405, 2023 (14). Reko, B. P. 3475 (14). Renson, C. 167 (14). Reverchon, J. ed (19). ло J. з.п., 58 (6). Rose, J. N. ди (15); 2987 (16). Rose, J. №. & J. 8. 21887 (14). Rose, J. N. & Hay, R. 5690 (15). Rose, J. N. & Hough, W. 4972 (2). Rose, J. N. & Painter, J. H. 6722 (13). Rose, J. N., Painter, J. H. & Rose, J. В. (15). 2490 9243 Rose, J. N. & Russell, x G. 24160 (6). 3). W.& в Е. 988 (14). Sehlottmann, s.n. (6). Schott, A. s.n. (6). Sehrenk, Н. von. s.n. (19). Schulz, E. D. 7 Seler, C. & E. je 3273, 5379 (14). Standley, P. C. 62922 (14). 1942] BRINKER— MONOGRAPH ОЕ SCHOENOCAULON Stanfield, S. W. s.n. (19). Stanford, L. R., Retherford, K. L. & North- 227, 32025 (14). Wilkins, H. 2026 (19). 309 Wilkinson, E. s.n. (19). Williams, C. B. 50 (19) ; 91 (6). Wiliams L. 12398, 18604 (14). „Е. О. г.п. (19). wien с. И Се n. (7). Ше M. ы з.п. (19). Yuncker, Т. G., Dawson, В. Е. & Youse, H. R. 6385 (14). INDEX TO GENERA AND SPECIES Accepted names are printed in Roman type; synonyms in italics ; combinations, in bold face type Page As „л ERS e ors EM p cir rm 293 COT OCOSQIG Met c E 302 СОЛО орат. ее 296 О ТЬ т 307 О о зо атоо 307 ОН. 302 CONAAT Ol са eR SS tres 305 THIClOnAas Parao с dst. 293 СА 298 ОПОПаИЗ ое 302 ISU GOAL о ое 293 ООО OT a V eiit сеге 296 Coulier$ лро 297 ОТО CL оро 298 Иса ОА о 298 (Uu ит парола ала даи зара 302 Sehoenocaulo n NN униле claret 292 LR ILLE: ew cn E SD E a a 298 Е uM OTT 295 Sabina uM 295 alm ce OE 296 меп eA be es eec Eee ак үе, 297 (ОГ Цар л асо зл: 297 ТОТОП 297 DIT UMNO TAU ueris ere etes 305 new names and new Page GUI O E E BS 294, 298 Ghiesbreghbg O a 99 I IDS 298 intermedi covers 306 ООН е crete dae ае 300 TIACIOCATDIUDAS neuro. 300 терага ааа сезонни orm 300 MERTE esee mte cris йы а: 300 МОГ e e ER Ie 301 obtusum e оо. 301 остао Pacers. nee. ea ОС ОЕ 301 ETIN IOI X OM ее 303 те Шатев пот i ener esos 304 BUG ааа ааа аи rs 304 вее OC л жалал sve трета 305 пиша hee. 305 УЛОВ. 1 ризата ре 307 SKOMOLON Ле. ieir e илуу: 298 Stenanthium Раши ..... v. 307 това Та. eere ert Sims 307 VAI SIT а. 293 СООО ео 296 dieta SO ОООО 307 OMONE rr ат 302 itn Р е CE Senecae: 305 310 [Vor. 29, 1942] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 27 Center: Generalized flower of Schoenocaulon, x 4%. Figs: A-G: Capsules. Fi ig. A. S. tenue, x 4. Fig. B. S. — ا‎ x ad 8. regulare, x 514, 8. officinale, x 5, S. texanum, x 5. Fig. C. S. Pringeli, Fig. D. S. maer ыо 416. Fig. E. S. tenuifolium, x 414. Fig. F. S. Conzattii, x 4, S. dubium, x 4, S. jaliscense, x 6, S. megarhiza, x 6, S. ob- tusum, x 4. Fig. × G. 8. caricifoliwm, x не; 8. comatum, x 6. "- 1-13: Perianth-segment Fig p $939 sms po 8. regulare, x 3. 8. yucatanense, x 415. S. Pringlei, x 4. S. obtusum, x 4, S. tenue, x 5. S. Mortonii, x 416 S. dubium, x 4. caricifoliwm, x 4. „в. а А 10. S. Ghiesbreghtii, x 4. . 11. S. macrocarpum, x 5; 8. texanum, X 4; S. jaliscense, x 5; S. calcicola, x 4; . Conzattii, x 3; S. megarhiza, x 6; S. Coulteri, x 4. в E A fenuifolism, х 4. . 13. S. comatum, x 4. ANN. Мо. Вот. GARD., Vor. 29, 1942 PLATE 27 LI WO) E 12 G o BRIN KER—-MONOGRAPH OF SCHOENOCAULON | | | 3 | à PLATE 28 ANN. Mo. Вот. GARD., Vor. 29, 1942 S. tenue 29) EXPLANATION OF PLATE PLATE 29 Maps showing geographical distribution of species of Schoenocaulon. E fe № C sae 3 N | к PLATE 29 Conzattii . macrocarpum 15. Ghiesbreghtii ANN. Мо. Bor. GARD., Vor. 29, 1942 BRINKER—MONOGRAPH OF SCHOENOCAULON CONTRIBUTIONS TOWARD A FLORA OF PANAMA}? VI. COLLECTIONS CHIEFLY By Н. von WEDEL IN Bocas DEL Токо ROBERT E. WOODSON, JR. Assistant Curator of the Herbarium, Missouri Botanical Garden Associate Professor, Henry Shaw School of Botany of Washington University AND ROBERT W. SCHERY earch Assistant, Missouri Botanical Garden Instructor, б, Shaw School of Botany of Washington University GRAMINEAE (Jason R. Swallen) Cryptochloa Swallen, gen. nov. Monoecia; spiculis masculis et femineis in eadem inflorescentia dispositis ; spiculae masculae: glumae et lemma sterile nulla; lemma acutum vel acuminatum, 1-nervio ; palea lemma aequans ; stamina 3; spiculae femineae: gluma prima nulla; gluma secunda et lemma sterile subaequalia acuta vel acuminata, 3—5-nervia, nervis laterali- bus approximatis ; fructus stipitatus, subcylindricus, firmus, albus vel fuseus, marginibus non involutis ; palea lemma aequans. Peren- nis culmis gracilibus, planis, et laminis latis, planis, breve petiolatis. Low monoeeious perennial grasses with broad flat blades usually crowded toward top of the slender wiry eulms. Inflorescence small, usually partly hidden in the upper sheaths, each bearing both stam- inate and pistillate spikelets in no definite arrangement ; staminate spikelet : glumes and sterile lemma wanting ; lemma and palea acute or acuminate, thin, the lemma 1-nerved ; stamens 3; pistillate spike- let : first glume wanting ; second glume and sterile lemma acuminate, subequal, 3- or usually 5-nerved, the lateral nerves approximate, finely transversely veined; fertile floret subcylindrical, raised on the enlarged and thickened EE of Ше rachilla ; lena firm, subin- durate, smooth and shining, gradually narrowed to the blunt tip, the margins not inrolled ; palea as long as the lemma, similar in texture. A genus of southern Mexico and Central America. Type species: C. variana. Name from кролтос, hidden, хАба, grass, referring to the partly enclosed panicles and also its forest habitat. 1 Issued December 18, 1942. (317) [Vor. 29 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN This genus is closely related to Raddia Bertol. (Strephium Schrad., type S. distichophyllum) and has been confused with it. Raddia, however, has the staminate and pistillate spikelets in sep- arate inflorescences (the staminate terminal or from the upper nodes, the pistillate axillary) ; the fertile floret is not raised on the lengthened and thickened segment of the rachilla; the fruit is much smaller, oval or ovate rather than subcylindrical, and the lateral nerves of the second glume and sterile lemma are evenly spaced rather than approximate. Diandrolyra Stapf and Olyra L. are also close kin. The first is distinguished by the spikelets arranged in pairs, one staminate and one pistillate, the fruit is not raised on an enlarged segment of the rachilla, and the lateral nerves of the second glume and sterile lemma are evenly spaced rather than approximate. There are also only two rather than three anthers. In Olyra the panicles are all terminal on the main eulm and branches, not axillary, the pistillate spikelets at the ends of the branches and the staminate below. The second glume and sterile lemma are usually attenuate, with the lat- eral nerves evenly spaced, not approximate. The fruit is sessile, relatively broad and thick, with the margins of the lemma inrolled, rather firmly clasping the palea. Key ro SPECIES Culms 10-30 em. high, slender; staminate spikelets 2.5-3 mm. long. Blades 3-5 at the summit of each culm, 3-5 em. long, not conspicuously pp i DUELO EA UK EET EAT и OCTET eee eee ee Te ҮКҮ ҮП C. variana Blades 10-20 at the summit of each culm, 1.5-3 ст. long, conspicuously ойнаи: Cp MGA FSW Wale ia V кит и ара Каида С FEY'S PEL GIF ER ERE A UP RE TES 2. C. concinna Culms 20-50 em. high, at least some of them more than 30 em., relatively coarse; staminate spikelets 4.5—5 mm. lon Panieles, or at least some of thew. са on long slender peduncles; blades ob- long, 3.5-5.5 em. long, mostly 12-20 mm. wide; second glume and sterile lemma of pistillate spikelets smooth; lem of staminate spikelets acute..3. С. strictiflora Panieles, all of them, partly e hri in the sheaths, the peduncles short; blades —Ó 6-7.5 em. long, 10-13 mm. wide (occasionally as much as 10 em. long 7 mm. wide); second glume and sterile lemma of fertile spikelet granular- качва lemma of staminate spikelet subattenuate............ 4. C. granulifera 1. Свуртоснтол variana Swallen, sp. nov. Culmi caespitosi, 10-20 em. alti, infra nodos pubescentes ; vaginae carinatae margini- bus pubescentibus; ligula 1-3 mm. longa, obtusa, pubescens ; lami- пае 3—5 em. longae, 8-13 mm. latae, oblongo-lanceolatae, acutae vel acuminatae, minute pubescentes ; spieulae masculae 2.5 mm. longae, lemmate subacuto; antherae 1 mm. longae; spiculae femineae 7-8.5 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 319 mm. longae, gluma secunda et lemmate sterili acutis vel acuminatis, 3-nerviis, glabris ; fructus 6.5-7 mm. longus, subeylindricus, fuscus. Culms in small dense tufts, wiry, 10-20 cm. tall, erect to spreading, often geniculate at the densely pubescent nodes, glabrous, or pubes- cent below the nodes, the lower internodes somewhat elongate, the upper ones much shorter, completely hidden by the overlapping sheaths; sheaths keeled, pubescent toward the summit and on the margins, nearly glabrous on the back, often auriculate, the auricle Cryptochloa variana: Plant, x 1% ; staminate spikelet, fruit, and pistillate spi fused with the ligule, the lower ones bladeless or with very much reduced blades ; ligule 1-3 mm. long, membranaceous, obtuse, puber- ulent on the back or nearly glabrous; blades 3-5 on each culm, crowded toward the summit, 3-5 em. long, 8-13 mm. wide, oblong- lanceolate, rather abruptly narrowed to an acute or acuminate tip, broad and rounded at the base, with a densely pubescent petiole about 1 mm. long, minutely pubescent on both surfaces or some- times only obscurely puberulent; inflorescences terminal and axil- lary, partly enclosed in the sheaths, 2-3 em. long, bearing 1-6 pistil- late spikelets, the branches elosely appressed, usually pubescent; staminate spikelet 2.5 mm. long, the lemma subacute, the palea as [Vor. 29 320 ANNALS OF THE MISSOURI BOTANICAL GARDEN long as or slightly longer than the lemma; anthers 1 mm. long; pistillate spikelet 7-8.5 mm. long, the pedicel 4-10 mm. long, rela- tively stout, much enlarged toward the summit; second glume and sterile lemma 3-nerved, with a few fine transverse nerves, glabrous or obscurely seaberulous, the second glume acute, the sterile lemma acute or subacuminate, slightly exceeding the second glume; fruit 6.5-7 mm. long, 2 mm. broad, subeylindrieal, broadest above the middle, gradually narrowed to a blunt tip, smooth, shining, gray- greenish, at maturity mottled with darker drab; rachilla segment between the sterile lemma and the fruit enlarged and elongated, 1-1.5 mm. long, about as thick as the base of the fertile floret, whit- ish, somewhat soft or waxy in appearance; caryopsis 4.5 mm. long, light brown. PANAMA: CANAL ZONE: near bank of Madden Reservoir, Muenscher 12212. сос: hills north of Е Valle de Antón, alt. 1000 m., July 14, 1940, Allen 2201 (U.S. National Herb., түрк). The name variana, meaning vari-colored, refers to the mottled fruits. 2. CnvProcuroA concinna (Hook f.) Swallen, comb. nov. (Olyra concinna Hook. f. Bot. Mag. [Curtis] III, 52: pl. 7469. 1896 ; Raddia concinna Chase, Proc. Biol. Soc. Washington 21: 185. 1908.). Culms 15-30 em. high, slender, erect, or geniculate at the lower nodes, the lower and especially the middle internodes much elongated, the upper ones very short; sheaths keeled, glabrous or pubescent; ligule truncate, 0.3-1 mm. long; blades 1.5-3 em. long (usually 2 em.), 5-9 mm. wide, crowded on the upper third of the culm, 10-20 on each culm, conspicuously distichous ; inflorescences small, almost entirely hidden in the uppermost sheaths; fertile spikelet 8-10 mm. long, the second glume and sterile lemma subequal, subacuminate, 3—5-nerved, with fine transverse veins, glabrous; fruit about 8 mm. long, subeylindrieal, scarcely broadened above the base, gradually narrowed to the blunt or rounded tip. Wet forests at low altitudes, Nicaragua, Costa Rica, and Colom- bia. Nicaraaua: Sandy Вау, 1922, Hamilton s.n. Costa Rica: 1лмбх: Hamburg Finca, Río Reventazón below Cairo, Standley 48661, 48783, 48854; Oesterr. Biol. Costarica-Expedition 680, 1930, (coll. Cufodontis ). COLOMBIA : BOLÎVAR: Norosi-Tiquisio trail, Lands of Loba, Curran 129. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI J2 3. Свуртосньол strictiflora (Fourn.) Swallen, comb. nov. (Strephium strictiflorum Fourn. Bull. Soc. Bot. Belg. 15: 465. 1876; Olyra strictiflora Hemsl. Biol. Cent. Am. Bot. 3: 510. 1885; Raddia strictiflora Chase, Proc. Biol. Soc. Washington 21: 185. 1908.). Culms 20-50 em. high, erect or geniculate at the lower nodes, pubescent, more so on one side than the other, the nodes retrorsely pubescent, the intermediate internodes elongate; sheaths keeled, glabrous or pubescent; ligule 2-3 mm. long, obtuse, fused with the auriculate summit of the sheath; blades scarcely crowded, sometimes not at all, oblong, 3.5-5.5 em. long, 12-20 mm. wide, abruptly narrowed to an acute tip, glabrous or minutely puberulent, the margins scabrous; inflorescences terminal and axillary from the middle and upper sheaths, at least some of them on long slender exserted peduncles ; staminate spikelets 4.5 mm. long, the lemma acute; fertile spikelet 11-12 mm. long, the second glume and sterile lemma acuminate, subequal, 5-nerved, very faintly transversely veined; fruit 7 mm. long, subeylindrical with nearly parallel margins, ivory-white. Gulf region of Mexico. Mexico: veracruz: Mirador, Liebmann 266; Hacienda de Jovo, Inebmann 267; Cordoba, 1865-66, Bourgeau. 4. Свуртоснгол granulifera Swallen, sp. nov. Culmi graciles, ereeti, ad 50 em. alti, nodiis pubescentibus; vaginae carinatae gla- brae vel ad apieem pubescentes; ligula 0.5-2 mm. longa, obtusa; laminae lanceolatae, 6-7.5 cm. longae, 10-13 mm. latae, acuminatae, marginibus seabris; paniculae 2—5 em. longae, contractae; spiculae maseulae 4.5-5 mm. longae, lemmate subattenuato; spiculae femi- neae 11-12 mm. longae, gluma secunda et lemmate sterili acumi- natis, 5-nerviis, granulosis; fructus 7-8 mm. longus, albus. Culms slender, erect, occasionally somewhat geniculate at the lower nodes, as much as 50 em. high, pubescent in a line on one side, the nodes retrorsely pubescent, otherwise glabrous; sheaths keeled, glabrous, or pubescent toward the summit, mottled with dark spots ; ligule 0.5-2 mm. long, obtuse, pubescent, fused with the auriculate summit of the sheath; blades broadly lanceolate, the upper ones 6—7.5 em. long, 10-13 mm. wide (sometimes 20 mm.), rather gradu- ally narrowed to an acuminate tip, glabrous on both surfaces, the margins scabrous; panicles 2—5 em. long, contracted, partly enclosed in the upper sheaths; staminate spikelets 4.5—5 mm. long, the lemma acuminate or subattenuate, appearing as if awned; fertile spikelet 11-12 mm. long, the second glume and sterile lemma subequal, acu- [Vor. 29 322 ANNALS OF THE MISSOURI BOTANICAL GARDEN minate, 5-nerved, finely transversely nerved, granular-roughened ; fruit 7-8 mm. long, ivory-white, narrowed at the summit to a blunt tip. Forests, Mexico, Honduras and Guatemala. Mexico: VERACRUZ: Fortuña, Coatzacoalcos River, Williams 8378. CHIAPAS: Finca Irlanda, Ригриз 7403. Hoxpunas: Puerto Siena, forest along Tela River, Feb. 4, 1903, Wilson 325 (U. S. National Herb., туре). GUATEMALA: SAN MARCOS: Río Mopá, below Rodeo, Standley 68769. This specimen differs from the type in having blades as much as 10 em. long and 27 mm. wide. CYCLANTHACEAE CARLUDOVICA Druper Mast.—CFHIRIQUÎ : vicinity of Puerto Armuel- les, alt. 0-75 m., July 28-31, 1940, Woodson & Schery 910. Originally deseribed from plants of Colombian origin cultivated at Kew. CarLupovica ENSIFORMIS Hook.f.—currievi: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 625; Casita Alta to Cerro Copete, Voleán de Chiriquí, alt. 2300-3300 m., July 10, 1940, Woodson € Schery 373. Previously considered to be an endemic of Costa Rica. CanLUuDovicaA integrifolia Woodson, sp. nov. Planta ut videtur gracilis scandens. Folia longiuscule petiolata membranacea ; lamina 22-30 em. longa 4-6 em. lata elliptico-oblanceolata apicem versus late acuta ibique levissime crenulata deinde abrupte angusteque subeaudato-acuminata basim versus (са. 2/3 longitudine) integra gradatim acuta; petiolo 12-15 em. longo ca. 2/3 longitudine vagi- nato. Pedunculus 5-6 em. longus tenuis, spatharum nodiis ca. 3 sat distantibus. Spadix in fructu immaturo fusiformi-cylindricus ea. 2.0— 2.5 em. longus basi ca. 0.4 em. crassus; floribus femineis parvis concrescentibus ca. 0.5 em. diam., stigmatibus sessilibus parvis, lobis perigonialibus tenuissimis vix manifestis. Spathae deciduae ut videntur ca. 3 parvae pedunculi in parte dimidia superiore ges- tae.—parién : along the Sambü River, southern Darién, above tide limit, Feb., 1912, Pittier 5560 (U. S. Nat. Herb., түрк). As far as Тат aware, only three species of Carludovica previously have been described with entire leaves: C. diversa Drude, C. Traili- ana Drude, and C. heterophylla Mart., all from Brazil. The three species previously described apparently were found with both entire and bifid leaves (the entire leaves all of outline significantly differ- ent from ours), with larger spadices, and with only 1 or 2 spathes. 1942] ; WOODSON & SCHERY—FLORA OF PANAMA. VI 323 Our type specimen bears 6 leaves, all entire and so uniform as to indicate that to be the normal condition of the foliage. Another entire-leaved Carludovica was collected by P. C. Stand- ley in the garden of C. W. Powell at Balboa, C. Z., in 1925. 'The specimen (in U. S. Nat. Herb. no. 1252076) consists of a single leaf of thesame general outline as those of C. integrifolia but much larger (about 50 em. long, 12 em. broad), and is accompanied by the follow- ing remarks: ‘‘Said to be from the nearby woods. Acaulescent. Leaves all simple." No similar plants have been encountered in the Canal Zone since that time, and the cultivated plant apparently was lost when Mr. Powell's collections were removed by the Mis- souri Botanieal Garden. CARLUDOVICA Кпллри Standl—pnanmiÉw: Cerro de Garagará, Sambü basin, southern Darién, alt. 500-974 m., Feb. 7, 1912, Pittier 5658. Originally deseribed (Field Mus. Publ. Bot. 22: 65. 1940) from the region of Buenaventura Bay, Colombia. I have not been able to check the Pittier specimen with that of Killip, but Schery, who has seen both, regards them as probably conspecific. CARLUDOVICA MICROCEPHALA Hook.f.—socas DEL Toro: Water Val- ley, Sept. 23, 1940, H. von Wedel 921. Previously known to occur in the Greater Antilles, Honduras, and Costa Rica. CARLUDOVICA MICROPHYLLA Oerst.—CHIRIQUÎ: Rio Chiriquí Viejo valley, April 8, 1938, G. White 75; vicinity of Bajo Mona and Que- brada Chiquero, alt. 1500 m., July 18, 1940, Woodson & Schery 567. Previously considered as an endemic of Costa Rica. White 75 is quite typical of the species as exemplified by other herbarium speci- mens. However, Woodson Ё Schery 567 has very slender leaves attaining 60 em. in length and seems to correspond perfectly with the diagnosis of C. stenophylla Standl. (Fl. Costa Rica, p. 130. 1937) which was described from sterile plants. Our flowering and fruiting specimens have inflorescences quite conformable with those of C. microphylla, particularly with regard to the persistent spathes. Since similarly sharp leaf variation has been found to occur in other species represented by a number of herbarium specimens (cf. C. ensiformis and C. Oerstedii), we are tentatively regarding С. stenophylla as a synonym of C. microphylla. CanLupovica Pittieri Woodson, sp. nov. Planta mediocris ut vide- tur subaeaulis. Folia longe-petiolata membranacea ; lamina ca. 28 em. longa medio са. 15 em. lata apicem versus са. 1/4 longitudine bifida, segmentis late ovato-trigonalibus late acutis; petiolo 19 em. longo evaginato. Peduneulus 5 em. longus prope medium 2 nodiis [Vor. 29 324 ANNALS OF THE MISSOURI BOTANICAL GARDEN braetealibus (bracteis deciduis). Spadix in fructu globosus са. 2.0- 2.5 em. diam.; floribus femineis sat magnis concrescentibus ca. 0.8 em. diam., stigmatibus sessilibus, lobis perigonialibus depressis stigmata vix aequantibus.—san BLAS: high hills back of Puerto Obaldía, alt. 50-200 m., Aug., 1911, Pittier 4312 (U. S. Nat. Herb., TYPE). This species is conspicuous amongst the Central Ameriean Car- ludovieas because of its broad, scarcely divided leaves, recalling those of C. latifrons Drude of Brazil. The latter, however, has more deeply divided leaves of a different shape, and the stigmas are sup- ported by rather slender styles. CanLUDOVICA ROTUNDIFOLIA Wendl.—cuirigui: Quebrada Velo, Volcán de Chiriquí, alt. 1800 m., July 8, 1940, Woodson & Schery 248; vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson £ Schery 675. Originally described from Costa Rican plants grown at Kew. C. rotundifolia has usually been regarded as a synonym of the widespread C. palmata R. & P., but it probably should be maintained from the latter on the basis of the more elongate fruit- ing perigonial lobes and foliaceous outer spathes of the inflores- cence. COMMELINACEAE CocHLIOSTEMA ODORATISSIMUM Lem.—socas DEL токо: Isla Lobo, Chiriquí Lagoon, Sept. 6, 1941, H. von Wedel 2627. **Wercklé has published the following statement: ‘In the mountains south of Tur- ruvares a gigantic Cochliostema is abundant, and covers the thick trunks of trees. It is a very beautiful epiphyte.’ He reports it also from the Cordillera de Dota, as a plant 2 meters in height. The genus is known only from Ecuador, but probably it is represented also in Costa Rica.” [Standl. Fl. Costa Rica, p. 163. 1937]. This is one of the most interesting discoveries in Panama during recent years, and it comes as a distinct surprise, also, to find it growing so near sea-level, instead of in the mountains where it might have been expected. Although the species is so infrequently collected that I have not been able to compare our plant with other herbarium specimens from South America, the determination appears to be established by reference to the numerous published icones. Of С. odoratissimum, Sir J. D. Hooker wrote: ‘‘This superb plant cer- tainly ranks amongst the grandest stemless Monocotyledons known, combining the foliage of a gigantic Anthurium with masses of in- florescence which, for size, delicacy, and beauty of tints, cannot well be surpassed.’’ [Bot. Mag. pl. 5705. 1868.] 1942] WOODSON & SCHERY—-FLORA OF PANAMA. VI 329 LILIACEAE Ecueanpia prolixa Woodson, sp. nov. Herbae perennes 6-10 dm. altae omnino glabrae; rhizoma brevi recta; radicibus multis carno- sis tuberos elongato-fusiformes gerentibus. Folia plurima rad- icalia late linearia 60-95 em. longa ca. 2 em. lata, caulina 1-2 minora. Inflorescentia prolixa plus minusve procumbens paniculato-racemi- formis ; ramis 3-6 saepissime 2-4 ex axilla unica; bracteis scariaceis minimis. Flores parvi in fasciculis aggregati; pedicellis 1.0-1.5 em. longis sub medio articulatis; perianthii segmentis anguste oblongo- linearibus 1.0-1.2 em. longis са. 0.1 em. latis albis patulis; staminis antheris oblongo-sagittatis 0.6 em. longis, filamentis rugosis aequi- longis. Capsulae trigone obovoideo-oblongoideae apice truncatae vel leviter emarginatae basi attenuatae 0.7-0.8 em. longae са. 0.4 ст. latae.—PaNAMÁ: vicinity of Bejuco, alt. about 20 m., Sept. 7, 1942, P. H. Allen 2962 (Herb. Missouri Bot. Gard., түре). ‘Соттоп weedy herb growing in lax clumps on rocky hilltops. Inflorescence weakly procumbent in most cases. Flowers white with yellow stamens.”’ This species is conspicuous amongst described Echeandias be- cause of its rank growth. It is most nearly allied to E. macrophylla Rose, but material of that species which is available for study shows plants which are smaller in general stature, with anthers about 1 em. long and filaments about 0.6 em. long, as well as broader almost exactly ovoid capsules. The genus Echeandia has previously been known only from southern Mexico and northern Central America, with the exception of a single specimen (Fendler 1549) from north- western Venezuela which possibly represents E. prolixa. An addi- tional Echeandia is known from western Panama: EcnHeanpia venusta Woodson, sp. nov. Herbae perennes са. 3—4 dm. altae omnino glabrae; rhizoma brevi recta; radicibus multis сагпоѕіѕ. Folia plurima radicalia late linearia 12-30 em. longa 1.0- 1.5 em. lata multinervia albomarginata, caulina 1-2 multo minora. Inflorescentia racemiformis saepissime simplex; floribus in fascicu- lis aggregatis vel solitariis; braeteis exterioribus valde foliaceis spathaceis 2-7 cm. longis; pedicellis 1.5-2.0 cm. longis sub medio articulatis; perianthii segmentis aureis nervis 3 nigris oblongo- ellipticis 1.5-2.0 em. longis 0.4-0.5 em. latis; staminis antheris an- guste oblongo-sagittatis ca. 0.6 em. longis, filamentis 0.5 em. longis rugoso-erispatis.—cnH1iRIQUÍ: Potrero Muleto, Volcan de Chiriquí, alt. 3500 m., July 13, 1940, Woodson & Schery 379 (Herb. Missouri Bot. Gard., TYPE). [Vor. 29 326 ANNALS OF THE MISSOURI BOTANICAL GARDEN E. venusta is closely related to the group of rather dubious species centering about E. reflexa (Cav.) Rose, but differs from all in its conspicuously spathaceous bracts and somewhat larger flowers with longer pedicels. The species was very abundant at the type locality, and as charming as a planned floral display; it is well worthy of cultivation in northern greenhouses. SMILACACEAE (C. V. Morton!) Smax chiriquensis Morton, sp. nov. Liana 7.5 m. longa, caulibus conspicue et argute quadrangularibus, pallide lutescentibus, glabris, parce aculeatis, aculeis rectis vel eurvatis, basi latis ; petioli elongati, usque ad 6 em. longi, glabri, medio vel supra medium articulati ; lami- nae foliorum ovatae, usque ad 19 em. longae et 12 em. latae, apice breviter apiculatae, majores basi cordatae, minores basi truncatae, omnes integrae, non lobatae, papyraceae, pallide virides, glabrae, 9-nerviae, nervis extimis marginalibus, venis secundariis perspicue reticulatis, utrinque elevatis; umbellae masculae in ramis axillari- bus brevibus foliis suffultis, vel foliis valde reductis umbellis pseu- doracemosis ; pedunculus 1-3 em. longus, glaber, complanatus, quam petiolus longior; receptaculum parvum; pedicelli 5-11 mm. longi, glabri; perianthium viride, segmentis linearibus, 8-9 mm. longis, ea. 1.5 mm. latis, recurvis, glabris; filamenta gracilia, ca. 6 mm. longa, antheris parvis, ca. 1.5 mm. longis; flores feminei ignoti.— CHIRIQUÍ: valley of the upper Río Chiriquí Viejo, March 22, 1940, Peggy White 348 (U. S. Nat. Herb., no. 1,791,114, түрк); same locality, April 3, 1938, Gene White 59; Bajo Mona, Boquete, alt. 1350 m., April 2, 1938, M. E. Davidson 478. This species belongs to the section Hispidae of the revision of Killip and Morton. It may be distinguished from all the continental American species of that group by its sharply quadrangular stems, large, long-petioled leaves, large flowers, and minute anthers. The Davidson specimen was distributed as S. Ведейш Killip & Morton, to which it is not closely allied. S. Regelii has the perianth segments only 3.5-5 mm. long (rather than 8-9 mm., as in S. chiriquensis ) and the anthers are longer than the very short (1.2 mm.) filaments. In S. chiriquensis the anthers are very small, much shorter than the elongate filaments, these about 6 mm. long. 1 Published by permission of the Secretary of the Smithsonian Institution. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 327 DIOSCOREACEAE (C. V. Morton) DrioscongA STANDLEYI Morton—cuiriavt: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson € Schery 652 and 664. Previ- ously known from Costa Rica. IRIDACEAE NEOMARICA CAERULEA (Ker-Gawl.) Sprague—socas DEL TORO: Lit- tle Bocas, July 16, 1941, H. von Wedel 2546. This species occurs nat- urally from the Guianas to southern Brazil, according to J. G. Baker (Handb. Irid.). Out of this range, it has been cited from cultivation in highland Costa Rica by Standley (Fl. Costa Rica). Consequently, whether the species is indigenous or an escape is open to question, although Mr. von Wedel usually is careful to limit his collections to apparently indigenous plants. The plant resembles a gigantie Sisyrinchium with scapes 1.5-2 m. tall and violet-blue flowers 5-6 em. in diameter. MUSACEAE HELICONIA PsrrrACoRUM L.f. (Н. hirsuta L.f. Suppl. 158. 1781; H. cannoidea L.Rich. Nova Acta Acad. Nat. Cur. 15, suppl.: pls. 9-10. 1831; H. aurantiaca Ghiesbreght ex Lem. Ill. Hort. pl. 332. 1862; H. straminea (Griggs) Standl. Fl. C. Z. 75. 1928).—Unless one shares the rather naive faith in the constancy of Heliconia species sug- gested by some recent authors, one must assume great complexity for this species, particularly with regard to such characters as length of the peduncle and color of bracts and flowers. Fortunately, the species is common in Panama, and has been collected abundantly. H. psittacorum and H. hirsuta were published in the same work by the younger Linnaeus (Suppl. 158. 1781), and since they have not been combined previously, I am adopting the former as more ex- pressive of the aspect of the plants. Нетлсохта gosrRATA В. & P. (Н. pendula Wawra, Oesterr. Bot. Zeitschr. 13: 8. 1861 ; H. curtispatha Petersen in Mart. Fl. Bras. 38: 15. 1890; H. longa Griggs, Bull. Torrey Club 31: 446. 1904).—It is difficult to see why modern authors have failed to recognize the es- sential similarity of the plants bearing these names to the suggestive illustration published by Ruiz and Pavon (Fl. Peruv. 3: pl. 305. 1802). HELICONIA sUBULATA В. & P. Fl. Peruv.3: pl. 303. 1802. (Н. acum- (Vor. 29 328 ANNALS OF THE MISSOURI BOTANICAL GARDEN inata L. C. Rich. Nova Acta Acad. Nat. Cur. 15, suppl. : pl. 11. 1831; Н. psittacorum L. f. var. В. subulata (В. & P.) Baker, Ann. Bot. 7: 199. 1893).—4 thoughtful interpretation of the illustrations, crude as they are, makes their association as synonyms quite inescapable for me. HELICONIA VELLERIGERA Poeppig, Reise Chile 2: 295. 1836; Peters. in Mart. Fl. Bras. 33: 18. 1844; K.Sch. in Engl. Pflanzenreich 445: 37. 1900. (Bihar vellerigera (Poeppig) O.Ktze. Rev. Gen. 2: 685. 1891; Griggs, Bull. Torrey Club 4? : 318. 1915).—cocr£: El Valle de Antón, Woodson & Schery 205, Allen 1818; Las Minas, Allen 2707; PANAMA: Río Boquerón, Hunter Ё Allen 659. This is one of the most striking species of Heliconia, attaining a height of 3 meters, the long inflorescences pendulous and clothed with very dense and brilliantly ferruginous hairs about 1 em. long (the brilliant scarlet bracts and upper peduncle occasionally somewhat glabrate). The identification of the Panamanian plants with Poeppig’s has not been entirely an easy matter, although a guiding principal in our dealings with highland Panamanian plants has been liberal consul- tation of the disused species of Peru enumerated by Ruiz & Pavon and Poeppig. The complicating factor in this instance was that H. vellerigera is not represented by exsiccatae in American herbaria which we have consulted, and was known to Poeppig, Petersen, and Schumann only from a fragment of an inflorescence in the herbarium at Vienna. Schumann undertook to key the species from others on the basis of a supposedly erect inflorescence, which was repeated by Griggs. Nevertheless, we were struck with the fact that both Petersen and Schumann took pains to describe the indument of the inflorescence as “pili. .ad 1 cm. longi" amplified by the former by the remark, ‘‘Species incomplete cognita, sed vellere ad omnibus celeris Нейсопиз distinctissima," and we hopefully labelled our Panamanian specimens as Н. vellerigera. Faith in our rather intuitive use of the early Peruvian authors has recently been vindicated in this case by examination, through the kindness of Dr. Standley, of two photographs of H. vellerigera from Peru; one specimen, the type of Poeppig, preserved in the herba- rium at Vienna, and the other, a recent collection (Weberbauer 6764) in the herbarium at Berlin. There can be scarcely a shadow of doubt that the plants of Panama and those of Peru are quite con- specific. Unfortunately, the label of Weberbauer’s plant does not indicate whether the inflorescence was erect or pendulous, but the latter almost certainly must have been the case. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 329 Hxaricowia уплозА Kl. (Heliconia nutans Woods. Ann. Missouri Bot. Gard. 26: 276. 1939).—cumiqvuí: Volcán de Chiriquí, alt. 1500- 2000 m., Woodson, Allen Ё Seibert 968 ; сост : north of El Valle, alt. 1000 m., Allen 2167. Intensive study of the Panamanian Heliconias has convinced me of the folly of continuing the tenuous specific dis- tinetions eurrent at present. The species undoubtedly are extremely variable in all but the most conservative criteria, and hybridization may be suspected in several instances. The specimens cited above agree in all essential particulars with the original description of H. villosa and the illustration by Petersen in the ‘Flora Brasiliensis’ (33: pl. 4. 1890). Particularly is this true with regard to the speci- men from the Province of Coclé. The type of H. nutans (Woodson, Allen & Seibert 968) is merely a specimen with somewhat more numerous and smaller bracts. The indument is rather variable, and nearly glabrous specimens have been observed, particularly in the northern range of the species which appears to extend as far as Honduras. ZINGIBERACEAE Reneatmia (Scaposae-Racemosae) Arundinaria Woods. sp. nov. Herba gracilima omnino glabra. Culmi foliiferi gracillimi 3-4 dm. alti. Folia longiuscule (0.5-1.0 em.) petiolata angustissime ob- longo-lanceolata basi apiceque attenuata 7-15 em. longa 1.0-1.7 em. lata superne minora; ligula angusta 3—6 cm. longa truncata vel obscure auriculata purpureo-marginata. Panicula racemiformis pauciflora; culmis gracillimis 5-7 em. longis, vaginis ca. 5-6 oblongo-ovatis obtusis 1.25-1.5 ст. longis membranaceis ; pedunculis 2.5-3.0 em. longis 8-12-floris; bracteis ovatis vel ovato-lanceolatis 0.5-0.7 em. longis pallide roseis. Flores ignoti. Capsula ovoidea 0.6-0.7 ст. longa coccinea; calyce anguste turbinato са. 0.5 em. longo; pedicello са. 0.5 em. longo.—parién: foothills of Garagará, Sambü basin, southern Darién, alt. 30-500 m., Feb., 1912, Pittier 5597 (U.S. Nat. Herb., туре; Gray Herb., 1sorype). Because of its narrow, grass-like leaves and short, few-flowered racemiform inflorescences, this species does not coincide with any other previously published with which I am familiar. Costus Евтконтснзехи О. С. Peters. Bot. Tidsskr. 18: 260. 1893. — "This name should apply to the plants assigned to C. argenteus R. & P. in the third of this series (Ann. Missouri Bot. Gard. 26: 277. 1939), as a result of my perhaps overly zealous eagerness to resusci- tate the disused species of the *Flora Peruviana.' The illustration [Vor. 29 330 ANNALS OF THE MISSOURI BOTANICAL GARDEN by Ruiz and Pavon certainly bears great similarity to our plants. Since the publication of my note, however, I have received for identi- fieation a Bolivian specimen (Krukoff 10489) which agrees even better with the illustration and which has such a strikingly dis- tinetive silvery indument that Ruiz and Pavon's epithet is the in- voluntary one to deseribe it. Most fortunately, about the same time I received on loan the type specimen of C. Friedrichsenti from Ber- lin. The latter undoubtedly is the plant from Panama so frequently confused with C. villosissimus. Costus LAEVIS В. & P. Fl. Peruv. 1:3. 1798 (C. giganteus О. Ktze. Rev. Gen. 2: 687. 1891, non Ridl.; C. maximus K. Sch. in Engl. Pflanzenreich, 456: 405. 1904; C. splendens Donn. Sm. et Tuerckh. Bot. Gaz. 32: 260. 1902; C. Skutchii Morton, Jour. Wash. Acad. Sci. 27: 306. 1937).— The identification of this name of Ruiz and Pavon with the magnificent plants of Panama and southern Central America is made through an examination of the type specimen in the herbarium at Madrid and notes kindly supplied by Dr. Loesener of Berlin. Costus RUBER Griseb. Cat. Pl. Cub. 256. 1866 (C. formosus Mor- ton, Jour. Wash. Acad. Sei. 27: 305. 1937; С. spicatus Jacq. accord- ing to many authors; C. spiralis (Jacq.) Roscoe, according to K. Sch. in Engl. Pflanzenreich, 448: 400. 1904).—ocmrimiqví: Puerto Armuelles, alt. 0-75 m., Woodson & Schery 857; San Bartolomé, Woodson € Schery 886; сос: El Valle, alt. 800-1000 m., Allen 1825; CANAL ZONE: Gold Creek, Seibert 584; Barro Colorado Isl., Woodson Schery 993 ; parié : Pinogana, alt. 20 m., Allen 938. The identification of these plants with C. ruber is made quite positive by an examination of Wright 1514, from eastern Cuba, cited by both Grisebach and Schumann. The species is one of the most attractive and frequent of southern Central America, well characterized by its ordinarily red flowers and bracts, the latter with margins densely ciliate, and rather large obovate or oblanceolate leaves. It is dif- ficult to understand how Schumann was able to harmonize C. ruber as represented by Wright 1514 with Roscoe’s excellent illustration of C. spiralis (Monandr. Pl. pl. 79. 1828), which shows with partic- ular detail the characteristic obovate-oval leaves with subcordate- aurieulate base of the latter. In addition, C. spiralis is quite gla- brous in all parts. Cosrus зслвев В. & P. Fl. Peruv. 1: 2. pl. 3. 1798.—cocr£: near Cerro Turega, 650-700 m., Woodson € Берегу 202; panamá: hills above Campana, 600-800 m., Allen 1873. Our plants agree closely 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 381 with a photograph of Ше type specimen in Ше Madrid herbarium, and with the published illustration. I believe that this species ex- tends to the Chiriquí region, and probably hybridizes with C. nutans K. Sch. throughout its range. Costus SPIRALIS (Jacq.) Roscoe, Monandr. Р]. pl. 79. 1828.—n50cas DEL TORO: Old Bank Island, Chiriquí Lagoon, Н. von Wedel 2000; Isla Colón, von Wedel 2939; Isla Bastimentos, von Wedel 2899. These specimens agree strikingly with Roscoe's illustration, par- tieularly with regard to the leaf shape, as has been discussed in a preceding paragraph. CANNACEAE CANNA FLACCIDA Salisb.—socas DEL TORO: Isla Colón, March 30, 1940, H. von Wedel 78. A new record for Central America. Previ- ously known to occur in coastal South Carolina, Georgia, and Flor- ida; also in Cuba and Hispaniola. The islands of the Chiriquí La- goon apparently eontain numerous Antillean elements. ANNA GLAUCA L.—socas DEL TORO: Old Bank Island, Chiriquí Lagoon, Feb. 5,1941, H. von W edel 2001. Apparently the first record for Central America. Previously known to occur in the Antilles, the Guianas, and northern Colombia. MARANTACEAE CALATHEA VILLOSA Lindl. Bot. Reg. 31: pl. 14. 1845 (C. hirsuta Standl. Jour. Wash. Acad. Sei. 15: 4. 1925).— This correction serves to re-emphasize by so much the affinities of the Panamanian flora for that of northeastern South America. САТАТНЕА (Pseudophrynium-Scapifoliae) Allenii Woods. sp. nov. Planta 1 m. alta. Folia longiuscule petiolata; lamina oblongo- elliptica basi rotundata apice abrupte subcaudato-acuminata 20—45 em. longa 8-15 em. lata supra nervo medio excepto glabra subtus nervo medio praecipue minute puberula, petiolo ad 25 cm. longo pilosulo parte superiore ad 5 cm. callosa tereti, vagina ad 10 cm. longa 4 em. lata dorso margineque pilosulis. Spica late fusiformis 11-13 em. longa 3.0-3.5 em. crassa sessilis vel pedunculo dense pilosulo ad 4 em. longo, bracteis (ca. 20-25) dense imbrieatis ob- longis inferne late ovalibus apice late rotundato-emarginatis fre- quenter euspe minuto ad medium munitis 5.0—5.5 em. longis 1.5-3.5 em. latis luteis marginibus apiceque praecipue pilosulis glabratisve. Flores ascendentes fasciculati bracteolis oblongo-linearibus ad 4 em. longis; ovario са. 0.4 em. longo apice villoso; sepalis anguste [Vor. 29 332 ANNALS OF THE MISSOURI BOTANICAL GARDEN oblongo-lanceolatis acutis ca. 3 em. longis glabris ; corolla lutea extus minute sparseque pilosula, tubo angustissimo ca. 3.5 em. longo, lobis staminodioque са. 1.0-1.3 em. longis.—panamMA: summit of Cerro Campana, alt. 800-1000 m., Sept. 1, 1940, P. H. Allen 2218 (Herb. Missouri Bot. Gard., TYPE). The rather narrowly fusiform, sessile or very shortly peduncu- late spikes of this species are very distinetive, as are the peculiarly emarginate braets. CALATHEA ALLOUIA ( Aubl.) Lindl. Bot. Reg. 14: sub pl. 1210. 1828. (Maranta АПоша Aubl. Hist. Pl. Guian. 1: 3. 1775; C. grandifolia Lindl. Bot. Reg. 14: pl. 1210. 1827; Phrynium cylindricum Roscoe, Monandr. Pl. pl. 40. 1828; Calathea cylindrica (Roscoe) K. Sch. Engl. Pflanzenreich, 4*5: 83. 1902; C. macrosepala К. Sch. loc. cit. 84. 1902).—After intermittent consideration for several years, I have come to the firm conviction that the showy plants of southern Central America and northern South America usually identified as C. violacea (Возс.) Lindl. and C. macrosepala K. Sch. are quite conspecifie and represent merely minor varieties with blue or pale yellow varieties respectively; nor is there much doubt in my mind that Aublet's name should be applied to them. Upon numerous occasions I have collected the two varieties growing intermixed, and Mr. Allen has confirmed my observations independently. The typical variety is that with the pale yellow flowers; that with blue flowers may be indicated as follows: CALATHEA ALLOUIA (Aubl.) Lindl. var. violacea (Roscoe) Woods., comb. nov. (Phrynium violaceum Roscoe, Monandr. Pl. pl. 37. 1828). CALATHEA (Pseudophyrnium-Scapifoliac) foliosa Rowlee, sp. nov. Planta submetralis et humilior. Folia folio 1 eaulino excepto sub pedunculo ca. 7-10 dense rosulata, lamina oblongo-lanceolata apice breviter acuminata basi obtusa ad 32 em. longa 8 em. lata nervo medio subtus minute puberulo caeterum glabra, petiolo 1 em. longo tota longitudine calloso, vagina angusta obtusa 15-20 em. longa glabra. Spica globosa ca. 6 em. diametro, pedunculo ca. 9 em. longo glabro, bracteis ca. 25 plus minusve imbricatis late ovatis breviter acuminatis 2-3 em. longis extus intusque dense tomentellis. Flores ut videntur gilvi fasciculati; ovario glabro; sepalis lanceolatis 3 em. longis; corollae tubo 1.5 em. longo lobis ellipticis 1 em. longis staminodio elliptico 0.7 em. longo.—socas DEL TORO: Farm 6, near Almirante, Sept. 23, 1920, N. W. Blair 1016 (U. S. Nat. Herb., TYPE). 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 333 This manuseript species of the late Prof. Rowlee appears to be most closely related to C. indecora, described in a previous para- graph of this report. The lower stature, acuminate bracts, and particularly the rosulate leaves render it distinctive. Rosulate cau- line leaves have been reported previously for the genus in C. Pearcet Rusby, of Bolivia, with which C. foliosa can claim little close rela- tionship. САТАТНЕА (Pseudophrynium-Scapifoliae) indecora Woods., sp. nov. Planta valida 2.0-2.5 m. alta. Folia longe petiolata, lamina ob- longo-elliptica basi rotundata apice breviter acuminata 40-65 em. longa 14-22 cm. lata supra nervo medio puberulo caeterumque glabra subtus minute puberula, petiolo 20-45 em. longo minutissime pilosulo parte superiore ad 4 ст. longa callosa tereti dense papil- lata, vagina 11-20 em. longa haud auriculata. Spica late ovoidea 5-8 em. longa, peduneulo valido 10-18 em. longo superne puberulo, braeteis (ca. 15-30) latissime ovatis subrotundatis са. 2 em. longis latisque pallide viridibus dense pilosulis ad anthesim laceratis laxe patulis. Flores ad anthesim patuli; ovario 0.2 em. longo glabro; sepalis anguste oblongis acutis 2 em. longis glabris superne min- utissime pilosulis ; corollae lacteae glabrae tubo angustissimo ca. 2 em. longo, lobis obovato-ellipticis 1.5 em. longis, staminodio 1 ст. longo.—sBocas DEL TORO: Isla Colón, Aug. 15, 1940, H. von Wedel 476 (Herb. Missouri Bot. Gard., түрк) ; Old Bank Island, Feb. 15, 1941, von Wedel 2102; Isla Colón, Oct. 18, 1940, von Wedel 1229; Water Valley, Sept. 11, 1940, von Wedel 712. Dr. Sehery and I were with Mr. von Wedel when the type speci- mens were collected in fairly low woods on Isla Colón. On the basis of dried plants alone I might otherwise have been deceived into identifying them as somewhat anomalous specimens of C. Allouia, with which it is doubtless closely related. The fine stand of plants that we saw, however, did not at all recall that species, so distinctive were they. C. indecora is a much taller, stouter plant, and the aspect of the spikes, with their roughly reflexed and spread- ing bracts and flowers, from which the specific adjective is derived, is quite distinctive amongst the species of the subgenus Pseudo- phrynium with which I am familiar. САТАТНЕА (Pseudophrynium-Scapifoliae) lagunae Woods. sp. nov. Planta submetralis. Folia brevissime petiolata, lamina late rarissime anguste oblongo-elliptica non raro ovali 11-45 em. longa 7-12 em. lata utrinque glabra, petiolo 1.5-6.0 em. longo tota longi- tudine ealloso minute papillato, vagina 6-15 em. longa obtusa haud [Vor. 29 334 ANNALS OF THE MISSOURI BOTANICAL GARDEN aurieulata glabra. Spica ovoidea 3-6 em. longa, peduneulo 10-17 em. longo superne excepto glabro, bracteis са. 10-15 imbricatis subreniformi-ovatis latissime obtusis vel rotundatis 1-2 em. longis luteis praecipue basi dense pilosis. Flores ad anthesim ascendentes ; ovario 0.2 em. longo glabro; sepalis oblongo-ovalibus obtusis 2 em. longis glabris ; corollae albae glabrae tubo angusto 2 em. longo, lobis late elliptieis 1.5 em. longis, staminibus staminodioque ea. 1 em. longis.—Bocas DEL TORO: Western River, Sept. 19, 1941, von Wedel 2706 (Herb. Missouri Bot. Gard., түрк); Isla Colón, Oct. 23, 1940, von Wedel 1328. Most closely related to C. picta Hook. f. (Bot. Mag. pl. 7674. 1899) of previously published species. This Brazilian species, published from a plant cultivated at Kew, however, is stated to be glabrous in all parts, and other discrepancies might be added. CALATHEA MICROCEPHALA (Poepp. & Endl.) Koernicke, Bull. Soc. Nat. Moscow 35': 125. 1862 (Phrynium microcephalum Poepp. & Endl. Nov. Gen. & Sp. 3: 20. pl. 128, figs. a-b. 1838; Maranta micans Mathieu, Cat. 1853; C. micans (Mathieu) Koernicke, loc. cit. 126. 1862; C. albicans Brongn. ex К. Sch. in Engl. Pflanzenreich 4*5 : 112. 1902).—I have taken the opportunity to study this species upon several collecting trips to Panama, and have found it to be quite variable in all the key characters used by Petersen and Schumann to separate C. microcephala, C. micans, and C. albicans, notably height of plant, shape and size of leaves, and color of staminodia. Biologi- eally speaking, I feel quite confident that a single species is represented. CALATHEA PICTA Hook. f. Bot. Mag. pl. 7674. 1899.—cocr£: north of El Valle, alt. 1000 m., Allen 2331; Panamá : summit of Cerro Cam- pana, alt. 800-1000 m., Allen 2219. Mr. C. V. Morton and I ean find no characters to separate Mr. Allen's specimens from the illustra- tion of C. picta. 'This is somewhat embarrassing, since the species previously has been known only from a plant, supposedly from Brazil, which was cultivated at Kew. STROMANTHE LUTEA (Jaeq.) Hichl. Abhandl. Akad. Berlin 1882: 81. 1883. (Maranta lutea Jacq. Collect. 4: 117; Icon. pl. 201. 1794; Myrosma Guapilesense Donn. Sm. Bot. Gaz. 23 : 251. 1897).—I have been unable to distinguish Capt. Smith's species, ranging from Guatemala to Panama, from that of Jacquin, which occurs in Colombia, Venezuela, and northern Brazil. The former was over- looked by Schumann in his account of Marantaceae for the * Pflanzen- 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI ӨӨ) reich.’ The problem of whether our plant is a Myrosma or a Stro- manthe as applied by Schumann resolves largely into a question of whether the leaves are ‘‘ап горле’? or ‘‘homotropiec.’’ This distinc- tion of Eichler appears to me as extremely deceptive, if not arti- ficial, as is shown by Schumann’s rearrangement of Eichler's species amongst the same genera, upon the same criterion. I am inclined to disregard this dubious character in favor of others more easy to apply. From the same standpoint, I feel that Stromanthe Tonckat (Aubl.) Eichl. is very much better left in Maranta where it was placed by Aublet since its entire aspect and structure, exclusive of ‘‘homotropic’’ or ‘‘antitropic’’ orientation of the leaves, is indeli- bly suggestive of M. arundinacea L. M. Tonckat has been reported in Costa Rica, and is to be expected in Panama as well. Myrosma dasycarpa (Donn. Sm.) Woods., comb., nov. (Calathea dasycarpa Donn. Sm. Bot. Gaz. 31: 123. 1901; Ctenanthe dasycarpa (Donn. Sm.) К. Sch. in Engl. Pflanzenreich, 4*5: 153. 1902).—I сап scarcely call myself an authority on Marantaceae, and feel a becom- ing sense of modesty in contradicting the generic concepts of an authority of Dr. Schumann’s ealibre. Nevertheless, I feel very strongly that several of his generic conceptions amongst the Maran- teae in the ‘Pflanzenreich’ are extremely impractical. Foremost of these, as I have remarked in a previous paragraph, is the distinction between ‘‘homotropic’’ and ‘‘antitropic’’ leaves, of which Schu- mann even was somewhat wary (Pflanzenreich, loc. cit.). I do not have a large collection of living Maranteae available for observa- tion; but whatever the situation in life, the leaf character is entirely inapplicable in the herbarium, and I am unwilling to perpetuate it in the ‘Flora of Panama.’ Therefore I am considering Ctenanthe Eichl. as a synonym of Myrosma L. f. I am not aware of any valid morphological characters to distinguish them. Іѕснхоѕірнох Pittieri (Rowlee) Woods., comb. nov. (Pleiostachya Pittieri Rowlee, ex Standl. Jour. Wash. Aead. Sei. 15: 5. 1925).— Since I have made a start toward reforming the genera of Maran- taceae as represented in the microcosm of Panama, the job might as well be made consistent. Pleiostachya has no observable distine- tion from Ischnosiphon save the compression of the bracts, as far as I am aware; other morphological characters appear to be fairly coherent. Fortunately, the other species currently treated as Pleiostachyas are both provided with combinations in Ischnosiphon: I. pruimosus (Reg.) Peters., and I. Morlaei Eggers. [Vor. 29 336 ANNALS OF THE MISSOURI BOTANICAL GARDEN BURMANNIACEAE APTERIA APHYLLA (Nutt.) Barnhart—socas DEL токо: Old Bank Island, Feb. 17, 1941, H. von Wedel 2111. Previously known to occur from the southeastern United States to Bolivia. ORCHIDACEAE (Louis O. Williams) HABENARIA MONORRHIZA (Sw.) Reichb. f.—coc.é: moist roadside banks, dry hills south of El Valle de Antón, alt. 600—800 m., flowers white, Nov. 13, 1941, Allen 2771. A not uncommon species found from Guatemala through Central America, south to Peru and in the West Indies. Apparently not reported for Panama. SoBnALIA Allenii L. О. Williams, sp. nov. (pl. 30, figs. 1-3). Plan- tae eaespitosae, epiphyticae, usque ad 5 dm. altae. Folia elliptico- lanceolata vel anguste elliptica, acuminata, plicata, 7-nervia. Inflo- rescentia terminalis, uniflora. Sepalum dorsale oblanceolatum, apieulatum. Sepala lateralia lineari-oblonga, apieulata. Petala oblanceolata, acuta, serrulata. Labellum oblongo-ovale, truneatum vel leviter retusum, lacerato-dentatum ; discus carinis et callo bipar- tito ornatus, pubescens. Columna generis. Caespitose epiphytie plants up to about 5 dm. tall. Stems about 1-2 mm. in diameter, slender, bearing 1—3 leaves toward the apex, leafless below or the leaves reduced to sheaths. Leaves 13-18.5 em. long, 1.8-2.5 em. broad, elliptic-lanceolate to narrowly elliptic, acu- minate, plicate, with 7 principal nerves, lepidote on the lower sur- face especially along the nerves at the base, glabrous above or essentially so; leaf-sheaths closely appressed to the stem, lepidote. Inflorescence terminal, 1-flowered, flowers small, white with a pale yellow lip. Dorsal sepal about 3.5 em. long, 7 mm. broad, oblanceo- late, apiculate, 7-nerved. Lateral sepals about 3.5 em. long, 7 mm. broad, linear-oblong, apiculate, 7-nerved. Petals about 3.5 em. long, 6-7 mm. broad, oblanceolate, acute, terminal half serrulate, 7-nerved. Lip about 3.5 em. long, 1.8 em. broad, oblong-oval, trun- cate or shallowly retuse, terminal half lacerate-dentate; dise with several inconspicuous longitudinal carinae and with a small bipar- tite callus thickening at the base, pubescent longitudinally along the middle, especially toward the apex. Column of the genus, about 1.5 em. long.—cocr£: epiphytic, trail to La Mesa, hills north of El Valle de Antón, alt. ca. 1000 m., (sepals and petals white, labellum pale yellow), Aug. 31, 1941, Allen 2686 (Herb. Ames, TYPE). 1942] WOODSON & SCHERY—-FLORA OF PANAMA. VI 381 Sobralia Allenii is most closely allied to S. mucronata A. & S. from which, however, it is distinguished by the pubescent lip which is lacerate-dentate in the terminal half. The leaf-sheaths and leaves of S. Allenii are lepidote, while those of S. mucronata are glabrous or essentially so. Description and illustration from a dried specimen and a flower in liquid. SoBRALIA DECORA Batem. var. aerata Allen & Williams, var. nov. (pl.31). E specie planta parviore et floribus aeratis differt.—cocr£: bad lands south of El Valle de Antón, alt. са. 500 m., Sept. 18, 1941, (flowered in collection of A. M. Bouché, Pedro Miguel, C. Z.), Allen 2755 (Herb. Ames, түрк) ; ravines in bad lands south of El Valle de Antón, alt. ca. 600 m., (flowered in collection of Mr. and Mrs. Bar- rett, Bas Obispo, C. Z.), Allen 2846. In the field this plant appears to be amply distinct from Sobralia decora Batem., but herbarium study indicates that it is perhaps only a variety. The plants average about 5 dm. tall while S. decora is commonly 2-2.5 m. tall. The flower color is quite distinctive: sepals greenish on the outer surfaces, very near Brick Red within; petals a washed Brick-Red with lighter margins and median line dorsally ; lip Briek Red dorsally with a white median line and white margins, —the inner surface is very nearly Acajou Red with upper margins white and with an Amber Brown median stripe; the column is pure white. [Ridgeway colors]. The local name for this variety is ‘¢Bronze Sobralia,’’ hence the varietal name, meaning ‘‘ornamented with bronze. SprraNTHES navarrensis (Ames) L. О. Williams, comb. nov. (Stenorrhynchus navarrensis Ames, Sched. Orch. 9: 13, t. 3. 1925).— CHIRIQUÍ: cloud forest, Cerro Horqueta, alt. ca. 2000 m., (flowers yellow), June 2, 1940, von Hagen & von Hagen 2111. New to the flora of Panama, previously known from Costa Rica. Closely allied to Spiranthes speciosa (G. F. Gmel.) A. Rich. Sprrantues Woodsonii L. О. Williams, sp. nov. Herbae palustres, terrestres, usque ad 5.5 dm. altae. Caules graciles, basi foliosi. Folia oblongo-elliptiea vel ovalia, acuta vel obtusa. Sepalum dorsale lanceolatum, aeuminatum. Sepala lateralia in mentum saccatum producta, lanceolata, acuminata, arcuata. Petala elliptica vel angus- te oblanceolata, acuta acuminatave. Labellum lineari-oblongum, panduratum et caudatum, apice leviter expansum. Columna sec- tionis ($ Sarcoglottis). Terrestrial, palustrine herbs from underground rhizomes, up to [Vor. 29 338 ANNALS OF THE MISSOURI BOTANICAL GARDEN 9.0 dm. tall. Rhizome slender, rooting at most of the nodes, with searious sheaths arising from the nodes. Stem slender, with well- developed leaves at the base which become bract-like above. Leaves 3-10 em. long, 1.3-2.8 em. broad, oblong-elliptic to narrowly oval, aeute or obtuse, largest near the base of the stem and reduced to amplexicaul bracts above. Inflorescence up to 10 em. long, congested in flower, becoming more open in fruit; braets up to 4 em. long, 1.6 em. broad, lanceolate, acuminate, pubescent dorsally. Flowers large, similar to those of S. acaulis. Dorsal sepal 16.5-19 mm. long, 3.5-4 mm. broad, lanceolate, acuminate, 3—5-nerved, densely pubes- cent dorsally. Lateral sepals long-decurrent on the ovary, 35-40 mm. long from the apex to the base of the saccate mentum; free part 16.5-19 mm. long, 4—4.5 mm. broad, lanceolate, acuminate, arcuate, spreading, densely pubescent dorsally, 3-5-nerved. Petals 15-18 mm. long, 2-2.5 mm. broad, narrowly elliptic to narrowly oblanceo- late, acute or acuminate, arcuate, adherent to the dorsal sepal, pubescent on the margins, the basal half ciliate. Lip 28-32 mm. long, 6-7 mm. broad (apical lobe), linear-oblong, caudate, some- what expanded and pandurate, with two extremely pubescent, con- verging callus-ridges on the terminal third, the basal part densely pubescent and the remainder, except the glabrous apex, less pubes- cent, glabrous below except at the base; terminal lobe transversely oval or transversely rhombic; caudae about 5 mm. long, retrorse, fleshy but flattened. Column (free part) about 8 mm. long, pubes- cent at the base ventrally ; rostellum oblong-lanceolate, obtuse, flat- tened. Ovary densely pubescent.—cuirigui: vicinity of Boquete, alt. 1200-1500 m., (flowers pale green), July 24-26, 1940, Woodson Ф Schery 753 (Herb. Ames, түре); in swampy meadows, Finca Lérida to Boquete, alt. ca. 1300-1700 m., (flowers pale yellowish green), July 8-10, 1938, Woodson, Allen & Seibert 1148. Spiranthes Woodsonii is not very closely allied to any Central American species of the genus. It belongs in the section Sarcoglot- tis. The species is particularly distinctive in having a creeping underground rhizome or stem instead of the usual fascicle of fleshy roots. In flower structure Spiranthes Woodsonu approaches 8. acaulis (J. E. Sm.) Cogn. (5. picta (Anders.) Lindl.) but in detail is amply distinct. бтетлв Allenii L. О. Williams, sp. nov. (pl. 32, figs. 1-3). Herbae eaespitosae, epiphytieae, usque ad 4 dm. altae. Folia elliptica vel elliptico-ovalia, acuta vel obtusa. Inflorescentia elongata; bracteae ovato-lanceolatae, acutae vel acuminatae, infundibuliformes. Sep- 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 339 alum dorsale lanceolatum, acutum, cucullatum. Sepala lateralia connata, suborbiculari-ovata, acuta vel obtusa, cucullata et gibbosa. Petala late flabellata vel transverse ovalia. Labellum flabellatum, truneatum, apice callo transverso ornatum. Large caespitose epiphytic herbs up to 4 dm. tall. Secondary stems 7-15 em. long, 0.25-0.35 em. in diameter, covered with 2 or 3 loose sheaths which soon disintegrate, shorter than the leaves. Leaves 10-19 em. long, 3.5-7 em. broad, elliptic to elliptic-oval, acute or obtuse, coriaceous, attenuated into a short petiole at the base. Inflorescence up to 30 em. long, floriferous to the base, 1 or more borne from the apex of the stems (if more than 1 then presumably borne in different years); sheaths up to 2.5 ст. long, cucullate, ample; bracts 2-18 mm. long, reduced upward, ovate-lanceolate, acute or acuminate, infundibuliform. Flowers largest of the genus. Dorsal sepal 14-16 mm. long, 5-6 mm. broad, lanceolate, acute, 11-13-nerved, cucullate. Lateral sepals connate to their apices, together 10-12 mm. long and 8-10 mm. broad, suborbicular-ovate, acute or obtuse, cucullate and gibbous at the base, many-nerved. Petals about 1 mm. long, 1.5 mm. broad, broadly flabellate to trans- versely oval, the apex much thickened. Lip 0.75-1 mm. long, 1-1.4 mm. broad, about 0.75 mm. thick at the apex, flabellate, truncate, with a transverse callus at the apex of the lip,—very like the petals but slightly smaller.—cocré: hills north of El Valle de Antón, alt. 800 m., (flowers nearly black), April 10, 1942, Allen 2952 (Herb. Ames, TYPE). Stelis Allenii is perhaps the most distinctive species of this diff- cult genus in Central America and seems to be the largest-flowered species of the genus. There are no near allies in Central America but the species seems to belong to Lindley's section Dialissa, a sec- tion with but a few species in the Andes. Зтетлз atrorubens L. О. Williams, sp. nov. (pl. 32, figs. 48). Herbae epiphyticae, caespitosae, parvae, usque ad ca. 18 em. altae. Folia petiolata; lamina elliptica vel elliptico-oblanceolata, obtusa. Inflorescentia densiflora, quam folia longiora. Sepala basi eonnata, rotata, triangularia, acuta. Petala suborbicularia, carnosa. Label- lum obscure trilobatum, oblongum vel oblongo-ovale, carnosum. Small caespitose epiphytic herbs up to about 18 cm. tall. Secon- dary stems 1.5-3.5 em. long, slender, covered with sheaths. Leaves 6-9 em. long, petiolate, much longer than the secondary stems; lamina about 4-6.5 ст. long, 0.5-1 em. broad, elliptic to elliptie- oblanceolate, obtuse, fleshy; petiole 2-3 em. long. Inflorescence up [Vor. 29 340 ANNALS OF THE MISSOURI BOTANICAL GARDEN to 15 em. long, upper half more or less densely flowered; bracts about 2 mm. long, infundibuliform, acute, scarious. Sepals connate at the base, rotate, 2-2.5 mm. long, 1.5-2 mm. broad, triangular, acute, 3-nerved. Petals about 0.75 mm. long, 0.75 mm. broad, subor- bicular, 1-nerved, the terminal part thickened, fleshy. Lip 1-1.5 mm. long, 0.5-0.6 mm. broad, obscurely 3-lobed, oblong or oblong-oval, basal part of the lip oblong, somewhat concave, fleshy, terminating into two small, subereet lateral lobes,—terminal lobe of the lip about 0.5 mm. long, suborbicular, strongly concave, fleshy.—cocr£: vicin- ity of El Valle, alt. 600-1000 m., Dec. 8, 1938, Allen 1234; hills north of El Valle de Antón, trail to Las Minas, (flowers maroon), Dec. 2, 1941, Allen 2876 (Herb. Ames, TYPE). Stelis atrorubens is allied to S. montana L. Wms., but differs in several details. The flower parts of S. montana, a species not pre- viously illustrated, are figured in pl. 32, figs. 9-12, for comparison. STELIS MONTANA L. Wms. in Ann. Mo. Bot. Gard. 27: 272. 1942. — Figures of this distinctive and unusual species are given in pl. 32, figs. 9-12. CRYPTOPHORANTHUS lepidotus L. О. Williams, sp. nov. (pl. 30, figs. 4-7). Herbae caespitosae, epiphyticae, usque ad ca. 17 em. altae. Folia oblanceolata vel anguste obovata, obtusa vel acuta, coriacea, petiolata. Sepala basi et apice connata; sepalum dorsale oblongo- oblanceolatum, cucullatum, carnosum; sepala lateralia usque ad apicem connata. Petala late ovato-lanceolata, acuta vel acuminata. Labellum hastatum, unguiculatum; lamina verrucosa vel lepidota. Columna generis. Caespitose epiphytic herbs up to about 17 cm. tall. Secondary stems 2-7 em. long, 1-2 mm. in diameter, covered with 4—5 perga- meneous, infundibuliform sheaths which soon disintegrate. Leaves oblanceolate to narrowly obovate, obtuse or acutish, coriaceous, contracted into a distinct petiole; lamina 3.5-10 em. long; petiole 1-2 em. long, conduplicate. Inflorescence a fascicle of 1-6 long pedunculate flowers at the apex of the secondary stem; the peduncle with 1-3 short infundibuliform sheaths. Sepals joined at the base and at the tip leaving a small opening between the dorsal and lat- eral sepals; dorsal sepal about 15-20 mm. long, 6-7 mm. broad, oblong-oblanceolate, strongly cucullate, fleshy, 7-nerved, ridged dorsally, the ridges verrucose ; lateral sepals connate to their apices, about 12-18 mm. long and together 6-8 mm. broad, fleshy, each about T-nerved, with verrucose ridges dorsally. Petals 5-6 mm. long, 2.5-3.5 mm. broad, broadly ovate-lanceolate, acute or acuminate, 1942] WOODSON & SCHERY—FLORA OF PANAMA, VI 341 3(—5)-nerved. Lip 5-6 mm. long, the lamina about 4 mm. long, 1.5-2 mm. broad, hastate, unguiculate, 3-nerved, verrucose or lepidote, with two longitudinal, lamellate calluses extending from the auri- cles to about the middle; auricles about 1 mm. long, retrorse, subu- late; claw 1.5-2 mm. long, 1-1.5 mm. broad, verrucose-scurfy or lepidote toward its apex. Column of the genus.—cocr£: epiphytic, trail to Las Minas, hills north of El Valle de Anton, alt. ca. 1000 m., (lower sides of many of the leaves deep purple, flowers white with very heavy purple stripes), Sept. 1, 1941, Allen 2718 (Herb. Ames, TYPE). Cryptophoranthus lepidotus is apparently somewhat allied to C. beloglottis Schltr. from Ecuador but it is easily distinguished by the smaller, differently shaped leaves as well as by the details of the flowers. Cryptophoranthus Endresianus Krünzl. is not well described but the available record indicates that it differs from C. lepidotus. Kranzlin’s monograph of Cryptophoranthus (т Fedde Repert. Beih. 34: 220-232. 1925) omits several species of this genus published prior to the date of the monograph. LgPANTHES LINDLEYANA Oerst. & Reichb. f.—'This name has been taken up for a not-uncommon species of Costa Rica and Panama. Reichenbach illustrated the species but his illustration does not coincide too well with the plants referred here. The plant usually identified as L. Lindleyana has a name which seems to refer to it exactly, Lepanthes chiriquensis Schltr. In the original description of L. Lindleyana Ше petals are described as ‘‘Tepala dimidiata tri- angula angulo inferiori obtusata, ciliolata,’’ and in the illustration the petals might be considered to fit this characterization rather loosely. The petals, if bipartite, have a large anterior lobe and no posterior lobe; the dorsal вера! is as broad as the combined laterals; the lip shows no apiculation or mid-lobe. PLEUROTHALLIs antonensis L. О. Williams, sp. nov. Herbae caes- pitosae, epiphyticae, usque ad 3 dm. altae. Folia lanceolato-cor- data vel late cordata, acuta vel acuminata, coriacea. Inflorescentia fasciculata, uni- vel pauciflora. Sepalum dorsale elliptico-obovatum, obtusum vel acutum. Sepala lateralia connata, ovata vel late ovata, obtusa vel acuta. Petala lineari-oblonga, acuta, serrulata, arcuata. Labellum unguiculatum; lamina cordata vel oblongo-cordata, den- tieulata; unguis brevis. Columna generis. Caespitose epiphytic herbs up to about 3 dm. tall. Secondary stems slender, with one or two scarious sheaths at the base, naked above. Leaves 4—9 em. long, 1.5—4.7 em. broad, lanceolate-cordate to [Vor. 29 342 ANNALS OF THE MISSOURI BOTANICAL GARDEN broadly cordate (young leaves elliptic), acute or acuminate, coria- ceous. Inflorescence a 1-several-flowered fascicle from the apex of the secondary stems, much shorter than the subtending leaves. Dor- sal вера! 6-7.5 mm. long, 3-4 mm. broad, elliptic-obovate, obtuse or acute, 3-nerved. Lateral sepals connate to their apices, 5.5-7 mm. long, 4-5 mm. broad, ovate or broadly ovate, obtuse or acute, 6-nerved. Petals 3.8-4.5 mm. long, 0.6-1 mm. broad, linear-oblong, acute, serrulate, strongly arcuate, 1-nerved. Lip unguiculate; lam- ina 2.5-3 mm. long, 2-2.5 mm. broad, cordate to oblong-cordate, denticulate, fleshy, the surface obscurely verrucose or smooth, with a small central cavity near the base; claw short. Column about 1 mm. long.—cocrÉ: hills north of El Valle de Antón, alt. са. 1000 m., (flowers light brown), July 23, 1940, Allen 2156; same locality, July 14, 1940, Allen 2194; same locality, Nov. 21, 1940, Allen 2267; same locality, Sept. 1, 1941, Allen 2701 (Herb. Ames, rvPE) ; mountains beyond La Pintada, alt. 400-600 m., Feb. 17, 1935, Hunter $ Allen 594. Pleurothallis antonensis has been referred to P. phyllocardia Reichb. f., a rather obseure and poorly described species. We have an analysis of P. phyllocardia which shows the lip to be about one- fourth the length of the lateral sepals, while in P. antonensis the lip is half as long as the sepals. The lateral sepals of P. phyllocardia are about as broad as the dorsal sepal and the petals are only slightly arcuate, while in P. antonensis the lateral sepals are broader than the dorsal sepal and the petals are strongly arcuate. The specific name recalls Е Valle de Antón which, apparently, has a very large number of endemic species of orchids. PLEUROTHALLIS ARISTATA Hook. in Ann. & Mag. Nat. Hist. 2: 229, t. 15. 1889. (P. dichotoma Ames, Sched. Orch. 6: 58. 1923, non Schltr. ; P. divexa Ames. Sched. Orch. 7: 20, t. 5. 1924). Costa Rica, Panama, the West Indies and British Guiana. Pleurothallis Urbaniana Reichb. f. has been considered to be the same as the two synonyms cited (Schweinfurth in Bot. Mus. Leafl. Harv. Univ. 6: 36. 1938) but no authentic material or record is available and the description shows some discrepancies. Known only from Chiriqui in Panama (Davidson 187). Рувовотнатллз BricHamit S. Wats. in Proc. Am. Acad. 23: 285. 1888; Ames, Sched. Orch. 2: 18. 1923; 7: 19, t. 7. 1924. ( P. periodica Ames, Sched. Orch. 7: 21, fig. 4. 1924; P. acrisepala A. & 5. Sched. Orch. 8: 22. 1925; P. barboselloides Schltr. in Fedde Repert. Beih. 17: 18. 1922; 59: t. 29, fig. 113. 1931). 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 343 In attempting to separate and describe the various species cited above for the account of the Panamanian flora it was found that, as a whole, they formed a most complete series. Pleurothallis pyrosodes Reichb. f. is of this alliance. It is known to me only by the description and a rather rough analysis of the type but it may prove to be the same and if so it will include all of the above names as synonyms. Pleurothallis Brigham, as now con- stituted, is a species of low elevations, perhaps not exceeding 400 m., which extends from Guatemala to Panama. PLEevRoTHALLIS cardiochila L. О. Williams, sp. nov. (pl. 33, figs. 8-10). Herbae caespitosae, epiphyticae, usque ad ca. 21 сп. altae. Folia lanceolato-cordata, acuminata, coriacea. Inflorescentia fasci- culata, uni (-pauci)-flora. Sepalum dorsale oblongo-ovale, obtusum, cucullatum. Sepala lateralia connata, ovata, acuta. Petala lineari- oblonga, acuta, basi subauriculata. Labellum unguiculatum ; lamina suborbiculari-cordata, obtusa; unguis oblongus. Columna generis. Small eaespitose epiphytic herbs up to about 21 cm. tall. Sec- ondary stems up to about 11 em. long, slender, with 1-2 loose, char- taceous sheaths toward the base, naked above. Leaves 9-10 ст. long, 3.6-3.9 em. broad, lanceolate-cordate, acuminate, coriaceous. Inflorescence a 1(-few ?)-flowered fascicle subtended by a charta- ceous sheath about 1-1.5 em. long, much shorter than the leaves. Flowers very large for the group, yellowish with the dorsal sepal tinged with dark red, lip deep orange. Dorsal sepal about 20 mm. long, 12.5 mm. broad, oblong-oval, obtuse, with 7-9 principal nerves, strongly cucullate. Lateral sepals connate to their apices, about 18 mm. long, 10 mm. broad, ovate, acute, with 7-9 principal nerves. Petals about 13 mm. long, 2.5 mm. broad, linear-oblong, acute, arcu- ate, subauriculate on the posterior margin at the base, entire, 3-nerved at the base, the posterior nerve short, anterior nerve branched near the base,—hence the apical part of the petal 3-nerved. Lip unguiculate ; lamina about 6 mm. long, 5.5 mm. broad, suborbic- ular-cordate, obtuse, callus-thickened along the sinus and below at the subapiculate apex, 3-nerved; claw about 2 mm. long, oblong. Column about З mm. long; clinandrium lacerate-digitate.—CHIRIQUÎ : rain forest, Bajo Chorro, Boquete District, alt. ca. 1800 m., Jan. 13, 1938, Davidson 119 (Herb. Ames, TYPE). Pleurothallis cardiochila is most nearly allied to P. palliolata Ames, from which it is easily distinguished by the suborbicular- cordate lip and by the 3-nerved, entire petals. [Vor. 29 344 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ръвсвотналтллв ellipsophylla L. О. Williams, sp. nov. (pl. 33, figs. 1-7). Herbae parvae, repentes vel caespitosae, usque ad 2 dm. altae. Folia late elliptica vel elliptico-oblonga, acuta vel obtusa. Inflores- centia uni- vel pluriracemosa, racemis quam folia brevioribus. Se- palum dorsale anguste lanceolatum vel elliptico-oblanceolatum, acutum vel acuminatum. Sepala lateralia in laminam connata; lamina lanceolata, acuta vel subaristata. Petala elliptica vel ellip- tico-lanceolata, acuta vel acuminata, prope medium denticulata vel denticulato-lacerata. Labellum lanceolatum, acutum vel subarista- tum, basi biauriculatum; unguis perbrevis; lamina prope basem callo V-formi ornata. Small repent or caespitose epiphytic herbs up to about 2 dm. tall. Secondary stems up to 9.5 em. long, prominently angled when dry, with one or more loose sheaths covering the lower part. Leaves 4—9.5 em. long, 0.7—2.7 em. broad, elliptic to elliptie-oblong, acute or obtuse, epetiolate, coriaceous. Inflorescence 1 or several short, few- flowered racemes from the axil of the leaf, approximately half as long as the leaf. Dorsal вера! 9-14 mm. long, 1.5-2.5 mm. broad, nar- rowly lanceolate to elliptic-oblanceolate, acute or acuminate, eucul- late, 3-nerved, puberulent dorsally. Lateral sepals 10-13 mm. long and together 3-4 mm. broad, connate except at the tip, the lamina lanceolate, acute or subaristate, puberulent dorsally, gibbous and with a mentum at the base, 6-nerved. Petals 5.5-6.5 mm. long, 1.4-2 mm. broad, elliptie or elliptic-lanceolate, acute or acuminate, areu- ate, denticulate or denticulate-lacerate toward the middle but the base and apex usually entire, 1-nerved. Lip 4-5 mm. long and 1-1.3 mm. broad, lanceolate, acute or subaristate, somewhat fleshy, prom- inently biauriculate at the base, claw very short, lamina with an inconspicuous V-shaped callus on the basal third. Column about 3 mm. long, with a narrow wing and two erect teeth at the apex; eolumn-foot very short.—socas DEL TORO: epiphyte, southwest of Bocas at Maccaw Hill, Isla Colón, alt. 0-125 m., (flowers greenish- brown), Aug. 25, 1940, H. von Wedel 560 (Herb. Ames, түрк); Rio Crieamola, between St. Louis and Konkintóe, alt. ca. 10-15 m., (**labellum and hood greenish-yellow, striped with brown’’), Aug. 12-16, 1958, Woodson, Allen € Seibert 1884. Pleurothallis ellipsophylla is allied to P. vittata Lindl. and to P. geminicaulina Ames. From the former it is distinguished by larger flowers with comparatively narrow lip which lack lateral auricles or lobes; from P. geminicaulina it is distinguished by the compara- 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 345 tively longer and narrower lip with more prominent basal auricles, but lacks the parallel carinae of the lamina. PLEUROTHALLIS OCTOMERIAE Schltr. in Fedde Repert. Beih. 17: 21. 1922. (P. cerea Ames, Sched. Orch. 4: 19. 1923; 7: 26, fig. 4 and t. 8 in part. 1924).—cocr£: epiphytic, hills north of El Valle de Antón, trail to Las Minas, alt. ea. 1000 m., (sepals and petals pure white with narrow bright red margin, anther-cap red), Sept. 1, 1941, Allen 2719. Pleurothallis octomeriae and P. cerea have flowers that are iden- tical but the leaves on the types of the two species differ markedly in size, those of P. octomeriae varying from 4 to 14 em. long and from 1 to 2.5 em. broad while the single imperfect leaf known for Р. cereais “17 em. or more long, 6.3 em. wide.’’ In both the shape of the leaf is essentially elliptie. In the collection cited above ( Allen 2719), of which there are four excellent specimens, the extremes of leaf sizes are 8 and 17 em. long and 1.7 and 6 em. broad. This indi- cates that the leaf size in the two proposed species is not specific and that they should be treated as one. EPIDENDRUM pHysopes Reichb. f.—socas DEL Toro: terrestrial, vicinity of Chiriquí, Oct. 9, 1940, H. von Wedel 1126. New to Panama, previously known from Mexico, Guatemala and Costa Riea. CATTLEYA SkINNERI Batem. var. autumnalis P. Н. Allen, var. nov. Herbae epiphyticae vel saxicolae. Labellum concolor vel sub- concolor. E specie planta robustior et floribus in autumno differt. Differs from the species in that it is a somewhat more robust plant which flowers in the autumn. The lip usually lacks the white spot.—PawAMÁ: vicinity of Bejuea, alt. ca. 30 m., Aug. 15, 1941, Allen 2668 (Herb. Ames, түрк); east of [Panama] City, fall 1915, Powell 16; CANAL ZONE: along Caño Quebrada, Oct. 30, 1914, Pittier 6828. Endemie to Panama where the species, a plant of higher eleva- tions, does not oceur (unless possibly in the mountains of Chiriquí). Platyglottis L. Wms. gen. nov. Platyglottis L. O. Williams gen. nov. (Orchidaceae-Monandrae- Acrotonae-Kerosphaereae-Acranthae-Ponereae). Herbae epiphy- ticae, caespitosae, e rhizomate perbrevi. Caules graciles, non pseu- dobulbosi , indurati, simplices. Folia disticha, coriacea. Inflore- scentia terminalis (vel subterminalis ?), racemosa. Sepala petalis [Vor. 29 346 ANNALS OF THE MISSOURI BOTANICAL GARDEN subaequalia vel petala latiora. Labellum apice columnae pedi arti- culatum ; lamina late ligulata, integra. Columna brevis, clavellata, exalata, basi in pedem brevem producta ; anthera terminalis, oper- eulata, incumbens, sex-loculata; pollinia 6( ER aequalia cerea. Epiphytie eaespitose herbs from a very short rhizome. Stems slender, non-pseudobulbous, indurated, simple. Leaves alternate, distichous, coriaceous, plane, deciduous, leaf-sheaths tightly enfold- ing the stem. Inflorescence terminal (or subterminal?), racemose. Sepals subequal; dorsal sepal free; lateral sepals adnate to the short column-foot at their bases and with it forming an inconspicu- ous mentum. Petals similar to the sepals or broader. Lip articu- lated to the apex of the column-foot; lamina broadly ligulate, entire. Column short, clavellate, wingless, produced into a short foot at the base ; anther terminal, operculate, incumbent, 6-loculate ; pollinia 6, equal,—four basal in one laterally compressed series,—two termi- nal in a second laterally compressed series, ceraceous. A single species known only from the vicinity of El Valle de Antón, Coclé Province, Panama. The Ponereae is a small tribe of orchids the genera of which are limited to the American tropics or subtropies. The twelve genera of the tribe which seem to be recognizable are, with one exception, small with about six or fewer species each. Vegetatively these genera are divisible into two approximately equal series: (1) those with elongated stems and distichous leaves scattered along the stem ; and (2) those with swollen or pseudobulbous stems upon which the leaves are terminal. These two groups may be further divided by various means. The Brazilian genus Orleanesia Rodr., which I know only from a plate and descriptions, seems to approach Platyglottis in vegeta- tive charaeters more than do the other genera of the tribe. Thus Orleanesia, which has two small and two large pollinia and appar- ently other technieal characters, is easily separated from Platy- glottis. The other genera which approach Platyglottis in vegetative char- acters, Jacquiniella Schltr., Ponera Lindl., Neourbania Fawe. & Rendle, Isochilus В. Br. and Octadesmia Benth., may be quickly distinguished by the technical characters of the pollinia,—all having four pollinia except Octadesmia which has eight. Platyglottis is the only genus of this alliance which has six pollinia. However, six pollinia are not uncommon in the genus Scaphyglottis Poepp. & 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 347 Endl. (sens. lat.) which belongs in the series of the tribe having thickened stems or pseudobulbs. PuatyGLottis coriacea L. О. Williams, sp. nov. (pl. 34). Herbae epiphyticae, caespitosae, usque ad ca. 4.5 dm. altae. Folia anguste ligulata, obtusa, coriacea. Inflorescentia racema brevis, pauciflora. Sepalum dorsale lanceolatum, acutum vel acuminatum. Sepala late- ralia lanceolata, acuta vel aeuminata, leviter obliqua. Petala ellip- tico-oblanceolata, leviter obliqua. Labellum late ligulatum, subpan- duratum, integrum, apice rotundatum. Columna generis. Epiphytie eaespitose herbs up to about 4.5 dm. tall. Stems 5-8 mm. in diameter, slender, leafy, becoming naked with age. Leaves 2.5—6 em. long, 1-1.5 em. broad, narrowly ligulate, obtuse, obscurely and unequally bilobed at the apex, coriaceous, distichous, deciduous. Inflorescence a short, few-flowered raceme, terminal (or subtermi- nal?); bracts 7-20 mm. long, elliptic-lanceolate, acute, cucullate. Flowers large for the Ponereae, sepals and petals lavender, the lip lavender with greenish margins. Dorsal sepal about 10 mm. long, 3.5 mm. broad, lanceolate, acute or acuminate, 5(—7)-nerved. Lat- eral sepals about 10 mm. long, 3 mm. broad, lanceolate, acute or acuminate, slightly oblique, 5-nerved. Petals about 10 mm. long, 3.0 mm. broad, elliptic or elliptic-oblanceolate, somewhat oblique, obscurely constricted near the apex, 5(- 7)-nerved. Lip about 10 mm. long and 5 mm. broad near the apex, broadly ligulate, subpan- durate, entire, apex rounded, margins thin and plicate laterally; dise fleshy, with a pair of inconspicuous, subumbonate calluses at the base. Column of the genus, about 4 mm. long.—cocr£: region north of El Valle de Antón, alt. ea. 1000 m., Feb. 20, 1942, Allen 2936 (Herb. Ames, TYPE). Described and illustrated from a herbarium specimen and flowers preserved in alcohol. COELIOPSIS HYACINTHOSMA Reichb. f.—cocrí: epiphytic, hills north of ЕІ Valle de Antón, alt. са. 1000 m., (inflorescence pendant, sepals and petals ereamy white, labellum ereamy white with yellow blotch near base of column), April 12, 1941, Allen 2402. The original speeimens were grown in England from material presumed to have been collected in Panama. Three collections, at least, have been made subsequently in Costa Rica but this is the first speeimen known from a definite locality in Panama. KrcarELrgLLA Houtteana (Reichb. f.) L. О. Williams, comb. nov. ( Kegelia Houtteana Reichb. f. in Bot. Zeit. 10: 670. 1852; Xenia (Vou. 29 348 ANNALS OF THE MISSOURI BOTANICAL GARDEN Orch. 1: 45, Ё. 20, I, 1-7. 1854).—cocr: vicinity of La Mesa, region north of El Valle de Antón, alt. 1000 m., Oct. 10, 1941, Allen 2759. К egeliella is a genus so extremely rare that but a single specimen previously had ever been received at the Ames Herbarium, and that from Jamaiea. Mansfeld (Fedde Repert. 36: 60. 1934) described an additional species of the genus from Costa Rica and called atten- tion to the fact that Kegelia Reichb. f. was a homonym. It is appar- ent from Mansfeld's diseussion that the genus was unrepresented in the Berlin herbarium until he received that specimen. Kegeliella Houtteana was originally described from garden mate- rial that was said to have been discovered in Surinam. In repeating his description two years later in ‘Xenia Orchidacea’ Reichenbach wrote ‘‘ Es ist sehr wahrscheinlich, dass die Mutterpflanze—von der Tracht einer Cirrhaea—aus Surinam stammt und von Herrn Kegel entdeckt wurde," which may indicate that he has some reason to doubt his previous statement of locality. Reichenbach’s drawing of his Kegelia Houtteana, like so many of his drawings, shows enough to recognize the genus but is not definite enough for one to be able to state that a specimen in hand is, with- out doubt, conspecific. There are some differences in the flowers of the Allen specimens and the only other specimen available (Jamaica, Skinner & Robinson 7432) but until more material is available and Reichenbach’s type is better understood it is perhaps better not to try to separate the two. The genus is new to Panama. MAXILLARIA ARACHNITIFLORA А. & S.—cocré: epiphytic; trail to Las Minas, hills north of El Valle de Antón, alt. 1000 m., Dec. 2, 1941, Allen 2875. Previously known from Costa Rica where it is rare. 'The present specimen is smaller than the type specimen. Млхил.АвтА conduplicata (A. & S.) L. О. Williams, comb. nov. Ornithidium conduplicatum A. & S., Sched. Orch. 8: 66, fig. 1925).— cHIRIQUÎ: Palo Alto Hill, alt. 1300-1600 m., Sept.-Oet. 1923, Powell 341. Known only from Panama. MAXILLARIA ENDRESII var. ANGUSTISEGMENTA (А. & S.) C. Schweinf. —cocth: hills north of El Valle de Antón, alt. 1000 m., Dec. 2, 1941, Allen 2870.—Previously known from Costa Rica. 'The present speci- men has blunter sepals and petals than usual and the peduneles are shorter. Maxtuuaria neglecta (Sehltr.) L. О. Williams, comb nov. (Orni- thidium anceps Reichb. f., Beitr. Orch. Centr.-Am. 75. 1866, non Maz- illaria anceps A. & S.; O. neglectum Schltr. in Fedde Repert. 19: 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 349 242. 1923).—cANAL ZONE: on trees, hills north of Frijoles, Dec. 19, 1923, Standley 27669; cHtRIQUi: epiphyte in shade; valley of Upper Río Chiriquí Viejo, alt. 1300-1900 m., July-Aug. 1937, White Ф White 37. The species is quite a common one and extends from Hon- duras to Panama. Maxitiaria Pittieri (Ames) L. О. Williams, comb. nov. (Orni- thidium Pittieri Ames, Sched. Orch. 2: 35. 1923).—cnumiqví: rain forest, Bajo Chorro, Boquete District, alt. ca. 2000 m., Davidson 117. Costa Rica and Panama. CRYPTOCENTRUM CALCARATUM Schltr. in Fedde Repert. 12: 214. 1913. (Pittierella calcarata Schltr. l.c. 3: 81. 1906).—cocr£: region north of El Valle de Anton, alt. ca. 1000 m., (flowered at Ancon, C. Z.), Nov. 5, 1941, Dunn 2762. This is the third species of Crypto- centrum to be found in Panama recently. Previously known from Costa Rica where it is not uncommon. TRICHOPILIA TURIALBAE Reichb. f.—veEracuas: epiphytic; moun- tains west of Azuero, alt. 600 m., (flowered in collection of Mr. A. G. B. Fairchild), Sept. 10, 1941, Dunn s.n.—Previously known from Costa Rica. Trichopilia turialbae is allied to T. tortilis Lindl. and seems to be distinguished by its smaller, differently colored flowers with lateral sepals connate to about the middle. The sepals and petals are 3 em. long in our specimen, while the lip is 4 em. long. The sepals and petals of the type of T. turialbae, of which we have a photograph, are nearly as long as the lip. Mesospinip1um Warscewiczir Reichb. f. in Bot. Zeit. 10: 929. 1852; Xenia Orch. 1: 36, $. 16, figs. I, 1-11. 1854.—cocrf: epiphyte; El Valle de Anton, alt. 650 m., (sepals and petals greenish yellow spot- ted maroon, lip pale yellow with minute red spots; flowered at San Francisco de la Calita), Aug. 30, 1941, A. G. B. Fairchild (Allen 2685). Apparently the first collection of this species since the origi- nal was made some ninety years ago. The determination has been made on the basis of the original description and figures I and 1 cited above. Figure 1 is exactly like the plant in hand, and figure I is fairly good; figures 4 and 5 are erroneous and misleading; the remainder of the figures are fairly good. The generic status of the plant is open to some question. It is very close to ce and especially to Lindley’ s group ** Муат мит,” an aggregation of small-flowered species not unlike M DAMAGE HE in habit and flower structure. However, in Meso- spinidium Warscewicezii there seems to be a short but distinct col- [Vor. 29 350 ANNALS OF THE MISSOURI BOTANICAL GARDEN umn foot and, consequently, a mentum. These conditions are not known to me to occur in Odontoglossum. LOCKHARTIA MICRANTHA Lindl.—cocrÉ: region north of El Valle de Antón, alt. ca. 1000 m., Jan. 13, 1942, Allen 1942. This collection is peculiar because of the very short lateral lobes of the lip, in this case almost lacking. Lockhartia integra A. & S. is probably a synonym. DicnaEgA Мовнтвп Fawe. & Rendle.—cocr£ : epiphytic; trail to Las Minas, hills north of El Valle de Antón, alt. ca. 1000 m., (sepals and petals striped with deep lavendar, lip deep lavendar), Dec. 2, 1941, Allen 2874. Previously known from the Costa Rica and the West Indies. CHLORANTHACEAE HEDYOSMUM NUTANS Sw.—socas DEL TORO: Fish Creek Mts., Apr. 29, 1941, H. von Wedel 2362.—Known from the West Indies. Judg- ing from the original description of H. Brenesii Standl. (Fl. Costa Rica, p. 371), that species may be synonymous with H. nutans. There seems to be no valid distinction between the von Wedel speci- men and West Indian plants determined as H. nutans, and no justi- fieation in assuming a plant to be different from West Indian species merely because it is found on the Central American mainland. LACISTEMACEAE Lozania Kuve Mansf.—socas DEL токо: Old Bank Island, Feb. 18, 1941, H. von Wedel 2121.—Previously known from Colombia and Peru. No specimens of L. pedicellata (Standl.) L. B. Smith were available for comparison, but in the von Wedel specimen the racemes are not always solitary in the leaf axils, the charaeter sep- arating L. pedicellata in Smith's key (Phytologia 1: 138. 1935) from the L. Klugii group. The specimen agrees well with South American specimens of L. Klugii. This genus has also been included in the Flacourtiaceae by many systematists. MORACEAE ( P. C. Standley) РЕВЕВЕА hispidula Standl., sp. nov.—Ramuli gracillimi brunnes- centes vel ochracei dense hispiduli vel fere hirsuti, internodiis elon- gatis; stipulae caducae lineares ca. 1.5 em. longae extus hirsutae; folia majuscula brevissime petiolata membranacea, petiolo erasso hirsuto vix ultra 4 mm. longo; lamina oblonga vel elliptico-oblonga 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 351 14—23 em. longa 4.5-8 em. lata caudato-acuminata, acumine angusto interdum fere lineari usque 2 em. longo, basi paullo inaequali angu- ste rotundata vel obtusissima, remote saltem supra medium serrato- denticulata, supra sublucida ad costam sparse hirsuta aliter glabra, subtus paullo pallidior ad costam nervosque hispidula vel hirsuta ad venas puberula et in areolis minute seaberulo-asperata, costa tenui prominente, nervis lateralibus utroque latere ca. 16 angulo lato di- vergentibus leviter areuatis prope marginem arcuato-conjunetis, venis prominulis laxe reticulatis; receptaculum femineum axillare fere sessile in statu fructifero 2 em. latum, bracteis paucis imbricatis inaequalibus latis obtusis vel apiculatis, interioribus vix ultra 4 mm. longis, omnibus extus dense albido-strigosis; drupae ca. 8 sessiles ovali-globosae са. 7 mm. longae et 5-6 mm. latae apice basique ro- tundatae, ubique dense hispidulae; stylus crassus cum ramis vix 1 mm. longus, ramis stylo brevioribus crassis acutis.—BOCAS DEL TORO : vieinity of Chiriquí Lagoon, H. von Wedel 1935 (Herb. Missouri Bot. Gard. TYPE). In leaves and inflorescence this tree, of somewhat doubtful generic position, is quite unlike any species of Perebea that has been re- corded from Central America, and it is not very closely related to any of the South American ones of which material is available. Probably it is referred correctly to Perebea, but further material will be necessary to decide this point, the available specimens being unfortunately rather inadequate and incomplete. CAPPARIDACEAE Capparis Clara Schery, sp. nov. Arbuseula 5 m.; ramulis novellis aureobrunneis cum squamis minutis peltatis stellatis, internodiis ea. 2 em. longis; foliis alternatis ellipticis basi acutis apice attenu- atissime Шоо; petiolis brevibus са. 5 mm. longis angulatis supra anguste canaliculatis, laminis 15-20 ст. longis 4.3-6.3 em. latis subtus argenteo-lucidis medio nervio prominente nervis late- ralibus secondariis subprominentibus 7-12 paribus supra planis lepi- dotis glabris glaucescentibus; pedunculis ca. 10 em. longis termi- nalibus vel subterminalibus in axilles foliorum aureo-brunneis squamis pilisque minutis stellatis, pedicellis ca. 6, ca. 2 cm. longis umbellatis ; floribus cine MEAE calycibus Бае ниво extus argenteo-stellato-] bus 4-lobatis lobis triangularibus 1 mm. longis, tubo 1 mm. alto; Sc oblongis ea. 8 mm. longis 3.5 mm. latis extus minutissime stellato-pubescentibus intus subglabris; staminibus ca. 20, filamentis glabris ca. 2 cm. longis, antheris [Vor. 29 352 ANNALS OF THE MISSOURI BOTANICAL GARDEN oblongis ca. 2 mm. longis, gynophorio glabro ca. 2.5 em. longo; ovario pluriovulato oblongo 3 mm. longo stellato-pubescenti, stig- mate sessili obtuso placentis 3 quisque cum 2 ordinibus ovularum; fructu ignoto.—socas DEL TORO: Fish Creek Mts., Apr. 16, 1941, Н. von Wedel 2235 (Herb. Missouri Bot. Gard., түрк). Ар tly this species is distinct from any published species known from Panama, Costa Rica, or Colombia. Its affinities are prob- ably with C. detonsa Tr. & Pl., judging from the description of that species. In Dugand’s clarifying synopsis of Capparis in Colombia (Caldasia 2: 37. 1941), the von Wedel plant keys out to C. detonsa, but differs from it especially in having smaller flowers and a less pronounced pubescence. C. clara can be distinguished from most or all other species by the very long apical attenuation of the leaf. Characters also distinctive are the silvery sheen of the lower leaf surface, the slender, graceful inflorescence, and the minute golden- brown peltate scales of the very young branches. The leaves are perfectly elliptic, acute at the base, short-petiolate. The flowers are of medium size, 4-parted, umbelliform from an elongate peduncle. The elongate gynophore bears terminally an ovary with 3 placentae, each placenta with 2 rows of ovules. About 20 stamens surround the gynophore. LEGUMINOSAE CozoBA CATENATA (D. Sm.) Britton & Rose.—socas DEL Toro: Isla Colon, July 29, 1940, H. von Wedel 199. Previously known from Costa Rica. DrrnEMEXA LIGUSTRINA (L.) B. & R.—socas DEL токо: Isla Colón, Aug. 23, 1940, Н. von Wedel 517. Apparently new to the continent; previously known from the West Indies. Inca PnEvsu Harms—socas DEL toro: Water Valley, Sept. 9, 1940, H. von Wedel 672. Previously known from northern Central America. MELIACEAE GvanEA Снтснох C. DC.—zocas DEL roro: Cocoa Cay, vicinity of Chiriquí Lagoon, Oct. 26, 1941, H. von Wedel 2874. Previously known from Mexico and British Honduras. MALPIGHIACEAE НЕТЕКОРТЕВТВ PLATYPTERA DC.—socas DEL токо: Old Bank Island, Feb. 8, 1941, H. von Wedel 2027. Previously known from South America. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 353 DICHAPETALACEAE DıcHAPETALUM axillare Woodson, sp. nov. Arbuscula ca. 5 m. alta; ramulis novellis dense cinereo-tomentellis tandem glabratis. Folia brevissime petiolata oblongo-oblanceolata apice obtuse acu- minata basim versus acute attenuata eum petiolo ca. 0.3 em. longo 9-15 em. longa 2.5-5.0 em. lata firmiter membranacea venis venu- lisque utrinque cinereo-pilosulis petiolo simile. Inflorescentia aut terminalis aut axillaris numquam petiolo adnata subcapituliformis pauciflora ca. 1-2 em. longa et lata dense cinereo-tomentella. Flores ca. 0.1 em. longi pedicellati albi; sepalis ovatis ca. 0.2 cm. longis extus dense cinereo-pilosulis intus glabris; petalis aequilongis anguste ovatis medio bifidis ; ovario ovoideo са. 0.1 ст. longo dense pilosulo.—cocrÉ: hills north of El Valle de Antón, alt. about 1000 m., July 14, 1940, P. H. Allen 2202 (Herb. Missouri Bot. Gard., TYPE). Undoubtedly a close relative of D. Donnell-Smithii Engl., which has been reported as ranging from western Panama (probably Chiriquí) to British Honduras. This latter species, however, appar- ently always has more diffuse inflorescences with peduncles strongly adnate to the petiole of a subtending leaf, the leaves broader and less attenuate toward the base, and the indument less shaggy and conspicuously ferruginous. DicHAPETALUM NEVERMANNIANUM Standl.—socas DEL TORO : Water Valley, Oct. 30, 1940, H. von Wedel 1424. Previously known only from the type collection from Costa Rica. EUPHORBIACEAE (L. Croizat) Pera aperta Croizat, sp. nov. Arbor 6-metralis; innovationibus plus minusve conferte erustaceo-lepidotis, indumento haud laeto; foliis more generis in sicco atro-discoloribus, supra costa excepta glabris, subtus lepidibus argillaceis dissitis adspersis, oblongis, apice breviter acuminatis, basi rotundato-cuneatis, 1.0—5 em. longis, 2-4 em. latis, margine revoluto subintegro, venis primariis plus mi- nusve patentibus са. 7-jugis, sat obscuris ; petiolo eglanduloso, 1.5-2 em. longo; inflorescentiis axillaribus, $ tantum visis; pedunculo communi gracili, lepidoto, ad 1 em. longo, apice bracteolis binis ca. 2 mm. longis terminato, floribus $ pernudis, scilicet in sacco e brac- teolis more generis efformato haud inclusis, glomerulatim 2—5, eca- lyculatis vel subecalyculatis, ovario lageniformi, ca. 2.5 mm. longo, stylo in stigmatibus 3 latis, papillosis, obtriangularibus evadente; (Vor. 29 354 ANNALS OF THE MISSOURI BOTANICAL GARDEN loculis in ovario (videtur) 4, semine valde immaturo earuneulato.—- сос: hills south of El Valle de Antón, May, 1941, P. Н. Allen 2506, (U. S. National Herb. түрк). The peculiar characters of the new species here described under Mutis's genus Pera suggest brief considerations of a preliminary nature on the limits of the genus. The region in which the type material was collected is aptly char- acterized in a letter of P. H. Allen to C. V. Morton, quoted in the latter's work on the Gesneriaceae from Panama, and this writer readily agrees with Allen’s comment: ‘‘I think I ean guarantee that you will get some interesting plants’’ (Ann. Mo. Bot. Gard. 29: 35. 1942). Pera aperta, indeed, is an extremely interesting plant, the first of its kind ever seen by this writer. Aside from the technicalities of its flowers, Pera is supposed to be characterized by a peculiar involucre which surrounds the inflo- rescence. This involucre (see Pax & Hoffmann in Engler & Prantl’s Nat. Pflanzenfam. 19c: 154, fig. 78 b,c,d,e. 1931) is a bracteolate, bag- like strueture which, towards anthesis, breaks open to expose the é and 9 flowers. The perianths proper are much reduced, and the entire arrangement may be defined as a coarctate lateral spicate or subglomerulate inflorescence, uni- or bisexual, surrounded by a braet or fused braets which become open at the time of flowering. On account of the bracteate inflorescence, Pera has been maintained by Pax & Hoffmann as the type of the tribe Pereae, placed next to the Dalechampieae in the с! tion of Engler & Prantl, a dispo- sition which is probably more artificial than natural but is not to be challenged here. Pera is the subject of an unusually large literature which deals with its sectional or presumedly sectional units. In one of the early studies of this group, Klotzsch (in Wiegm. Arch. 7: 176. 1841) pub- lished the tribe Prosopidoclineae, typified by Pera and charaeter- ized by: **Ovarii loculi uniovulati. Semina arillata, exalbuminosa. Involuera subgloboso-vesicaeformia, hine hiantia, deinde plus minusve explanata, demum decidua, 3,4-6 flora, bracteis suffulta. Flores dioeci, apetali. Regarding Klotzsch’s later publication of the Peraceae as a family, this writer may remark that Prosopido- clineae is an illegitimate tribal name, because it is not derived from the name of a genus under the group (Art. 24, Internat. Rules Bot. Nomencl., 1935). Non-existent as a taxonomie name under the Rules (Art. 7) are Baillon's **Péridées" (Étude Gén. Euphorb., 433. 1858), improperly cited by Pax & Hoffmann (in op. cit., 153) as 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 300 Perideae. The legitimate tribal" name for the Pera plexus, conse- quently, remains Mueller's Pereae (Linnaea 34: 144. 1865). Klotzsch recognized under the Prosopidoclineae: Schismatopera Kl., Зрапа Leandro do Sacr., Pera Mutis, and Peridium Schott. This arrangement was criticized by Baillon (op. cit., 268-272), who maintained as a genus only Pera with two sections (op. cit., 433- 434), Eupera and Schismatopera, characterized, respectively, by **Étamines au nombre de 2-6 (ou plus), unies inférieurement dans une étendue peu considérable," and **Étamines 4-8 monadelphes ; filets soudés en une colonne cylindrique plus longue.’’ This separa- tion rests upon an essential differenee apparent in many euphor- biaceous flowers, some of which have the androecium connate to form a staminal eolumn (Jatropha, for instance) ; others, on the con- trary, have solute stamens (Manihot being an example). Whether this difference holds good in Pera, as claimed by Baillon, this writer does not know ; its basis is sound in theory, but it may be found in practice that the much reduced perianths of Pera bear both androe- cia set up into eolumns or more or less solute, within the same natural affinity. Tn his elaboration of the Euphorbiaceae (in DC. Prodr. 15? : 1025- 1031. 1866), Mueller maintains five sections, Schismatopera (КІ.) Baill., Eupera Baill., Spixia (Leandro do Sacer.) Baill., Neopera Griseb., Peridium (Schott) Muell.-Arg. Another section, Diplopera, is added by Mueller in Martius, ‘Flora Brasiliensis’ (11? : 425. 1874), which Mueller describes as follows: ‘‘Involucra basi unibracteo- lata, bisexualia, simul flores masculos 3-4 centrales et femineos pe- riphericos gerentia. Calyx masc. evolutus . . . Sectio insigniter pecu- liaris. . . . Six sections, too, are maintained in the most recent classification of Pax & Hoffmann (op. cit., 154), as follows: Diplo- pera Muell.-Arg., Perula (Schreb.) Pax & Hoffmann, Spixia (Lean- 1The proper definition and correct consequent use in modern taxonomy of the units between the genus and the family, and the genus and the species, published by early or comparatively en кора is beset with difficulties on account of the loose and conflict- ing manner in hese authors were wont to interpret or define these units. For in- stance: Mueller - ein Mt Pereae, but designated and consequently used them as a subtribe, not as a tribe. Since Art. 24 of the current Rules declares that the name of a tribe ends in -eae and that of a subtribe in -inae, there is confliet between Mueller's name and his own designation of it. This writer believes that, im principle, such confliets are to be defined in the sense that the designation of the rank by the author of the name is irrelevant, under Art o long as it is contradieted by the name's ending. us, the Mueller, d б legitimate ending of a tribal name, are to be aecepted in pog ains as a tribe, despite Mueller's insistence that they are a subtribe. This prineiple is general and important enough to deserve speeial mention here [Vor. 29 356 ANNALS OF THE MISSOURI BOTANICAL GARDEN dro do Saer.) Muell.-Arg., Schismatopera (Kl.) Baill, Neopera Griseb., Peridium (Schott) Muell.-Arg. This classification, which does not differ from that of Mueller in essential characters, is erro- neous in two respects: (1) It substitutes Perula Pax & Hoffm. for Eupera Baill., under the evidently mistaken assumption that Schre- ber’s Perula, being earlier than Eupera Baill., must be used as the basis for a new combination. Pax & Hoffmann probably are not aware of the fact that new combinations are required (Art. 53) only when names are transferred without a change of rank, it being ille- gitimate to effect a transfer, when the rank is changed and the posi- tion is preoccupied (Art. 16, Art. 61 [1]).! (2) It mistakenly credits the combination of Spixia to Mueller, while this combination was effected by Baillon (in Adansonia 5: 222. 1864-1865). Under the systematic viewpoint, none of the current subgeneric units of Pera even approach the characters of P. aperta, for in this peculiar plant the involucre is absent, or at least reduced to minute bracteoles at the base of the gynoecium. In view of the emphasis placed throughout classification upon the involucre as a generic character it should be possible to elect P. aperta as the type of a new genus. This writer does not believe that the erection of a new genus is advisable, at least until much better material is available, because: (1) The Euphorbiaceae are identified by a tendency towards reduction and recombination of the floral organs rather than by set morphological characters. Accordingly, tendencies count in their classification far more than do characters; (2) The wood characters of P. aperta are those of Mutis’s genus, as this writer has kindly been informed by Prof. I. W. Bailey of the Bio- logical Laboratories of Harvard University, and so its gross mor- phology is that of Pera. (3) It is not to be overlooked that certain species in the vicinity of P. aperta may have partially developed involucres, intermediate between those of this species and the ones of P. arborea and its immediate allies. In conclusion, the emphasis placed upon the involucre as a generic character of Pera is not completely justified, and rather than a new genus, an emended description of Pera, the Pereae, and two new subgenera are required, as follows: Prra Mutis in Svensk. Vetensk. Akad. Handl. Stockholm 5: 299. 1784; Muell.-Arg. in DC. Prodr. 15? : 1025. 1866, et in Mart. Fl. Bras. *On account of a faulty interpretation of Art. 58, some botanists believe that older names must be maintained even when the rank is changed. This fallacy is refuted in a paper on the trinomial typicus, by this writer, now in course of publication. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 357 11?: 421. 1874; Pax & Hoffmann in Engler & Prantl’s Nat. Pflan- zenfam. 19c: 153. 1931; Столе. descr. emend. Flores utriusque sexus in involuero alabastriformi inclusi (subg. Eupera), involucro saltem 9 interdum (subg. Gymnopera) nullo vel subnullo. This emendation modifies also the characters of the Tribe Pereae, which is essentially based upon Pera, as follows: РЕВЕАЕ Muell.-Arg. in Linnaea 34: 144. 1865, et in DC. Prodr. 15?: 1025. 1866; Pax & Hoffmann in Engler & Prantl’s Nat. Pflan- zenfam. 19c: 153. 1931; Croiz. descr. emend. Flores utriusque sexus saepissime in involuero alabastriformi inclusi. Pera subg. Eupera (Baill.) Croiz., st. nov. (Pera Sect. Eupera Baill. in Étude Gén. Euphorb., 434. 1858, p.p.typ.; Pera Sect. Perula [nomen serius Шедійтит| Pax & Hoffm. in Engler & Prantl’s Nat. Pflanzenfam. 19c: 154. 1931). Floribus utriusque sexus in involueris alabastriformibus inclusis. Typus nomenclaturalis : Pera arborea Mutis. This writer does not know whether all the sections of Mueller and Pax & Hoffmann properly belong to a single subgenus, although he accepts as most probable at this writing that a separation can be established between Pera ssp. in general and P. aperta on the strength of the involuere character. In Martius’, ‘Flora Brasilien- sis’ (op. cit., 11?: 421. 1874) Mueller accepts Pera ‘‘sensu Baill.,’’ a disposition which this writer is neither prepared to challenge nor to endorse without qualification at this time. Pera subg. Gymnopera Croizat. Floribus saltem 9 involucro alabastriformi carentibus, ad brac- teolas minimas reducto. Typus nomenclaturalis : Pera aperta Croizat. HIPPOCRATEACEAE HIPPOCRATEA cELASTROIDES HBK.—socas DEL TORO: Old Bank Island, Feb. 8, 1941, H. von Wedel 2038. Known from northern Cen- tral America and recorded from Colombia. Specific delimitation in this genus is at present so uncertain that it cannot be stated with surety that the von Wedel specimen is correctly determined. How- ever, it matches other specimens so labeled in the herbarium, and seems distinct from H. volubilis L. growing in the same locality. HIPPOCRATEA ovara Lam.—socas DEL TORO: vicinity of Chiriquí Lagoon, Oct. 7, 1940, H. von Wedel 1059. Known from South [Vor. 29 358 ANNALS OF THE MISSOURI BOTANICAL GARDEN America. This Wedel specimen in fruit may prove to be the same as H. volubilis or H. celastroides, but best matches specimens in the Missouri Botanical Garden herbarium labeled H. ovata. ICACINACEAE (Е. A. Howard) DiscopHora MONTANA Howard—socas DEL токо: Fish Creek Mts., Apr. 14, 1941, H. von Wedel 2225. Previously known only from Colombia. SAPINDACEAE Cupania HIRSUTA Radlk.—panamA: Vicinity of Capira, Feb. 12, 1939, P. H. Allen 1687. Previously known from South America. No satisfactory separation seems possible between this species and speeimens in the herbarium labeled C. fulvida. The Allen speeimen is intermediate between South American forms represented in the herbarium as C. hirsuta, and Costa Rican forms represented by specimens determined as C. fulvida 'Tr. & PI. PAULLINIA CALOPTERA Radlk.—socas рег токо: Water Valley, Sept. 23, 1940, H. von Wedel 904. Previously known from South America. QUIINACEAE (P. C. Standley) Lacunaria panamensis (Standl.) Standl., comb nov. (Quiina panamensis Standl. Field Mus. Publ. Bot. 4: 236. 1929).—Cocré: hills north of El Valle de Antón, trail to La Mesa, alt. about 1000 m., Aug. 31, 1941, P. H. Allen 2689. A tree 15 m. tall; fruits woody, brown. Quiina panamensis was based upon G. Proctor Cooper 609, from Buena Vista Camp on Chiriqui trail, Prov. Bocas del Toro, and consisted of specimens in bud. Study of the type and of the recent collection obtained by Mr. Allen shows that the tree belongs to the genus Lacunaria, segregated in 1925 from Qwiina by Dr. Ducke. Rather numerous Quiina species of South America have been transferred to Lacunaria in recent years, and it is not sur- prising to find a member of the genus in Panama. The only other member of the family known from North America, Quiina Schippii Standl. of British Honduras, is properly referable to Quiina. Mr. Allen's material includes a mature fruit, which may be described as follows: Fruit depressed-globose, in shape resembling that of Hura crepitans, about 8 em. broad and 5.5 em. high, slightly depressed at the apex, densely costate vertically with low rounded 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 359 ribs, filled with pulp, the pericarp hard and woody, 3 mm. thick; seeds about 1 cm. long, very densely covered with long brown hairs. TILIACEAE (P. C. Standley) Мовтохторехрвох hirsutum Standl., sp. nov.—Frutex metralis, ramis gracilibus ochraceis teretibus, novellis pallide viridibus sat dense pilis longis patentibus fulvis hirsutis; folia breviter petiolata membranacea, petiolo crassiusculo 4-5 mm. longo hirsuto; lamina anguste oblonga vel lanceolato-oblonga 10.5-17 em. longa 3-4.5 em. lata, apice longissime lineari-attenuata, basi insigniter obliqua, latere interiore acuta, exteriore late rotundata vel subeordata, supra in sicco viridis ad costam sparse hirsuta, aliter glabra, venis promi- nulis laxe reticulatis, subtus fere concolor ad costam nervosque hir- suta, costa tenui elevata, nervis lateralibus utroque latere 8-9 te- nerrimis arcuatis, venis prominulis laxe reticulatis, basi trinervia ; inflorescentia (tantum in statu fructifero visa) terminalis longipe- duneulata ut videtur pauciflora, rhachi hirsuta, pedicellis fructiferis са. 12 mm. longis crassiusculis; capsula depresso-globosa lutea са. 2 em. alta atque 2.5 em. lata valde rugosa, valvis crassis ca. 12 mm. latis ; semina in quoque loculo 2 crassa obtuse angulata 6 mm. diam. —BOCAS DEL TORO: Water Valley, vicinity of Chiriquí Lagoon, Nov. 21, 1940, H. von Wedel 1694 (Herb. Missouri Bot. Gard. TYPE). The genus Mortoniodendron Standl. & Steyerm., published in 1938 and based upon a Panama plant, has grown surprisingly in species numbers during the past few years. In 1940 two species were de- scribed from Honduras and Guatemala, the present being the fourth known representative of the genus. M. hirsutum may be recognized at once as distinct from the other Panama species, M. anisophyllum (Standl.) Standl. & Steyerm., by its fairly abundant pubescence of long, stiff, straight spreading hairs, a character which distinguishes it also from the more northern species. BOMBACACEAE Сетва Allenii Woodson, sp. nov. Arbor epiphytica ramosissima ut dieitur ea. 10-12 m. alta; ramulis erassiusculis rimosis sparse acu- leolatis. Folia desunt. Flores solitarii vel bini magni speciosi ; pedi- cellis crassiusculis ca. 2 em. longis; calyce late urceolato са. 2 cm. longo ostio ca. 1.5 em. diam. obscure irregulariterque 2- vel 5-lobato coriaceo glabro vel indistincte papillato; petalis 5 oblongis rotun- datis minute emarginatis 5-6 em. longis са. 1.5 em. latis carnosis [Vor. 29 360 ANNALS OF THE MISSOURI BOTANICAL GARDEN patulis extus papillatis marginibus densius saturate roseis intus lacteis ; staminibus 5, filamentis 3.5-4.0 ст. longis са. 16 longitudine connatis, antheris linearibus dorsifixis ca. 3.5 em. longis basi bifidis. —cocLÉ : hills north of El Valle de Antón, alt. about 1000 m., May 10, 1942, Р.Н. Allen 2924 (Herb. Missouri Bot. Gard. түрк). Ceiba Allen? falls readily into Schumann "в section Eriodendron, previously represented by two species of central Brazil. From our species C. Rivieri (Dene.) K. Sch. differs because of its shorter, truncate calyx, and C. Erianthos K. Sch. because of its more slender habit and much smaller flowers. Itis very unfortunate that the plant bloomed in a leafless condition. Mr. Allen describes it as a ‘‘stran- gler tree"! similar to species of Ficus. STERCULIACEAE TurosnoMa asclepiadiflorum Schery n. sp. Arbor ut dieitur 30 m. alta; ramis glabris brunneis; foliis magnis 30—40 em. longis 10-13 em. latis ellipticis utrinque omnino glabris apice cus- pidatis basi acutis in petiolis subdecurrentibus subtus costa 1 pro- minentissima nervis lateralibus 24—34 areuatis prominentibus; peti- olis ca. 1.5 em. longis supra subcanaliculatis; inflorescentiis cymi- formibus, pedunculis brunneis stellato-tomentosis 3—4-pli-divisis in multos pedicellos cum bracteis parvis ad locos divisionum; floribus coccineis magnis ; sepalis anguste ovato-lanceolatis ea. 12 mm. longis 3—4 mm. latis erassis reflexis extus stellato-tomentosis intus glabris margine puberulentis basi pilis erassis glandulosis luteis; petalis obovatis cochleatis 6 mm. longis 4 mm. latis interne glabris externe apice pubescentibus cum ligulis rotundatis lepidotis ; staminibus et staminodiis basi cohaerentibus in tubum glabrum 2 mm. altum, fila- mentis glabris 3 mm. longis cum 2 antheris biloeularibus, stami- nodiis anguste lanceolatis 10 mm. longis basi 2 mm. latis puberulen- tis; ovario ellipsoideo tomentoso 5-lobato eum stylo 5-partito; fructibus ignotis.—sBocas DEL токо: Water Valley, Nov. 8, 1940, Н. von Wedel 1535 (Herb. Missouri Bot. Gard. түрк). Although fruiting material of this species is lacking, floral and vegetative characters distinguish it sufficiently to warrant descrip- tion as a new species. As in T. Cacao the branchlets and leaves are entirely glabrous, a character which separates it from the typical T. bicolor. The elliptic leaves are not asymmetric at the base and are subdecurrent on the petiole. The inflorescence is a many-flow- ered eyme about 5 em. long, branching 3 or 4 times in a di- or tri- chotomous fashion. The large red flowers, with their reflexed sepals, 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 361 cochleate petals resembling hoods and long narrow staminodes resembling horns, superficially simulate asclepiadaceous flowers. The ligules of the petals are more or less sessile, not clawed as in T. Cacao. The long thin staminodes appear tentacular, not folia- ceous as in T. simiarum. The form of the ovary suggests that the fruit is shallowly 5-lobed. DILLENIACEAE (C. V. Morton) SAURAUJA PAUCISERRATA Hemsl.—cocré: trail to Las Minas, north of El Valle de Antón, alt. 1000 m., May 10, 1941, Paul H. Allen 2464. This species was described from the Volcán de Fuego, Guatemala. The present specimen agrees with the description and with a photo- graph of the type. Several specimens from Guatemala and El Sal- vador have been referred to this species, some of which may be incorrectly identified. The recently described S. Seibertw Standl., which also has glabrous leaves, differs (from deseription) in its larger sepals and petals, longer inflorescence and pedicels, and larger, many-veined, more strongly serrate leaves. OCHNACEAE SauvaGEsia ELATA Benth.—socas DEL Toro: Old Bank Island, Feb. 8, 1941, H. von Wedel 2029. Previously known from South America. MARCGRAVIACEAE MancGRAVIA NEPENTHOIDES Seem.—sBocas DEL TORO: Water Valley, Sept. 17, 1940, H. von Wedel 818; Chiriqui Lagoon, Oct. 15, 1940, H. von Wedel 1192. Previously known from northern Central America, possibly reported from Costa Rica. The von Wedel specimens check well with published illustrations of this species, although differing somewhat from the only specimen in the herbarium under this name. NorANTEA ALBIDO-ROSEA Gilg, ex char.—socas DEL токо: Old Bank Island, Feb. 8, 1941, H. von Wedel 2035. Known from Costa Rica. This specimen may well prove to be the same as №. Вгепези Standl., also known from Costa Rica, corresponding within the limits of variability with specimens of the latter in the herbarium. However the original description of N. Brenesw does not fit the von Wedel specimen as accurately as that of N. albido-rosea. The plant can be distinguished from N. subsessilis (Benth.) D.Sm. by its sub- umbelliform inflorescence. [Vor. 29 362 ANNALS OF THE MISSOURI BOTANICAL GARDEN GUTTIFERAE Сълувта longipetiolata Schery, п.зр. Arbuscula circ. 30 m. alta; ramis glabris irregulariter nodulatis internodiis 0.5-1.5 em. longis; folis magnis longe-petiolatis glabris crassis coriaceis; laminis ellipticis vel obovato-oblongis basi cuneatis apice rotundatis mar- gine leviter subrevolutis, nervis lateralibus multis (60-90) cire. 80? a nervo medio divergentibus utrinque prominentibus ; petiolis 4.5— 8.0 em. longis crassis in sicco longitudinaliter striatis subteretibus vel angulatis sed non alatis basi lacunis semierateriformibus ; inflo- rescentiis terminalibus cymiformibus 2—3-floris ; pedicellis principa- libus 0.5-1.0 em. longis apice cum bracteis 2 deltoideis oppositis; pedicellis secondariis subteretibus crassis 1-2 ст. longis item apice cum 2 bracteis deltoideis oppositis; floribus flavo-albis magnis cire. 5—6 em. latis; 2 lobis calycis exterioribus basi concretis patellifor- mibus fulgidis, 5 aliis lobis late ovato-rotundatis coriaceis; petalis submembranaceis ovato-rotundatis, cire. 2.5 em. longis, 3 em. latis; toro cylindrico 6-7 mm. alto ovarium cireumdante; ovario 5-carpel- lato pyramidali stigmati sessili ; flores masculae ignotae.—nocas DEL TORO: vicinity of Chiriquí Lagoon, alt. near sea-level, Oct. 12, 1940, Н. von Wedel 1136 (Herb. Missouri Bot. Gard. түре). This species is distinguishable especially by its large cream-col- ored flowers, very long unwinged petioles, and thick coriaceous leaves. The torus (fused staminodia of Vesque) is cylindric-cam- panulate, completely surrounding the 5-carpellate ovary to a height of 6 or 7 mm. The relationship of this species is probably with C. rosea, C. Cooperi, C. stenophylla, etc., from the same general region. Although belonging to a genus in which **probable new species’’ of distinctive appearance are continually turning up, this von Wedel specimen could not be satisfactorily matched in the herbarium, nor in Vesque’s monograph nor with recent descriptions. VIOLACEAE (C. V. Morton) Іомірісм Тнтемет Donn. Smith—coocr£: north rim of El Valle de Antón, June 4, 1939, Alston & Allen 1859. This collection was dis- tributed as Hybanthus parietariifolius (DC.) Loes., a synonym of the annual species, H. attenuatus (H. & В.) С. К. Schulze. Ionidium Thiemei (previously known from Mexico to Honduras) is a peren- nial, and differs in many ways from H. attenuatus. It is to be referred to the genus Hybanthus also, of course, but I refrain from 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 363 making a combination at the present time, inasmuch as I am pre- paring a paper to be published elsewhere on the North American species of this genus. FLACOURTIACEAE Lunania Pirtiert Standl.—socas DEL Toro: Fish Creek, moun- tains, May 7, 1941, H. von Wedel 2396. Previously known only from the type collection from Costa Rica. TURNERACEAE (C. V. Morton) ERBLICHIA ODORATA Seem.—cocrÉ: trail to Las Minas, north of El Valle de Antón, alt. 1000 m., May 10, 1941, Paul H. Allen 2468. This genus was long considered monotypie, but in a recent revision Standley and Steyermark (Field Mus. Publ. Bot. 22: 351—357. 1940) have recognized three species and one variety. According to their treatment, the present specimen woud key at once to E. Standleyt Steyerm. of Oaxaea and Honduras. However, both Standley and Steyermark have studied Allen's specimen and are inclined to place it with E. odorata, the type species of the genus, which was origi- nally collected in Panama but has not since been found there, although it is rather common in Costa Rica. Allen's collection throws some doubt on the validity of E. Standleyi as a species, since it differs from typical E. odorata in the same characters (glabrous ovary and branchlets) as E. Standleyi does from E. xylocarpa. CARICACEAE CARICA DOLICHAULA D. Sm.—socas DEL TORO: Chiriquí Lagoon, Oct. 8, 1940, H. von Wedel 1084. Previously known from Costa Rica and ranging to British Honduras. CACTACEAE (L. Cutak) EPIPHYLLUM MACROPTERUM (Lemaire) Britton & Rose—socas DEL toro: Water Valley, Sept. 12, 1940, H. von Wedel 736; vicinity of Chiriqui Lagoon, Oct. 16, 1940, H. von Wedel 1164. Previously known from Costa Rica. Евтрнутллум Pirriert (Weber) Britton & Rose—socas DEL TORO: vicinity of Chiriqui Lagoon, Oct. 21, 1940, H. von Wedel 1294. Pre- viously known from Costa Rica. [Vor. 29 364 ANNALS OF THE MISSOURI BOTANICAL GARDEN THYMELIACEAE ScHOENOBIBLUS PANAMENSIS Standl.—socas DEL тото: Isla Colón ?, H. von Wedel 410, Aug. 12, 1940. This is apparently the first record of the genus for North America. Species have previously been known from South America and the West Indies. ONAGRACEAE J USSIAEA LATIFOLIA Benth.—socas DEL Toro: Water Valley, Н. von Wedel 781, Sept. 14, 1940; Chiriqui Lagoon, H. von Wedel 1384, Oct. 18, 1940. Previously known from South America. LOGANIACEAE РотатлА AMARA Aubl.—socas DEL токо: Fish Creek Mts., Арт. 30, 1941, H. von Wedel 2369. Previously reported from Costa Rica and South America. Although to be expected from Panama, this collec- tion is of interest in that it apparently is only the third or fourth for the genus from North America. The ovary of the von Wedel plant is four-carpellate in contrast to the two-celled condition reported by Standley (Fl. Costa Rica, p. 921. 1938) and the three-celled condi- tion reported by Aublet (Pl. Gui. 1: 394. pl. 151. 1775) in the original description for the species. APOCYNACEAE SrEMMADENIA lagunae Woodson, sp. nov. Arbor vel arbuscula ca. 5-12 m. alta omnino glabra; ramulis dichotomis gracillimis, inter- nodiis са. 1.0-2.5 em. longis. Folia opposita inaequalia breviuseule petiolata elliptica apice abrupte subcaudato-acuminata basi in peti- olum attenuata 4.5-14.0 em. longa 1.5-9.0 em. lata; petiolis 0.3-0.8 em. longis. Inflorescentiae terminales vel subterminales bostrycino- racemosae pauci- vel pluriflorae ; pedunculo simplici vel basi dichot- ото 1.0-2.5 em. longo omnino minute bracteato cicatricosoque; pedicellis ea. 0.5 em. longis. Calycis laciniae oblongae vel oblongo- oblanceolatae acutae vel obtusae valde inaequalia 1-2 cm. longae subfoliaceae. Corolla pallide aurea infundibuliformis tubo proprio 2.5-3.0 em. longo basi са. 0.2-0.3 em. diam. superne contorto angus- tioreque ibique staminigero faucibus conicis abrupte dilatatis 1.0— 1.25 em. longis ostio ca. 0.5-0.7 em. diam. lobis oblique obovatis acuminatis 1.5-2.0 em. longis patulis. Follieuli ignoti.—szoca4s DEL TORO: Careening Cay, July-Aug., 1940, H. von Wedel 570 (Herb. Missouri Bot. Gard., rypr); Water Valley, Sept. 6, 1940, Н. von Wedel 587 (Herb. Missouri Bot. Gard., сотурЕ). 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 365 This species of the S. Alfari complex is interesting because of its occurrence near the coast of the Chiriquí Lagoon, the other species inhabiting higher altitudes. It apparently is most closely allied to S. Allenii Woods., which has much shorter calyx lobes (0.6-0.9 em.) and corollas of different proportions and size (proper tube 1.0-1.25 em. long; throat about 2 em. long). Рвеѕтохтл Wedelii Woodson, sp. nov. Frutex volubilis; ramulis dense ferrugineo-tomentosis. Folia brevissime petiolata ovato- elliptica apice acuminata basi obscure cordata 13-16 em. longa 6-10 em. lata utrinque subtus densius minute ferrugineo-tomentella, peti- olo 0.2-0.5 em. longo tomentoso. Inflorescentia dense bostrycino- racemosa multiflora; pedunculo 2-15 em. longo ferrugineo-tomen- toso, pedicellis 0.5—0.7 em. longis similiter vestitis; bracteis anguste lanceolatis ca. 1 ет. longis foliaceis pilosis. Calycis laciniae oblongo-lanceolatae longe acuminatae 3.0—3.5 em. longae foliaceae pilosae ; squamellis deltoideis apice emarginatis vel laceratis sparse pilosulis. Corolla luteo-laetea extus dense ferrugineo-tomentella tubo infundibuliformi ca. 4 em. longo basi ca. 0.2 em. diametro tertia parte superiore staminigera ibique conico-dilatata ostio ea. 0.7-0.8 em. diametro, lobis oblique ovatis acuminatis ea. 1.2 em. longis patulis; annulo faucali bene manifesto ca. 0.1 em. lato con- tinuo haud lobato. Ovarium glabrum са. 0.2 em. altum; nectario conico-annulato ostio 5-lobato glabro ovarium paulo superante.— BOCAS DEL TORO: Water Valley, vicinity of Chiriquí Lagoon, Oct. 26, 1940, Н. von Wedel 1353 (Herb. Missouri Bot. Gard., TYPE). Superficially similar to P. Allenit, but with shorter pubescence and obscurely cordate leaves. The flowers of P. Allen, also, are larger, and the faucal annulus of the corolla consists of 5 discrete, round lobes. Bonarousta Sananuo (К. & P.) Mgf.—cocr£: vicinity of La Mesa, region of El Valle de Antón, alt. ca. 1000 m., Nov. 12, 1941, P. H. Allen 2804. Previously known from western Colombia (Bogotá), eastern Ecuador, eastern Peru, and western Brazil. This discovery is of particular interest since it represents an element of Табегпае- montana previously thought to be entirely South American. Other South American species have been collected by Mr. Allen in the region of ЕП Valle, and this element probably is considerable in its ora. ASCLEPIADACEAE SaRCOSTEMMA ODORATA Hemsl.—socas DEL Toro: Water Valley, Nov. 23, 1940, H. von Wedel 1755. Previously collected in Guatemala [Vor. 29 366 ANNALS OF THE MISSOURI BOTANICAL GARDEN and Costa Rica. Our plants differ somewhat from those of Guate- mala in being practically glabrous. Сохоговоѕ Ophioglossa Woodson, sp. nov. Frutex volubilis; ramulis gracilibus laxe pilosulis. Folia opposita longiuscule petio- lata heterophylla oblongo- vel obovato-ovalia apice acuminata basi aut obtusa aut obscure sagittata aut profunde cordato-sagittata 4—10 em. longa 1.5-4.5 em. lata membranacea glabra; petiolo 1-2 em. longo minute pilosulo. Inflorescentia extra-axillaris subumbel- lata pluriflora ; pedunculo 0.3-0.7 em. longo glabro ; pedicellis tenui- bus ca. 2 em. longis glabris ; bracteis minimis. Flores virides ; calycis laciniis ovato-lanceolatis acuminatis са. 0.3 em. longis glabris; corolla rotata ca. 3 em. diam. omnino glabra, lobis late ellipticis ea. 1.5 em. longis margine albo ; corona acute 5-gona plana laevi corollae annulum earnosum minute puberulo-papillatum aequante; gyno- stegio stipitato acute 5-gono ca. 0.35 em. diam., antherae appendici- bus dorsalibus anguste ligularibus ca. 0.2 em. longis apice furcatis patulis.—cocré: vicinity of La Mesa, north of El Valle de Antón, alt. about 1000 m., April 12, 1941, P. H. Allen 2366 (Herb. Missouri Bot. Gard., TYPE). This species of the subgenus Eugonolobus is outstanding not only because of the ligular, forked anther appendages which suggest its name, but also because of the variable leaves. BORAGINACEAE BOURRERIA SUPERBA Johnston var. glabra Schery, var. nov. Ab specie differt foliis utrinque glabris et calycibus glabris loborum marginibus intus tomentosis exeeptis.—nocas DEL токо: Isla Colón, June 3, 1941, H. von Wedel 2472. This variety resembles the species except that the lower leaf surface and the inside of the calyx tube are entirely glabrous. The margins of the calyx lobes, however, are white-tomentose on the inside. The large size of the flowers (about 4 em. long, 4 em. wide) distinguishes this species from other species previously known from the region. B. superba is known from Mexico (Michoacan). SOLANACEAE Capsicum STANDLEYANUM Morton—socas DEL TORO: vicinity of Chiriquí Lagoon, Oct. 23, 1940, H. von Wedel 1316. Previously known from Costa Rica. This species as represented by the von Wedel specimens may not be distinct from C. stenophyllum Morton & Standley. The leaves are strongly dimorphie, resembling those of Solanum diphyllum. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 367 BIGNONIACEAE ScurEGELIA fastigiata Schery, sp. nov. Frutex scandenti-epiphy- ticus; ramulis teretibus brunneis rugosis parvis cum lenticellis pro- minentibus ovalibus albis; foliis magnis glabris coriaceis ellipticis apice aeutis basi rotundatis vel acutis in petiolo decurrentibus, nervis reticulatis supra aliquid impressis inconspicuis subtus pro- minentibus nervis lateralibus arcuatis ad margines, aliquid confluen- tibus, 12-25 em. longis, 4-10 em. latis; petiolis crassis, supra subeanalieulatis, са. 1 ст. longis, 3-4 mm. latis; inflorescentiis terminalibus vel subterminalibus crassis condensatis fastigiatis ses- silibus, 2-7 cm. longis, 5-10 cm. latis, plurimis subeorymbiformibus racemis compositis; pedicellis 4-8 mm. longis, 0.5-1.0 mm. latis; calycibus plus minusve 4-lobatis vel subtruncatis cylindrico-cam- panulatis reticulato-subrugosis glabris, 7-10 mm. longis; corolla glabra, 1.0-2.5 em. longa, 5-lobata, lobis ca. 2.5 mm. longis, ovato- triangularibus subhastatis imbricatis, tubo plus minusve cylindrico, ea. 8 mm. longo; staminibus 4, ca. 5 mm. longis, 14 longitudine tubi adjunctis; staminodio 1; filamentis glabris aequalibus linearibus basi latioribus; antheris bilocularibus, loculis longitudinaliter de- hiscentibus basi divergentibus; stylo crasso, са. 2 mm. longo apice aliquid баі; ovario sessili biloculari pluriovulato basi crasso-car- noso; fructibus globularibus purpureis lepidotis, circiter 14 mm. diametro.—nB0cas DEL TORO: Water Valley, Sept. 14, 1940, H. von Wedel 773 (Herb. Missouri Bot. Gard., түре) ; Isla Colón, July 26, 1940, H. von Wedel 154 (Herb. Missouri Bot. Gard., coryrr) ; Water Valley, Oct. 31, 1940, H. von Wedel 1447 (Herb. Missouri Bot. Gard., COTYPE). This species seems to be most closely allied to S. lawrancei Standl. and S. dariensis Sandw. (ex char.). Like S. dariensis it seems to be intermediate between short, lateral inflorescence types (sect. Pa- rantanaecium K. Schum.) and long terminal inflorescence types (sect. Euschlegelia K. Schum.). The species differs from S. lawran- cei chiefly in having a very characteristic inflorescence and more narrowly elliptic leaves, and from S. dariensis (ex char.) especially in having elliptic rather than broadly ovate, more or less cordate leaves. The most striking and characteristic feature of this species is the dense fastigiate inflorescence which resembles ‘‘witches broom."' The many pedicels and peduncles are so congested that it would be impossible for each to bear a flower simultaneously. Leaves are elliptic, short-petiolate, acute to rounded at the base and somewhat [Vor. 29 368 ANNALS OF THE MISSOURI BOTANICAL GARDEN decurrent into the petiole, acute apically. The flowers are reported as red (calyx) and white (corolla). The calyx is subtruncate or shallowly 4-lobed, glabrous within and without. The corolla is gla- brous, small, 1.0-2.5 em. long, 5-lobed, the lobes being imbricate, ovate-triangular, subhastate. The tube is about 8 mm. long and bears at about 3 mm. from the base 4 stamens and one staminode. The filaments are glabrous, linear, slightly broader at the base than above. The anthers are bilocular, the locules dehiscing longitudi- nally and diverging from one another basally. The style is stigma- tose and slightly bifid apically. The ovary is 2-celled, many ovules being borne in each cell from a central placenta. ARRABIDAEA CHICA (Н. & B.) Verl.—socas DEL TORO: Water Valley, Sept. 19, 1940, H. von Wedel 860. Previously known from British Honduras and Guatemala, possibly other countries. LENTIBULARIACEAE (C. V. Morton) UTRICULARIA Expresu Rehb. f.—cocré: hills north of El Valle de Antón, alt. 1000 m., Sept. 1, 1941, Paul H. Allen 2704. Previously known only from Costa Rica. RUBIACEAE (P. C. Standley) ISERTIA HYPOLEUCA Benth.—socas DEL TORO: Nances Cay Island, H. von Wedel 580, Sept. 2, 1940. Previously known from Colombia and other South American countries. GUETTARDA CRISPIFLORA Vahl.—zocas DEL Tono: Isla Colón, H. von Wedel 508a, Aug. 20, 1940. Previously known from Costa Rica and the West Indies. ALLENANTHUS EnYTHROCARPUS Standl.—This well-marked genus has been known from the single type collection, and still has been collected only in the region of the type. An additional collection of it is worth reporting: cocré: Region of El Valle de Antón, trail to Las Minas, alt. about 700 m., Sept. 1941, P. H. Allen 2713. A tree of 8-25 m. ; fruits pink to red; trees abundant and conspicuous along the upper reaches of the Río Antón. Корсвл Sxutcun Standl—cocrí: hills north of El Valle de Antón, trail to La Mesa, alt. about 1000 m., Aug., 1941, P. Н. Allen 2699. A shrub 3 m. tall, the flowers white. Described from the vicin- ity of El General, Costa Rica, and known heretofore only from the original collection. 1942] WOODSON & SCHERY—FLORA OF PANAMA. VI 369 CUCURBITACEAE SELYSIA PRUNIFERA (Poepp. & Endl.) Cogn.—socas DEL TORO: Water Valley, Nov. 13, 1940, H. von Wedel 1596. Standley reports that this is the first record of this genus north of South America. The species was previously known from Peru. [Vor. 29, 1942] 310 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 30 Figs. 1-3. Sobralia не І. po fig. 1, plant, x №; fig. 2, flower with segments expanded, x 1; fig. 3 Figs. 4-7. ааны а. L. Wms.: fig. 4, plant, х №; fig. 5, lip, х 5; fig. 6, lip and eolumn from the side, x 5; fig. 7, petal, x 5. ANN. Mo. Bor. GARD., Vor. 29, 1942 PLATE 30 \ SOBRALIA „Аф ели * 7! tf}: /{ ty 1 4 \ ү! ? 4 ¢ NM Г. Кл "S 1 IN SS hs "MG r УСА Á (S 1 N y K А К 5 К | бе р “ЖМ, = S \ ( Ж 24 Z \ NAN | " f SSS \ S | —S WAS - в D DN RAN THUS ` с -— >= J - я xm ў WOODSON & SCHERY—FLORA ОЕ PANAMA [Vor. 29, 1942] 312 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 31 Sobralia decora var. aerata Allen & Williams: fig. 1, plant, x !4; fig. 2, lip expanded, X 1; fig. 3, lateral sepal, x 1; fig. 4, petal, x 1; fig. 5, dorsal sepal, x 1. PLATE 31 ANN. Mo. Bor. GARD., Vor. 29, 1942 — 772 А $ MOM ars Bate E Ss SA aeng га ЗОВЕАПТА десога var. WOODSON & SCHERY—FLORA OF PANAMA 374 [VoL. 29, 1942] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 32 Figs. 1-3. Stelis Allenii L. Wms.: fig. 1, PS X №; fig. 2, column, petals and lip seen from the front, x 8; fig. 3, flower expanded, x 2 Figs. 4-8. Stelis atrorubens * Wms.: fig. 4, plant, x №; fig. 5, flower expanded, x 5 fig. 6, lip from ihr x 10; E Bus v X 10; fig. 8, petal, x 10. Figs. 9—12. Wm : fig. 9, flower exp: "P: x 215; above, x 10; fig. 11, lip from the ifie x Ст; & 12, petal, x 5. Stelis mo ntana a ; fig. 10, lip from ANN. Mo. Bor. GARD., Vor. 29, 1942 PLATE 32 7. цу SLE ZZ 4 ЖЛ, / М LL Ае АД JN (b Si 3 Ñ кае Уу" . б NN in Н 1 [И | T | / + Ili Ц I IUE d Г ПЕ: л И ; | 1 Lyi Y | i |, / H 7, \\ ЛУ YS. montana Да | \ И j РА WOODSON & SCHERY—FLORA OF PANAMA [Vor. 29, 1942] 316 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 33 Figs. 1-7. Pleurothallis ellipsophylla L. Wms.: fig. 1, plant, x %; fig. 2, flower from the side, x 215; fig. 3, dorsal sepal, x 214; fig. 4, lateral sepals, x 2%; fig. 5, petal, x 2%; fig. 6, lip and column from the side, x 5; fig. 7, lip from above, x 5. Figs. 8—10. "nc aui cardiochila L. Уф fig. 8, plant, x 15; fig. 9, flower expanded x 1%; fig. 10, lip, x ANN. Mo. Bor. Garp., Vor. 29, 1942 PLATE 33 PLEUROTHALLIS /f f, ا‎ / p ОД аз ре M И «=. = N 7 Р cardiochila < ASI misa ` Y Kit An N WOODSON & SCHERY—FLORA OF PANAMA (Vor. 29, 1942] 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 34 Platyglottis coriacea L. Wms.: fig. 1, plant, x 1; fig. 2, column and lip from the side, x 3; fig. 3, column from front with the lip bent down, x 3; fig. 4, dorsal sepal, x 3; fig. 5, petal, x 3; fig. 6, lateral sepal, x 3; fig. 7, anther from below, x 10. ANN. Mo. Bor. GARD., Vor. 29, 1942 PLATE 34 > SN Po | A Sz j = СТЯ йо WOODSON & SCHERY—FLORA OF PANAMA GENERAL INDEX TO VOLUME XXIX New scientific names of plants and the final members of new combinations are printed in bold-face type; synonyms and page numbers having reference to figures and plates, in italics ; and previously published names and all other matter, in ordinary type A Achania, Ыы; oe a, Е coccinea, 209; concin cordata, 231; floridana, 231; Ma eni 209; mollis, 20 losa, 209; LI 231; stylosa, 931; е of, Апо етае panam ensis, 36; е 9i и carboniferous floras, ae ns to our ni pele of: I. Scler Бела gen. no oat on and E ORNL IL Lepiioeerpon, LOTES Stipi- topteris, 59; IV. A new spec ie s of Lepi- dodendron, 245; V. Berop, 275 0 and Hugh C. Cutler aces of Zea B MIS I. Their вов and classification, 69 Contr inia to our erous ras. II. Lepidoearpon, она Arancarian wood 19; A from western Wy , 283 ЖОЛДА. 46; chihuahuensis, 147; flori- bunda, 15 52; geniculat > 147; Greenmanii, т Karwi rwinskyana, ; linearis, 148; pulehella, 148 Ankyropters is corr ugat Anotea, 204; chlorantha, E flavida, 231 Кол аи pube from NES ак Wyoming, A. fossil, 28 Zona rici а wyomingense, 284, 28 oretum, see Missouri Botan nieal Gar- den Wildflower Reservation Archacologieal research on prehistoric n, E ide кч 298; caracasana, 302 29 6; fri igida, 307; longifolia, 3 309; tenuifolia, 305 epee n аа 365 Athyrocarpus, 150 Pe folia, 07; offici- B Besleria Allenii var. paucivenia, 38; bar- bensis var. hirsuta, 38; crassicaulis, 39; 40 rai RPO Panamanian, 366 Botrychioxylon, 3; paradoxum, 4 Bourreria superba var. glabra, 366 Brenner, Louis G. Jr. The environmental variables of the Missouri Botanical Gar- den ildflower Summit, 103; and Ral and Jose eph Wraight. Dolomitie glades of east-central Missouri, 89 Brinker, ма К. Мой ак of Schoe- nocaulon, 28 Bumelia, A r sd ision of the genus, in the United States, 155 Bumelia, A атт ма, 178; аси 177; an а, 169; arachnoidea, 163 borea, 162; po aS chrysophyt loides, 168; onfertiflor ‚ 178; ida, 165; lucida, 172, cid. Iyeioides, 170, var. ellipso oidalis, 172, ar. reclinata, 1 73, var ; macroca нау 173; carpa, 173; ранет 167 £s Schotti, 77; rata, serrulata Sm. ED spi- nosa на и Are. tenax, 168; ' texan a, 175; tomentosa, 163; undu- ata, 179 Burmanniaceae, Panamanian, 336 C Caetaceae, Panamanian, 363 а Allenii, 331; ' Allouia Mar 2; dasycarpa D се 333; lagunae, 33 3 M rcu 154; fr OAM 154; grandiflora, viol a, 232; Gan aceae, Panamanian, 331 Capparidaceae, Panamanian, 351 Capparis clara Carboniferous floras, American, Contribu- tions to our know ledge of: I. Scleropte- ris, gen. nov. Mesoxylon and Amyelon, Dp Lepidocarpon, 19; TII. St tipitopte- ris, 59; IV. An W species o of Lepidoden- dron, das, У. Het eterangium, 275 Caricaceae, Pan namanian, 36 Carludovica integrifolia, ыы о 323 Cattleya Skinneri var. au alis, 34 5 aulopteris, 61; varians, aL Ceiba Allenii, 359 (381) 382 Centrosolenia MN 41 Chloranthaceae, manian, 350 sion gg phyllum ARRA, 168; ludovician- Citronella, mem nag ê kT I پیر‎ on, Clar ыы Brown. A revision of t gen sid. in the United States, iss с торт а, 362 — uw of Illinois. flora of, 1, 19, 59, 275 Cochllostem ma, 149 Columnea, 35; Allenii, 42; arguta, 43; cit- Е es crassa, 45; darienensis, 46; dissimi- lis, 47; MN ри pu ; incam rt i 40; мае rubra, 48, 52; silvarum ‚53% Commelina, 149; anomala, 149; Blainii, 150 Commelinaceae: Commentary on the North American genera of, 141; Panamanian, 324 Commelinantia, 149; anomala, 149 Contributions to our knowledge of Am at — floras: I. ora , Mesoxylon and Amyelon, 1; II. Lepidoearpon 19; m LL Stipitopteris, 59; IV. A new species ‘of Lepidodendron, 245; . Heterangium, 275 Contributions toward a flora of Panama, VI. Collections qr by H. von Wedel —-— Morphology. of the plant, 72; Guate- Big Grains, 81; Guate malan Trop- i midal, 80, 82; Pima-Papago, 34 Croizat, L., Panamanian Писна RR de- termine d by, 353 Cryptochloa, Т, concinna, 320; granuli- fera, 321; strictiflora, 321; variana, 817, 319 Cryptophoranthus lepidotus, 340, 370 Ctenanthe dasycarpa, 33 eurbitaeea e, Panamanian, 369 on Li, Адна manian Cactaceae deter- by, Cuth beri 151; rosea, 1 Cutler Вл A C., Mar Anderson and. R Ta V I. Their recognition and ква test, Cyclanthaceae, Panamanian, 322 Cymbopogon Nardus subs sp. genuinus and C. citratus, porn бет on, 137 D Descantaria, 150; amplexicaulis, 152; an- gustifolia, 152; cumanensis, 152; ' Dis- grega, 152; el ongata, 152; minuta, 153; multiflora, ЕЙ Palmeri, 153; procum- 3; saxicola, 15 Асен с, ае, (зада И 353 Diehapetulum axillare, 353 [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN Diehorisandra, 145, 146; hexandra, 146; longifolia, 147 радом, Рапатапіап, 361 Гіовеотеасеае, pice 327 Disgrega, d mexicana, 152 Dithyrocarpus, 149 Dodge, Carroll William. Helminthosporium spot of citronella and lemon grass Guatemala, 137 Dolomitie glades of east-central Missouri, of, 9 89; map o Donne llia, 151; ; ГРЧ lata, 55; marmor a parvifolia, 56; rosea, 56; turrialbae, 5i E Echeandia it d 325; venusta, 325 Endosporites, 2 Environmental pane of the Missouri otanical Garden Wildflower Reserva- tion a Yu The, 103 түн Ralph О., Louis С. Brenner and eph Wraight, Tomis glades of Missouri, 89 SOMME Panamanian, 353 F Ferns, fossil, 3, 59, 275 Flacourtiaceae, i Ц 363 Floseo[ Fossil Arauearian wood from western Wy- omin Fossil 2 1, 19, 59, 245, 275 G Gametophyte, fossil, 3, 21, 2 Gesneriaceae, New, from лавина 35 Glades: associes at ur Summit, 132; dol- omitie, of east-central Mies, 89, 95, 96, flora of, 90, geology of, 92, map of, Gonolobus Кен a т ramineae, Pana rass, lemon b pe reo С Helmintho- aper ium —— 7 mit, The environmental vari- ables of re. Missouri Botanical Garden Wildflower Reservation at, Gua ar py Helmint hosporium spot of cit- ronella and lemon grass in, 137 Guttiferae, Panamanian, H Hapalanthus, 154 — вар ау. spot of citronella and emon grass in Guatemala, 13 Helminthosporiam Cymbopogi, 139, 140, Sacchar тон, 22; я она, 298; (o us 302 Hem a, 150; multiflora, 1 Heterangium, 3, | americanum 276, 277, 282; subgen. Eu- angium, 280; “Grievii, 975; Hop petondterl, 281; 1942] INDEX Kukuki, 281; subgen. IS Del 280; shorense, 281; tili o Е , 281 me "hg croftianu ‚ 231; cocci- neus, 230; Dru ; floridanus, 231; pura 2307 лы, 230; Malva- viscus, 209; Po oeppigii, 231; racemosus, 210; rosa-sinensis, 230 Hibiscus, 203 Hippocrateaceae, rna 357 How namanian Щаба cud by, 358 I Icacinaceae, more 358 Illiniocarpon, 1 9 ‚ 19 Illino КЕЧ, йога ОК 1, 59, 245, 275 Inflorescence structure in Commelinaceae, 143 Iridaceae, Panamanian, 327 Ischnosiphon Pittieri, 335 K Kegelia Houtteana, 347 Kegeliella Houtteana, 347 L Lacistemaceae, ЕТ 350 1518, 358 Of "Le epidodendron scleroti- eum, 256 2o Lepidophyllum m, 258 Leguminosae, Panamanian, 352, 368 bn T Werl Panamanian, 368 Lenz, L. Wayne. bius tions to our erican earboniferous 22; Lomaxi, И 98-34, 259, gameto- " "Ti p of, 23; Wildianum, 22 oy Ф JS = > о >. Ф 5 Qu = o + 27 B EE MI x Ф e = Фф un © ов л xou cond Brownii, 250; в 254; fuliginosum, ' 249; Harcourtii, 24 2 in "ice 24 9; Johnson p obovatum hodumnen , 25 sum, 254; ' sclero rope 249, 25 E 262 24; sela egi oides, 249, 255; Volkmannianum, 256 Lepidophloios 256; larieinus, 256, вр., 274; sck um 949, 2: Lepidophyllum , 2; Thomasi, 258 Lepidostrobus " Bertrandi, 2 0; Brownii, 249, 250; Coulteri, 26; Veltheimianus, 22 Leptorhoeo, 51; filiformis, 152; flo ri- 152 тано ааа. 27 Loganiaceae, P 1, 364 к OH RA. EA lanuginosum, ; spinosum, 170; tena, 168 fd 22, 24; obscur um, 21 Lycopods, 2, 19, 24 5 Lyginopteris, T M Maize: ge qm of, 69; morphology of, 72; ear a tassel characters for five races Uu 82, 88; Guatemalan Big Grains, 81; ears deu tassels, $8; Guatemalan 383 M nd Flints, 81, ears and tassels, 88 ; D n Pyramidal, 80, ears and tassels, Papago ‚ 84; Pueblo ‚ 84 Е Panamani 52 Malvaviscus, dU ie 0 е. 183 Malvaviseu , 203; pulcensis, 209; var. Drummondii, 215, 2 pilosus, Sloanei, 219. atus, 219; bre evipes, 219; can 207 ; : ciliatus, 219; chloranthus, 231; cine- reus, 230; coccineus, 230 Cokeri, 214; concinnus, 209, 230; Coneattii, 223; cor- tus, 210; cordifolius, 209; cuspidatus, 218; Cutteri, 210; Drumm ondii, 215; ele- ‚ 218; flavidus, 2 231; tres tid glabr Malvavis 222; ход 208; И ‚ 223; ovid дайыр; 219; pi- 210; pleurantherus, 231; ppigii, 231; Polakowskyi, 219; popultfoliue , 230; po opulneus, 230; i, ; pulvinatus, 219; puniceus, 230; тим daris, 219; rosa-sinensis, 230; бадтаеатиз, 219; Wem, 6; spathula- из hi ; Ulei, 218; velu- tinus, 210; Williamsii, 226 Marantacene, Panamanian, Ar Mar e, Р: Maxillaria conduplicata, E Nn 348; Pittieri, 349 Meli aceae, Panam Mesoxylon, 3, 7, атруту of, 3, 16 ; Nauerti- anum, Microsporangiate cones, fossil, 24 Missouri, east-central, Dolomitic glades of, 89 Missouri Botanical маш, cune Reservation at Gray Sum tions and associes, 124; Pate n 114; Environmental variables of the, 103; geology, 104; history of m. и of the area, 118; soil survey of, vegetation of, 121 мое: Maxonii, 58; panamensis, 58; puberu мро "Panamanian, 350 Morton, C. New Ead M Panama, 35 "Panamanian plant : 361; Di Lentibulariaceae, reaceae, 327; 384 рона», 326; Turneraceae, 363; Violaceae, 362 Mo rtoniodendron hirsutum, 359 147 vee “Pa anamanian, 327 роке and the origin of roots, 5 Myrosma dasycarpa, 335 N Neodonnellia, 151, grandiflora, 153 Ochnaceae, Sage an 361 320; нь, 321 ier fecta fails of, 89 P Panama: Contributions toward a flora of, VI, 317; New Gesneri ean саг- 0 II. езана 19; A foss Калма un wood from western Wyomin ng, 283 Pavonia, 227; amplifolia, 230; coccinea, 231 Dru иттопди, 216; Ягш iflora, мө, 28; longifolia, 230; montana, 230; йога, 230; Ра almeri, 229, 230; no. ши 353; aperta, 353; subg. Eupera, 357; subg. Gymnoptera, 357 Perebea hispidula, 350 Pe Ria E о 255 Petalox Petioles,. ен; ‚ 59 Phaeosphaerion, 150 Pima and Papago Indians, eorn grown by, RECTA 345, ши coriacea, 347, 378 Pleiostachya Pittie ‚ 335 — крака ‚ 341; cardiochila, 343, 3 Кенет 344, 376 ченче нй fossil, 275 Pueblo Indians, corn grown by, 78 Quiinaceae, Panamanian, 358 Races of Zea Mays, 69 Raddia concinna, 320; strictiflora, 321 Rectanthera, 154; fragrans, 1 Renealmia P naria, 329 oots, fossil, w Rubiaceae, а, 368 [Vor. 29 ANNALS OF THE MISSOURI BOTANICAL GARDEN 8 Sabadilla Mr Rp 296; Coulteri, 297; Drummo , 298; dubia, 298; officina- rum, Sapindaceae, Panamanian, 358 Sauvallea, 149; Blainii, 150 Sehery, Robert Ла Monograph of Malva- viscus, 183; t E. penc Jr. = Contributions o. flora of 'Pan VI. Collections chiefly by Н. уоп Wedel in Boeas del Toro, 317 Sehlegelia fastigiata, 367 Sehoenoeaulon, Monograph o Schoenoeaulon, 292; aletroides prot ealei- eola, 295; caricifo 295; comatum, 296; Conzattii, 297; Obiter, 597; Drum- mondii, 297; Drw mmo пай ‚ 305; домат, 208; Ghiesbreghtii, 299; graci 03; nue, 304; tenuifolium, 305; nein i 305; yuca atan nense, 307 Sclerocladus tenaz, 1 Scleropteris, gen. nov., Mesoxylon and Amyelon, Scleropteris, 3; MES 8, 12, 14, 16 Sclerozus tenax Sideroxylon bette: 3 168; deoan- drum, 170; laeve, 170; lanuginosum, 161 lucidum, 178; lycioides, 1705 ide iie. 173; sericeum, 16 пах, 161, 168 Rigillaria Boblayi, 250 Skoinolon, 293 Smilacaceae, Panamanian, 326 Smilax chiriquensis, 326 Sobralia Allenii, 336, 870; decora var. aerata, Soil survey of Missouri Botanical Gardes r Reservation, 109 ceae, San manian, 366 басу ab T. Spiranthes e 337; Woodsonii, 33 Spironema, 154; fragrans, 154; Warsce- wiczianum, 154 Standley, Р. С., Panamanian plants deter- mined by: Moraeeae, 350; Quiinaceae, 358; [Ne Aen 368; Tiliaceae, 359 Stelis ' Allenii, : 338, 374; atrorubens, 374; montana, 37 74 Stemmadenia — 364 ee Е 339, фа п, 2 65, 245, 251, 262- Stenanthiu рет 1 307 tereuli: incene, arama Bo 360 Stipitopteris, 59; americana, 63, 65, 66, 68 1942] INDEX Strephiwm qe 321 Swallen, Jas , Panamanian Gramineae retarted mo 317 ay Loisirs eei абе. 360 Thapa a populnea, 230 Thyme "A aceae, у 364 19, 147; cumanensis, 152; ke mis, 152; fl 14: 147; а 148; 147; Lundellii, 153; macrophylla, 53; multiflora, 153; Pal- meri, 153; procumbens, 153; pulchella, о Tripogandra, 150; amplexicaulis, 152; an- gustifolia, 152; cumanensis, 2; 'Dis- grega, 152; elongata, 152; floribunda, 152; grandiflora, 153; Lundellii, 153; minuta, n multiflora, 153; Palmeri, 153; rose 153; saxicola, 154; Warsce- wicziana, “154 Turneraceae, Panamanian, 363 у Variability, inter-racial, in maize, 72 Variables, The environmental, ar the Mis- 385 souri а Garden Mi ad Res- пра ion at Gray Summit m 908; Цена, 396; ; frigidum, 307; Да оси 302 tenuifolium, 05 Violaceae, Panamanian, 362 W von, d rra chiefly by, i F Toro Wildflower i ud at Gray Summit, аа f the, 103 manian Orchida- W ied 236 1 Arauearian, from western Wood foss Wyoming, 2 Woodson, Robert Commentary on the North American gen of Commeli aceae, 1 2 and W. Sehery, Con tributions toward a flora os ш VI. Collections. Hs E Ll H. von Wedel in Boeas del Toro Wraight, оа "Ralph O. Erickson, Louis G. Brenn d Dolomitie glades of east- central мъжа 1, 89 Wyoming, ester A fossil Araucarian wood from, 283 X Xylem in wood of Lepidodendron seleroti- cum, 247 Zea Mays, Races of: I. Their recognition and classification, 69; ears and tassels of three races, aud tassel branches Pip Peru ,76 ‚апа from Me ingiberaceae, [MANU 329 2 par is, 3, 7