Dodota JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER Axis eee ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. ^ VOLUME 70 1968 The Nef England Botanical Club, Inc. ' Botanical Museum, Oxford St., Cambridge 38, Mass. FARLOW REFERENCE LIBRAR X 72 JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL P STUART KIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER | Associate Editors ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. Vol. 70 January-March, 1968 No. 781 CONTENTS: Comparisons of Sexual and Apogamous Races in the Fern Genus Pellaea. Alice F. Tryon .............. certet 1 On the Names in Fraser's 1813 Catalogue. JOMES d geet 5 iaire teni e aN ELEME S 25 A Cytotaxonomic Study of the Herbaceous Species of Smilax Section Coprosmanthus. (to be continued) "E e ET E EE TEPORE ER 55 Variation in Cone Morphology of Balsam Fir, Abies balsamea, Donald T. Lester ......... eese 83 A Cytotaxonomic Study in Verbesina (Compositae). James TE, COLMAN eoe ee másaoetonertha enin ati guten EE 95 New Records of Disjunct Arctic-Alpine Plants in Montana. S. A. Bamberg and BR. H: Pemble Kasasi 108 De Plantis Toxicariis E Mundo Novo Tropicale Commenta- tiones II. Richard Evans Schultes and Bo Holmstedt .... 118 The Nem England Botanical Club, Inc. Botanical Museum, Oxford St., Cambridge, Mass. 02138 Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 70 January-March, 1968 No. 781 COMPARISONS OF SEXUAL AND APOGAMOUS RACES IN THE FERN GENUS PELLAEA ALICE F. TRYON Apogamy in ferns is of the obligate type involving changes in the gametophyte eliminating fertilization by absence or non-function of the archegonia. This is correlated with a compensating system in the sporophyte, circumventing the ređuction in chromosome number by changes in the mitotic or meiotic divisions in the sporangium, prior to spore for- mation. Th*s both gametophyte and sporophyte stages have the same chromosome number. This synchronized apogamous system is particularly remarkable for it evident- ly has been independently initiated more than once in 25 of some 209 genera of the Filicopsida which are cytological- ly known. This study in Pellaea, on some morphological modifica- tions which are associated with apogamy, was made with the purpose of understanding the evolutionary role of the mechanism in this group, against a background of rapidly increasing information on the frequency of the phenomenon in the Pteridophyta. Apogamous ferns are considered to be of hybrid origin by Manton (1950, 1961) and Walker (1966) and there is support for this view in the following survey of chromosome levels and the apogamous condition. Data are drawn from the reports of chromosome numbers in the Filicopsida by Chiarugi (1960) and Fabbri (1963, 1965). An analysis of these records shows apogamy in 32 diploids in 9 genera, 1 2 Rhodora [Vol. 70 66 triploids in 15 genera, 15 tetraploids in 6 genera, 5 pentaploids in four genera and one hexaploid and one octo- ploid. About 73% of these apogamous forms are at polyploid levels while 27% are apparent diploids, although the latter may not represent base numbers of the genera. The large proportion of apogamous forms at polyploid levels supports the hypothesis of a relationship of apogamy with hybridization. However, the percentage of diploids is a significant one and the origin of apogamy in these is not apparent. In some genera, as Pteris, apogamy is especially frequent, and it is reported by Walker (1962) in 16 diploids, 14 tri- ploid and 9 tetraploid forms. In P. cretica apogamy occurs in diploid, triploid and tetraploid forms and it also occurs in P. biaurita at diploid and triploid levels and in triploids and tetraploids in P. quadriaurita. In Adiantum, apogamy is reported in five diploid, three triploid and one hexaploid member. In Dryopteris five diploids and 16 triploids are apogamous, and in D. Borreri it is known at diploid, triploid, tetraploid, and pentaploid levels. In contrast to these, there are some large genera such as Asplenium, in which apogamy is infrequent with reports of only three triploid, one pentap- loid, and two octoploid members, and it is not, or doubtfully, reported in Blechnum, Cyathea and Thelypteris. Thus, while apogamy is prevalent in some groups and absent in others it is notable that it occurs in a number of genera which are only distantly related. Some information on the origin of apogamy in ferns comes from experimental studies on induced apogamy in gametophytes in which there appears to be a genetic pre- disposition for the condition. In these, apogamy is influenced by environmental changes such as drying, light intensity and in increased concentration of certain sugars. In some of these studies, as those by Whittier (1964), it is recognized that the induction of apogamy is more readily accomplished in certain “races” of a species than in others. The genetic control of apogamy as a dominant is demonstrated by ex- periments on Pteris by Walker (1962) where crosses made 1968] Pellaea — Tryon 3 between apogamous diploids and sexual species produce only apogamous forms. Dominance may be effective in per- petuating and increasing the number of apogamous forms by crosses involving antherozoids from apogamous plants. This may account for the frequency of apogamy in certain groups as Pteris, Adiantum and Pellaea. There have been several studies of apozamy in Pellaea and it is known in six of the 15 species in section Pellaea. The triploid species P. atropurpurea, reported by Manton (1950) and a triploid variety in P. sagittata, noted by Tryon & Britton (1958) appear to be of hybrid origin. Apogamy is reported here in three species of section Pellaea belonging to the “light-stiped” group which represent a separate evo- lutionary line in the section. This record of apogamy in a distinct species-group is in harmony with reports of its occurrence in different genera and is indicative of inde- pendent origins of the mechanism. These three closely re- lated species — P. andromedifolia (Kaulf.) Fée, P. interme- dia Mett. ex Kuhn and P. ovata (Desv.) Weatherby are exceptional in the ferns for each has both sexual diploid and apogamous triploid races. The triploid chromosome level of the apogamous types shows a relationship to hybridiza- tion. The morphological similarity of the sexual and apog- amous races in each of the species is suggestive of a close relationship between the putative parents. Comparative studies were made of the chromosome numbers, gameto- phytes, juvenile leaves, guard cell size and spore size for the two forms in each species. Comparisons are also drawn with a sexual diploid and sexual tetraploid in P. ternifolia (Cav.) Link var. ternifolia (later referred to by the species name only), also a member of Section Pellaea but belonging to a *dark-stiped" group, as a basis for comparison of polyp- loidy without apogamy. Geographic distributions of the races are compared as a measure of the evolutionary success of the different forms. Most of the apogamous members in Section Pellaea are widely distributed in North America or also South America while most sexual members have more limited ranges. Plate 1371. : E E : Fig. 1. a-c. Pellaea andromedifolia, squash preparation of meiosis. a. Sexual diploid, n = 29, San Gabriel Mts. Calif., X 650. b. Apoga- mous triploid, “n” = 87, Humboldt Co., Calif., X 1650. c. Apogamous 1968] Pellaea — Tryon 5 Material and Methods Measurements and cytological reports in this study were obtained from original collections, or plants grown from spores obtained from them, maintained in a greenhouse for a period of two to six years, Field studies and collections were made with R. M. Tryon in the western United States and Mexico in 1950, 1956 and 1957. Collections from the first trip are deposited at the Missouri Botanical Garden. Some duplicates of these, those from the last two trips, as well as cytological vouchers, are deposited at the Gray Herbarium. Meiotic chromosome counts reported here were obtained by the standard fixation in 3:1 alcohol and acetic acid and stained in acetocarmine as described by Manton (1950). The same fixative and stain were used for the mitotic cells from root tips and this material was treated with snail cytase following the procedure reported by Fa- bergé (1945). The snail enzyme breaks down the middle lamella and allows the chromosomes to spread freely when pressure is applied. Distribution records were plotted from specimens deposit- ed at the Gray Herbarium, the New York Botanical Garden, the United States National Herbarium, the California Academy of Sciences, Pomona College, Rancho Santa Ana Botanic Garden, San Diego Natural History Museum, Stan- ford University and the University of California, Berkeley. Comparisons of Apogamous and Sexual Forms Chromosome Numbers Plants of the sexual type of Pellaea andromedifolia, from the San Gabriel mountains in southern California (Ait 3) and from Humboldt County in northern California (Tryon & Tryon 5557) have the normal complement of 16 spore mother cells and show at diakinesis that n = 29 with com- triploid, “n” = 87, Point Mugu, Calif., X 1300. d-e. P. intermedia, d. Mitotic cell from root tip, 2n = 58, Saltillo, Mexico, X 1300, e. Squash preparation of meiosis, apogamous triploid, “n” — 87, Chiri- cahua Mts., Arizona, X 2500. f-g. P. ovata, squash preparations of meiosis, f. Sexual diploid, n = 29, Gillespie Co., Texas, X 2000. g. Apogamous triploid, “n” — 87, San Luis Potosí, Mexico. X 650. 6 Rhodora [Vol. 70 9x * r me 4t e 7 Y d 7 P 4 d SC Jake w T eg, $ X “e sy Ke a e x 7 "e i e9 a b 5 Y > TA C 2 Ver S AX (op T$ t^ d N YW NO e. E e» $ "e Phry 7 A e KT +38 Ai ” e e Ha A » LS WE t d x ER) “4º, A + De: aw Sei d + E 4 a Ge E Ze eg oF ^ eg TH at 2$4* `. Let de "e ES "E Ka $ f g Fig. 2. Explanatory, enlarged diagrams for Fig. 1 showing chromo- somes in focus in black and white and the remainder in outline, a-c. Pellaea andromedifolia, a. Sexual diploid, n = 29. b. Apogamous 1968] Pellaea — Tryon 7 plete and regular pairing (Figs la, 2a). Apogamous plants from Humboldt County in northern California (Harris 21651) resemble the sexual plants in leaf morphology, have sporangia with eight spore mother cells and at diakinesis show that “n” = 87 with complete and regular pairing of the chromosomes (Figs. 1b, 2b). Apogamous plants from Ventura County in southern California ( Kiefer 1132) differ from apogamous plants from Humboldt County in having narrower leaves and pinnae of a more coriaceous texture. They have eight spore mother cells per sporangium and at diakinesis “n” = 87 with the chromosomes completely paired (Figs. 1c, 2c). Plants of Pellaea intermedia grown from a collection from northern Mexico, east of Saltillo (Rollins & Tryon 58144) have 64 spores per sporangium. Mitotic cells from the root tips show chromosome figures of 2n — 58 (Figs. 1d, 2d). Apogamous plants, morphologically similar to the sexual ones, grown from a collection from the Chiricahua mountains in southern Arizona (Tryon & Tryon 5089) show "n" = at diakinesis and complete pairing (Figs. le, 2e). No 16-celled sporangia were observed. Plants of Pellaea ovata of the sexual type which were grown from a specimen collected in Gillespie County in cen- tral Texas (Tryon & Tryon 5029) have the normal comple- ment of 16 spore mother cells and the haploid chromosome number is — 29 (Figs. 1f, 2f). Apogamous plants grown from a collection from northern Mexico, near San Luis Potosí (Rollins & Tryon 58222) in leaf morphology closely resemble plants of the sexual type; only eight spore mother cells were observed and at diakinesis “n” = 87 chromosomes (Figs. 1g, 2g). These cases of apogamy in triploid plants of Pellaea are similar to the classic example reported for Pellaea atropurp- urea by Manton (1950) and also by Tryon and Britton triploid, “n” == 87. c. Apogamous triploid, “n” = 87. d-e. P. inter- media, d. Sexual diploid, 2n = 58. e. Apogamous triploid, “n” = 87. f-g. P. ovata, f. Sexual diploid, n — 29. g. Apogamous triploid, n" rus 87. 8 Rhodora [Vol. 70 (1958) in P. sagittata. They are, however, unique in the occurrence of both sexual diploids and apogamous triploids which are of generally similar morphology, in each of the species. In P. sagittata the diploid var. cordata and the trip- loid var. sagittata are morphologically distinct — and in P. atropurpurea the triploid is the only known form, In these apogamous taxa, the problem which Manton alluded to as the “triploid dilemma” was explained in the case of P. atropurpurea as possibly arising as a back-cross between an allotetraploid and one of its diploid parents, largely on the basis of pairing relationships in the 16-celled sporangia. In the three species reported here each sexual diploid form has most probably served as a parent in the cross producing a morphologically similar triploid. The second parent might be a tetraploid as suggested by Manton or possibly an autotetraploid originating from different races of the same species. Tetraploids are not known in these three species although they are known in the section and two of these are apogamous. Another possible parent is an apogamous diploid, as was reported in the formation of apogamous triploids in Pteris by Walker (1966). Plants of this constitution are not known in Pellaea. This record of both sexual diploid and apogamous triploid plants in three closely related species is of special interest for it appears that the parental forms in each of them share a capacity for generating the synchronized mechanism necessary for obli- gate apogamy. Gametophyte and Juvenile Leaves Illustration of the sexual and apogamous forms for com- parisons of these stages in the life history is shown in Pellaea andromedifolia (Fig. 3). More detailed accounts of particular structures have been made by Atkinson, in Tryon and Britton (1958), Tryon (1960) and Nayar and Bajapi (1964). Spores of both sexual and apogamous types germinate rapidly, usually between 6-15 days, and the prothallia also develop quickly with as many as 26 cells and a small plate in ten days. The mature structure is cordate and erect. In 1968] Pellaea — Tryon 9 a b Fig. 3. Tracings of gametophytes and juvenile leaves of sporo- phytes, of Pellaea andromedifolia, at 6 months, X 5. a. Sexual diploid, gametophyte at base with several irregular lobes and the first two symmetrically lobed leaves of the sporophyte, from Tryon & Tryon 5557. b. Apogamous triploid, gametophyte and early leaf of sporophyte at base, the fifth leaf with deep, asymmetrical lobes, from Tryon & Tryon 5556. the sexual type antheridia are produced first, usually after 15 days, and may be formed on plants with as few as 20 cells. They are confined largely to the lower surface in the posterior region, nearest the spore, and are abundant among the rhizoids. Archegonia are formed later than the antheri- dia, on a cushion of cells 2 or 3 layers thick, on the lower surface in the anterior region, a few cells behind the central notch. Among the rhizoids of large prothallia there may 10 Rhodora [Vol. 70 be several small prothallia bearing only antheridia. These seem to be of the form recognized by Stokey (1951) as ameristic and their development may promote outcrossing in these plants. Juvenile leaves on sexual plants have symmetrical lobes; the initial ones bilobed and later ones with more and deeper sinuses (Fig. 3a). The prothallus may persist after the formation of several leaves with continued production of antheridia on new lobes. The apogamous type germinates slightly earlier than the sexual one. One or both sex organs may be suppressed but usually antheridia are formed. Antherozoids are abundantly produced and appear normally developed. Archegonia are usually absent but non- functional ones were observed in P. andromedifolia, after young “embryos” were developed. “Embryos” develop, in- dependent of sex organs, after two or three months, usually in the central portion of the prothallus, adjacent to the notch. They may be surrounded by slender, hyaline scales attached to the prothallus and young leaves. The first leaves of apogamous plants are often asymmetrical, and pinnate leaves resembling the mature form are produced earlier than on sexual plants (Fig. 3b). The different time of development and morphology of these structures in the apogamous forms appear to represent a repatterning of the similar growth systems in the sexual forms. Precise ecological data are lacking for most collections of these species; however, in Humboldt County, California where both sexual and apogamous plants of P. andromedi- folia grow together; the apogamous ones grow among rocks in exposed situations and the sexual often grow in humus in more shaded sites. The apogamous condition, eliminating the need for fertilization, would be advantageous to these species in occupying more xeric habitats. The evolutionary potential of the apogamous type is indicated by its adapta- tion to a different ecological niche. Epidermis — Stomatal Guard Cell Size The epidermal cells are of particular interest in these species since there is a correlation between the apogamous, polyploid condition and larger guard cell size. The epidermal 1968] Pellaea — Tryon 11 > > c f Fig. 4. Tracings of epidermal cell walls in Pellaea, X 80, of the abaxial surface with stomata, except c, adaxial surface. a-c. Pellaea ovata. a. 3X, Correll 22792. b. 2X, Tryon & Tryon 5029. c. 3X, Rollins & Tryon 58222. d-f. P. andromedifolia. d. 3X, Harris 21651. e. 2X, Alt 3. f. 2X, Tryon & Tryon 5557. g-h. P. intermedia. g. 3X, Tryon & Tryon 5089. h. 2X, Rollins & Tryon 58144. i. P. myrtillifolia, Wagenknecht 946, Chile. j-k. P. ternifolia. j. 4X, Tryon & Tryon 5141. k. 2X, Tryon & Tryon 5105. l. P. rufa, Schelpe 4939, South Africa. cells of the abaxial surface, are photosynthetic and the guard cells which are confined to this surface, may develop against either one or two walls of the stomata] initial and are dense- ly packed with chloroplasts. In the three species in Pellaea examined the abaxial epidermal cells have peculiar anticlinal walls that are sharp-angled with irregularly thickened peaks and sinuses. This is one of the features, along with the slender, creeping rhizome with bicolorous scales and small, ovate segments, characterizing these three species and two others allied to P. andromedifolia. Tracings of the walls from a microprojector are illustrated (Fig. 4) for all five species in this group and also for P. ternifolia representing 12 Rhodora [Vol. 70 Fig. 5 a-d. Length of longer guard cell in 100 stomata; abscissa- guard cell length in micra; ordinate-frequency and %. a. Pellaea andromedifolia, solid line, sexual diploid, Tryon & Tryon 5557; broken line, apogamous triploid, Tryon & Tryon 5556. b. P. intermedia, solid line, Rollins & Tryon 58144; broken line, apogamous triploid, Tryon & Tryon 5089. c. P. ovata, solid line, sexual diploid, Tryon & Tryon 5029; apogamous triploid, Rollins & Tryon 58222. d. P. ternifolia, solid line sexual diploid, Tryon & Tryon 5105; broken line, sexual tetraploid, Tryon & Tryon 5141. another group in the genus. The greater size of the guard cells in the apogamous and polyploid plants in the top row, compared to those in the center row, can be readily seen. The pattern of epidermal cells on the adaxial surface is also 1968] Pellaea — Tryon 13 illustrated for P. ovata (Fig. 4c). These have a different form with shallowly undulate walls, which are uniform in thickness and the cells are mostly longer than broad. In Fig. 5 a-c measurements showing the guard cell length are based on the larger member in each of 100 stomata, for both sexual and apogamous plants, plotted to the first decimal. In each species these illustrate the smaller guard cell size in the sexual diploid race as compared to the apogamous triploid. In the text the ranges are given by summation to the nearest whole number. In P. andromedifolia the range in length of the guard cells (Fig. 5a) in a sexual diploid plant from northern California is 18-40», with a mean of 27. The range of an apogamous triploid plant from the same area is 24-42, with a mean of 34, which is 26% larger than the sexual. In P. intermedia the range in guard cell length (Fig. 5b) of a sexual diploid plant from northeastern Mexico is 22-40 with a mean of 31. An apogamous triploid plant from southern Arizona has a range of 28-48, with a mean of 39, which is about 26% greater than the sexual type. In P. ovata the range in guard cell length (Fig. 5c) of a sexual diploid plant from central Texas is 22-37, with a mean of 30. An apogamous triploid plant from north- central Mexico ranges from 29-48, with a mean of 38, which is about 27% larger than the diploid. The data from these apogamous and polyploid races may be compared with that from P. ternifolia, in which a sexual polyploid race is known. In Fig. 5d the measurements of the guard cells are given for a sexual diploid and for a sexual tetraploid of P. ternifolia. The sexual diploid plant from southcentral Mexico ranges from 29-44, with a mean of 37. The sexual tetraploid from the same region ranges from 35-51, with a mean of 42, which is 14% larger than the diploid. The increase in size of guard cells in the polyp- loid over the diploid in the sexual P. ternifolia, with a 1:2 ratio in chromosome number of diploid to tetraploid, is significantly less than the 26% or 27% increase in the apogamous polyploids over the other three species having a 2:3 ratio of chromosome number in the diploid to triploid 14 Rhodora [Vol. 70 levels. This suggests that apogamy may be a factor in the increase in the size, augmenting, in these examples, the well known effect of polyploidy. The epidermal patterns for two species of the group in which apogamy is not reported are also included (Figs. 4 i, 1). The average guard cell size of P. myrtillifolia, an en- demic in north Chile, is 394 and that in P. rufa, which occurs in the South African Karroo, is 384. The chromosome num- bers are not reported but both species have the normal com- plement of 64 spores. The average size of the guard cells in these corresponds to that of the polyploid forms in the previous sets and suggests that they may be polyploid. Spores The apogamous condition in ferns is most readily detected by 32 spores per sporangium which is half the normal complement in most leptosporangiate species. In apogamous plants having well developed spores this number is con- sistent at either diploid, triploid or tetraploid levels. The mechanism aecounting for this departure from the normal is the premeiotic doubling of chromosomes along with arrested cytoplasmie cleavage, prior to the formation of spore mother cells. These changes result in eight rather than the normal number of 16 spore mother cells. The spores of both the sexual and apogamous forms in P. andromedifolia, P. intermedia and P. ovata are generally morphologically similar. They are tetrahedral-globose, trilete and, in amb, biconvex. They are pale yellow with excrescences in the outer layer of the spore wall forming rugae which fuse into irregular reticula. The rugae are irregular, flattened and wing-like in surface view and appear as moderately sharp projections in profile, Three com- missural ridges protrude slightly from the surface and extend to the equatorial region. However, spores of apogamous plants differ from those of sexual plants in having more numerous and more prominent rugae as well as a larger size and a smaller number per sporangium. Spores of apogamous plants may also show marked varia- tion in the form of the commissural ridges. They may be 1968] Pellaea — Tryon 15 trilete, as in spores of sexual plants, or they may have only a single ridge or a dense rugose patch asymmetrically placed on the proximal face. The monolete spores which frequently occur in apogamous plants appear to become dissociated from the tetrad early in their development. Measurements of the greatest diameter of 100 spores from plants of each of the races of these three species are presented in graphs (Fig. 6 a-c). In these, size has been plotted to the first decimal. In the text the decimal has been elminated by summation to the nearest whole number. In P. andromedifolia the range in spore size (Fig. 6a) of z b soh ot AND, INT, 25) oh | | Hn - A Op A LN . V d - Aw A - i + H 1 4 H H je gi 1 y E d o- y ta + wi »- oe! 3 z 4 di n x H ene ` H 4 1 1 1 1 1 L 1 Pl 1 1 DEE, 65 so 70 75 c d zo | OVT. TER, om - A / - , t 4 E i ty ` ÁN ! M X L v-a h ! EN L ; Vë - sz Fig. 6 ad Length of longest diameter in 100 spores; abscissa- length in micra; ordinate-frequency and %. a. Pellaea andromedi- folia, solid line, sexual diploid, Alt 3; dashed line, apogamous triploid, Harris 21651; dotted line, apogamous triploid, Kiefer 1132. b. P. intermedia, solid line, sexual diploid, Rollins & Tryon 58144; broken line, apogamous triploid, Rollins & Tryon 58306. c. P. ovata, solid line, sexual diploid, Tryon & Tryon 5029; broken line, apogamous triploid, Rollins & Tryon 58222. d. P. ternifolia, solid line sexual diploid, Rollins & Tryon 58218; broken line, sexual tetraploid, Tryon & Tryon 5141. 16 Rhodora [Vol. 70 a sexual plant from southern California is is 34 - 524, with a mean of 44. Spores from an apogamous plant from the southern portion of the state range from 59-83,, with a mean of 70 which is 59% larger than the sexual. Spores from an apogamous plant from the northern part of Cali- fornia range from 63-83, and have a mean of 71 which is 61% larger than the spores of the sexual plant. In P. intermedia (Fig. 6b) spores of the sexual plant range from 44-59, with a mean of 52, while those of the apogamous plant range from 56 - 79, with a mean of 66, which is about 27% larger than those of the sexual plant. In P. ovata (Fig. 6c) spores of the sexual plant range from 34 - Div with a mean of 41 and those of the apogamous plant range from 52-794, with a mean of 66 (61% larger than the sexual plant). In the graphs illustrating these ranges in spore size, it may be seen that there is no or little overlap in sizes of the sexual and apogamous members in each of the three species. These appear in contrast to the graph showing spore sizes in P. ternifolia (Fig. 6d). Here spores from a sexual diploid plant range from 35-53, with a mean of 44, while those of the sexual tetraploid range from 46 - 59,4, average 54 (23% larger), with a clear overlap. The influence of apogamy on the size of the spores is not evi- dent. The ratios of cell size, based on chromosome level, in the diploid to triploid is 1 : 3 and of the diploid to tetra- ploid is 1 : 2. Thus the effect of apogamy on spore size is obscured by differences in the chromosome levels. The extent of variation in spore size of apogamous plants is clearly greater than that shown by the sexual plants. This contrasts with the condition in the sexual diploid and tetraploid and suggests that the process of spore formation is not as strongly controlled in the apogamous plants. The larger spore size in the tetraploid P. ternifolia may be useful in distinguishing different cytotypes in the her- barium specimens. In the apogamous plants differences in spore number and size have utility in determining the ranges of cytotypes from herbarium material and such ranges may supply useful information on the geographic centers and migration of the species. 1968] Pellaea — Tryon 17 Fig. 7. Distribution of Pellaea ovata with leaf of sexual diploid (Fig. 1f) general distribution within solid line, stars at localities of 64-spored, sexual plants; dots at localities of 32-spored, apogamous plants. Geographic Distribution The general distribution of the species is shown in Figs. 7-9 by solid lines; the dots representing localities of 18 Rhodora [Vol. 70 Fig. 8. Distribution of Pellaea intermedia with leaf of sexual diploid (Fig. 1d) general distribution within solid line, stars at locali- ties of 64-spored, sexual plants; dots at localities of 32-spored, apogamous plants. apogamous plants and stars the sexual ones, based on spore counts of material with sporangia intact. The silhouettes show the general shape of the leaves for each species and are taken from cytologically determined diploid plants, It is expected that additional material and field study will extend the ranges. However, these data are adequate to represent the general patterns. In these species, as in others where such distributions have been plotted, the apogamous races are generally more wide ranging. In P. ovata (Fig. 7), which has one of the most extensive ranges in the 1968] Pellaea — Tryon 19 AO pd bot — Fig. 9. Distribution of Pellaea andromedifolia with leaf of sexual diploid (Fig. 1a) general range within solid line, stars at localities of 64-spored, sexual plants; dots at localities of 32-spored apogamous plants. DS Eri de 120 genus, plants having the normal 64-spored sporangia are limited to central Texas and adjacent northeastern Mexico. All specimens from South and Central America, the Carib- bean area and southern Mexico have 32 spores and are of 20 Rhodora [Vol. 70 the apogamous form. The range of P. intermedia (Fig. 8) is restricted to the southwestern United States and adjacent northeastern Mexico, Plants with 32-spored sporangia are widely distributed in Texas, New Mexico, Arizona and northern Mexico while plants with 64 spores are known only from the region around Saltillo, Mexico. The distribu- tion of sexual and apogamous plants in P. andromedifolia (Fig. 9) contrasts with the ranges of the previous species. Plants with 32 spores are disjunct, occurring in northern California in Humboldt County, and in southern California near Palm Springs and along the coast near San Diego south into Baja California and in adjacent Cedros Island. Both sexual and apogamous races occur within the same region and in Humboldt County they grow a few meters apart. The ranges of the apogamous races of these three species may be compared with other apogamous pellaeas to provide a broader basis for an evaluation of the phenomenon. One measure of the evolutionary success of apogamous species is the extent of their geographic range. This is particu- larly applicable in comparison with the ranges of closely related sexual species. Apogamy is reported in six of the 15 species in section Pellaea and in four, the apogamous races have wide ranges in North America or in both North and South America, In two species the apogamous types have been distinguished as varieties and are far wider ranging than the sexual counterpart. Thus, on the basis of their more extensive distributions, four of the apoga- mous forms are more successful than related sexual ones. The distribution of P. intermedia is not extensive, but within it the apogamous race has a broader distribution than the sexual race. There are some collections of P. inter- media from the Chisos mountains in Texas, within the range of the apogamous type, bearing on a single leaf some sporangia with 32 well-developed spores and others with a larger number of undeveloped spores. In P. andromedi- folia the apogamous race is disjunct and not as abundant 1968] Pellaea — Tryon 21 as the sexual one. This contrasts with the distributions shown in the other species in which the apogamous race predominates, and suggests the possibility of a different evolutionary history in P. andromedifolia. SUMMARY Comparisons were made of epidermal cell and spore size of sexual diploid and apogamous triploid forms in three closely related species of Pellaea in a study of the effect of apogamy on morphological characters. Comparisons were also made with sexual diploid and sexual tetraploid forms in another species and with other members of Pellaea. Size of stomatal guard cells of the apogamous triploid plants relative to the sexual diploids in each of the species, was greater than would be expected on the basis of the chromo- some levels alone. The same effect was not evident in com- parisons of spores. In these cells the large size of the apogamous triploids relative to the sexual diploids was of the order that would be expected from the different chromo- some levels. Comparisons of the geographic distribution patterns of the sexual and apogamous races and with those of other members of the genus show that the apogamous types are wider ranging. However, in P. andromedifolia apogamous plants have a disjunct distribution and are not as abundant as the sexual ones, The evolutionary potential of apogamous plants is shown in their production of functional antherozoids, their adapta- tion to different ecological niches, and their broad distri- butions. In these pellaeas differences in the ranges of the apogamous and sexual types are interpreted here as reflect- ing the dynamics of these reproductive systems in natural populations. [Vol. 70 Rhodora 22 81299 URL] ‘WH Y 841° 7D CN XP gc jenxos oa “1S070q] SINT UBS SOTS UOL] VPN D0ftu49]. “IRA XZ 66 jenxos ODIXƏW 'eoeAtu.ron;) quit (ABD) vyofiuta} "qd HI BIJ — ZELE UOL JC A P SUNY "OD Ce XE L8 snowrsody ODIXƏW '1$0304d SINT UBS JL "Sid — 660€ vii ^4 ^y PW "eI XZ 6c [enxes sexo[ “09 erdsoe[[u5 Aqiauyjeom (Casa) mao ‘q eT “Sl — 680€ us] “AV PW SUOZLIK Xs 18 snowesody ‘SUL nyeor (ge = ug PL ‘Std — ffI8€9 Wofap "WM WOLF) A sumoyg "DA XZ 6c [enxeq OoIXe]N ‘OMHELS uyny xo "ah ?rpowu23w "d dT ‘SI — ZETT 49 fory "T €IUIOgI[G)) "oi Xs L8 snoure3od y vinjuoA NIN 3utoq 9966 uos, '4 "V X "W A X& L8 snouresody "09 3p]oqun H 4966 uofi4] “AV PW "8I XZ Dé [enxas OD 3p[oquiH qr “SIX — [6912 SDH 'S €IUI0JI[9; ) XE L8 snowesody "09 3pjoqumH el "uq — E NV mwy €TU.I0JT[9;) XG 66 [enxes "SIW [PLqLO ues aq CHNEX) vgofipowo4pun ‘q Aptolg S1oquinu uornonp uorjoe[[o) pue Aye o sated BULOSOULO.L YD -o1do. NOTA Jo 9pow vdv'TId4d JO SHdHW/üN WAWOSONWONHO 1968] Pellaea — Tryon 23 Acknowledgements I am grateful to the officers in charge of the herbaria which were cited earlier, who have made specimens avail- able for this study. I am especially indebted to Donovan Correll, Karen Alt Grant, Stuart Harris, Paul Hutchison, Larry Kiefer, Thomas Pray and particularly to Reed Rollins and Rolla Tryon for collections of living plants or fixed material examined in this study. Donald M. Britton has been most helpful in his critical review of the manuscript and Rolla Tryon has helped immeasurably, from the first field trips to the final manuscript, with stimulating and constructive ideas. LITERATURE CITED CHIARUGI, A. 1960. Tavole cromosomiche delle pteridophyta. Caryo- logia 13: 27-150. FABBRI, F. 1963. Primo supplemento alle tavole cromosomiche della pteridophyta di Alberto Chiarugi. Caryologia 16: 237-335. 1965. Secondo supplemento. Ibid. 18: 675-731. FABERGÉ, A. C. 1945. Snail stomach cytase, a new reagent for plant cytology. Stain Technology 20: 1-4. MANTON, IRENE. 1950. Problems of Cytology and Evolution in the Pteridophyta. Cambridge University Press. 1961. Evolution in the Pteridophyta. In a Darwin Centenary, Conference Report Bot. Soc. Brit. I. Ed. P. J. Wan- stall, No. 6: 105-120. Nayar, B. K. & N. BAJPAI. 1964. Morphology of gametophytes of some species of Pellaea and Notholaena. J. Linn. Soc. 59: 63-76. STOKEY, A. C. 1951. The contribution by the gametophyte to classi- fication of the homosporous ferns. Phytomorphology 1: 39-58. TRYON, ALICE F. 1957. A revision of the fern genus Pellaea sec- tion Pellaea. Ann. Mo. Bot. Gard. 44: 125-193. . 1960. Observations on the juvenile leaves of Pellaea andromedifolia. Contrib. Gray Herb. 187: 61-68. TRYON, A. F. & D. M. BRITTON. 1958. Cytotaxonomic studies on the fern genus Pellaea. Evolution 12: 137-145. 24 Rhodora [Vol. 70 WALKER, T. G. 1962. Cytology and evolution in the fern genus Pteris L. Evolution 16: 27-43. 1966. Apomixis and vegetative reproduction in ferns. In Reproductive Biology and Taxonomy of Vascular Plants, Conference Report Bot. Soc. Brit. I. Ed. J. G. Hawkes, No. 9: 152-161. WHITTIER, D. P. 1964. The influence of cultural conditions on the induction of apogamy in Pteridium gametophytes. Amer. J. Bot. 51: 730-736. GRAY HERBARIUM, HARVARD UNIVERSITY ON THE NAMES IN FRASER’S 1813 CATALOGUE JAMES L. REVEAL The 1813 Fraser’s Catalogue (essentially accurately reprinted in Greene, 1890) and fully entitled “A Cata- logue of New and Interesting Plants, Collected in Upper Louisiana and Principally on the River Missourie, North America,” which was published prior to August 1813, is known to nearly every botanist who has had the privilege of working with the names of Thomas Nuttall, the famous English taxonomist who made his greatest contributions in this country as a botanist, ornithologist, and explorer. This small document has been the source of controversy, neglect, and dispute for as many years as it has been in existence, Not only is its author questioned, its very nature as a place of publication for the species that appear in it is doubted by some. Contrary to the beliefs of some, the author must have been Thomas Nuttall. The interested reader on this subject is invited to refer to the series of arguments presented by Shinners (1949, 1955, 1956) for Nuttall not being the author and the Catalogue as an invalid source of publica- tion; Graustein's (1956) contentions that Nuttall was the author but that he considered the species not validly pub- lished because of an agreement with his sponsor, Benjamin Smith Barton of Philadelphia, not to publish any new species without Barton's consent; and Cronquist, Keck, and Maguire (1956) who believe that Nuttall was the author and that the Catalogue is a valid place of publica- tion. It has been common knowledge that Nuttall was at least associated with the names found in the Catalogue. Pennell (1936) and Graustein (1967) point out that he was in England and associated with the Fraser Brothers' Nursery at the time of publication. From Nuttall's two subsequent publications (1817, 1818) which followed shortly after the Catalogue, we know that he considered at least some of the 25 26 Rhodora [Vol. 70 names as his own. Several of the early authors attributed the names in the Catalogue to Nuttall, and they certainly could have asked Nuttall if they had any doubt as who the author was. Pennell (1936) gives the impression that Nut- tall was a rather shy and inhibited person. It would seem out of character for a man of such temperament to assume responsibility for the names in Fraser’s Catalogue if he was not in any way responsible for them. It is inconceivable to me that someone else could have assigned names to Nut- tall’s own collection without Nuttall stating this fact later. If Nuttall was not associated with the entities, why should he later accept some of them as his unless he actually had given the names to the Fraser Brothers, helped someone who was employed by them to prepare the Catalogue, or perhaps have written the text himself. It should be noted here, however, that I do not say that Nuttall was the author of the paper, although Greene (1890) suggests this after seeing the copy of the Catalogue in Philadephia. What I do contend is that Nuttall is the author of the names in the paper, and thus, as provided by the Internationa] Code (1966), the species should be cited as “Nutt. in Fras.” Shinners (1956) goes to great length to show that several of the names in Fraser’s Catalogue were not claimed by Nuttall in his later publications, and while this is true, I suspect that it was for reasons other than those given by Shinners. Some species were found to have been adequately described between 1813 and 1818 with names acceptable to Nuttall, but for those that were not, he used his own names that he had published in the Fraser’s Catalogue. Nuttall occasionally cited only ‘Frasers Catalogue" and did not give himself credit for the name. Shinners uses this argu- ment to show that Nuttall was not the author. The species with which I am most familiar that was published in the Catalogue is Eriogonum flavum. The name is credited to “Fras. Catal. 1818" and not starred as a new species in Nuttall's 1818 book, The Genera of North American Plants, and for this reason, Shinners suggested that Nuttall did not consider this species as his. Nothing is further from 1968] Fraser’s Catalogue — Reveal 27 the truth. In a detailed paper on this species in particular, and the genus Eriogonum in general, which Nuttall pub- lished in 1817, he makes a point of stating that he published the name E. flavum in Fraser's Catalogue. The reason why Nuttall simply stated “Fras. Catal. 1813.” instead of “T. N. in Fras, Catal. 1813.”, as he sometimes did, was probably to save space. To say that Nuttall was not the author of Fraser’s Catalogue I believe is on more tenuous bases than to say that he was indeed the sole author. As no one else has come along to claim authorship of the species published in the list, this, I believe, proves the point that only Nuttall could have given the names to his own collection. The International Code of Botanical Nomenclature (1966) states in Article 34 that “A name is not validly published . .. when it is not accepted by the author who published it ..." As Nuttall did not specifically reject the names in the Catalogue, they must be considered as pub- lished. The arguments put forth by Graustein (1956) are immaterial. She states that under the terms of a contract, Nuttall’s journals and observations became the exclusive property of Barton, and Nuttall was not supposed to do anything else but what was specifically stated in the con- tract. Thus, Graustein believes that Nuttall could not legally publish any new species in Fraser's Catalogue, for if he did, he would be breaking the terms of the contract. How- ever, as McKelvey (1955) has pointed out, Nuttall's mere presence with the Overland Astorians as they ascended the Missouri River was breaking his contract with Barton, as was the shipping of his plants to England instead of Phila- delphia. Certainly one more step in the breaking of the contraet by publishing his findings would not be totally surprising. Therefore, it is my desire in this paper to discuss each species listed in Fraser's 1813 Catalogue, basing the validity of the names on the adequacy of the descriptions and subse- quent identifiability of each entity, considering the nomen- clatural status and history of each, and making necessary adjustments and suggestions where necessary. 28 Rhodora [Vol. 70 ACKNOWLEDGEMENTS I wish to thank the United States National Herbarium and the Smithsonian Institution which sponsored my Pre- doctoral Internship program in Washington, D.C., from September 1966 to February 1967, where this paper was prepared. I would like to acknowledge the Academy of Natural Sciences in Philadephia who kindly provided me with a photostat copy of the Catalogue that is deposited in their library. This paper, as well as the others written during this program, were submitted to the Department of Botany, Brigham Young University, in partial fulfillment of three credit hours of Special Problems given Fall Semester of 1966-1967. I wish particularly to acknowledge the assist- ance of C. V. Morton who was not only my supervisor on the Internship, but gave me several hours of his time in answering my many questions and assisting me in gaining a greater knowledge in the field of plant taxonomy. He has kindly commented, as has Dr. Arthur Cronquist of the New York Botanical Garden, upon the problems of each entry in the Catalogue, but the final disposition of each name is my own responsibility. 1. Allium reticulatum Nutt. in Fras., nom. nud. This name was published in 1827 as A. reticulatum Nutt. ex G. Don in Wern., Nat. Hist. Soc. Mem. 6: 36, but it was a homonym of A. reticulatum J. & C. Presl, Fl. Cech. 73. 1819. Nelson and Macbride later renamed this species A. textile in Bot. Gaz. 56: 470. 1913. 2. Allium stellatum Nutt. in Fras., nom. nud, This species was a short time later described by Sims, referring to Nuttall’s name in the Fraser’s Catalogue, and should be cited as A. stellatum Nutt. ex Sims, Bot. Mag. 38: pl. 1576. Aug. 1813. 3. Alyssum petraeum Nutt. in Fras., nom. nud. I have been unable to determine the identity of this nomen nudum. The name has not been treated by any subsequent author. 1968] Fraser’s Catalogue — Reveal 29 4. Amorpha canescens Nutt. in Fras., nom. nud. Pursh published this same name in Fl. Am. Sept. 2: 467. 1814, but made no reference to the Fraser’s Catalogue. 5. Amorpha nana Nutt. in Fras. Catal. 1813. In my opinion this species is validly published in the Catalogue and should be accepted. Nuttall gives the following key points: * . . elegant dwarf shrub, with highly odorous purple flowers . . . It is perfectly glabrous, dentures of the calyx all acuminate, and the legume one seeded." Pursh cites A. nana as a synonym of his own A. microphylla Pursh, FI. Am. Sept. 2: 466. 1814. 6. Astragalus crassicarpus Nutt. in Fras. Catal. 1813. Barneby (1964) in his recent monograph on Astnagalus succinctly states the problem on the Catalogue and this species as follows: “The status of the name A. crassicarpus is some- what controversial, being vulnerable to criticism from two directions. The Fraser Brothers' Catalogue has been attacked and defended in recent years... as A valid medium of publication, neither side to the ques- tion having gained a conclusive victory. A graver fault is the extreme brevity of the original description and the fact that Nuttall himself [1818] adopted Pursh's A. carnosus in place of his own (or what is presumed to be his own) proposition. However no reasonable doubt has ever been entertained as to the identity of A. crassicarpus, which was already known to botanists contemporary with Nuttall." The brief description noted by Barneby is just that: "Fruit about the size and form of A. physodes, but thick and suc- culent.” Although it is brief, there is no question as to the species Nuttall had in mind. Pursh described A. carno- sus in Fl. Am. Sept. 2: 740. 1814 from a Bradbury collection of Sophora (S. nuttalliana B. L. Turner) and fruits taken (according to Nuttall) from the Nuttall collection of A. crassicarpus. The species was later described as A. caryo- 30 Rhodora [Vol. 70 carpus Ker in Bot. Reg. 2: 176, cum icone. 1816, from plants cultivated in Lambert’s Garden from seeds that Nuttall had provided for the Fraser Brothers’ Nursery. 7. Astragalus melanocarpus Nutt. in Fras., nom. nud. Nuttall apparently renamed this species A. missouriensis Nutt., Gen. N. Amer. Pl. 2: 99. 1818, but made no refer- ence to the Fraser’s Catalogue. 8. Astragalus gracilis Nutt. in Fras., nom. nud. Nuttall described this species in Gen. N. Amer. Pl. 2: 100. 1818, stating that his species “. . . appears to be A, tenellus of Pursh and also Ervum multiflorum, Suppl. 2. p. 739. It is likewise Dalea parviflora PH. 2. p. 474. as I have examined the specimen so marked in Herb. Lambert." As A. tenellus is a totally different species, Barneby (1964) comments: “Nuttall cited his own A. gracilis of Fraser's Catalogue, a nomen nudum, and Dalea parviflora Pursh . .. The name A. gracilis is therefore best treated as a legitimate substi- tute for Dalea parviflora . . ." 9. Artemisia cernua Nutt. in Fras, nom. nud. Nuttall described this species and ascribed the name in Fraser’s Catalogue to himself in Gen. N. Amer. Pl. 2: 143, 1818, but by then Pursh had described the same species as A. dracunculoides Pursh, Fl. Am. Sept. 2: 742. 1814. When Pursh proposed his new species, he cited the name from Fraser's Catalogue, but as A. nutans instead of A. cernua, no doubt an error, in synonymy of his new species, It is generally agreed today that the American plant and the European one are conspecific and should be called A. dra- cunculus L. 10. Aristida, This grass genus is cited without a specific ephithet. Nuttall cites A. pallens Cav. in his Genera, but according to Hitchcock (1951), this is A. oligantha Michx. Possibly this is the grass cited. 1968] Fraser’s Catalogue — Reveal 31 11. Batschia fimbriata Nutt. in Fras., nom. nud. Pursh described this species as B. longiflora in Fl. Am. Sept. 1: 132, 1814, and Nuttall redescribed the species by the same name in Gen. N. Amer. Pl. 1: 114. 1818, adding B. decum- bens. In no instance is B. fimbriata referred to. The species is now known as Lithospermum incisum Lehm., Asperif. 303. 1818. The name B. longiflora is rejected because of L. longiflorum Salisb., 1796, while B. decumbens is rejected due to L. decumbens Vent., 1800. 12. Bumelia confertiflora Nutt. in Fras., nom. nud. This species is stated as having been collected near Saint Louis, Missouri. The only Bumelia found in that area is B. lanu- ginosa (Michx.) Pers. var. oblongifolia (Nutt.) R. B. Clark. Although Nuttall did not mention B, confertiflora when he described B. oblongifolia in Gen. N. Amer. Pl. 1: 135. 1818, it certainly must be the same plant. Neither Clark (1942) or Cronquist (1945) indicated the possible position of this nomen nudum. 13. Bartonia decapetala Pursh in Sims, Bot. Mag. 36: pl. 1487. Aug. 1812. When Nuttall wrote the treatment for the Fraser's Catalogue, he indicated that this was a new species, and probably for him it was. Pursh apparently took notes from living material grown in England from seeds presented to the Fraser Brothers by Nuttall, and rushed into publication in Sims’ Botanical Magazine the name of the new genus Bartonia. This action was probably not intended to beat Nuttall to the publication, but to please Benjamin Smith Barton, the teacher and sponsor of both Pursh and Nuttall. Needless to say, the name Bartonia had already been published by Muhlenberg for an entirely different plant. Bartonia decapetala later proved to belong to an already published genus, and is known now as Ment- zelia decapetala (Pursh in Sims) Urban & Gilg, Ber. Deut. Bot. Ges. 10: 263. 1892. A more detailed history of the feud between Nuttall and Pursh is given in McKelvey's book (1955). 32 Rhodora [Vol. 70 14. Bartonia polypetala Nutt. in Fras., nom. nud., as polly- petala. This species name has not been treated by any sub- sequent author, but is probably the same plant later named B. nuda Pursh, Fl. Am. Sept. 1: 328. 1814. This species is now called Mentzelia nuda (Pursh) Torr. & Gray, Fl. N. Amer. 1: 535, 1840. 15. Chloris mucronata Michx., Fl. Bor. Amer. 1: 59. 1803. This grass was included on the list in the Catalogue as one of the more interesting North American grasses available from the Fraser Brothers’ Nursery. It is known presently as Dactyloctenium aegyptium (L.) Beauv. 15. Chloris mucronata Michx., Fl. Bor. Amer, 1: 59. 1803. 1803. Like the above grass, this one was cited in the Cata- logue as being available from the Fraser Brothers’ Nursery. It is known presently as Bouteloua curtipendula (Michx.) Torr. 17. Cyperus. An unspecified species of this genus was listed as available at the Fraser Brothers’ Nursery. It was not indicated as being a new species. 18. Cynoglossum glomeratum Nutt. in Fras., nom nud. This species was described the following year by Pursh at which time he cited the Fraser Brothers' Catalogue as the source of his name. The citation should thus read: C. glomeratum Nutt. ex Pursh, Fl. Am. Sept. 2: 729. 1814. This species is now known as Cryptantha celosioides (Eastw.) Pays. Ann. Mo. Bot. Gard. 14: 299. 1927. 19. Calymenia nyctaginea (Michx.) Nutt. in Fras. Catal. 1813, comb, nov. This new combination was later made in, and is usually dated from, Gen. N. Amer. Pl. 1: 26. 1818, by Nuttall. As there was no requirement in 1813 that a basionym be cited, the combination should be considered as made in the Catalogue. The species is now known as Mirabilis nyctaginea (Michx.) MacM. 1968] Fraser's Catalogue — Reveal 33 20. Calymenia pilosa Nutt. in Fras., nom. nud. Nuttall in Gen. N. Amer. Pl. 1: 26. 1818 published this species, but by then Pursh had published Allionia hirsuta Pursh, FI. Am. Sept. 2: 728. 1814, from a Bradbury collection. The species is presently known as Mirabilis hirsuta (Pursh) MacM. 21. Calymenia angustifolia Nutt. in Fras., nom. nud. Pursh cited this nomen nudum when he proposed Allionia linearis in Fl. Am. Sept. 2: 728. 1814, based on a Bradbury collec- tion. Nuttall, in his Gen. N. Amer. Pl. 1: 26. 1818, de- scribed as new his C angustifolia, citing not only “T. Nut- tall in Fraser’s Catalogue. 1813." but the Pursh name as well. The species is now known as Mirabilis linearis (Pursh) Heimer. 22. Cactus viviparus Nutt. in Fras, Catal. 1813. This species is certainly described here, although Nuttall re- described it in Gen. N. Amer. Pl. 1: 295. 1818. The de- scriptive points given by Nuttall in 1813 are: “This species has much the appearance of C. mamillaris, but produces a large red flower, like C. Flagelliformis, and a greenish edible fruit, about the size of a grape.” The species is now called Coryphantha vivipara (Nutt. in Fras.) Britt. & Brown, Ill. Fl. 2nd Ed. 2: 571. 1913. It should be noted that Cactus mamillaris of Nuttall is Coryphantha missouriensis (Sweet) Britt. & Rose. 23. Cactus ferox Nutt. in Fras., nom. nud. Nuttall later described this species in Gen. N. Amer. Pl. 1: 296. 1818, but was not aware that this name was a later homonym of C. ferox Willd., 1813. In 1828 de Candolle substituted Opuntia missouriensis DC., Prodr. 3: 472, for the Nuttallian species. The name O. polyacantha Haw., Syn Pl. Succ. Suppl. 82. 1819, however, was already available for the species. 34 Rhodora [Vol. 70 24. Cactus fragilis Nutt. in Fras., nom. nud. As with the above species of cactus, Nuttall also described it in 1818, and this place of publication is the basis for Opuntia fragilis (Nutt.) Haw., Syn. Pl. Suec. Suppl. 82. 1819, the generally accepted name for this species. 25. Cheiranthus asperus Nutt. in Fras., nom. nud. Pursh did not attempt to ascertain the nature of this species in 1814, and thus this enabled Nuttall to describe the species as C. asper in Gen. N. Amer, Pl. 2: 69. 1818. This species is now called Erysimum asperum (Nutt.) DC., Syst. Veg. 2: 505. 1821. 26. Cytisus rhombifolius Nutt. in Fras., nom. nud. This species was described by Pursh in Fl. Am. Sept. 2: 741. 1814, at which time he cited the Fraser's Catalogue as the source of his name. Nuttall transferred the species from Cytisus to Thermia as T. rhombifolia (Nutt. ex Pursh) Nutt., citing both his name in the Catalogue and Pursh's place of publication. The species has since been trans- ferred to Thermopsis, the citation being T. rhombifolia (Nutt, ex Pursh) Nutt. ex Richards. in Frankl. 1st. Journ. App. 737. 1823. 27. Diodia virginica L. This species was mentioned in the Catalogue as a plant that was available from the Fraser Brothers' Nursery. 28. Dracocephalum cuspidatum Nutt. in Fras. Catal. 1813. In my opinion the statement given “A Verticillate species” is enough to describe this species of mint. Nuttall later named this species as D. parviflorum Nutt., Gen. N. Amer. Pl, 2: 35. 1818, the name commonly used today. Of the species of Dracocephalum from the United States, as under- stood and known to Nuttall at that time, none but this has flowers that are verticillate, and this point is the first de- scriptive statement he used in describing D. parviflorum. The generic problem of whether or not to separate Draco- 1968] Fraser’s Catalogue — Reveal 35 cephalum and Physostegia is not for me to discuss, but if D. cuspidatum should be transferred to Physostegia, I can find no conflict. 29. Dalea aurea Nutt. in Fras., nom. nud. This species was described the following year by Pursh as D, aurea Nutt. ex Pursh, Fl. Am. Sept. 2: 740. 1814. Pursh based his description on a Bradbury collection, but cited the Fraser's Catalogue as the source of his name. Nuttall later redescribed this species using the same name in Gen. N. Amer. Pl. 2: 101. 1818, citing both the Fraser's Catalogue and Pursh's flora. 30. Dalea enneandra Nutt. in Fras. Catal. 1813. The de- scription of this species is one of the more complete. Nuttall stated that the “. . . stem is tall and erect, sending out nu- merous slender waving branches towards its summit, ter- minated by racemes of white flowers, and silky calyces." Pursh renamed this species D. laxiflora in Fl. Am. Sept. 2: 741. 1814, and Nuttall accepted this name in his Gen. N. Amer. Pl. 2: 101. 1818, with no mention of his own species published five years before. It is difficult to determine whether Nuttall was rejecting his own name because of some taxonomic reason, or simply because the Pursh name seemed better. Nevertheless, Nuttall’s name can not be rejected solely because Nuttall himself did not later accept IE 31. Elaeagnus argentea Nutt. in Fras., nom. nud., as argenteus. Pursh described this species under the same name in Fl. Am. Sept. 1: 114. 1814. As this name appeared in the first volume, I believe that Pursh was selecting the same name on the same basis as Nuttall selected the name. The name, however, is a homonyn of E. argentea Moench, Meth. Pl. 638. 1794 which is a superfluous substitute for E. angustifolius L., a different species entirely. The species is presently known as E. commutata Bernh. ex Rydb., Fl. Rocky Mts. 582. 1917. It should be noted that the citation 36 Rhodora [Vol. 70 “E. argentea, Colla, 1791," which is sometimes given as the earliest E. argentea is an error, Colla's name was published in Hort. Rip. in 1824. 32. Evolvulus sessiliflorus Nutt. in Fras., nom. nud. Pursh described this species as Evolvulus argenteus Pursh, Fl. Am. Sept. 1: 187. 1814, but made no reference to Fraser's Catalogue. This name, however, is a later homonym of E, argenteus R. Br., Prod. Fl. Nov. Holl. 489. 1810. Nuttall proposed a substitute name, E. pilosus in Gen, N. Amer. Pl. 1: 174. 1818, which is generally accepted as the valid name. Although E. pilosus was an alternative name which is now illegitimate under the Code, the rule is not retro- active. In no case is the nomen nudum mentioned. 33. Echites puberula Michx., Fl. Bor. Amer. 1: 120. 1808. This species was listed in the Catalogue as being available from the Fraser Brothers' Nursery. It is presently known as Trachelospermum difforme (Walt.) A. Gray. 34. Eriogonum flavum Nutt. in Fras. Catal. 1813. In my opinion this species is adequately described even though the only character given is “Fl. bright yellow.” All of the other species in the genus known in 1813 had white flowers, and this note of distinction would certainly have held up if no other species were ever discovered. Pursh described this species as E. sericeum in Fl. Am. Sept. 1: 277. 1814, but the Nuttall name is commonly used. 35. Erysimum montanum Nutt. in Fras., nom. nud. I have been unable to discover what plant this nomen nudum is referable to. It has not been treated by any subsequent author. 36. Ferula pubescens Nutt. in Fras., nom. nud. This name was not adopted by Nuttall in 1818, as he named this plant F. foeniculacea Nutt., Gen. N. Amer. Pl. 1: 183. 1818. This species is now commonly called Lomatium foeniculaceum 1968] Fraser’s Catalogue — Reveal 37 (Nutt.) Coult. & Rose, Contr. U.S. Nat. Herb. 7: 222. 1900. In a recent revision of the species in this complex by Theo- bald (1966), the nomen nudum is not mentioned. 37. Fritillaria americana Nutt. in Fras., nom. nud. This species was described as F. alba Nutt., Gen. N. Amer. Pl. 1: 222. 1818, but with no reference to Fraser’s Catalogue, The name F. alba is currently considered a nomen dubium. Beetle (1944) in her revision of the genus, states that the descrip- tion is mainly that of a Fritillaria, and that the only species in the genus in the area where Nuttall visited is F. atropur- purea Nutt., Journ. Acad. Phil. 7: 54. 1834. Nevertheless, the description of the flowers states that they are “... rather large and white . . ." Ownbey (1940) suggests that this part of the description refers to Calochortus nuttallii Torr., which also occurs in the area that Nuttall collected in 1811. It might be well to comment in this connection that the flowers of F. atropurpurea, when poorly prepared, blacken as do several other white-flowered species, and it might have been that Nuttall simply forgot their true color. Until the type can be located, the name had best be considered a nomen dubium, but I feel that a case could be made to re- move that title if it were seriously attempted. 38. Gonolobus hirsutus Michx., Fl. Bor. Amer. 1: 119. 1803. This species was listed as one of the interesting plants that could be obtained from the Fraser Brothers! Nursery. Donald J. Drapalik of the University of North Carolina, informs me that he believes this name, a new name for Vincetoxicum acanthocarpus Walt., is referable to Matelea carolinensis (Jacq.) Woodson. 39. Gaura coccinea Nutt. in Fras., nom. nud. The first place of valid publication for this species is G. coccinea Nutt. ex Pursh, Fl. Am. Sept. 2: 733. 1814, and the description is based on a Bradbury collection, but with the citation of the Catalogue as the source of the name. 38 Rhodora [Vol. 70 40. Hypericum surculosum Nutt. in Fras., nom. nud. I have been unable to determine what species from along the Mississippi River this nomen nudum refers to. 41. Hyssopus anethiodorus Nutt. in Fras. Catal. 1813. When Nuttall described this species, he pointed out the unique feature of the flower color when he said “Both calyx and corolla blue,” for at that time, only two other species were known from the United States. These species, Agasta- che nepetoides (L.) Kuntze with greenish-yellow flowers and A. scrophulariifolium Willd.) Kuntze with purplish flowers, were both well known to Nuttall. By pointing out the obviously distinguishing feature, the species, in my opinion, is validly published. Pursh described the same spe- cies as Stachys foeniculum Pursh, Fl. Am. Sept. 2: 407. 1814, and Nuttall redescribed the same species as H, anisa- tus in Gen. N. Amer. Pl. 2: 27. 1818, citing only the Pursh name in synonymy. Inasmuch as H. anethiodorus is validly described, the proper name should be A. anethiodora (Nutt. in Fras.) Britt. in Britt. & Brown, Ill. Fl. 3: 85. 1898. 42. Jussieua angulata Nutt. in Fras., nom. nud. This spe- cies was described as J. leptocarpa Nutt., Gen. N. Amer. Pl. 1: 299. 1818, but with no reference to the name in Fraser's Catalogue. This species is currently called Ludwigia lepto- carpa, (Nutt.) Hara, Jour. Jap. Bot. 28: 292. 1953. 43. Linum perenne L. var. americanum Nutt. in Fras., nom. nud. This same plant was named L. lewisii Pursh, FI. Am. Sept. 1: 210. 1814, but by the time Nuttall wrote his treatment of the genus for his flora, he considered that the differences between the American and European plants were not great enough to warrant even varietal separation. Since then, however, the trend has been to recognize the American phase as L. perenne L. ssp. lewisii (Pursh) Hult.’ ‘Like many American taxonomists, I have been slow in accepting the reality that L. lewisii was not as distinct from L. perenne as one should prefer distinct species to be. By having an opportunity to study 1968] Fraser's Catalogue — Reveal 39 44. Lilium andinum Nutt. in Fras. Catal. 1813. In my opin- ion the description given by Nuttall is adequate enough to determine what species of lily he had in mind, and thus the name should be considered as validly published. He states that this species *. . . has affinity to L. Catesboei, but the stem is usually 3-5 flowered." 'The following year, Pursh published L. umbellatum in Fl. Am. Sept. 1: 229. 1814, citing only a Nuttall specimen but making no reference to Fraser's Catalogue. When Nuttall republished L. andinum in Gen. N. Amer. Pl. 1: 221. 1818, he cites his name from the Catalogue. The species is closely related to L. philadelphicum L., and should be considered as a variety, var. andinum (Nutt.) Ker., Bot. Reg. 7: pl. 594. 1821. Wherry (1947) states that the combination proposed by Ker-Gawler was of subspecific rank, but as one may see, Ker-Gawler distinctly says that his new combination was at the varietal rank when he wrote: "We have thought it safer to keep the two [i.e. L. phila- delphicum and L. andinum] under the same specific title, assorting to each its particular synoymy below the head of its variety." 45. Liquiritia lepidota Nutt. in Fras., nom. nud. This name was first published in Pursh, and may be cited as Glycyr- rhiza lepidota Nutt. ex Pursh, Fl. Am. Sept, 2: 480. 1814. When Nuttall redescribed this species in Gen. N. Amer. PI. 2: 106. 1818, he stated why he no longer used the name Liquiritia when he said: “This plant appears to destroy the herbarium material in the major herbaria across the United States during my Internship at the U.S. National Herbarium, I came to realize that Hultén's treatment of our plant was most acceptable. This means, however, that a rare, high alpine ecotype of this species from Nevada must have its nomenclature adjusted, and is to be now known as L. perenne ssp. lewisii var. saxosum (Maguire & Holmgren) Reveal, comb. nov., based on L. lewisii var. saxosum Maguire & Holm- gren, Leafl. West. Bot. 4: 265. 1946. During the summer of 1964, Noel H. Holmgren, of the New York Botanical Garden, and I found a single specimen of this variety on top of Mt. Jefferson in the Toqui- ma Mountains, Nye Co., Nevada. This extended the known range of this variety south from its type locality in the Ruby Mountains. 40 Rhodora [Vol. 70 artificial distinctions by which Glycyrrhiza and Liquiritia have been separated; as it can be equally referred to either one or the other." 46. Lathyrus myrtifolius Muhl. in Willd., Sp. Pl. 5: 1091. 1803. This plant is listed in the Catalogue as available from the Fraser Brothers’ Nursery. 47. Lactuca oblongifolia Nutt. in Fras. Catal. 1815. In con- sidering the species of Lactuca that Nuttall was familiar with from the United States, and the plants which Nuttall considered to belong to this genus, all had yellowish flowers. Therefore, his characterization of his new species as having “Fl. blue.” is, in my opinion, a valid description. Within the related genus Sonchus, there are several blue-flowered spe- cies including some from Missouri, but as Nuttall placed them in Sonchus, this does not in any way invalidate the Nuttall species published in the Catalogue. Although these blue-flowered species of Sonchus have subsequently been transferred to Lactuca, Nuttall did not, at that time, con- sider them to be members of the genus Lactuca. The following year, Pursh described Sonchus pulchellus in Fl. Am. Sept. 2: 502. 1814. Later, Nuttall decided to rename both as L. integrifolia in Gen. N. Amer. Pl, 2: 124. 1818, at which time he cites his name from the Catalogue, and Pursh's 1814 species. Thus, as L. oblongifolia was a validly and properly described species, it should replace L. pulchella (Pursh) DC., Prodr. 7: 134. 1838, which is pres- ently the commonly used name for this species. 48. Liatris graminifolia (Walt.) Willd., Sp. Pl. 3: 1636. 1803. This species was listed as one of the more interesting species that could be obtained from the Fraser Brothers' Nursery. 49. Liatris paniculata Willd., Sp. Pl. 3: 1637. 1803. This was listed as one of the more interesting species that could 1968] Fraser’s Catalogue — Reveal 41 be obtained from the Fraser Brothers’ Nursery. Its exact identity, however, was questioned. 50. Mentzelia oligosperma Nutt. in Fras., nom. nud. This nomen nudum was first published by Sims as M. oligosperma Nutt. ex Sims, Bot. Mag. 42: pl. 1760. 1815, citing Fraser’s Catalogue as the source for his name. It is interesting to note that while the Catalogue was not signed, or in any way credited directly to Thomas Nuttall who by 1815 had not in print credited himself with any of the names in the Cata- logue, Sims had no question as to who the author of this species was. 51. Malva coccinea Nutt. in Fras. Catal. 1813. Nuttall de- scribes this species as follows: “Flowers scarlet, produced in dense spikes," a sufficient description to consider Fraser's Catalogue as the original place of publication. Pursh made the combination Cristaria coccinea in Fl. Am. Sept. 2: 453. 1814, citing “Malva coccinea, Fraser. catal." This species was again described by Nuttall in Gen. N. Amer. Pl. 2: 81. 1818, at which time he cited his original description and indirectly referred to Pursh's new combination. He states rather briskly that “This plant has no sort of affinity to Cristaria, and by the fruit is a genuine Malva." In con- sidering the Catalogue as the place of publication, typifica- tion of the species must be changed from the Lewis collec- tion cited by Pursh to the Nuttall collection, This species is currently known as Sphaeralcea coccinea (Nutt. in Fras.) Rydb., Bull. Torr. Bot. Club 40: 58. 1913. 52. Oenoplea volubilis (Linn. f.) Nutt. in Fras., comb. nov., as Oenoplia. This new combination is based on Rhamnus volubilis Linn. f., Suppl. 152. 1781. The combination here was made before that of Roemer & Schultes who proposed the same combination in Syst. Veg. 5: 332, 1819. The species is now known as Berchemia scandens (Hill) K. Koch, Dendrol. 1: 602. 1869. This latter combination might be noted as it is usually attributed to Trel., Trans. Saint 42 Rhodora [Vol. 70 Louis Acad. 5: 364. 1889. The species was listed as one of the more interesting plants that could be obtained from the Fraser Brothers’ Nursery. 53. Oenothera cespitosa Nutt. in Fras. Catal. 1813. This species is amply described by Nuttall when he states that “This species is more perfectly stemless than Oe. acaulis, of Cavanilles, from which it is perfectly distinct. Flowers very large and white, with dilated obcordate petals.” Pursh described the same species from a Lewis collection as Oe. scapigera in Fl. Am. Sept. 1: 263. 1814, but as the species is adequately described in the Catalogue, it should be cited as the original place of publication. 54. Oenothera albicaulis Nutt. in Fras., nom nud. When Pursh described Oe. albicaulis in Fl. Am. Sept. 2: 734. 1814, he cited the Fraser's Catalogue as the source of his name, but did not have the same Evening Primrose in mind that Nuttall had when he proposed the name. As the Nuttall name in the Catalogue is a nomen nudum, the cita- tion of the Pursh species must be given as Oe. albicaulis Pursh without any reference to the Catalogue. Nuttall later redescribed his Oe. albicaulis from his material in Gen. N. Amer. Pl. 1: 245. 1818, and pointed out the error that Pursh had made in associating his species with that of Nuttall's. As the Nuttall name was a later homonym, we refer to his distinct species as Oe. nuttallii Sweet, Hort. Brit. ed. 2, 199. 1830. 55. Oenothera serrulata Nutt. in Fras. nom. nud. This species was later described without reference to the Cata- logue as Oe. serrulata Nutt., Gen. N. Amer. Pl. 1: 246. 1818. 56. Oenothera macrocarpa Nutt. in Fras. Catal. 1813. Al- though Nuttall adequately described this species there has been a reluctancy to take up the name because of the more commonly used name Oe. missouriensis Sims, Bot. Mag. 1968] Fraser’s Catalogue — Reveal 43 39: pl. 1592. Nov. 1813. Nuttall described the species as follows: “The fruit of this species readily distinguishes it from every other, being remarkably compressed, and furn- ished with four broad wings." It is my opinion that Pursh was not describing as new Oe. macrocarpa in Fl. Am. Sept. 2: 134. 1814, as generally assumed, but was giving the Nut- tall name to the plant that had been adequately described, and putting the Sims name in synonymy. It seems certain that Pursh was taking his name from Fraser's Catalogue even though it is not cited. The name “macrocarpa” is not one that one would just happen to come up with. When Nuttall wrote his Genera in 1818, he proposed an entirely different name for this species, Oe. alata Nutt., Gen. N. Amer. Pl. 1: 248. He cited “Oe. macrocarpa Ph" as a synonym, but as the name in the Catalogue was validly pub- lished, it can not be discarded, even by its author. In accepting this name, which has been recently accepted in Gleason & Cronquist (1963), the following new combina- tions are proposed following the recent treatment of the family by Munz (1965) : Oe. macrocarpa Nutt. in Fras. var. incana (A. Gray) Re- veal, comb. nov., based on Oe. missouriensis Sims var. incana A, Gray, Bost. Jour. Nat. Hist. 6: 189. 1850. Oe. macrocarpa Nutt. in Fras. var. oklahomensis (J. B. S. Norton) Reveal, comb. nov., based on Megapterium okla- homense J. B. S. Norton, Rep. Missouri Bot. Gard. 9: 153. 1898. 57. Orobus diffusus Nutt. in Fras., nom. nud. I have been unable to determine the exact identity of this nomen nudum. It is probably the same as O. dispar Nutt., Gen. N. Amer. PI. 2: 95. 1818, which is a synonym of Astralagus tenellus Pursh, Fl. Am. Sept. 2: 473. 1814. This species is often diffuse. Index Kewensis (Jackson, 1895) says that the nomen nudum is referable to Vicia americana Muhl. ex. Willd., Sp. Pl. 3: 1096. 1802, but Hermann (1960) makes no mention of this name in his revision of the genus. 44 Rhodora [Vol. 70 58. Oxytropis acuminata Nutt. in Fras., nom. nud. As Barneby (1952) points out, this nomen nudum must be ref- erable to O. lambertii Pursh, Fl. Am. Sept. 2: 740. 1814. The species was described with no reference to Fraser's Catalogue. Nuttall accepted the Pursh name without com- ment. 59. Phalangium esculentum Nutt. in Fras. Catal. 1813. Nuttall based this name on two elements which are now separated as distinct species. Both species have blue flowers, the character used by Nuttall to distinguish it from the other yellow-flowered species in the United States. The fact that we know he had in mind both species can be seen both in the original description in which he states that “The roots of this plant is eaten by the Savages of the Missourie and Columbia Rivers," and later in his Genera in which he discusses the wide distribution of the single species as he conceived it. The element of the species that Nuttall himself collected was illustrated by Sims, and a new combination proposed, as Scilla esculenta (Nutt. in Fras.) Sims, Bot. Mag. 38: pl. 1574. Aug. 1813.' This illustration was made from plants grown in the Fraser Brothers’ garden from seeds collected, according to Nuttall (1818), near the Huron River and also from around Saint Louis. The following year, Pursh described the second element of the Nuttall species, and I doubt that Nuttall ever saw this plant before 1813. Lewis had collected P. quamash Pursh in Fl. Am. Sept. 1: 226. 1814, from Clearwater Co., Idaho (Gould, 1942), and this was the plant Nuttall referred to in his comments about the Columbia River Indians who used the root for food. This information probably came to Nuttall from the fur traders and trappers that he was travel- ling with in 1811. The Columbia River plant was later col- lected by David Douglas, and from cultivated plants grown in England, Camassia esculenta Lindl., Bot. Reg. 18: pl. 'On the basis of this reference, the Catalogue is considered to have been published prior to August, 1813. 1968] Fraser’s Catalogue — Reveal 45 1486. 1832, was described and Lindley placed P. quamash in synonymy under his own new species. Lindley was ex- plicit in stating that his C. esculenta was not P. esculenta of Nuttall or S. esculenta Sims, even though he was using the same specific name “esculenta” for his new species. Gould (1942) handles the nomenclature properly, calling the eastern species that Nuttall had originally named P. esculentum, Camassia scilloides (Raf.) Cory, Rhodora 38: 405. 1936. He called the western species which Pursh had named P. quamash and later renamed C. esculenta by Lind- ley, C. quamash (Pursh) Greene, Man. Bot. Bay Reg. 313. 1894. 60. Podalyria mollis Michx., Fl. Bor. Amer. 1: 264. 1803. This species is questionably identified in the Fraser Cata- logue. It must have been the same plant that Nuttall later described as Baptisia leucophaea in Gen. N. Amer. Pl. 1: 282. 1818, which Nuttall stated was collected near Saint Louis. Podalyria mollis is now.known as Thermopsis mollis (Michx.) M, A. Curtis ex A. Gray. 61. Potentilla mollis Nutt. in Fras., nom. nud. I have been unable to determine the identity of this nomen nudum as it is not mentioned by Pursh or Nuttall. Nor has it been treat- ed in either of the two Rydberg revisions of the genus (1898, 1908). I suspect that it may possibly be the same as P. humifusa Nutt., Gen. N. Amer. Pl. 1: 310. 1818, not Willd., 1813, which is now known as P. concinna Richards. in Frankl. 1st. Journ. 739. 1823, since that was the only new Potentilla described by Nuttall. 62. Pycnanthemum dichotomum Nutt. in Fras., nom nud. It has been impossible to determine exactly what species of Pycanthemum Nuttall had in mind when he proposed this name. He described two new species in his Genera, and of these two, he apparently collected only P. pilosum Nutt. him- self. This species, published in Gen. N. Amer. Pl. 2: 33. 1818, was grown at the Fraser Brothers’ Nursery, and the nomen nudum may be referred to this species. 46 Rhodora [Vol. 70 63. Penstemon angustifolium Nutt. in Fras., nom. nud., as Pentstemon. This species was subsequently published as P. angustifolium Nutt. ex Pursh, Fl. Am. Sept. 2: 738. 1814, at which time Pursh cites the Fraser Catalogue as the basis for his name. Nuttall chose to reject this name stating that he could not refrain from adopting P. caeruleum Nutt., Gen. N. Amer. Pl. 2: 52. 1818, over P. angustifolium *. .. which I had formerly given . ." The species is now called P. angustifolium. 64. Penstemon grandiflorum Nutt. in Fras. Catal, 1813. I am inclined to consider this species as published and that the comments given by Nuttall are distinctive enough to make an unquestionable recognition of this species. Nuttall states: “Flower about the size and form of Digitalis pur- purea, and spontaneously subject to great variation in color.” Later, under P. erianthera, Nuttall comments: “P. grandiflorum is nearly the commonest species in that coun- try, and it was first met with near the confluence of the river Platte, from whence it continues to the Andes, fre- quently occurring in vast fields, together with the scarlet mallow, which form a brilliant object, visible at some miles distance.” Of the known species of Penstemon in 1813, this was cer- tainly the only large flowered species, and the similarity of the flower to Digitalis is apparent. Pursh described the same species as P. bradburii Pursh, Fl. Am. Sept. 2: 738. 1814, and this species name would have to be adopted if the Nuttall name should be deemed invalid. 65. Penstemon cristatum Nutt. in Fras., nom. nud. Nuttall described this species as P. cristatum in Gen. N. Amer. Pl. 2: 52. 1818, but by that time, Pursh had described the spe- cies as P. erianthera Pursh, Fl. Am. Sept. 2: 737. 1814. In attempting to apply the Nuttall name P. erianthera, Pursh mistakenly applied the name to the plant that Nuttall had called P. cristatum. Therefore, the name of this species must be P. erianthera Pursh. 1968] Fraser’s Catalogue — Reveal 47 66. Penstemon teretiflorum Nutt. in Fras., nom. nud. Nut- tall described this species as P. albidum in Gen. N. Amer, PI. 2: 53. 1818, citing his nomen nudum as a synonym. 67. Penstemon erianthera Nutt. in Fras., nom. nud. As noted above, under P. cristatum, Pursh misapplied the Nutt- all name to that species, and the plant that Nuttall had called P. erianthera, Pursh proposed P. glaber in Fl. Am, Sept. 2: 738. 1814. When Nuttall described his P. erianthera in Gen. N. Amer, PI. 2: 52. 1818, he placed the Pursh name in syn- onymy in an attempt to correct the Pursh error. However, the rules of nomenclature do not allow such superfluous sub- stitutes, and this species must be called P. glaber Pursh. 68. Passiflora lutea L. This species was listed as one of the more interesting plants which was available at the Fraser Brothers’ Nursery. 69. Petalostemon violaceum Michx., Fl. Bor. Amer. 2: 50. 1803. This species was listed as an interesting plant which was available at Fraser Brothers’ Nursery. The correct name for this species is Dalea purpurea Vent., Hort. Cels. pl. 40. 1800. 70. Petalostemon candidum (Willd.) Michx., Fl. Bor. Amer. 2: 49. 1803. This species was listed as an interesting plant which was available at the Fraser Brothers' Nursery. The correct name for this species is Dalea candida Willd., Sp. Pl. 3: 1337. 1802. 71. Psoralea macrorhiza Nutt. in Fras., nom. nud, Pursh described this species, naming it P. cuspidata Pursh, Fl. Am. Sept. 2: 741. 1814, and cited the Nuttall nomen nudum in synonymy. The Pursh name was accepted by Nuttall in his Genera without comment. 72. Psoralea esculenta Nutt. in Fras., nom. nud. Pursh pub- lished this species as P. esculenta in Fl. Am. Sept. 2: 475. 48 Rhodora [Vol. 70 1814, without reference to the Fraser Catalogue. Nuttall later, in Gen. N. Amer. Pl. 2: 102. 1818, credited himself with the name from Fraser’s Catalogue, and put the Pursh citation in synonymy. 73. Psoralea incana Nutt. in Fras., nom. nud. Pursh de- scribed this species as P. argophylla in Fl. Am. Sept. 2: 475. 1814, the name this species may be known as today. Later, Nuttall published his name P. incana in Gen. N. Amer. PI. 2: 102. 1818, and placed the Pursh name in synonymy. 74. Ribes longiflorum Nutt. in Fras., nom nud, According to Torrey & Gray, Fl. N. Amer. 1: 552. 1840, this species is R. aureum Pursh, Fl. Am. Sept. 1: 164. 1814. In 1834, Nutt- all claimed authorship of this name when he cited Wyeth's collection of Ribes aureum. Nuttall wrote: “Ribes longi- florum, Nuttall,in Fraser's Catalogue, 1813. So named one year previous to Pursh's publication." 75. Rudbeckia columnifera Nutt. in Fras. Catal. 1813. In my opinion, this species is adequately described in Fraser's Catalogue, as Nuttall states: “Spontaneous varieties of this plant sometimes occur with bright fulvous flowers, coloured like Tagetes patula: the stem is simple, seldom produc- ing more than three flowers, which are of an un- common length, appearing like a column of flosculi, subtended by 5-8 neutral florets, and a simple calyx.” This species was also described as R. columnaris Sims, Bot. Mag. 39: 1601. 1813, from cultivated plants given to Sims by the Fraser Brothers who had grown the species from seeds collected by Nuttall. The Pursh name, R. columnaris Pursh, Fl. Am. Sept. 2: 575. 1814, is an illegitimate substi- tute for Nuttall's R. columnifera, and is thus an exact syn- onym of it. The species is now commonly known as Ratibida columnifera (Nutt. in Fras.) Woot. & Standl., Contr. U.S. Nat. Herb. 19: 706, 1915. 1968] Fraser’s Catalogue — Reveal 49 76. Rudbeckia purpurea L. var. serotina Nutt. in Fras. Catal. 1813. In my opinion, this variety is validly published. Nuttall redescribed this variety in Gen. N. Amer. Pl. 2: 178. 1818, but with no reference to Fraser's Catalogue. In the Catalogue, the descriptive comments are: “Stem some- what branching and hirsute, fls. brighter and more numer- ous." This form is usually referred to Echinacea purpurea (L.) Moench. 77. Saccharum brevibarbe (Michx.) Pers., Syn. Pl. 1: 103. 1805, This species was listed in the Catalogue as one of the interesting plants that could be obtained from the Fraser Brothers’ Nursery. It is now known as Erianthus brevibar- bis Michx., Fl. Bor. Amer. 1: 55. 1803. 78. Seseli lucidum Nutt. in Fras., nom. nud. This name was never published, and Nuttall mentions it as a synonym under S. divaricatum Pursh, Fl. Am. Sept. 2: 732. 1814. Accord- ing to Mathias & Constance (1944), the nomen nudum is referable to Musineon divaricatum (Pursh) Nutt. ex Torr. & Gray var. hookeri Torr. & Gray. 79. Stylosanthes racemosa Nutt. in Fras., nom. nud. I have been unable to determine the identity of this nomen nudum. 80. Sideranthus integrifolius Nutt. in Fras., nom. nud. The name Sideranthus was proposed as a new genus, but as it lacks a description, it was not validly published. It was not until 1840 that Nuttall finally accepted the name as his. The identity of this species has not been definitely ascer- tained. Pursh, on the basis of a Bradbury specimen, placed S. integrifolius under his Amellus villosus in the Addenda et Corrigenda section of the Flora on page 750. Steudel, Nom. 775, 1821, also places S. integrifolius under A. villosus. The Steudel publication is often cited as the first place to indicate what the Nuttall nomen nudum refered to. 50 Rhodora [Vol. 70 81. Sideranthus pinnatifidus Nutt. in Fras., nom, nud. This name was referred to Amellus spinulosus Pursh, Fl. Am. Sept. 2: 750. 1814, an opinion in which Hall (1928) concurs under Haplopappus spinulosus (Pursh) DC., Prodr. 5: 347. 1836. 82. Sycios angulata. The genus name Sycios is an alternate spelling for Sicyos. The Fraser Brothers were offering for sale Sicyos angulatus L. through their Nursery. 83. Troximon ciliatum Nutt. in Fras., nom. nud. Pursh published this species as T. cuspidatum in Fl. Am. Sept. 2: 742. 1814. Later, Nuttall described T. marginatum in Gen. N. Amer. Pl. 2: 128. 1818, citing the Pursh species in syn- onymy. The species is now known as Microseris cuspidata (Pursh) Schultz-Bip., Pollichia 22-24: 309. 1866. 84. Troximon glaucum Nutt. in Fras., nom. nud. This spe- cies was described as T. glaucum Pursh, Fl. Am. Sept. 2: 505. 1814, but with no reference to Fraser’s Catalogue. Index Kewensis (Jackson, 1895) indicates that the name should be cited as Nutt. ex Pursh. The plant is now called Agoseris glauca (Pursh) Raf., Herb. Raf. in Extra of Atl. Journ. 6: 39. 1833. 85. Thuraria herbacea Nutt. in Fras., nom, subnud, Cron- quist (1957) stated that in his opinion this species is a nomen dubium, and this opinion he still holds. In corres- ponding with Dr. Cronquist on this species, he has written: *Neither Gutierrezia nor Euthamia can be excluded on the basis of the original publication alone, although one might have a good suspicion that the plant would turn out to be a Grindelia. The point is that a good suspicion is not enough. When it is not clear that the author is even trying to provide a botanical description, then his comments must make the application of the name unmistakably plain if the name is to be considered validly published." 1968] Fraser’s Catalogue — Reveal 51 The species was described as having “Fl. yellow, and the calyx resiniferous.” Nuttall also stated that the species occurred in open ground along the Missouri River. Dr. Alfred E. Schuyler of the Academy of Natural Sciences in Philadelphia has informed me that there is a collection of this species in the Academy Herbarium, but under the name Donia squarrosa. Pursh described D. squarrosa in Fl. Am.. Sept. 2: 559. 1814, which Nuttall later accepted in his Gen- era without comment or reference to the name published in Fraser's Catalogue. Although I feel certain that Nuttall had Grindelia squar- rosa (Pursh) Dunal, Mém, Mus. Par. 5: 50. 1819, in mind when he proposed his new genus Thuraria, he did not ade- quately separate it from related genera, and thus it can not be considered as validly published. 86. Uniola latifolia Michx., Fl. Bor. Amer. 1: 70. 1803. This plant was listed as one of the interesting new grasses that was available at the Fraser Brothers Nursery. 87. Virgilia grandiflora Nutt. in Fras. Catal. 1813. Nuttall described this species as follows: “This fine species differs essentially from V. bicolor, in being perennial, and produc- ing a simple stem, terminated by one to three large flowers: its entire pappus also distinguishes it from V. fimbriata." Pursh published this species as Gaillardia aristata Pursh, Fl. Am. Sept. 2: 573. 1814, thinking the plant to be an an- nual on the basis of the Lewis collection. On page 570, Pursh cites V. grandiflora from the Catalogue, and corrects his error, indicating that the species is perennial. Shinners (1956) stated that if V. grandiflora is consid- ered as validly published, then a new combination would be necessary. Cronquist (1957) however, pointed out that there is already a Gaillardia grandiflora Hort. ex Lemaire, Ill. Hort. 4: pl. 139. 1857, and thus the proper name for our plant is still G. aristata Pursh. 88. Vitis campestris Nutt. in Fras., nom. subnud, The only descriptive term used in the statement given by Nuttall is 52 Rhodora [Vol. 70 “bushy,” and it is my opinion that this does not validly distinguish this species from V. riparia Michx., Fl. Bor. Amer. 2: 231, 1803. The Nuttall name is referable to the Michaux species. 89. Yucca glauca Nutt. in Fras. Catal. 1813. Nuttall ade- quately described this species which was later named Y. angustifolia Pursh, Fl. Am. Sept. 1: 227. 1814. Later, how- ever, “Nuttall did not take up his own name, but chose that of Pursh, and made no reference to the earlier name in Fraser’s Catalogue. The original description is: “Leaves narrow, and filiferous; capsule dry, coriaceous, and large as that of Y. gloriosa." DEPARTMENT OF BOTANY BRIGHAM YOUNG UNIVERSITY PROVO, UTAH 84601 LITERATURE CITED BARNEBY, R. C. 1952. “A revision of the North American species of Oxytropis DC.” Proc. Calif. Acad. Sci. IV: 7: 177-312. 1964. “Atlas of North American Astragalus.” Mem. New York Bot. Gard, 13: 1-1188. BEETLE, D. E. 1944. *A monograph of the North American species of Fritillaria.” Madroño 7: 133-159. CLARK, R. B. 1942. “A revision of the genus Bumelia in the United States.” Ann. Mo. Bot. Gard. 29: 155-182. CRONQUIST, A. 1945. “Studies in the Sapotaceae III. Dipholis and Bumelia.” Journ. Arn. Arb. 26: 485-471. 1957. “Frasers Catalogue again." Rhodora 59: 100 ,D. D. Keck, & B. MAGUIRE. 1956. “Validity of Nut- tal's names in Fraser’s Catalogue." Rhodora 58: 23-24. GLEASON, H. A. & A. CRoNQUIST. 1963. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. Princeton, New Jersey. GOULD, F. J. 1942. “A systematic treatment of the genus Camassia Lindl.” Am. Midl. Nat. 28: 712-742. 1968] Fraser’s Catalogue — Reveal 53 GRAUSTEIN, J. E. editor. 1951. “Nuttalls travels into the old Northwest. An unpublished 1810 diary.” Chronica Botanica 14: no. %: 1-88. 1956. “Nuttall’s quarrel with Pursh.” Rhodora 58: 20-22. 1967. Thomas Nuttall, Naturalist. Explorations in America. 1808-1841. Harvard Univ. Press, Cambridge. GREENE, E. L. 1890. “Reprints of Fraser’s Catalogue.” Pittonia 2: 114-119. HALL, H. M. 1928. “The genus Haplopappus.” Carn. Inst. Wash. Publ. 389: 1-391. HERMANN, F. J. 1960. “Vetches in the United States — native, natu- ralized, and cultivated." U.S.D.A. Agric. Handb. 168: 1-84. HITCHCOCK, A. S., 2nd Ed. rev. by A. CHASE. 1951. “Manual of the grasses of the United States." U. S. Dept. Agric. Miscell. Publ. 200. JACKSON, B. D. 1895. Index Kewensis ... . London. LANJOUW, J., editor. 1966. International Code of Botanical Nomen- clature. Utrecht. MATHIAS, M. & L. CONSTANCE. 1944. “Umbelliferae.” North Amer. Flora 28B: 43-160. Munz, P. A. 1965. “Onagraceae.” N. Am. Flora, Series II: 5. 1-278. NUTTALL, T. 1813. A catalogue of new and interesting plants, col- lected in Upper Louisiana and principally on the River Missourie, North America. London. Prior August 1813. Copy at the Phila- delphia Academy of Natural Sciences; reprinted in Greene (1890). 1817. “Observations on the genus Eriogonum, and the Natural Order Polygonaceae of Jussieu." Journ. Phil. Acad. 1: 24-31, 33-37. 1818. The genera of North American plants .... Philadelphia. In two volumes. 1854. “A catalogue of a collection of plants made chiefly in the valleys of the Rocky Mountains or Northern Andes, towards the sources of the Columbia River, by Mr. Nathaniel B. Wyeth, and described by T. Nuttall.” Journ. Acad, Nat. Sci. 7: 5-60. 1840. ‘Descriptions of new species and genera of plants in the natural order Compositae . . . ." Trans. Amer. Phil. Soc. n.s. 7: 283-356. OwnBy, M. 1940. “Monograph of the genus Calochortus.” Ann. Mo. Bot. Gard. 27: 371-560. 54 Rhodora [Vol. 70 PENNELL, F. W. 1936. “Travels and scientific collections of Thomas Nuttall.” Bartonia 18: 1-51. PunsH, F. 1814. Flora Americae Septentrionalis .... London. In two volumes. RypBERG, P. A. 1898. “North American Potentilleae." Mem. Dept. Bot. Col. Univ. 1: 1-223 with 112 plates. 1908. “Rosaceae.” North Amer. Flora 22: 239-388. SHINNERS, L. H. 1949. “The genus Dalea (including Petalostemum) in North-central Texas. Field & Lab. 17: 85-89. 1955. “Non-validity of Nuttallian names in Fraser’s Catalogue." Rhodora 57: 290-293. 1956. “Nuttall not the author of Fraser's Catalogue.” Rhodora 58: 281-289. THEOBALD, W. L. 1966. “The Lomatium dasycarpum - mohavense - foeniculaceum complex." Brittonia 18: 1-18. Wuerry, E. A. 1947. “A key to the eastern North American lilies.” Bartonia 24: 5-8. A CYTOTAXONOMIC STUDY OF THE HERBACEOUS SPECIES OF SMILAX: SECTION COPROSMANTHUS* JOSE K. MANGALY The herbaceous species of Smilax (section Coprosman- thus), widely distributed both in the Old and the New World, have their greatest concentration in the eastern half of the United States. The genus has been conventionally placed in the family Liliaceae. Ventenat isolated it from the Liliaceae and placed it under Smilaceae. Rafinesque removed the herbaceous species from Smilax on the basis of their non-prickly stems and the presence of staminodes in pistillate flowers, and placed them under a new genus Nemexia. Later, Torrey relegated these taxa to a new section, Coprosmanthus within the genus Smilax, Kunth, in turn, raised section Coprosmanthus to the level of genus. The naturalness of the new genus has often been questioned and most recent authors refer to the herbaceous species *Part of a dissertation in partial fulfilment of the requirements for the degree of Doctor of Philosophy at the Catholic University of America. This study has been financed in part by a research grant from The Catholic University of America. I express my gratitude to Dr. Robert A. Davidson for his constant encouragement and constructive criticism. Also, my thanks are due to Dr. Ross H. Arnett, Jr., for making available his trailer truck for extensive field work; to Mr. Lee H. Herman for his assistance in collecting in the Southeastern States during the spring of 1963; to Mr. George B. Stebbins for the Latin description; and, to Rosalie A. Dunn for all her assistance. The problem was first suggested by Robert F. Thorne of Rancho Santa Ana Botanic Garden, Claremont. I gratefully acknowledge the assistance provided by the curators of the following herbaria: Botanischer Garten und Museum, Berlin- Dahlem; Conservatoire et Jardin Botaniques, Genéve; Muséum National d'Histoire Naturelle, Paris; Royal Botanic Gardens, Kew; and all the others mentioned in the text. Their assistance in making available the herbarium specimens, including types or their photo- graphs, was invaluable in completing this study. Voucher specimens and slides are deposited in the herbarium of The Catholic University of America. x 55 56 Rhodora [Vol. 70 under a section of Smilax. Small retained the genus name Nemexia in his Flora of the Southeastern United States and recognized eight species. Pennell, in his revision of the herbaceous species of Smilax, recognised nine species. Fer- nald, in Gray’s Manual Ed. 8, and Gleason in Manual of Vascular Plants, recognised only three species and two varieties. Thus, there is a diversity of views as to the number of species within this section and as to the circumscription of species and varieties. In many of the herbaria checked, mis- applied names are often the rule rather than the exception. Moreover, no cytological work has been done within these taxa except for Smilax herbacea. The necessity of a revision is thus apparent. The present study is an attempt to bring together all available information in order to clarify existing problems, especially at the specific level, with the hope that this synthe- sis may help to illuminate sectional relationships, Breeding experiments were not performed since seeds failed to germ- inate in culture. Acquaintances with local and regional populations, their habitats, and variability have been of extreme importance to the study. Chromosomal information has been proved to be a signifi- cant aid, especially in elucidating the matter of dioecism in these taxa. Except for one taxon, chromosome numbers are uniform throughout section Coprosmanthus. I feel that many of the problems existing in this complex have been solved by this attempt. The genus Smilax was proposed by Tournefort and later was included in Systema Naturae (1735) by Linnaeus. The genus included dioecious plants with flowers composed of six perianth parts, six stamens (more than six in the section Pleiosmilax), superior ovary of three carpels and 1-2 ovules in each locule. Rafinesque (1825) circumscribed the herbaceous species with perennial rhizomes and annual herbaceous, non-prickly aerial shoots in a new genus, Nemexia. Torrey (1843) classified Smilax herbacea (S. peduncularis, S. pulverulenta 1968] Smilax — Mangaly 57 and S. lasioneuron were recognised as synonyms of S. her- bacea) under the section Coprosmanthus, omitting the two other binomials, S. pseudo-china and S. tamnifolia, already published at that time. Kunth (1850) raised the section Coprosmanthus to the level of genus. De Candolle (1878) divided the genus Smilax into four sections, Nemexia, Eus- milax, Coilanthus, and Pleiosmilax, these being mainly based on the number of ovules, nature of perianth at an- thesis and number of stamens. Nemexia was characterized by recurved perianth, six stamens, and biovulate carpels. However, S. china and its relatives, with woody stems and uniovulate carpels, were relegated to the section Nemexia. Koyama (1960) suggested that de Candolle included S. china under the section Nemexia thinking that the section is char- acterized by a single ovule to a locule but, in fact, de Candolle mentioned the biovulate locule in his description of the sec- tion: “Flores musculi hexandri, sepalis petalisque plus minus excurvatis; feminei loculis biovulatis . . ." Most European authors, e.g., Bentham and Hooker (1883) and Krause (1930), followed de Candolle's system. How- ever, most American authors have recognised two sections, Eusmilax and Nemexia, for the North American species. These sections generally divide the woody from the herbace- ous taxa. Koyama (1960) divided Smilax (world wide basis) into six sections, Pleiosmilax, Smilax, Macranthae, Coilanthus, China and Coprosmanthus. 'The section Copros- manthus is based mainly on the somewhat reflexed nature of the tepals at anthesis, presence of staminodes in pistillate flowers, and the fact that the petioles and blades fall off together. By rules of priority, Coprosmanthus is to be re- tained for the section that includes all the herbaceous species in the United States. Linnaeus (1753) recognised two herbaceous species, S. herbacea and S. pseudo-china, from the North American Continent. Michaux (1803) added two more species, S. pulverulenta and S. tamnifolia. In the present paper, S. tamnifolia is recognised as a synonym of S. pseudo-china. Hooker (1840) added another species S. lasioneuron. Engel- 58 Rhodora [Vol. 70 mann (1835) recognised a new variety, S. herbacea var. ecirrhata. However, the new varietal epithet was first valid- ly published by Kunth (1850) as Coprosmanthus herbaceous var. ecirratus, Gray (1848) recognized two species and one variety in the section Coprosmanthus; Watson (1890) raised the variety S. herbacea var. ecirrhata to the rank of species in the 6th ed. of Gray's Manual. Small (1898) circumscribed two new species, S. tenuis and S. diversifolia, but later (1903) he re-established the genus name Nemezia for the herbaceous species Smilax and recognised two more species, N. hugeri and N. biltmoreana, The taxonomic position of S. tenuis is not clear, Even though the stem and leaves are slender, it resembles S. lasioneuron in many respects and representative specimens are ex- tremely rare in most herbaria checked. In the present treat- ment, S. diversifolia is included within S. lasioneuron. Pennell (1910) described a new variety, S. herbacea var. crispifolia, based on its relatively thick leaves, a character common among plants of this species growing in direct sun- light; in the present treatment, S. herbacea var. crispifolia is considered as an ecological variant. Pennell in 1916 at- tempted a taxonomic revision of the section and added one more species, S. leptanthera. This also, in the present treat- ment, is regarded as an ecological variant. Ahles (Radford, et. al. 1964) reduced S. hugeri and S. biltmoreana to two varieties of S. ecirrhata. The section Coprosmanthus has a wide distribution on the eastern half of the North American Continent extending from the Gulf area to the Lake region and northward into Canada. In all eight species recognized, the annual shoots develop very early in the spring. Flower buds are already developed by the time the shoots emerge from the soil. The flowering period generally lasts about two weeks for each species; however, there is considerable variation within and between species with regard to the time of flowering. Most species flower between early May and late June, Therefore, a major part of the field work was done during these months 1968] Smilax — Mangaly 59 in 1963, 1964 and 1965. Representative specimens and cyto- logical material were collected from Florida, South Carolina, North Carolina, Virginia, Maryland, Pennsylvania, Michi- gan, Illinois, and Wisconsin. Most of the materials for cytological studies (especially flower buds) were collected in the field. The most favorable time of fixation was found to be between 11:30 A.M. and 12:30 P.M. Materials collected at other times showed prac- tically no mitotic stages. In a few cases, however, good meiotic stages were obtained from flower buds collected early in the morning. Attempts were made to transplant rhizomes to the greenhouse. Shoots were cut back and the rhizomes were dug out and transplanted in pots and then transported to the greenhouse. Most specimens of S. ecirrhata complex, all with short erect shoots, survived while most specimens of S. herbacea complex, all with long climbing shoots, died in the greenhouse. More than 20 random population samples of Smilax her- bacea were collected from Virginia, Maryland, and Pennsyl- vania. A few population samples of S. lasioneuron and S. pulverulenta were also obtained. Smilax hugeri, S. biltmore- ana, and S. pseudo-china were represented by widely spaced individuals rather than by populations. Therefore, only single specimens of these species could be acquired. Meas- urements were made mostly on herbarium specimens. Care was taken to select the specimens from wide geograpnic ranges. The list of herbaria consulted is provided elsewhere. CHROMOSOMAL STUDIES Because of the high morphological and ecological varia- bility and wide distribution of the members of this section, it was thought early in the study that chromosomal infor- mation might be of considerable value in dealing with taxonomic problems. It soon became apparent that the num- ber n = 13 was constant for nearly all members of the sec- tion, the exception being S. pseudo-china. 60 Rhodora [Vol. 70 Methods: Root tip and flower bud squashes were used for chromo- somal studies. Both materials were pre-treated with 0.002 mol/L solution of 8-hydroxyquinoline (ortho) and were fixed in modified Carnoy’s fixative (4 chloroform: 3 absolute ethanol: 1 glacial acetic acid). The materials were kept in Carnoy’s fixative for 24 hours and then preserved in 70% ethanol or were used immediately after thorough washing. Initial use of aceto-orcein and aceto-carmine procedures was abandoned since the meristematic cells became highly re- fractory and because certain granular bodies became as heavily stained as chromosomes. A satisfactory staining procedure resulted from modify- ing Marimuthu and Subramaniam's (1960) hematoxylin squash technique. Both flower buds and root tips were hydro- lysed in a mixture of concentrated HCI and absolute ethyl aleohol (1:1) for about one minute, then washed thoroughly in distilled water and mordanted in 2% ferric ammonium sulphate for about 5-10 minutes. The material was then washed in distilled water and stained in a time-ripened solu- tion of hematoxylin for 10-15 minutes. (Other materials may require a longer staining period ; overstaining is seldom a serious problem.) The tissue was softened and destained in 45% acetic acid at 60°C. for about one minute, Both hydrolysing and destaining time depend on the size and tex- ture of the material. The material was then mounted in a drop of the same solution of 45% acetic acid and squashed under the coverslip. A quick survey of these temporary mounts was made, and good preparations were made perma- nent, using the CO, freeze technique, and mounted in green euparal. The chromosomes were stained dark blue with the background almost colorless. Drawings were made with the aid of a camera lucida. In some cases chromosomes were respaced to avoid over- lapping. Results: Only one previous chromosomal count was reported for this section, S. herbacea, n — 13 (Lindsay 1930). In the 1968] Smilax — Mangaly 61 following list, counts are recorded for plants of all 8 species comprising Coprosmanthus. All plants of the 8 species except those of S. pseudo-china have chromosome numbers 2n — 26 (Figs. 1-6 and 9-14). Two specimens of S. pseudo- china, both from South Carolina, have a chromosome num- ber 2n — 30 (Figs. 7 and 15). Within karyotypes, chromo- somes exhibit great variability with regard to their size and morphology. The longest chromosomes measure 6 microns np wey evo Ca e VAC uS C A z B 3 ) MS — ^Y bd Sa : | Käre, AN, tae’, faye" AR AS 0,0 e oe? qQ Sy 3 i \ P. So 1 3 a^. eq 7 a% a es Ze zb 4 a (A ZA Uw, "ON ANE, AM, 7, “48, die ME Oh deii St AN uw 9 wis, a, CM \ 8 | “q, * A Figs. 1-8. Mitotic chromosomes of Smilax species. — 1. S. lasio- neuron. — 2. S. herbacea. — 3. S. illinoensis. — 4, S. pulverulenta. — 5. S. hugeri. — 6. S. biltmoreana. — T. S. pseudo-china. — 8. an appa- rent hybrid of S. biltmoreana and S. hugeri, showing the extra chromosome fragments. Figs. 9-16. Meiotic chromosomes of Smilax species. — 9. S. Lasio- neuron — 10. S. herbacea. — 11. S. illinoensis. — 12. S. pulverulenta. — 13. S. hugeri.—14. S. biltmoreana. — 15. S. pseudo-china. — 16. an apparent hybrid of S. biltmoreana and S. hugeri showing the lateral pairing of the bivalents. 62 Rhodora [Vol. 70 while the shortest, 2 microns and there is gradation between these two extremes (Fig. 17). In all taxa, all chromosomes display subterminal centromeres. S. pseudo-china appears to have the smallest and S. illinoensis has the largest. The plants of the genus Smilax are dioecious. Lindsay (1930) made an extensive study of the chromosomes of S. herbacea and reported the absence of any heteromorphic pair. On the other hand, she noticed at the periphery of the spindle an unusual pair of chromosomes with near termi- nal centromeres and practically free ends. From her figures, this chromosome pair appears to be one of the largest with- in the karyotype. She also suggested that the folding of the free end of a chromosome might result in the members of the pair appearing unequal in length. Such a case can only be verified when both the chromosomes of a homologous pair are seen in the profile view. An attempt was made to determine the chromosomal basis of sex differentiation in this section. Microsporocytes of Smilax species here investigated disclose the presence of a 12.324 56 789 10 11 12 Fig. 17. Karyotype of S. pulverulenta. 1-12 somatic chromosomes, and the other two apparently sex chromosomes. pair of unequal chromosomes during synapsis. This pair, one of the largest within the karyotype, is fairly distinct up to diakinesis (Figs. 9, 11, 12, and 13). Even at this stage the difference in size of the two chromosomes is not very prominent, and gradually it becomes indistinct during late prophase and metaphase as the chromosomes become further contracted. It is likely that this pair of unequal chromo- somes with long free arms is the same that Lindsay observed in the periphery of the spindle. The presence of this hetero- morphic pair during synapsis in microsporocytes suggests the possibility that the sex differentiation in Smilax, section 1968] Smilax — Mangaly 63 Coprosmanthus, is by X and Y chromosomes, where XX is female and XY is male. Such distinct heteromorphic pairs are noticed only in four species where very good stages of early prophase of the 1st meiotic divisions were obtained. In all other species this could not be verified. Therefore, at the present time this observation is far from conclusive. Two specimens with perfect flowers were found in the herbarium materials checked. Both specimens were collect- ed from Wisconsin. Attempts to collect bisexual living materials from their original localities were not successful. Therefore, no cytological investigation could be made. CHROMOSOME NUMBERS n 2n Coll. Mi- Taxon No. County Source PMC tosis Smilax VIRGINIA herbacea 219 1.5 mi, south of Lewis Mt., Shenandoah National Park — 26 2135 Lewis Mountain, Shenandoah National Park Grae DO 2187 4 mi. south of Lewis Mt., Shenandoah National Park 13 — 2136 4 mi. south of Lewis Mt., Shenandoah National Park 13 — PENNSYLVANIA 2002 Lycoming Williamsport, Pa. .. — 26 2073 Clearfield 2 mi. south of Coal Hill on Route 219, Pa. .. — 26 S. lasioneuron MICHIGAN 2145 Washtenaw East of Delhi Rd., Delhi — 26 ILLINOIS 2146 Cook 2 mi. west of Roselle on Route 72 e... — 26 2147 Cook 2 mi. west of Roselle on Route 72 tm. Cada MU 2148 Cook 2 mi. west of Roselle on Route 72 cuc. ur 20 64 Rhodora [Vol. 70 n 2n Coll. Mi- Taxon No. County Source PMC tosis WISCONSIN 2152 Iowa Jct. Rt. 18 and K, 1 mi. on K. 2... — 26 2154 Iowa Jet. Rt. 18 and K, 1 mi. on K. 2... — 26 2163 Iowa On Route 23 .. — 26 2166 Towa Jet. Rt. 23 and Rt. 14 13 — 2167 Iowa Jet. Rt. 23 and Rt. 14 13 26 2168 Iowa Jct. Rt, 28 and Rt. 14 13 26 2170 Iowa Jet. Rt. 23 and Rt. 14 — 26 S. pulverulenta VIRGINIA 206 Bath Doutat State Park — 26 210 Allegheny Strom, between Covington and Eagle Rock .. — 26 MARYLAND 231 Mont- Cabin John Park .. — 26 gomery 2123 Mont- Cabin John Park .. 13 26 gomery 2124 Prince Laurel, Rt. 1, 15 mi. Georges north of D.C. .. — 26 2126 Prince Laurel, Rt. 1, 15 mi. Georges north of D.C. .. 13 26 2127 Prince Laurel, Rt. 1, 15 mi. Georges north of D.C. .. — 26 S. pseudo-china SOUTH CAROLINA 270 Sumter 6.8 mi. s.w. of Sumter on Route 15 .. — 30 275 Orange- 2 mi. from Jet Rt. 176 burg and 15, on Route 176 .. 15 30 S. illinoensis MICHIGAN 2143 Washtenaw E. Delhi Road, Delhi .. 13 26 2144 Washtenaw E. Delhi Road, Delhi .. — 26 S. hugeri FLORIDA 282 Jackson Marianna Caverns, Fla. 13 26 1968] Smilax — Mangaly 65 n 2n Coll. Source PMC tosis Taxon No. County Mi- 285 Jackson Marianna Caverns, Fla. 13 26 291 Jackson Marianna Caverns, Fla. — 26 S. biltmoreana FLORIDA 281 Jackson Marianna Caverns, Fla. 13 — 290 Jackson Marianna Caverns, Fla. — 26 283 Jackson Marianna Caverns, Fla. — 26 S. ecirrhata WISCONSIN 2180 Iowa Jct. Rt. 23 and Rt. 14 — 26 CHROMOSOMAL ABERRATIONS IN THE SMILAX BILTMOREANA-S. HUGERI COMPLEX Exomorphologically intermediate forms of Smilax bilt- moreana (Small) J. B. Norton and S. hugeri (Small) J. B. Norton are collected from undisturbed mixed woodland habitats near Marianna, Florida. Careful examination of herbarium specimens of these species indicates that popula- tions from certain other localities are similarly confusing taxonomically. In these cases, misapplied names are the rule rather than the exception. This morphological evidence has suggested interspecific hybridization as a working hypothesis. A total of fourteen plants from the Marianna population collected in the spring of 1964 was examined cytologically. Root tip and flower bud squashes were used for chromosomal studies. Both materials were fixed and stained as described previously. Nine of these were morphologically typical of the species. These plants exhibited a diploid chromosome- number of 26, and meiotic behavior was normal for the approximately 150 PMCs examined. However, three of the remaining five intermediate forms exhibited extra-chromo- some fragments varying from two to eight, in addition to the normal chromosome complement (Fig. 8). Abnormal meiotic pairing is conspicuous in intermediate forms. The synapsed chromosomes show lateral connections 66 Rhodora [Vol. 70 Plate 1372. Figs. 18 and 19. Photomicrographs of pollen mother cells of Smilax hugeri — S. biltmoreana hybrids showing the unusual synaptic configurations. 18. thirteen bivalents in two small rings connected together. 19. the bivalents becoming free— six pairs still inter- connected. (1200 X). between adjacent bivalents and without exception are ar- ranged in rings or chains (Figs. 16, 18, and 19). Usually ten pairs form an outer ring with the other three remaining inside, interconnected. In most cells examined the later stages of meiotic divisions were apparently normal. Occa- sionally, however, lagging chromosomes were observed in anaphase. In anaphase I, the bivalents disengage and then continue normal anaphasic separation. Examination of pollen under the microscope indicated that 35% of the pollen was shrivelled up in these plants which showed the cytological ring formation. Attempts were made to test pollen germination, Incubation at room tem- perature in 3% sucrose resulted in pollen tube formation in only 17% of 1000 grains examined. The observed chromosomal aberrations in Smilax hybrids differ from the translocation systems in Oenothera and in Rheo by the presence of bivalent rings and chains. Although the available data are insufficient for a conclusive appraisal of the process they suggest that these aberrations result from a large number of reciprocal translocations of very 1968] Smilax — Mangaly 67 small units. Sterility of hybrids due to structural hybridity involving small chromosomal segments was suggested by Sax as early as 1933, for Campsis chinensis X C. radicans, and a large number of cases are known where hybrids show complete or partial sterility. The suggestion of Stebbins (1950) on the origin of translocations in connection with cryptic structural hybridity seems appropriate for Smilax. Equally important is the fact that when a translocation re- sults in a deficiency in one chromosome, it also results in a duplication in another chromosome. If the chromosome complements of the hybrid parents have more translocated duplications than deletions it is likely that these hybrids might show more of structural aberrations at synapsis and less of sterility. (A manuscript with details of these obser- vations has been submitted for publication. ) A few cytological problems remain unsolved. Two bi- sexual individuals collected from Wisconsin were found among the herbarium specimens. Attempts to collect such specimens in the spring of 1965 for cytological studies were not successful. All species of this section except S. pseudo-china produce only one peduncle from a node. But some specimens of S. pseudo-china, especially from South Carolina and coastal Virginia, have three peduncles at the same node, a central long one and two shorter lateral ones. Whether there is a cytological basis for this atypical expression has not been determined. MORPHOLOGIC AND TAXONOMIC CRITERIA It is very difficult to evaluate morphological characters which are used taxonomically in this section, as most features are extremely variable in relation to ecological and climatic variations. This is particularly true in the case of the S. herbacea complex where the relative length of the peduncle and the leaf vary on the same plant. Also there is considerable variation in leaf shape between young and older plants. Consequently, taxonomic distinctions are mainly based on a combination of morphological features. 68 Rhodora [Vol. 70 The characters listed below are useful principally in the determination of specific categories. Habit, Smilax ecirrhata and associates are all small erect plants with no tendrils or with a few rudimentary ones. Smilax illinoensis, an intermediate form between S. lasio- neuron and S. ecirrhata, is also erect and has a large number of tendrils mostly on the upper part of the stem. All others are fairly tall and climbing. All have perennial rhizomes with a few strong fibrous roots. The juvenile stage lasts about 3-6 years. Shoots from mature rhizomes of the S. herbacea complex often produce a few branches. Among older plants, more than one shoot usually appear from the same rhizome in one season. Leaves. The shape of the leaf, number of veins, presence or absence of pubescence, character of tip and base, relative length of petiole and blade were used to differentiate species. The floral characters in this group are evolu- tionarily conservative and, therefore, vegetative characters like the morphology of the leaf have been used in delineating species even though such characters are ecologically plastic. Leaf shape varies not only on different individuals of the same species, but on the same plant, depending on its loca- tion on the plant, and also among plants of different age. So extreme care should be taken to define species, with emphasis on the leaf shape. Another significant feature in the case of plants of the S. herbacea complex is that on older plants there is a gradual transition from scaly bracts at the base of the stem to the normal foliage leaves above. Pubescence and non-pubescence on the under-surface of the leaves of different species appear to be fairly constant. Of the three species with non-pubescent leaves, two grow mostly at higher elevations along the Appalachian Moun- tains and the third is restricted to the eastern Coastal Plain. Species with pubescent leaves extend to higher elevations too. This is especially true of S. lasioneuron which is dis- 1968] Smilax — Mangaly 69 tributed mostly in the central plains but is also found in the Great Smoky Mountain region and the eastern part of the Rockies, Inflorescence. The inflorescence is a few to many flowered umbel. Its place of origin, whether axillary to bracts or not, is considered as an important taxonomic character for separation of species. The number of umbels and the number of flowers on each peduncle are also useful. Perianth. The shape and the relative length and breadth of tepals are useful. In general, the inner ones are progres- sively narrower. Often the pistillate flowers have smaller tepals. In all species the tepals are light green and free. Stamens. In most species the ratio of length of filament to that of another is not uniform within the same flower, the outer ones with longer filaments and the inner ones slightly shorter. In general, however, relative length of anthers and filaments was found to be useful, In most species the anthers are shorter than the filaments; only in one are the anthers longer. These features are reliable only if the flowers are fully open. Fruits. The fruit is a one-seeded to six-seeded berry. It is relatively uniform in shape, but color varies from blue to black among different species. The fruit wall is either glaucous or not. The number of seeds in a berry does not appear to be a species character even though the short S. ecirrhata complex has fewer seeds per fruit. Branching. Only in the S. herbacea complex is branching prevalent and here the pattern of branching is almost the same in all species. Branches usually develop on the upper half of the stem. The leaves on the branches in general are smaller than those on the mature stem and vary more in shape. Thus, for most taxonomic purposes, only leaves on the stem were considered. 70 Rhodora [Vol. 70 FLOWERING Preliminary investigations suggested that flowering in this section is not photoperiodically controlled. In early spring, when the annual shoots sprout, the flower buds develop along with them, even before the leaves fully expand. Thus, flowering begins before the shoot is exposed to any photoperiodic influence. As expected, the southern forms of a given species flower slightly earlier than the northern ones. Smilax hugeri is the earliest to flower, beginning in early April and reaching its peak in the third week of April. Smilax pseudo-china, occurring sympatrically with S. hugeri in the Carolinas, reaches its peak flowering in May. Smilax biltmoreana, which occurs mostly in the mountains, has its peak flowering period during the second and third weeks of May. Smilax pulverulenta reaches its peak flowering in the second week of May and S. lasioneuron, with a farther northward distribution, has a peak flowering period from ' H LI a s a E) D eau? id MAY "yume Fig. 20. Bar diagram showing the flowering interval in Smilax species: 1. S. herbacea. — 2. S. lasioneuron — 3. S. pulverulenta. — 4. S. pseudo-china. — 5. S. illinoensis. — 6. S. ecirrhata. — T. S. h ugeri. The solid black line indicates the major interval of flowering and the dark spot in the middle represents the mode The dotted lines show the interval in which flowering is irregular. 1968] Smilax — Mangaly 71 the third week of May in the Mid-central Plains to the second week of June in the northern parts. These two species have been recognised by some authors as two dif- ferent varieties of S. herbacea. The difference in the flowering period between these two species might partly explain the rarity of intermediate forms as a result of partial or complete sexual isolation, even though both exist sympatrically in the Mid-central Plains. Smilax herbacea reaches its peak flowering in the first week of June in the Carolinas and Virginia; it extends to the end of June in the New England States. Smilax ecirrhata and S. illinoensis reach their peak flowering in the third week of May, but S. illinoensis has a slightly more extended flowering period, ranging from the second week of May (Fig. 20). POLLEN STUDIES Pollen morphology is fairly uniform in the section, The exine is slightly spinous, 1-2 microns in thickness, and with no distinct furrow. Pollen size varies from 16.8 to 32.2 microns. Smilax pseudo-china has the smallest pollen with a mean size of 23.0 microns; the largest is that of S. ecirrhata with a mean value of 26.24 microns Five plants were selected from each species and from each plant 50 pollen grains were measured at random. The mean value for each species was determined from these measurements. All together more than 2000 pollen grains from the eight species were measured. Species Mean Size of Pollen S. herbacea oi 24.75 microns S. pulverulenta SA 24.47 microns S. lasioneuron SH 24.48 microns S. illinoensis M 25.54 microns S. ecirrhata i 26.24 microns S. hugeri e 24.44 microns S. pseudo-china 23.00 microns Pollen measurements are of particular interest in three species. The mean size of pollen in S. lasioneuron is 24.48 72 Rhodora [Vol. 70 microns; in S. ecirrhata, 26.24 microns; in S. illinoensis, an apparent introgressant hybrid of the preceding two, 25.54 microns. The morphology of the tendrils in Smilax has long been controversial. The tendrils have been homologised with many other plant organs; modified stipules; bipartite liguils, each tendril representing a demiligule; meta- morphosed trichomes or emergences. and metamorphosed leaflets of à compound leaf, After extensive anatomical and morphological studies of the tendrils in S. herbacea, (Agnes) Arber (1920) demonstrated conclusively that these tendrils have arisen by chorisis or dedoublement of the petiole and so are equivalent to them. Working with S. macrophylla, Chakravarthi and Mitra (1948) have also reached the same conclusion that the tendrils are formed by chorisis or splitting of the petiole. MEASUREMENTS OF MORPHOLOGICAL CHARACTERS All measurements were made from dried herbarium specimens. In all cases, the largest leaf with its axillary peduncle, if any, was selected for measurement. The leaves on the branches were not used, as they differed from those on the main stem both in shape and size. SPECIFIC CONCEPT The section Coprosmanthus stands isolated from the woody sections by the presence of the annual herbaceous unarmed shoots arising from a perennial rhizome, and bi- ovulate carpels in contrast to the uniovulate carpels in the woody Smilax. Yet the differences between them are not sufficient enough to cireumscribe them under two separate genera as suggested by Rafinesque (1825) and accepted by Small (1903), and so the author prefers to keep the herba- ceous species under the section Coprosmanthus within the genus Smilax. The group as such has long been considered fairly distinct, the only problem being limited to the specific level. Observations of natural populations and greenhouse materials indicate that the shoots are highly sensitive to ecological variations, and show marked variation in ex- 1968] Smilax — Mangaly 73 ternal morphology, e.g., relative size of the leaves and peduncle, texure of leaves and height of the plant. In the presence of such ecological plasticity, it is likely that shoots developing from the same rhizome in different years might show morphological variation sufficient to result in the plants being considered as intraspecific categories. There- fore, prolonged observation under constant ecological con- ditions is necessary before attempting to designate such categories within species. No such intraspecific categories are recognised in this study. GENERIC RELATIONSHIPS AND TAXONOMIC POSITION The circumscription of the genus Smilax is generally accepted. The section Coprosmanthus represents a closely knit group of herbaceous species within the genus Smilax: smooth annual shoots from the perennial rhizomes ; presence of staminodes on pistillate flowers; and the presence of biovulate carpels. The uniformity of the floral and vege- tative characters is remarkable. The distinctions between the herbaceous and woody species are listed below: Herbaceous Smilax Woody Smilax 1. Shoots annual and herbaceous Shoots perennial and woody 2. Stem smooth and unarmed Stem armed (spinous) 3. Shoots short, erect, or up to Shoots generally much taller 3 meters high and climbing and climbing 4. Peduncles in general long Peduncles mostly short 5. Two ovules in each carpel One ovule in each carpel DISTRIBUTION, ECOLOGY AND EVOLUTION The section Coprosmanthus has a wide distribution which extends throughout the eastern half of the North American Continent. All species prefer very moist, shady woods. Smilax lasioneuron thrives well even in dry soil. Most species prefer the plains. Only two species, S. herbacea and S. biltmoreana, have their distribution almost exclu- sively on mountains. Smilax pseudo-china is restricted to the eastern Coastal Plain, extending from Georgia to New Jersey and southern New York. (Fig. 21). Smilax herbacea, like S. biltmore- ana, prefers higher elevation and is reported ag occurring 74 Rhodora [Vol. 70 - Fig. 21. Geographic distribution of Smilax: 1. S. herbacea. — 2. S. lasioneuron. — 4. S. pseudo-china. only along the Appalachian and Blue Ridge Mountains. Thus its distribution extends from the Carolinas to the New England States and southeastern Canada (Fig. 21). Smilax biltmoreana has a fairly restricted distribution. It is found in parts of North Carolina, South Carolina, Tennessee, Kentucky, Georgia, Alabama, and northern Florida (Fig. 23). Smilax hugeri has a more eastward dis- tribution, extending to the Atlantic Coast. This species is also found in Tennessee and Kentucky (Fig. 22). The distribution pattern of Smilax pulverulenta is rather unique. This species spreads across the Great Plains and reaches the Atlantic Coast through Virginia, Maryland, and New Jersey. It thus has a sympatric distribution with four other species of this section (Fig. 22). Smilax lasioneuron is widely distributed throughout the Coastal Plains and northward into Canada (Fig. 21). The major distribution range of S. ecirrhata is the northern Central Plains (Fig. 23) and southern Canada. 1968] Smilax — Mangaly 75 Fig. 22. Geographic distribution of Smilax: 3. S. pulverulenta. — 5. S. Illinoensis. —7. S. hugeri. Smilax illinoensis is distributed mostly in the Great Lakes Region, but it also extends to northern Missouri (Fig. 22). This species, a morphological intermediate be- tween S. lasioneuron and S. ecirrhata and identified in most 76 Rhodora [ Vol. 70 Fig. 23. Geographic distribution of Smilax: 6. S. ecirrhata. — 8. S. biltmoreana. 1968] Smilax — Mangaly q herbaria as either of these two species or both, exists sym- patrically with them. The greatest concentration of S. illinoensis is in Wisconsin, northern Illinois and Michigan and along the northern shores of the Great Lakes. — S. pulverulenta shoot | S. pseudo-china tall & climbing S. herbacea | | (S. herbacea L complex) | S. lasioneuron Section S. illinoensis Copros- | manthus shoot S. ecirrhata short & erect | (S. ecirrhata tí S. hugeri complex) E S. biltmoreana Both paleontology and comparative morphology strongly suggest the ancestral home of Smilax, both herbaceous and woody, to be in southeastern Asia, east of the Himalayas. The phylogeny of the section Coprosmanthus is not well understood. In eastern Asia, where the genus Smilax might have originated, the sectional barriers generally break down. Possibly, the Asian species S. riparia, widespread in Japan, Korea, and China, might give some insight into the probable ancestry of the section Coprosmanthus. Like all other herbaceous species, S. riparia produces annual shoots from the rhizome throughout Japan, Korea, and northern China, but in southern China the shoots are perennial. Basing his conclusion on fossil evidence and comparative morphology, J. B. Norton (1916) traced both the eastward and westward migration route of Smilax from Asia to North America, He suggested that the herbaceous species of Smilax reached continental North America by an east- ward migration along the path of S. hispida, by island hopping across the Pacific. These migrant ancestral forms might have been fairly well established in the central Cali- fornia region in Late Miocene or in Early Pliocene (Fig. 78 Rhodora [Vol. 70 24). Any suggestion as to the establishment, further migra- tion, and diversification of the group, in the absence of con- clusive fossil evidence, remains purely hypothetical. Fossil evidence of the vegetation both in Miocene and Mid-Pliocene indicate that in California both these epochs were comparatively dry and warm with a summer rainfall of 15-17 inches and the vegetation appears to have been similar to that of the present vegetation in this region. However, in the early Pliocene the climate was more or less equivalent to the present climate of eastern North America, and eastern Asia, with a minimum summer rainfall esti- mated to have been 25 - 30 inches (Axelrod, 1948). Such a moderate humid climate is ideal for Smilax and its presence in the California region during the Early Pliocene has been established by fossil evidence. During the Middle Pliocene, when the climate became progressively warmer and drier, a general eastward migra- tion might have occurred. Fossil evidence in western Kansas and Nebraska indicates clearly that the Middle Pliocene was the stage of greatest diversification of fossil types in contrast to the gradual changes in Miocene and Early Pliocene floras. Post Middle Pliocene was a time of large- scale mountain building, and as a result, a great diversifica- tion of the topography and habitat occurred. Added to that, the fluctuating Pleistocene climate kept the entire area in a continuous flux for a long time (Axelrod, 1948). This might have resulted in further diversification of the group that continues to the present time. Whether the divergence of the section Coprosmanthus into the tall Smilax herbacea complex and the short S. ecirrhata complex was completed in Asia itself or the di- vergence occurred in continental North America is un- certain. This differentiation was probably distinct by the time the herbaceous species became established in Kansas and Nebraska during Late Pliocene. Probably at this time or even earlier, further diversification had already begun in the S. herbacea complex, with one branch providing the precursors of S. pulverulenta (Fig. 24). This new species 1968] Smilax — Mangaly 79 S. HERBACEA S. LAS | ONE U RON S. PULVERULENTA S. PSEUDO-CHINA S. I LLI NOENSIS S. ECIRRHATA S. HUGERI S. BILTMOREANA > BS Fig. 24. Possible migration routes and diversification of herba- ceous Smilax on Continental North America. OO d OU WD — ` -1 gradually spread further eastward and northward, event- ually reaching the Atlantic Coast. The other branch of S. herbacea complex also continued the general eastward migration and soon bifurcated. One branch, the ancestral forms of S. lasioneuron, spread mostly northward into Canada as the Pleistocene glaciers receded. The other branch, progenitor of S. herbacea, reached the Blue Ridge Mountains and spread northward all the way to eastern Canada along the Appalachian Mountains. The phylogeny of Smilax pseudo-china is not clear. How- ever, the morphological resemblance of this species to S. herbacea and the presence of higher chromosome numbers provide a working hypothesis that S. pseudo-china origi- nated by interspecific hybridization of S. herbacea with some other Smilax species. The other major group, the ancestral forms of Smilax ecirrhata complex, also first diverged into two groups. One branch, the fore-runners of S. ecirrhata, slowly spread northward in the same direction of S. lasioneuron and be- 80 Fig. 25. Rhodora A. Smilax illinoensis and B. Smilax ecirrhata. [Vol. 70 1968] Smilax — Mangaly 81 came established in the northern central plains towards the end of Pleistocene. Hybridization might have been fairly common between these two sympatric species. In suitable habitats around the Great Lakes, these hybrids, by further introgression, formed S. illinoensis. The other branch of the Smilax ecirrhata complex con- tinued to migrate eastwards into Tennessee and Kentucky and slowly diverged into two groups. One group, the pro- genitors of S. biltmoreana, spread along the mountains in Tennessee, Kentucky, the Carolinas, and Georgia. The other group migrated further eastward into the Carolinas and evolved into S. hugeri. In western Tennessee and Kentucky, the species differences of the S. ecirrhata complex are not as sharp as elsewhere, suggesting that the divergence of this complex probably occurred in this region. Following is the list of herbaria from which specimens were borrowed during this study. The abbreviations used in citation of specimens are those recommended by Lanjouw and Stafleu (1963). Ann Arbor, Michigan; University Herbarium, University of Michigan (MICH) Athens, Georgia; Herbarium of the University of Georgia (GA) Austin, Texas; Herbarium of the University of Texas (TEX) Cambridge, Massachusetts; Arnold Arboretum of Harvard University (A) Cambridge, Massachusetts; Gray Herbarium of Harvard University (GH) Chapel Hill, North Carolina; Herbarium of University of North Carolina (NCU) Chicago, Illinois; Field Museum of Natural History (F) Gainesville, Florida; Herbarium, Agricultural Experiment Station (FLAS) Knoxville, Tennessee; Herbarium of the University of Tennessee (TENN) Laramie, Wyoming; Rocky Mountain Herbarium, University of Wyo- ming (RM) Madison, Wisconsin; Herbarium of the University of Wisconsin (wis) Morgantown, West Virginia; Herbarium of West Virginia University (WVA) 82 Rhodora [Vol. 70 New York, New York; New York Botanical Garden (NY) Philadelphia, Pennsylvania; University of Pennsylvania Herbarium (PENN) Philadelphia, Pennsylvania; Academy of Natural Sciences (PH) St. Louis, Missouri; Missouri Botanical Garden (MO) Tallahassee, Florida; Herbarium, Florida State University (FSU) Washington, D.C.; Herbarium, Catholic University of America (LCU) Washington, D.C.; National Museum, Smithsonian Institute (US) TO BE CONTINUED VARIATION IN CONE MORPHOLOGY OF BALSAM FIR, ABIES BALSAMEA" DONALD T. LESTER Balsam fir, Abies balsamea (L.) Miller, is a widely dis- tributed forest tree species in North America. The extent of its distribution is problematical in that Boivin (1959) has proposed the reclassification of the alpine fir of western United States and Canada, A. lasiocarpa, as a subspecies of balsam fir. Such reclassification would markedly extend the natural range and the amount of variation in taxonomic characteristics currently recognized within A. balsamea. Considering only the more eastern taxon (Figure 1) as A. balsamea (L.) Miller, recognition of variation in the species has been largely restricted to observations on the relation- ship between the length of cone scales and the length of the bract that subtends each cone scale. Taxonomic identi- fication of variation in balsam fir was proposed by Fernald (1909) with assignment of the varietal name phanerolepis to variants in which bracts were sufficiently long to be exposed in the mature cone (exserted). The variety phane- rolepis has been shown to be most common in the Maritime Provinces of Canada, the St. Lawrence Valley, and at higher altitudes on mountains of the northeastern United States (Fernald, 1950; Myers and Bormann, 1963). The scale- bract relationship has been the chief characteristic used in maintaining Fernald’s varietal distinction. Boivin (1959) noted the absence of discontinuities in the bract-length/cone scale length ratio in a sample of cones from Quebec but maintained the variety phanerolepis in his classification. Myers and Bormann (1963) reported altitudinal and longi- tudinal gradients in the scale/bract ratio and questioned the taxonomic validity of variety phanerolepis. *Work supported in part by the North Central Regional Project, NC-51, and the Wisconsin Conservation Department. Approved for publication by the Director, Wisconsin Agricultural Experiment Sta- tion. 83 84 Rhodora [Vol. 70 604 — ` o° Lo d 504 — 40” 10° Figure 1. Natural range of balsam fir (....), location of collec- tions (.), and score of bract length/scale length ratios. Scores repre- sent eight equal subdivisions of the array of ratios ranging from .44 to .80. The study reported here represents a westward extension of the area sampled in earlier studies, a more intensive study of variation in cone morphology of balsam fir, and analyses of possible associations between variation in cone morphology and climatic variables. This study is a part of investigations of variation in balsam fir which include seed source or provenance tests currently in progress. METHODS In the period 1960 to 1966, seed collections of balsam fir were made by personnel of the University of Wisconsin and by cooperators in the United States and Canada. One hun- dred populations representing much of the readily accessible natural range were sampled. As an adjunct to the collection of cones for seed, intact cones from 49 populations (Figure 1) were preserved in formalin-acetic acid-alcohol solution. Ten of the cone collections were made in 1960 (northern Wisconsin and Michigan), 41 in 1962, one (Iowa) in 1964, and 5 in 1966. Elevations of the collections were generally within the range of 500 to 2000 feet although one New York collection came from an elevation of 3600 feet. Two cones, randomly chosen from each tree, were preserved and 1968] Abies balsamea — Lester 85 from two to ten trees were sampled at each location. Sample trees generally were scattered over an area of a few acres although some sample trees were separated by distances of up to three or four miles. On each cone, cone width at the widest point and cone length were measured. Three cone scales were then removed from the middle of each cone for the measurements illus- trated in Figure 2. These measurements were made using an ocular micrometer in a dissecting microscope. Ten variables were studied. Measurements of length of cone, ovuliferous scale, seed wing, and bract were analysed F54 A. B Figure 2. Sketches of abaxial view of ovuliferous scales and bracts. A. Maine collection, bract-scale ratio is .85. B. Western Ontario collection, bract-scale ratio is .50. Indicated measurements are 1. scale length, 2. bract length, 3. stalk length, 4. scale width, and 5. bract width. directly. Due to the high frequency of zero, values for stalk length were transformed to \/x + 0.5 before analysis, Ratios of length/width were calculated and analysed for cones, scales, wings, and bracts, along with the ratio bract length/scale length as a measure of bract exsertion. Four sources of variation were hypothesized; variation associated with geographic location, with differences be- tween trees in locations, with differences between cones within trees, and with differences among measurements 86 Rhodora [Vol. 70 within cones. Analyses of variance (Table II) were used to estimate whether each of the proposed sources of variation contributed significantly to the total variation. The relative contribution of each source was estimated by extension of the analysis of variance to computation of variance compo- nents. Speculative associations between cone morphology and climatic variables were tested by linear correlation using mean values for each geographic location and for weather data from the station nearest to the collection location. Long-term averages were used; the period of recorded ob- servations being 30 years for temperature and precipitation, and generally 10 years or more for frost-free period. Prox- imity of weather stations to cone collections was sometimes poor with distances being up to 50 miles in northern and western regions. RESULTS VARIATION IN CONE MORPHOLOGY A sample of the arrays of average measurements for each location illustrate several points about variation in cone morphology in balsam fir (Table I). The arrays show a generally continuous pattern of variation with the excep- tion of one New Brunswick collection in which length of the ovuliferous scale was significantly greater (99% proba- bility) than in any other location, Although collections from the eastern portion of the sampled range were mostly among those with the longest cones, no clear geographic grouping was apparent for most of the characteristics measured. Cone length was chosen as the variable with which to organize the sample of arrays in Table I. It was expected that development of cone parts, with the possible exception of bracts, would be strongly correlated with cone length so that the arrays for all characteristics might be similar. From Table I it is apparent that although longer cones generally had longer scales, seed wings, and bracts, sub- stantial variation independent of cone length was present. For example, average cone length for the Maine collection was significantly longer than for either Wisconsin collec- 1968] Abies balsamea — Lester 87 tion and yet no significant differences were apparent in scale or seed wing length for the three locations. Ratios of length/width showed less relation to cone size than length measurements, suggesting that differences in the shape of cone parts were not primarily functions of cone length. The ratio of particular interest was bract length/scale length. For this ratio a geographic grouping of high values in the eastern portion of the range is apparent (Figure 1). The arrangement of scales in the fir cone is such that por- tions of the bract are generally visible when the bract length/scale length ratio exceeds 0.8 to 0.9. The highest average ratio for a location was 0.8 for the Nova Scotia collection. This value is appreciably less than the ratio of 0.9 suggested by Boivin (1959) or 1.0 suggested by Fernald (1909) for identifying the variety phanerolepis. However, the average ratio of measurements for one tree in the Nova Scotia collection exceeded 0.9 and the maximum value for a single measurement was 1.0. These observations emphasize the tree-to-tree variation as well as within cone variation for some cone characteristics. COMPONENTS OF VARIATION The analyses of variance indicated that locations, trees within locations, and cones within trees all contributed significantly (at 99% probability) to the observed variation in all measurements and ratios. The relative contributions of each source of variation are summarized in Table II. Variation in cone length was associated approximately equally with differences between locations, between trees within locations and between cones within trees. For the other measurements, the predominant sources of variation were differences between geographic locations and between trees within locations. Variation associated with location was pronounced for bract and stalk length while tree-to-tree variation was predominant in length/width ratios with the exception of the ratio for bracts. Variation associated with differences between cones from the same tree was relatively small in measurements of cone parts and in ratios. This suggests that except for measures 88 Rhodora [Vol. 70 of cone size, a small sample of cones from a given tree will accurately reflect the floral morphology of that tree. Variation associated with samples from the same cone was either small or large depending on the variable con- sidered. Restricting the sampling to the mid-portion of each cone unquestionably reduced sampling variation be- cause ovuliferous scales at either end of the cone are smaller than in the middle. Bract length/scale length ratios also vary somewhat from base to tip of the cone. Variation with- in cones is thus probably greater for all characteristics measured than variation between cones from the same tree. The predominance of geographic effects on variation in bract and stalk length suggests that bract development is of primary importance in relation to geographic distribu- tion of balsam firs with different bract/scale ratios. During measurement of cone parts, it seemed that variation in stalk length might be contributing in a major way to variation in bract length. To test this speculation, the contributions of variation in stalk length and in blade length to variation in total bract length were estimated. Average values for geo- graphic locations were used in multiple regression analysis. Contrary to impressions during measurement, the analysis revealed that variation in stalk length contributed only about 17% more than variation in blade length. CORRELATION WITH CLIMATIC VARIABLES The possible association of variation in cone morphology with certain climatic variables was tested by simple correla- tion. Frost-free period, mean annual temperature and pre- cipitation, and mean monthly temperatures for May, June, and July were the climatic variables used. Table IV sum- marizes the statistically significant correlations and illus- trates the association of bract development with climate. In view of the impossibility of obtaining weather data known to be representative of the collection location, the correlations must be interpreted with particular caution. Yet, the mod- erate correlation between bract characteristics and mean annual temperature and precipitation fits well with other analyses in which geographic location of the cone collection 1968] Abies balsamea — Lester 89 was shown to be the predominant source of variation in bract length. DISCUSSION Studies of variation in samples from native populations represent one type of approach to describing variation with- in species. These studies can complement experiments on comparative genetics and physiology conducted through the common-environment approach. Sampling in native popu- lations commonly emphasizes structures which are less easily modified by environment so that the observed pattern of variation may more closely resemble patterns of genetic variation. Floral characters are generally less subject to environmental modification, perhaps as a result of differen- tiation and development of primordia over relatively short periods of time (Stebbins, 1950). Cones have been the structures of chief interest in studies of variation in balsam fir. In the present study, samples from much of the readily accessible range of the species have shown generally con- tinuous variation in measurements on cones and cone parts, as well as in measures of shape by length/width ratios. The continuity of variation, however, did not reflect a completely random pattern of geographic distribution. The more east- ern collections were generally grouped near the top of some arrays, particularly those involving characteristics of the bract. The apparent association of variation in some cone characters with precipitation and temperature suggests that the grouping of eastern collections may be a reflection of climatic characteristics in the eastern portion of the range. It might be expected, therefore, that in other portions of the range, variation in local climates or years of unusually heavy rainfall at critical stages in cone initiation or develop- ment could produce similar variation. On the other hand, the major contribution of trees within locations to total variation suggests that genetic differences or local environment associated with soil differences, crown position, or stand density could cause much of the observed variation. Fernald (1909), in proposing the variety phan- 90 Rhodora [Vol. 70 erolepis, noted that trees bearing cones with exserted bracts were growing intermingled among trees of the species type at Percé Mountain on the Gaspé Peninsula. Myers and Bor- mann (1963) similarly found trees characterized by cone scale/bract ratios ranging from variety phanerolepis to variety balsamea within the St. Lawrence Valley, Gaspé Peninsula, and Maine, although from their data it is not clear whether the whole range of variation was present in any one small group of trees. Possible genetic control of variation in the bract/scale ratio is suggested by comparison of the cones of variety phanerolepis with A. fraseri, a species currently restricted to high elevations in the Southern Appalachian Mountains. Cones of A. fraseri are characterised by exserted and strong- ly reflexed bracts wherever the species is grown. In A. balsamea, altitudinal clines for the cone scale/bract ratio have been demonstrated, with exserted bracts being pre- dominant at the highest elevations (Myers and Bormann, 1963). In A. fraseri, gene fixation for bract development may have resulted from restriction of the species to high elevations following a xerothermic or hypsithermal period (Mark, 1958). In A. balsamea the high elevation gene pool remains connected to lowland populations. The relative con- tributions of environment and genetic differentiation to vari- ation in cone morphology of A. balsamea eventually should be estimable in the common-environment experiments now in progress. The influence of geographical location of cone collection was most evident in the variation in bract length. Although the correlations in Table IV must be viewed with particular caution, the apparent moderate association of bract length with mean annual temperature and mean annual precipita- tion coincide with the indicated influence of geographic location. Variation in the developmental pattern of bracts may thus be the basis for the variety phanerolepis. Fernald (1950) defined a variety as “a strongly fixed variation of a species with the essential reproductive parts unchanged but showing somewhat constant departures in 1968] Abies balsamea — Lester 9I size of parts, shape of leaves, or modification of the less fundamental parts of the flower, etc. and occupying a some- what segregated geographic area". If distinct discontinuity in the pattern of variation is not a prerequisite for varietal classification, Fernald’s definition seems most questionable on the point of “a strongly fixed variation .. . showing some- what constant departures ...”. The separation of genetic and environmental influences on bract development will be necessary to establish how strongly fixed and how constant is variation in bract development. Assignment of a bract length/cone scale length ratio of about 0.9, or, more simply, assigning trees bearing cones with exserted bracts to variety phanerolepis, will result in a geographic subdivision of the species range, Within that subdivision the frequency of trees classified as the variety phanerolepis will vary continu- ously along altitudinal and geographic gradients and per- haps from year to year. Varietal classification here serves merely to recognize an arbitrary, though visually prominent, point on a scale of continuous variation. ACKNOWLEDGEMENTS The following individuals provided samples of cones: C. M. Carl, Northeastern Forest Experiment Station, Burling- ton, Vermont; A. E. Grass, Department of Resources and Economic Development, Plymouth, New Hampshire; M. J. Holst, Petawawa Forest Experiment Station, Chalk River, Ontario; C. M. Hunt, New York State College of Forestry, Syracuse, New York; D. P. Fowler and H. G. MacGillivray, Canadian Department of Forestry, Fredericton, New Bruns- wick; L. Parrot, Laval University, Quebec, P.Q.; and S. S. Pauley, School of Forestry, University of Minnesota, St. Paul, Minnesota. D. O. Coffen and Miss S. J. Hewitt provided technical assistance. DEPARTMENT OF FORESTRY UNIVERSITY OF WISCONSIN MADISON, WISCONSIN 53706 [Vol. 70 (gg = d ta Geet "gun woz) Ayyiqeqord %66 ura FTO SO CO Pro SO 20 ZI etc Oz OTL SUSI Jp 3ueogrudrs ojewixorddy Ic OT rc L9 0% L'0 ST ott 9°6 0'€6 82.99 ' ,€0,LP —pIMsunig MƏN GG eI Lec cL Pé Lu SG OTT ER 9' IT ST.IOI' ,66,6€ UBMIYOJVYSES GU ot GG 94 9G 8'0 yc Val 601 OCP 08068 ' ,60.9T UISUOOSTIM 99 oT 66 EL ER? LO GC S TT Co Lrt 110006 < 200 UISUOISIM 08 OT 0% 9¢° Fé 8T 9'9 6'6 T8 SPP 0¥.09 * .€8.9F BIJOIS BAON ye ot Tg 89° 6% LO UG CTT V6 L'Sy 87.08 < ,OT.6F oneru cà LS SI 6 6 SL 8'3 8'0 69 O'STI UII 61S uve [E1940 e OL GT Tg LL 8S Cé 68 GI 8'II 9°9G LOoPL ' Spo} XIOA MON 3 IC ot 66 9L° 6$ 8'0 6'€ 6'8I Lit 6 LC 0T.6L ' ,96.9T or1g3u() S 84 SIT CG 69° 9G 9'6 0'6 OTI 9'II 06S ,S9.,8L ' GLP YOR MON = g9° ST Gë [JA LG VG CS c SI € 6I Fu 61.99 ' ,eg,Gp Yorlmsuntg MON GL LT Cé TL VS Cé L6 SCT OTT 6 69 18089 ' ,Sv.vT 9ure LS VI Cé L8 0'8 Tg cs 6ST V9I T89 M66.99 'N,89,Gp olasunig MON Wu] Peig Sum ojos quo) s[ejg rig ZuM ALF 9uo07) uorje»or orudea3oor) 9[eog peeg pees /yysueT perg pr /uq3uerq 92 sone 'suoreoo[ orgdexdoe3 10j on[e^ uvawt jo sere jo o[dureg (uu) sy}3uəq 'I qe 93 Abies balsamea — Lester 1968] 9T 91 86 OF mang CO L 9 Y ee Souo) UI 3uauiaInseo]y OT 8 ET g 08 6 8 LT IT 08 seal], UI Sou? 98 98 FP OF 6r 86 96 cv 6r 68 Uorjeoor[ ur soe1], 86 66 VI VI Tg Ly 8g SE 96 IS uoreooT ordv.rdoor) T XI€3S joeig JUM — e[eog 9euo) J[e3g pelg SuiMm ojrsg oquog /'] peig Daag UOTE A oey uU prM/u32uerq q3ue'T Jo SINOS "'99UU LIRA [240] Jo aëpiuaad e se posseidxo aouvliva Jo sjuauodwog "o 8'98 = uornje»o[ ied sjueureinseaur Jo goquinu uto] — “9D 6'G = 201} Jed sjueureinseeaur Jo Jequinu uve — “9 Oe = euoo Jed sjueureinsgaur JO ioqumu Uva — '9 "AJOAI9D9d sal SUOT}IOT put ‘saad, 'souoo 'squour -IINSBIU UJIM Pajridosse adUBIIVA JO sjuouoduioo = lo Lo Vo ip 6 6 Z z bës 90ZT $9U0^) UI sjuauraInsea]q Kon? + 662 Saal, UI Souo^) [9 p jer ee 992 ` UOTZBIO'TT UI Seal], p + er + oo + page gt uorjeoor orude13022) serenbg uva Uulopeadq UOILLA pəypədxy Jo see13e(q Jo amog "9UBLIBA JO sesÁA[eu? Jo Kieunung SIsA[vue JO ULI0J [Blouse "y ‘II Ste 94 Rhodora [Vol. 70 Table III. Correlation of cone measurements with measures of climate. Cone Annual Annual Measurement Precipitation Temperature Seale Length (L) —.01 —.01 Length/Width (L/W) .98** —.20 Bract L .35** .29* L/W .12 EK Stalk L .61** .A6** Bract L/Scale L .A3** .35** * correlation coefficient differs significantly from 0 with 95% probability ** correlation coefficient differs significantly from 0 with 99% probability REFERENCES Borvin, B. 1959 Abies balsamea (Linné) Miller et ses variations. Le Naturaliste Can. 86, 219-223. DUNCAN, D. B. 1955. Multiple range and multiple F tests. Biometrics 11, 1-42. FERNALD, M. L. 1909. A new variety of Abies balsamea. Rhodora 11, 201-203. 1950. Gray's Manual of Botany. Eighth Edition. American Book Co. p. VII. MARK, A. G. 1958. The ecology of the Southern Appalachian grass balds. Ecological Monographs 28, 293-336. Myers, O., JR. and BoRMANN, F. H. 1963. Phenotypic variation in Abies balsamea in response to altitudinal and geographic gradi- ents. Ecology 44, 429-435. STEBBINS, G. L., JR. 1950. Variation and evolution in plants. Colum- bia Univ. Press. p. 73. A CYTOTAXONOMIC STUDY IN VERBESINA (COMPOSITAE)! JAMES R. COLEMAN? In their comprehensive revision of Verbesina Robinson and Greenman (1899) recognized 109 species in 12 sections, 4 of which they considered as provisional, unnatural units. During this century Blake (1925) revised section Lipactinia and Coleman (1964, 1966a, 1966b) has recently treated 3 others. However since Robinson and Greenman’s revision the number of species assigned to Verbesina has more than doubled because of descriptions of new species and transfers made into the genus, As a result, the relationships of many species are uncertain and sectional limits have become in- creasingly obscured. In an attempt to better determine specific relationships and sectional limits, a study of chromo- some numbers and crossing relationships is being conducted. This paper represents the initial report on this study. MATERIALS AND METHODS The material utilized in this study is listed in Table 1. Voucher specimens are deposited in the herbarium of In- diana University. Crosses were made by rubbing together heads of the individuals being crossed for 2-4 consecutive days to insure that all florets had opened and were polli- nated. The crosses were made in insect-free greenhouses at Indiana University with 2-5 plants generally being in- volved in each series of crosses. Immature heads were fixed in a solution of 95% ethyl alcohol, chloroform and propionic acid (6:2:2) (Jackson 1962). Entire florets were smeared and stained with acetocarmine. Many slides were made permanent with venetian turpentine; selected cells were 'Based on part of a Ph.D. thesis submitted to Indiana University. This investigation was aided by funds from the National Science Foundation and Indiana University. *I wish to express my gratitude to Dr. Charles B. Heiser for sug- gesting this investigation and for reading the manuscript. Apprecia- tion is also expressed to Annetta M. Carter, L. N. Goodding, Dr. Robert K. Godfrey and Dr. Wilbur H. Duncan for providing seed material for this study. 95 96 Rhodora [Vol. 70 drawn with the aid of a camera lucida. Pollen stainability, taken as a measure of viability, is based on counts of not less than 200 randomly selected grains after having been stained in cotton blue for at least 24 hours. Grains which stained deeply were considered viable; those which stained lightly or not at all were considered inviable. CHROMOSOME COUNTS With the addition of the counts presented in this paper, the number of species of Verbesina counted is increased to 29, representing about 15% of the genus. Of the 11 species counts here presented, 4 are for previously unreported species. The counts determined for the material included in this study are given in Table 1. The count of n — 34 for V. alternifolia and n = 17 for V. occidentalis are in agreement with previous reports for these species (Heiser & Smith 1955) as are the counts of n = 17 for V. lindenii and V. serrata (Turner, Ellison & King 1961; Turner, Beaman & Rock 1961). Turner and Flyr (1966) report V. wareias n — ca 17. A definite report of n = 17 is here reported for V. sarei. The previously unreported species V. aristata, V. longifolia, V. rothrockii and V. virginica were determined as having n = 17. Of the species of Verbesina thus far reported, 1 has n = 16 (Turner, Beaman & Rock 1961), 2 have » — 18 (Turner, Ellison & King 1961; DeJong & Longpre 1963) and 3 are tetraploid with n = 34 (Heiser & Smith 1955; Coleman 1966b; Turner & Flyr 1966). All other reported species have n = 17. A single species, V. oligocephala, is reported to have B chromosomes (Coleman 1966a). It would appear very probable that » — 17 is in the primitive condition in Verbesina and that both aneuploidy and polyploidy have played a role in the evolution of the genus. CROSSING PROGRAM The source of the material utilized in the crossing pro- gram is listed in Table 1. The crosses made and the results obtained are listed in Table 2. Table 3 presents the pollen stainability and meiotic behavior observed in the hybrids. Intrasectional crosses: Only 2 sections, Pterophyton and 1968] Verbesina — Coleman 97 Verbesinaria, are represented by more than a single species in the crossing program. The 3 species of sect. Pterophyton, V. aristata, V. rothrockii and V. warei, were crossed in every possible combination. The cross V. sarei X V. aris- tata and its reciprocal both resulted in high seed set; how- ever those of the reciprocal were not grown to maturity. The hybrids displayed intermediacy for several characters. Verbesina aristata has an erect habit, rayed heads and numerous disk florets (ca 40-45) whereas V. warei is lax- stemmed, rayless and has much fewer disk florets (ca 13- 25). The hybrids were intermediate for erectness of stems and possessed rayed heads with numerous disk florets simi- lar to the male parent. The high pollen stainability of the hybrids as well as the occurrence of 17 bivalents at meiosis supports the morpho- logical evidence that these 2 species are quite closely related. Verbesina aristata occurs in southern Georgia and Alabama and in northern Florida, especially in the panhandle region. Verbesina warei is endemic to the litoral of the central pan- handle region of Florida. Although the ranges of the 2 species are contiguous, ecological isolation is evidently com- plete and the ranges apparently do not overlap. As would therefore be expected, an examination of herbarium speci- mens revealed no evidence of natural hybridization. Intersectional crosses: Nine species from 5 sections were utilized in intersectional crosses. Each of the 4 successful crosses was between species of sects. Pterophyton and Verbesinaria. Seeds from the cross V. aristata (Pterophyton) X V. occidentalis (Verbesinaria) germinated readily to produce robust hybrids. The leaves of the hybrids were intermediate between the petiolate, ovate ones of the male parent and the sessile, oblong ones of the female parent. The wingless stem of V. aristata dominated the winged condition found in V. occidentalis. The styliferous ray florets of V. occi- dentalis dominated the neutral ones of V. aristata. As might be expected for intersectional hybrids, meiosis revealed 34 univalents and pollen stainability was an 98 Rhodora [Vol. 70 accordingly low 5%. Although the species are sympatric, no hybrids were detected in the herbarium material ex- amined. The cross V. alternifolia (Verbesinaria) X V. aristata (Petrophyton) resulted in a single seed which germinated to produce an initially weak individual. Vegetative morpho- logical characters indicated this plant to be a hybrid. Like the male parent, it possessed wingless stems and the leaves were completely sessile and opposite rather than attenuate- based and alternate as in the female parent. Unfortunately this plant was inadvertently destroyed before material was taken for cytological examination. However pollen stain- ability was checked and found to be 4%. The low pollen stainability indicates a lack of chromosome pairing and further substantiates the hybrid nature of this individual. The species are sympatric, but no hybrids were encountered in the herbarium material examined. The hybrids produced by the cross V. warei (Pterophy- ton) X V. occidentalis (Verbesinaria) possessed the wing- less stems of the female plant and had leaves intermediate between the elliptic-oblong ones of the female parent and the larger, ovate or lanceolate ones of the male parent. The hybrid manifested none of the laxness of stems displayed by V. warei. Pollen stainability ranged from 1-4.5% and 34 univalents were observed at meiosis. No indication of natural hybridization was detected in an examination of herbarium material. Seeds from the cross V. rothrockii (Pterophyton) X V. longifolia (Verbesinaria) germinated readily to produce both selfs and hybrids. Verbesina rothrockii has lanceolate leaves which are mostly opposite below but soon become alternate whereas the linear leaves of V. longifolia are fre- quently subopposite, subwhorled and alternate on the same plant. The hybrids had leaves intermediate in shape which were mostly opposite, the uppermost becoming alternate and possessed linear phyllaries such as are found in V. longifolia rather than oblong ones as occur in V. rothrockii. Meiosis of the hybrids revealed 17 bivalents with no ab- normalities, Pollen stainability was correspondingly high, 1968] Verbesina — Coleman 99 ranging from 48-60%. Although the species are sympatric in southeastern Arizona, no natural hybrids were detected in the herbarium material examined. CONCLUSIONS The crossing program here reported involves too small a portion of the genus to permit broad generalizations. However some conclusions are in order. Although Robinson and Greenman (1899) retained V. alternifolia in the genus Actinomeris, there has been a tendency in recent years to treat this small genus of 2 species in Verbesina. The mor- phological characters, divergent pappus awns and globose receptacles, by which it is kept apart from Verbesina are weak. That an apparent hybrid was obtained between V. alternifolia and V. aristata further supports the morpho- logical evidence that there is no justification for maintain- ing the genus Actinomeris. The ease with which hybrids were obtained between sec- tions Verbesinaria and Pterophyton suggests an affinity on the part of some of the members of these sections. The low pollen stainability and the lack of chromosome pairing in 3 of these successful intersectional crosses does not indicate a particularly close relationship. However the high pollen stainability and the occurrence of 17 bivalents with no ab- normalities in the hybrids of the cross V. rothrockii X V. longifolia indicates a low genetic isolation and suggests the possibility that V. longifolia and its relatives may be closely related to some members of sect. Pterophyton. This also appears possible on morphological grounds. Robinson and Greenman considered sect. Verbesinaria to be unnatural, an opinion with which I concur. Additional crosses may be of aid in understanding the relationships of the species of this section. Since only sects. Pterophyton and Verbesinaria were represented by more than a single species in the crossing program, the failure to secure hybrids between other sections can not be taken as indicating that genetic isolation is complete between these sections. DEPARTMENT OF BOTANY UNIVERSITY OF GEORGIA, ATHENS [Vol. 70 Rhodora syunoo pagtodaiun ÁjsnoIadIg, 0£I9A “BIA “OD JMI “exqyyemom LT ug Aydoi q Avelyn ‘yy (1nN) totem ISI9A “el “OD uooT7 “oosseyeier «LI BIUTPIVAIYIO "Tt BOIULBALA ff] 'ooxo] 'oiejo1en() ‘oregon LT erjoueqnegs "ABD €1€.LI9S EETIA Zu “09 ƏSIYDOQ 'surejunopy ƏMW «LT uoyAydo19}g ‘WUdaal) Y "JOW ID[20.1430.I ££29/A “OOIXAT “VIULOTI[VD eleg *oprpuoos^] o31onq = LT e[O9LIOUOS uojsuuof e[eudooo3t[o OLE9A "€5 “OD YILI '[[€q931UA LT BLIBUIS9QIS A 'HEM (C'T) Sueyuoproo»o GEI9A "ZU ALIN y “09 ƏSIYIOQ ‘eonyoeny 34 «LIT €LISUIS9Q.I9 A (ÁvIS *y) vi[ogtduo[ SETIA 'OXIX9]N oyeg *oodoqusnge] jo snunps LT etjaueqnes (‘dig yog) uuoput[ 4II9A "al Jg XooH 9 “YU “09 UOSTIM ‘AsO LI ?IS9USUILX (Avg) soprorpooua 6169A “ZIV 3 Y400H 9 yug “og astmoS ‘eeqsIg LT eIsouswly ((AB)) Seprorjeoue EZTIA "e| “OD uosiogef ‘PLOI LT uojAudo193q IMH CUA) ESE ZET9A ‘PUI “09 eoiuopy ‘'uozsurwoog Pg BIIVUISAG.19 A yag (I) €r[ojgruxze3[e DINOS =u uornoeg soroodg SHIGALS 9NISSOHO ANV 'IVOIOO'IOLAO NI CASA 'IVIPHGULVIN T QLL 100 101 Verbesina — Coleman 1968] spriqAy spriqAq səs e[qeqo.id uorjeururtog ou SpliqAy pu? sj[os sps słļ@Ss e[qeqo.id Sjy[os uoneuruLrod ou SspriqAu e[qeqo.rd SpHqÁq ag 3ue1edde priq&q o[qeqo.rd BOIUISATA SI[€3U9p1290 IDporqod '€?je1sLre Sr[ogru roe e[eudeoo3I[o X Sijequaprooo x Seploljeoue X BOIUISILA I91€A X SI[£1uoproo0 X vI[ogI2uo| X Sopror[ooua X €)€]SLI? X IDN9OIU30. ID[oO1[)0d X sepror[ooua X v[eudeoodi[o €OIULZJIIA X BI[OFIULa}][e X sijejuaprooo IDPOIUjOoI X "r[ojgrguo[ '€OSIULJJIA X €[eudeooZi[o X sopror[oouo Totem X VOIULdJIIA X Hyo X se}łuəprooo X BIIOIIUL/IaUg X gyegzse TILM X BOULSIIA X Si[£3uoproo0 X eyeysue X eozruIəgzje X XXX X ema ITEM sj1ueutuio^) 0'0 79 Tet 89 0*0 Ez SEI Gg 0'0 86 0°0 0g1 LS Ss 0°0 80T Ly Gg 0'0 LL get 96T 0'0 OTT 0°0 06 E'L? GG 0*0 gg GOT 608 6TT TOT 8'8 TI 0°0 061 0*0 LOS ZG 29 0*0 601 0'0 661 Sor 691 8T Cat 0*0 18 0'0 Iv 0'0 ZA! 9°0 PLT (Sjo10B) PS Pos % sjduo3y [e10], $9880) WOHL »NISSON2 P STAO T 102 Rhodora [Vol. 70 Table 3 ANALSIS OF HYBRIDS Cross Pollen Stainability % Meiosis No. No. Range Average Plants Pairing Plants alternifolia X aristata — 4 1 0 aristata X occidentalis — 5 1 34 I 1 rothrockii X longifolia 48-60 53 3 17 II 2 warei X aristata 50-69 62 6 17 II 1 warei X occidentalis 1-5 2 2 34 I 1 LITERATURE CITED BLAKE, S. F. 1925. On the status of the genus Chaenocephalis with a review of the section Lipactinia of Verbesina. Amer. J. Bot. 12: 625-640. COLEMAN, J. R. 1964. A taxonomic revision of sections Pterophyton, Sonoricola and Ximenesia of the genus Verbesina (Compositae). Unpublished Ph.D. thesis. Indiana University. 1966a. A taxonomic revision of section Sonoricola of the genus Verbesina (Compositae). Madrono 18 (5): 129-137. 1966b. A taxonomic revision of section Ximenesia of the genus Verbesina (Compositae). Amer. Midl. Nat. 76 (2): 475-481. DEJowG, D. C. D. and E. K. LONGPRE. 1963. Chromosome studies in Mexiean Compositae. Rhodora 65: 225-239. HEISER, C. B. and D. M. SMITH. 1955. New chromosome numbers in Helianthus and related genera (Compositae). Proc. Amer. Acad. Sci. 64: 250-253. Jackson, R. C. 1962. Interspecific hybridization in Haplopappus and its bearing on chromosome evolution in the Blepharodon sec- tion. Amer. J. Bot. 49: 119-122. RoniNSON, B. L. and J. M. GREENMAN. 1899. A synopsis of the genus Verbesina, with an analytical key to the species. Proc. Amer. Acad. 34: 534-564. TURNER, B. L., J. H. BEAMAN and H. F. L. Rock. 1961. Chromosome numbers in the Compositae. V. Mexican and Guatemalan species. Rhodora 63: 121-129. , W. L. ELLISON and R. M. King. 1961. Chromosome numbers in the Compositae. IV. North American species with phyletic interpretation. Amer. J. Bot. 48: 216-223. and D. FLYR. 1966. Chromosome numbers in the Compositae. X. North American species. Amer. J. Bot. 53: 24-33. NEW RECORDS OF DISJUNCT ARCTIC-ALPINE PLANTS IN MONTANA S. A. BAMBERG! AND R. H. PEMBLE? Two recent studies dealing with Montana alpine vegeta- tion (Bamberg, 1964; Pemble, 1965) have produced new records as well as significant range extensions for several species representing a major floristic element in the alpine flora of the state, namely the arctic-alpine element. Four of these species which fill gaps in previously reported distri- butions are: Festuca baffinensis N. Polun., Juncus albescens (Lange) Fern., Juncus biglumis L., and Kobresia myosur- oides (Vill) Fiori & Paol. Three other northern species have their southern limits in the Rocky Mountains within Montana. These are: Cassiope tetragona (L.) D. Don, Dryas integrifolia Vahl, and Festuca vivipara (L.) Sm. FESTUCA BAFFINENSIS N. Polun. probably has a circum- polar distribution (Holmen, 1964) though Tolmachev (1964) does not believe that any of the Eurasian Arctic material can clearly be referred to a taxon separate from F. brachyphylla. In the Arctic F. baffinensis has a continu- ous distribution from Alaska across Canada to northeast Greenland, preferring areas with calcareous soil. It has also been recorded from Svalbard and Novaya Zemlaya (Holmen, 1964). The species is known from isolated sta- tions in the Rocky Mountains as far south as Colorado (Weber, 1961) and Wyoming (Johnson, 1962). Hultén (1962) suggests the probable presence of the species in northwest Montana although no specific stations are indi- cated from the state. Our collections were made in Glacier '"Present address: Desert Research Institute, University of Nevada, Reno. "We are indebted to the National Science Foundation for financial support of our studies. The senior author was supported under NSF Grant #11568 to the University of Montana and #23317 to the Uni- versity of California, Davis while the junior author was aided, in part, by a National Science Foundation Summer Fellowship for Teach- ing Assistants. 103 104 Rhodora [Vol. 70 National Park, one at Siyeh Pass (8 Aug. 1964, Harvey & Pemble 7135, MONTU) and one on Mt. Henry (15 Aug. 1964, Pemble & Harvey 128, MONTU). JUNCUS ALBESCENS (Lange) Fern. is regarded by Hultén (1962) as a Greenland-American, eastern Asiatic race of the circumpolar-montane J. triglumis L. Polunin (1959) indi- cates that J. triglumis and J. albescens form a circumpolar series, hybridizing where their ranges overlap. J. albescens has been reported from Wyoming (Johnson, 1962) and Colorado (Harrington, 1954). J. triglumis has been report- ed from New Mexico (Wooton & Standley, 1915) and from Utah (Tidestrom ,1925) although these early reports should be referred to J. albescens according to Fernald (1924). An early collection referred to J. triglumis was made in northwestern Montana and reported by Rydberg (1900). The occurrence of J. albescens (J. triglumis), previously suspected in Montana (Booth, 1950), is confirmed by two collections made in Glacier National Park, one at Siyeh Pass (3 Aug. 1963, Bamberg 825, COLO), and a second at Logan Pass (2 Aug. 1960, Schofield, MONTU 51409). JUNCUS BIGLUMIS L. is considered by Polunin (1959) and Hultén (1962) to be a fully circumpolar species. Hultén's distribution map for the species shows disjunct stations in Colorado, where the species was reported by Weber (1955), and near the Beartooth Plateau in Wyoming. Two Montana collections of this species have been seen, both from Logan Pass in Glacier National Park (2 Aug. 1960, W. B. Schofield, MONTU 51397 and 21 Aug. 1962, F. J. Hermann 18123, MONTU). This species has not previously been reported for the state but Dr. F. J. Hermann has given us permission to publish the report of his collection from Glacier Park. KOBRESIA MYOSUROIDES (Vill.) Fiori & Paol [K. bellardii (All) Degl.] is considered by Hultén (1962) to be circum- polar arctic-montane species with a continuous distribution from the North American Arctic south to the international border. From there the species is disjunct with stations in the Wallowa Mountains of Oregon (Peck, 1961), the Bear- tooth Mountains of northwest Wyoming (Johnson ,1962), 1968] Arctic-Alpine Plants — Bamberg, Pemble 105 Colorado where it occurs on scattered peaks (Harrington, 1954), Utah in the Unita Mountains (cf. Major & Bamberg, 1963) and New Mexico (Wooton & Standley, 1915). In addition, Major and Bamberg (1963) have reported its occurrence at Convict Creek in the Sierra Nevada of Califor- nia where it is found with four other disjunct cordilleran species. Our collections are from Glacier National Park (3 Aug. 1963, Bamberg 797, COLO) where the species occurs at Siyeh Pass. CASSIOPE TETRAGONA (L.) D. Don is probably circum- polar, being abundant almost everywhere in the Arctic ex- cept towards the highest latitudes (Polunin, 1959). 'The species has been reported as far south as Okanogan Co., Washington on the Pacific Coast and Glacier National Park in the Rocky Mountains (Hitchcock et el., 1959). Our col- lection of the species (27 Aug. 1965, Pemble 254, MONTU) was made on St. Mary Peak in the Bitterroot Mountains of western Montana about 175 miles south of Glacier Park. DRYAS INTEGRIFOLIA M. Vahl ranges throughout the Arctic regions of the western hemisphere from western Alaska to East Greenland (Polunin, 1959), across much of Canada south to southern British Columbia and Alberta and east to Newfoundland with an isolated station near Lake Super- ior (Hultén, 1959). This species has been reported once from Montana (Porsild, 1947) although Hitchcock et al. (1961) indicate that they have seen no material of D. integ- rifolia from the state. As the determinations of our Dryas specimens were made by E. Hultén we are including the collections from the Big Snowy Range (Bamberg 313, 565, 769, COLO) and from the Tobacco Root Mountians (Bamberg 429, COLO) which he referred to D. integrifolia ssp. integri- folia as confirmation of the species’ presence in the state. FESTUCA VIVIPARA (L.) Sm. is considered by Hultén (1958) to be a doubtful taxon with an uncertain geograph- ical range. Its taxonomic status in the Eurasian Arctic has been questioned by Tolmachev (1964) who indicates that vivipary occurs in almost sporadic fashion throughout the area of F. brachyphylla and F. ovina. Hultén's distribution 106 Rhodora [Vol. 70 map for the species shows it to be common in the Scandi- navian mountains, in south Greenland, and in Iceland. He also cites reports from Novaya Zemlaya and other places in Arctic Russia. Fernald (1950) considers it to reach from Greenland and Labrador to Alaska and south to west New- foundland, Anticosti, Gaspé Peninsula and Lake Mistassini, Quebec. Holmen (1964) indicates that F. vivipara (L.) Sm. is, in the sense of most authors, a complex probably consisting of a number of species and hybrids. Different chromosome races are known to occur although the taxono- mic status of these has not as yet been determined. He sug- gests on the basis of chromosome number (2n — 28) that the typical race includes the taxon common in the Scandi- navian mountains, in Iceland, and South Greenland, and that the viviparous Festuca with higher chromosome numbers, occurring in other parts of the Arctic, is probably closer to F. brachyphylla or F. baffinensis. Johnson and Viereck (1962) have indicated that their material from Alaska should be referred to the same taxon as that in Scandinavia. We also feel that our collections may be referred to the Scandinavian populations on the basis of external morph- ology, but we have made no chromosome counts. Several collections of this species have been made from Glacier National Park including localities at Logan Pass (11 Aug. 1949, L. H. Harvey 4030; 2 Aug. 1960, Schofield, MONTU 51406; 8 Aug. 1964, Harvey & Pemble 7091, MONTU) and Siyeh Pass (3 Aug. 1963, Bamberg 819, COLO; 8 Aug. 1964, Harvey & Pemble 7145, MONTU). These stations are con- siderably removed from the range shown by Hultén (1958) for the species in which the Alaskan stations are not in- cluded. If the Alaskan reports are included the gap extends from Alaska to the area of our present report in Glacier Park. In addition to the seven species just discussed which be- long to the Arctic flora two other extensions of alpine species are worth reporting here. Carex stenochlaena (Holm) Mack., previously known from Alberta and Alaska south to Washington and Idaho, was collected at Logan Pass in 1968] Arctic-Alpine Plants — Bamberg, Pemble 107 Glacier National Park by F. J. Hermann (26 Aug. 1962, F. J. Hermann 18300, MONTU) who has given us permission to include it as a new Montana record. A second interesting range extension is that of Saxifraga tolmiei T. & G. which is a typical member of the Pacific Coast-Cascade flora. This species was previously known in Montana only from St. Mary Peak in the Bitterroot Mountains although it is dis- tributed along the Pacific Coast from Alaska south through the Cascades and Sierra Nevada of California (Hitchcock et al., 1961). Additional Montana collections of this species were made on Trapper Peak (25 July 1964), Pemble & Harvey 58, MONTU), a southern extension of 50 miles, and in the Highland Mountains south of Butte, Montana, an eastern extension of approximately 80 miles. Disjunctions, which are common in Arctic and alpine plant species, are of interest to those concerned with the floristic history of any region. Such disjunctions may also have environmental causes. The disjunctions we have noted suggest a discussion of some recent phytogeographical work along these lines. Weber (1949, 1959, 1965) proposes that in the southern Rocky Mountains the richest alpine communities in terms of the Arctic relicts that they harbor are found in the most mesic areas, particularly where high ranges trending east- west lie between or connect the principal north-south ranges. In such ranges diurnal insolation through the summer months is less direct or of shorter duration, resulting in small bogs and lakes at high elevations. He indicates that the rare species may occur on either north- or south-facing slopes as well as in cold-air drainage basins. Johnson (1962) suggests that the disjunct occurrences of thirteen species in the Beartooth Mountains of Wyoming may be attributed to one of two factors: (1) that the bog environments are capa- ble of dissipating the excessive summer heat allowing cer- tain arctic-alpine relicts to persist, (2) that other arctic- alpine species can persist in the unstable environment of solifluction slopes. Dahl (1952, 1955) has pointed out that in Scandinavia the isotherms of the average yearly maxi- 108 Rhodora [Vol. 70 mum summer temperature at the highest points of the land- scape coincide closely with the distribution limits of a large number of arctic-alpine species. The importance of tempera- ture is difficult to analyze in the Rocky Mountain alpine areas where high altitude weather stations and climatic data are scanty. Fernald (1907) regarded the southern distribution of alpine species in eastern North America to be a function of the amount of calcium or potassium in the soil. In criti- cizing the “nunatak theory" later proposed by Fernald (1925), Wynne-Edwards (1937, 1939) suggested that most isolated occurrences could be better correlated with soils, particularly those rich in lime or combinations of magne- sium, lime and soda, than with the presence or absence of Wisconsin glaciation in a particular area. Rune (1953) has proposed that the main soil factor involved in the distribu- tion of alpine plants in Scandinavia appears to be the degree to which the humus colloids are saturated with cations, pre- dominantly calcium ions. In Montana, five of the seven arctic-alpine disjuncts occur in Glacier National Park on substrates derived from lime- stone and argillites of the Belt Series. This produces a clay- ey soil high in calcium. Dryas integrifolia in the Tobacco Root Mountains also occurs on mixed metamorphies where lime is present. Cassiope tetragona which was collected in the Bitterroot Mountains occupied a depressional area where snow often persists late into the summer, however the na- ture of the substrate was not determined. Intrusive igneous rocks predominate in the area although the possibility of a limestone outcropping is not ruled out, Tentatively we would say that the species has persisted in this locality due to con- ditions of the habitat such as those described by Weber (1949, 1959, 1965). Langenheim (1962) reports the occurrence of populations of certain disjunct arctic-alpine species in Colorado on metamorphosed limestones along a contact of intrusive igne- ous rocks while others occurred along a contact between metamorphosed sediments of siltstone. Major and Bamberg 1968] Arctic-Alpine Plants — Bamberg, Pemble 109 (1963) found disjunct species in the Sierra Nevada of Cali- fornia occurring on soils derived from calcareous parent material in conjunction with non-calcareous metamorphics. However, they also point out the presence of other smaller areas of caleareous metamorphic rocks in the Sierra Nevada where none of the disjunct species have been found, suggest- ing that there is no one-to-one correspondence here between edaphic specialization and floristic distinctiveness. They also point out the fact that their disjunct species are indiffer- ent outside of the Sierra Nevada as far as calcicoly is con- cerned. According to Bamberg (1964), and as the work of Mooney & Billings (1961) confirms, many of the wide- ranging arctic-alpine species have northern and southern ecotypes. In reviewing the literature dealing with disjunct or relict populations of arctic-alpine species it becomes obvious that such distributions cannot be attributed to any one factor. First, it is necessary to classify disjunct species as to their ecology throughout their continuous range of distribution as well as in the disjunct locality. We can then see how a population of Festuca baffinensis, found in open wind-swept areas in the Arctic, might persist in similar alpine situations, while such a locality does not harbor species of considerably different ecological requirements such as Cassiope tetragona, normally restricted in the Arctic to areas of snow accumula- tion. With this in mind we can next consider climatic limi- tations or edaphic specialization noting, for example, that soils high in calcium, or calcium and clay, such as the Mon- tana localities previously discussed, are in fact often corre- lated with disjunct arctic-alpine species provided the first criteria is satisfied. It is also not unexpected that other relict species are found where contact areas in the substrate pro- vide a unique edaphic situation satisfying the nutrient requirements of the relict species persisting there. Gankin and Major (1964) have reviewed the problem of disjunct populations, and suggested that their presence may often be attributed to limited competition resulting trom failure of the local, zonal vegetation to fully occupy the site. 110 Rhodora [Vol. 70 This in turn is often correlated with an unusual edaphic situation. Contrary to this observation are our disjunctions of Festuca baffinensis or Cassiope tetragona. The former occurs on open sites in Glacier Park as well as in Colorado (Weber, 1961) where competition from zonal vegetation is not a factor while the latter occurs intermixed with a luxuri- ant growth of zonal vegetation, Cassiope mertensiana, sug- gesting that it is not limited to the relict site by competition. It is suggested that while a correlation between disjunct or relict populations and local vegetation in an azonal state may exist at lower elevations, such a relationship does not necessarily apply to all alpine situations where Arctic relicts occur. Climatie conditions may be of importance directly, such as prohibitive maximum summer temperatures as suggested by Dahl and Johnson, or indirectly as they have modified a potentially suitable relict habitat. Thus, in analyzing dis- junct and relict situations we should be prepared to correlate the presence of the disjunct relict species not with one but with several factors of the ecosystem which may interact and thus provide the necessary habitat suitable for its persistence. DEPARTMENT OF BOTANY, UNIVERSITY OF CALIFORNIA, DAVIS 95616 LITERATURE CITED BAMBERG, S. A. 1964. Ecology of the vegetation and soils associated with calcareous parent material in the alpine region of Montana. Ph.D. Thesis. University of California, Davis. BoorH, W. E. 1950. Flora of Montana. Part I. Mont. State Coll., Bozeman. 280 pp. Dani, E. 1952. On the relation between summer temperatures and the distribution of alpine vascular plants in the lowlands of Fen- noscandia. Oikos 3: 22-52. DAHL, E. 1955. Biogeographic and geologic indications of unglaci- ated areas in Scandinavia during the glacial ages. Bull. cf the Geol. Soc. of Amer. 66: 1499-1520. 1968] Arctic-Alpine Plants — Bamberg, Pemble 111 FERNALD, M. L. 1907. The soil preferences of certain alpine and subalpine plants. Rhodora 9(105): 150-193. 1924. Juncus triglumis and its American repre- sentative. Rhodora 26(311): 201-203. 1925. Persistence of plants in unglaciated areas of boreal America. Amer. Acad. Arts Sci. Mem, 15: 237-242. 1950. Gray’s Manual of Botany. American Book Co., N.Y. 1632 pp. GANKIN, R. and J. MAJOR. 1964. Arctostaphylos myrtifolia, its biology and relationship to the problem of endemism. Ecology 45 (4): 792-808. HARRINGTON, H. D. 1954. Manual of the plants of Colorado. Sage Books, Denver. 666 pp. Hitcucock, C. L., A. CRONQUIST, M. OWNBEY, and J. W. THOMPSON. 1959. Vascular plants of the Pacific Northwest. Part 4. Univ. Wash., Seattle. 510 pp. HriTCHCOCK, C. L., A. CRONQUIST, M. OWNBEY, and J. W. THOMPSON. 1961. Vascular plants of the Pacific Northwest. Part 3. Univ. Wash., Seattle. 614 pp. HOLMEN, K. 1964. Cytotaxonomical studies in the arctic Alaskan flora: the genus Festuca. Bot. Notiser 117: 109-118. HULTÉN, E. 1958. The amphi-atlantic plants and their phytogeo- graphical connections. Kungl. Svensk Vetenskapsakademiens Handl. Fjarde Ser. 7: 1-340. — 1959. Studies in the genus Dryas. Svensk Bot. Tidskr. 53 (4): 507-542. — 1962. The circumpolar plants. I. Vascular crypto- grams, conifers, monocotyledons. Kungl. Svensk Vetenskapsa- kademiens Handl., Ser. 4, 8(5): 1-276. JOHNSON, A. W. and L. A. VIERECK. 1962. Some new records and range extensions of arctic plants for Alaska. Biol. Papers Univ. Alaska 6: 1-31. LANGENHEIM, JEAN H. 1962. Vegetation and environmental pat- terns in the Crested Butte Area, Gunnison Co., Colorado. Ecol. Monog. 32(3): 249-285. MAJOR, J. and S. A. BAMBERG. 1963. Some cordilleran plant species new for the Sierra Nevada of California. Madrono 17(4): 93-109. Mooney, H. A. and W. D. BILLINGS. 1961. Comparative physio- logical ecology of arctic and alpine populations of Oxyria digyna. Ecol. Monog. 31: 1-27. Peck, M. E. 1961. A manual of the higher plants of Oregon. Bin- fords and Mort, Portland. 866 pp. 112 Rhodora [Vol. 70 PEMBLE, R. H. 1965. ) x - WW o K ALS d o S . « 2 S + ® x E) o o D % may ine S E 3 c = o3 + a ` Y e + GI o 9 ~ PNY! È a f NS + Q * ( N 9 f (994910) v 71 ) t o xo Oly li ` Ge x v e = H Soon “oi Ol ~ l + / NEE CH DENK Ed * M Pd wei x N -_--- nn -4% \ 7 M - ‘sy 4777 C * «f? / 47 ^ NS en GS NAIN 4 ^N “eon (UA N N N YNYINƏS » jjnuS DJOJIA JO esn UMOUY JO D0810 jo euij|jnQo d DJOJA Ajqoqoud sjuojd asinos ejeum S8!j!|D20] \ HS 1188), eppu ueeq sou C (DIOJA) uO!jJD2!jIjUGp! JO2!UDJOQ 8J8UM S9141]0907 N + +++ x tte494 444 A (*9pixod) (0494ng) (140,0n040y) (gui409) SIINNS SNOSIVIILSIYAW 40 JSN DYIOM JI DYIOM EL) DYIDM DYIDM HIEL) DXIDM DXIDM Seqi4, uoipu| 91 SI vi £i el E Ol DUOMYƏA (D4!Dn959) OwIOM BADUINA Oyodisny oeqny ouDYNL DUNYONW DUDSD4DB OUDMID| oy Aey NM m sg ww nr OO o 1968] De Plantis Toxicariis — Schultes 117 prepared from the bark of a certain tree which, pounded up, is boiled in a small earthenware pot, until all the water has evaporated, and a sediment remains at the bottom of the pot. This sediment is toasted in the pot over a slight fire and is then finely powdered with the blade of a knife. Then the sorcerer blows a little of the powder through a reed (kuratá) into the air. Next he snuffs, whilst, with the same reed, he absorbs the powder into each nostril successively. The Aaküdufha obviously has a strongly stimulating effect, for immediately the witch-doctor begins singing and yelling wildly, all the while pitching the upper part of his body backwards and forwards." In connexion with this postulated identification of the Yekwana Indian bark-snuff as Virola, it should be borne in mind that Dr. John Wurdack (43), after field experience in the Río Ventuari in recent times, has written: “While Schultes believed, from Koch-Grünberg's description of a Ventuari narcotic, that Virola bark-exudate was used in the Ventuari drainage, we have found no such present-day evi- dence. Don Melisio Pérez of San Juan de Manapiare, who is the docent for Ventuari Indian lore, knows of no tribes in this drainage using bark-exudate." The Piaroa now in- habiting the same river do, according to Wurdack, prepare a snuff from beans of Anadenanthera peregrina, Living in a deeply forested area where Anadenanthera peregrina does not occur, these Indians make a long annual pilgrimage to secure seeds of this legume for snuff-making. In 1961, Dr. Helmut Fuchs of Caracas carried out field studies in the upper Río Ventuari. He mentioned (40) two snuffs, one of which has definitely, through botanical speci- mens, been determined as Anadenanthera peregrina. Of particular interest here, however, is his confirmation of Koch-Grünberg's statement concerning haküdufha. In a letter to Dr. Henry Wassén (October 24, 1961), Fuchs wrote: “The crushed bark of ai’-yuka (Piptadenia pere- grina?) (in the Venezuelan vernacular called ‘yono’) is known as a'ku:dwwha in the powdered form. In addition, the fruits of another plant are likewise consumed with 118 Rhodora [Vol. 70 wku:duwha, which, judging from explanation by Makiritare Indians, might also be a Piptadenia. There are, further- more, two other plants employed in the elaboration of the snuff: a broad-leaved one (ai’-yuku), the bark of which is ground up with the wku:duwha, and a small-leaved one (unknown to me), the fruit of which is crushed. These two plants are found only in the highest mountains, for which reason the natives could not, unfortunately, bring me speci- mens, I agree with you that it seems as though the designa- tions ať-yuku and a’ku: duwha have a generic meaning". If these data be accurate, they appear to indicate — be- yond the implication that these native terms are generic for snuffs — that the powdered bark, as well as the seeds or fruits, of Anadenanthera peregrina are employed. Although the literature usually indicates that it is the bean of Anadenanthera that is employed in snuff-making, there would seem to be no chemieal reason why the bark should not also be used since, like the seeds, it contains several psychoactive tryptamines (12). The possibility that Anadenanthera-bark may be so employed is a consideration that must constantly be borne in mind in our attempts at "identification" without botanical material of the plant sources of snuffs, One is likewise tempted to ask whether or not the broad-leaved plant described in Fuchs’ letter might be a species of Virola. The first definite association of a snuff with Virola was made in 1939 by the great Brazilian student of the Ama- zonian flora, Dr. Adolfo Ducke (8). In his monographic treatment of the Leguminosae of the Brazilian Amazon, Ducke wrote, in a footnote to a discussion of Piptadenia peregrina, that: “Martius and other writers attribute to this species the source of the narcotic parica employed by certain Amazonian Indians (the powder of the crushed seeds is inhaled through the nostrils). Notwithstanding, according to information which I obtained from the natives themselves in two localities in the upper Rio Negro, the paricá-powder comes from leaves of species of Virola of the Myristicaceae." All of the recent intensive research has not been able to 1968] De Plantis Toxicariis — Schultes 119 substantiate the statement that the leaves are used: all ethnobotanical field work and all recent reports of those who have observed the preparation of Virola-snuff coincide in indicating that the resin from the bark is the part of the tree employed by the natives. When one of the writers (Schultes) first went into the northwestern Amazon in Colombia — a region where he carried out continual field work in botany and ethnobotany from 1941 to 1953 — he fully expected to meet with the extensive use of yopo snuff prepared from the beans of Anadenanthera peregrina, The anthropological literature quite generally ascribed all narcotic snuffs of the Amazon not obviously prepared from tobacco to this leguminous tree. Amongst many tribes of the Colombian Amazonas and Vaupes, he met with the use — often excessive use — of tobacco snuff and, in several tribes, of a snuff made from coca (Erythroxylon Coca). He never met with any snuff called yopo or niopo nor with a snuff prepared from beans of a tree; furthermore, he failed to encounter, wild or culti- vated, a single tree of Anadenanthera peregrina. During the course of explorations in the Río Apaporis in Amazonian Colombia in 1951 and 1952, Schultes had, as assistants, several Puinave Indians from the Río Inírida: The Inírida is the highest Colombian affluent of the Orinoco, and it represents an area where the floras of the Río N egro- Vaupés and the upper Orinoco floras blend, Consequently, the Puinave helpers, from whom he first learned of the myristicaceous snuff, were familiar with many of the plants found in the Apaporis basin. They indicated two species of Virola — V. calophylla Warburg and V. calophylloidea Markgraf — from the resin of which a highly intoxicating snuff, called yá-kee in the Puinave language, could be elabo- rated (21, 22, 23). During 1951, the almost wholly uninhabited Río Apaporis was opened up at several points for tapping wild Hevea rubber, and natives from sundry tribes of the Colombian Vaupés and even from the Rio Uaupés of Brazil were trans- ported into the area by air. Schultes had, consequently, an 120 Rhodora [Vol. 70 unparalleled opportunity of studying the preparation of the narcotic by natives of tribes normally isolated from one an- other by great distances, and his field investigations dis- closed that a number of tribes knew of Virola-snuff, the use of which was restricted to their witch-doctors: the Puinaves of the Rio Inírida'; the Kuripako of the Río Guainía; the Kubeo of the Rio Vaupés and Rio Papurí; the Barasana and Makuna of the Río Piraparaná; and the Taiwano of the Rio Kananarí. Reports indicated that its use was known amongst Indians, especially the Tukano, of the Rio Uaupés in Brazil, and indirect evidence placed it likewise amongst the several tribes living on the Rio Issana. Virola-snuff is called yá-kee in Puinave, yá-to in Kuri- pako. Amongst the Tukano and linguistically related tribes, it is known as pa-ree-kd, a loan word from the Nheéngatu or Lingoa Geral, a Tupi-Guarani language once widely spoken in the northwest Amazon, especially of Brazil; this is the term by which snuff from Virola and from Anadenan- thera are known today by the civilized inhabitants of the Rio Uaupés — Rio Negro area of Brazil. In the Colombian Vaupés — its westernmost area of use — Virola-snuff is prepared rather simply. Schultes saw it prepared by Puinave, Tukano, and Kuripako Indians, and descriptions of its preparation from members of the Kubeo, Barasana, and Makuna tribes indicated no appreciable devi- ation from the process. In all cases, the same species were pointed out as sources of the snuff, although the possibility that other species of the genus could be used was indicated. ‘What almost certainly indicates the use of Virola-snuff by the Puinave of Laguna Negro on the lower Río Inírida in Colombia is a passing reference (19) to a snuff called ya-to, prepared from the inner bark of a vine known also as ya-to, resembling yasé [yaje?], and the ashes of a tree called matuire. Both are boiled, then dried and pulver- ized. How much reliability can be placed upon the botanical statement that a vine is the source of the snuff is open to question. Virola spe- cies, of course, are all trees, never vines, but was the source plant actually observed or is this report merely the uncritical repetition of a native's statement? We can say, however, that the vernacular name of the snuff in Puinave is ya-kee, not ya-to, which is the term applied to Virola-snuff by the neighboring Kuripako Indians (22). 1968] De Plantis Toxicariis — Schultes 121 The Indians strip the bark from the trees usually in the early hours of the morning, before the sun has begun to penetrate the forest canopy to heat up the trunk. Large strips of the bark, which peels easily from the cambium, are torn from the trunk and tied into loose bundles. Almost immediately upon separation of the bark from the tree, a profuse exudation or “bleeding” of a thick reddish resin- like liquid, which soon becomes viscous, oozes forth from the inner surface of the bark in small drops. The active princi- ple is contained in this exudation — called oom ( latex), or, specifically referring to these species of Virola, há-oom-tee- ét or yd-kee-oom, in Puinave. According to the natives, this exudation is greatly reduced in quantity and is weaker in its narcotic effects when the trunk of the Virola tree has re- ceived the warmth of the sun's rays. The bundles of bark are brought in and placed in water for about half an hour. They are then taken out, and the soft inner layer, on the surface of which the red exudation has congealed, is rasped off with a knife or machete. The shavings or raspings (yd-kee-tao in Puinave) are thrown into a large earthenware pot or enamel tray, and the rest of the bark is discarded. When enough shavings have been accumulated, a small amount of water is added, and the mass is thoroughly kneaded and squeezed by hand. The water becomes muddy and assumes a brownish or tan hue. This turbid liquid is strained several times, usually through a piece of finely hammered bark-cloth (prepared from Olmedia or Poulsenia), into a small-mouthed earthenware pot. The residual shavings, when as much of the water has been expressed as possible, are thrown away. Enough water is added to the strained liauid to fill the pot, which is then set to simmer over a slow fire. From time to time, a sordid foam, which rises to the surface, must be scraped off with a piece of bark. The boiling is allowed to continue for three or four hours, more water being added if evaporation is too rapid, until nothing remains except a thick, dark brown syrup at the bottom of the pot. This syrup must not be dried rapidly over a fire; the pot is set in the sun, and the syrup 122 Rhodora [Vol. 70 is permitted to solidify slowly. When nothing but a dry, brown crust is left, the residue is scraped free from the pot and is ground into a fine powder with a water-smoothed stone as a pestle, and the pot or an enamelware tray as a mortar. It is then ready to be mixed with ashes which have been made from the bark of a small wild cacao tree (Theo- broma subincanum Martius). Usually, equal amounts by volume of ashes and ya-kee powder are used, When they are thoroughly mixed, the product is put into a small bag made of finely hammered bark or cloth and is sifted through the bag by means of a gentle beating against the side of a small-mouthed receptacle. The resulting dust is the finished snuff. It is kept either in a small glass bottle, tightly corked, or else, more traditionally, in a type of jar made from a large snail-shell to which a hollow bird-bone tube has been fixed with pitch (usually from Symphonia spp. or Morono- bea spp.). This tube is stoppered with a plug of feathers glued together with pitch at the basal end to form a tight- fitting stopper. Amongst all the tribes of the Colombian Amazon, the consumption of Virola-snuff is limited to the medicine-men and is, therefore, small. Since it is said to lose its intoxicat- ing properties rather rapidly, even when in a tight con- tainer, it is made in small amounts and frequently. The next reference (2) to what seems to be the same myristicaceous snuff is found in an extraordinary article on the culture of certain Waiká Indians,* mainly those living "In view of the extreme confusion in terminology in current anthro- pological literature and in line with a more popular usage of terms, we intend, for the ethnobotanical purposes of this study of snuffs, to employ the term Waiká generically for many groups of Indians, apparently closely related, who dwell in southernmost Venezuela and northwesternmost Brazil, in the headwaters of the Orinoco and on northern affluents of the Rio Negro, respectively. The final classifica- tion of peoples, naturally, lies with the anthropologist. Pending, how- ever, a clearer classification for the area in question, specialists in other disciplines — such as ethnobotany — must have a simple nomen- clature, and, consequently, we are faced with the necessity of choosing one for use in this paper, pleading the while for a more logical, mean- ingful and standardized treatment of what are obviously closely 9 1968] De Plantis Toxicariis — Schultes 123 related, sometimes even identical, peoples. Present knowledge of the snuffs of South America seems to support a suspicion that the Waika (in our sense of this term) form the center of today’s most intensive use of the myristicaceous snuffs and may even represent the area of origin of the use of these preparations. It is this consideration that has made imperative our search for a simple terminology for use in the present summary. The Indians in question, who live in isolated hamlets or single communal malocas, are sometimes referred to inclusively as Yanonami, although the generic term Waiká is likewise very often employed to designate this large and dialectically somewhat heterogeneous group of natives. Actually, Waiká or variants of it represents the oldest term, having been first employed in the literature in 1779 (4). Some specialists have spoken of the Waiká and the Shirianá, although the name Shirianá (Xirianá) has also been used for groups in the Pari- ma Mountain area of British Guiana and along the Ríos Uraricapará and Uraricoera in the uppermost Orinoco drainage-area. Koch-Grün- berg (14) was apparently the first to conclude that the Waiká (Guai- cá), Kirishaná, Guaharibo and Shirianá, all of the Parima region, were probably one group of people. Métraux (17) considered the Shirianá and the Waiká as distinct. Zerries (47, 49) referred to the Waiká, Guaharibo and Shirianá Indians, which apparently may form *an identical group inhabiting the headwaters of the Orinoco of Venezuela and adjacent parts of Brazil" and used Waiká for the entire group called Yanomamó by later workers. Becher (3), who employed the term Yanonami, distinguished several sub-groups living along the Rios Aracá and Demini, tributaries of the middle course of the Rio Negro, as the Surára and Pakidái. Seitz (31) followed Becher in considering the Waiká as belonging to the Yanonami or Yanoama, although one of the groups that he visited on the upper Rio Marauiá he called the Karauetari tribe. Salathé (20) referred to the Karimé, culturally related to the Waiká and living in the head- waters of the Rio Caterinani in Venezuela. Barker (2), a missionary with many years of experience with these several peoples in the upper Orinoco, stated that the *. . . Guaiká [Waiká] tribe inhabit the upper Orinoco and its tributaries . . . and also extends to Brazil- ian territory, occupying the headwaters of various affluents of the Rios Branco and Negro" and that the Waiká and Guaharibo *. .. speak the same language, but with dialectical differences." The Gua- haribo, he reported, refer to themselves as the Shidishaná and Shiri- aná. Migliazza (18) did not accept Becher's inclusive use of Yanonami or Yanoama but preferred to speak of the "linguistic family Xirianá," living in a circular area the center of which is the Parima Mountain on the boundary of Brazil and Venezuela. They number, he stated, some 5000, divided into small groups of from 30 to 200 or more in- dividuals; these groups, each generally with its own dialect, are 124 Rhodora [Vol. 70 usually known as Shirianá, Waiká, Shamtari, Parahuri, Yanamani, etc. Wilbert (41), who has offered perhaps the most detailed treat- ment of the classification of these peoples, considers the Waiká as one sub-group of the Yanoama, suggesting that they are descended from very ancient inhabitants of South America. He considered the Surára-Pakidái (Sumatari or Samatari), Waiká, Karimé, Shirishaná (Casaparé), Warema (Sanema, Pubmatarí as all related and be- longing to the Shirishana or Guaharibo family. Most recently, Chag- non (5, 6) has accepted the term Yanomamo for “...a tribe in Vene- zuela and Brazil who practice a slash-and-burn way of horticultural life," a term that the employed “in a linguistic-cultured sense to define the tribe. The word, he maintained, “means true human begins” since “their conception of themselves as the only true ‘domestic beings’. . is demonstrated by the contempt with which they treat non-Yano- mamoó who... are ‘wild.’ In addition to meaning ‘people,’ Yanomamó also refers to the language . . . Sub-Yanomamoó groupings are based on language differences, historical separation and geographical loca- tion". We should point out parenthetically that tribal designations amongst South American Indians very commonly, in fact almost as a rule, mean only “human beings", “proper people", etc., to differentiate them from other linguistically non-accepted groups; it is not peculiar to the Waiká or Yanamamo. He asserted that “the Indians referred to themselves as Yanomamoó, which at once defines their mode of life, language, origin and includes all people who share this mode of life, language and origin". He considered the Sanema, Shamatari and Waika as “local groups". Two of these distinguishable groups, he pointed out, are the Waiká and Shamatari, both speaking nearly the same dialect but distinguished because of their separation and Shama- tari "differentiation" from the Waiká in the past seventy-five years: those moving “to the south side of the Orinoco came to be called Shamataris by those living on the north side, and the term is now applied to any village in this area, whether or not it can trace its origin to the first supporters of Shamatari" (a kind of hero who, following intra-tribal fights, led his followers south to the Orinoco). Chagnon's map (5), shading in the Yanomam6 area of southern Vene- zuela and northern Brazil, shows that his term is equivalent to what we prefer to call Waiká. He estimates their population to be up to 10,000, over 7,000 of them in Venezuela. It may be significant that none of the several contributors to the Handbook of South American Indians (9, 13, 15, 16, 17, 32, 42) who have had occasion to refer to these Indians have employed the terms Yanamamô or Yanonami. They have used Waiká, often however dis- tinguishing between the Waiká and the Shirianá. These examples from the divergent opinions of the several anthro- pologically oriented specialists who have worked with and/or written about these Indians illustrate the dilemma that faces the botanist 1968] De Plantis Toxicariis — Schultes 125 near the mission station of E] Platanal or Mahekodotedi on the uppermost Orinoco in Venezuela, not too distant from the area where Koch-Griinberg reported the bark snuff of the Yekwana tribe. In 1953, Barker, an American mission- ary with long experience in the region, offered a brief but telling sentence concerning the source of Waiká snuffs: “Different kinds of ‘yopo,’ of distinct strengths, are pre- pared from leaves, bark and ashes of bark." It should be noted that he made no mention of a bean or seed, thus, it would seem, ruling out the possibility that these people elab- orated their snuff from Anadenanthera peregrina. A year later, in 1954, the German anthropologist Dr. Otto Zerries (45, 46, 47, 48) began to publish a series of articles on the same and allied Indians: the Waiká of the village of Mahekodotedi and their relatives and neighbours whom he calls the Shi dishána of the Cantinama region of southern Venezuela. He wrote in great detail on the use and cultural significance of the narcotic snuff. His reference to the botanical sources of the snuff are, unfortunately, somewhat confused and often indecisive. Zerries referred to “yopo-snuff” as “a characteristic fea- ture of Waiká culture". This snuff, which the natives called ebéna and use “to be intoxicated and establish contact with the hekula, spirits of rocks and waterfalls’, appeared “to be made of the seeds of Mimosa acacioides or of a Pipta- denia species", a plant known to these Waiká as hisioma. Since Zerries worked with the same Indians on which Barker had based his report, one wonders, in view of Barker’s explicit enumeration of “leaves, bark and ashes of bark" and his omission of “seeds”, whether Zerries actu- ally observed the use of seeds. By Zerries’ acceptance of Cooper's distribution map detailing the use of Anaden- anthera (Piptadenia) peregrina for snuff, it would appear that realizing that in certain parts of the upper Orinoco drainage-area, Anadenanthera seeds formed the prime source of a narcotic snuff commonly called yopo or paricá, and pharmacologist. We trust that, under the circumstances, our course of action will be understood. 126 Rhodora [Vol. 70 he may conveniently have extended this generalization to his own group of Waika. It is quite apparent that Zerries’ identifications were not based on voucher botanical speci- mens. A botanist, had there been material to submit to a plant taxonomist for determination, would most certainly have known that the binomial “Mimosa acacioides" is mere- ly a synonym of Piptadenia peregrina, Consequently, by extrapolation, it seems wholly probable that this “identifi- cation” of the source of ebéna was neither made by a bota- nist nor based upon voucher specimens. In late papers, Zerries asserted that the “yopo” of these Waika is prepared basically from this leguminous tree. It may be interesting, if only to illustrate the complex role that these intoxicating plants play in Waika philosophy, to quote more fully one of Zerries’ discussions of certain ethno-botanical aspects of the sources of the narcotic snuff of the Waiká. “The snuffing of yopo-powder during which persons sitting in juxtaposition, snuff each other through a tube half a meter long, is an indispensable prerequisite for invoking the spirits, for only then can the hekula enter the medicine man’s chest and allow him to see the spirits in his yopo-intoxication. Sometimes, he is overcome with the drug and his own imagination and, more or less, loses his senses. This may sometimes be immediately manifest in a kind of seizure by the spirit that has laid hold of him, so that the medicine man cannot be exorcised by his col- leagues. Now and then, he loses consciousness, a condition that the Waika revealingly call by the word for “dying”. This effect is produced presumably by the alkaloid of a plant resembling a Mimosa, Piptadenia peregrina, and which is one of the numerous vegetal ingredients of the snuff. The plant hisioma, . . . from which this ingredient of the yopo-powder comes, . . . is subordinate to two hekula spirits, the ihamaliwa and kuhidiliwa, the lords of the sloth and of an unknown bird, who had, in primeval times, allowed the soul of the hisioma to come down and thus create the plant itself . . . Another ingredient, mashohara, derived from a species of Piperaceae, is also called yauardi- 1968] De Plantis Toxicariis — Schultes 127 hena (‘leaves of the forest spirit) . . . A further snuff powder which, unfortunately, cannot be botanically identi- fied, is called bolek-hena, signifying ‘leaves of the spirit of the dead’; this plant is subservient to hayacowardi, a female demon in the form of a wildcat... Whoever... is rubbed with this plant will always . . . be at the mercy of this female demon." A most revealing statement in this regard was made by Zerries in 1954(44). In speaking of other ingredients of ebéna, he mentioned masho-hara or yauardi-hena, which he described as the leaves of a piperaceous plant cultivated in the garden plots for this purpose and a powder made from an unknown plant known as bolek-hena (“leaves of the death spirits"). “The strongest ingredient is yacoana, the inner bark of a wild tree, which is dried, often toasted. I presume," he wrote, “that it is a species of Mimosa (Mimosa acacioides), which contains an alkaloid.” Here Zerries asserted that the strongest ingredient of the snuff is the inner bark of a wild tree, not a “seed”. Notwith- standing his presumption that it is Anadenanthera pere- grina, there can be little doubt that this “wild tree" is a Virola: the inner bark is the part of Virola that is em- ployed, and Virola is called nyakwana in other Waika groups. There are several other considerations in connection with possible identification of the plant ingredients of the Waika snuff. These Waikas are forest-dwelling Indians, and Ana- denanthera peregrina, definitely not a jungle tree, is found characteristically in open semi-savannah country. Zerries' descriptions and photographs of the snuffing equipment, snuffing ritual and physical effects of the narcotic are, in every detail, identical with those observed for Virola-snuff in other groups of these same peoples. Consequently, it is obvious that we are left with extreme uncertainty, insofar as the botany is concerned, by Zerries’ writings on the snuff. There are several possibilities: 1) that the snuff is prepared from seeds of Anadenanthera peregrina; 2) that it is elaborated from the inner bark of a 128 Rhodora [Vol. 70 species of Virola; 3) or that there are several kinds of snuffs made by the Waika — one basically active because of Ana- denanthera, the other, basically active because of Virola. In view of the later anthropological and botanical field work, we believe that the second possibility is the more probable and that the ebéna of these Orinoco Waika is a myristicaceous snuff. In 1960 the German anthropologist, Dr. Hans Becher (3), referred to the use of an intoxicating snuff, especially in connexion with adolescent initiation rites and endocanni- balistic ceremonies, by the Surára and Pakidái living on northern affluents of the Rio Negro in Brazil. These groups employ only two narcotics: tobacco, used hedonistically ; and epéna-snuff, used ceremonially. The narcotic epéna- snuff, he reported, is, like yopo or paricá, prepared from a number of ingredients. It is made, according to this reference, from the roasted and powdered paricá seeds ( Anadenanthera peregrina) to which the ashes of the bark of a “wild species of Mimosa", hekurahihena, and the pipe- raceous maxanaha are added. He suggested that the wild mimosoid plant might be “Mimosa acacioides", apparently not realizing that this binomial is a botanical synonym for Anadenanthera peregrina or Piptadenia peregrina, In one point in his discussion, Becher equated the “piperaceous” mazahara with the mashoraha reported by Zerries from the Waika of Venezuela. There is no indication that botani- cal specimens formed the basis of the identification of these plant sources, and one wonders whether or not Becher relied upon the identifications earlier published in Zerries’ papers. A recent popular article by Dr. Napoleon Chagnon (6) on the Waika and related Shamatari failed to refer to the preparation of their snuff, mentioning the narcotic only in a caption under an illustration showing a group of Indians snuffing: “In preparation for duels, the Indians take their hallucinatory drug, ebene ...”. In his unpublished doctoral thesis Chagnon (5) likewise did not consider the snuff. In 1960 Mr. George Seitz (29), who with his wife had, in 1957, visited a group of Waika and others whom he calls 1968] De Plantis Toxicariis — Schultes 129 Shamatari and Araraibo on the Rio Maia, a tributary of the Cauaburi, published a book on his trip to these people. He described the taking of snuff by these Waika but, as he wrote, “as regards the powder snuff, we could find out nothing. He [the Indian] said that it was prepared from the bark of a shrub which grows only in the montains. His description, however, was so general that we could not identify the plant.” Later, in a letter to Wassén (April 19, 1961), Seitz added significant data concerning the prepara- tion of the snuff, called epéna by these people: ‘The snuff, the elaboration of which we saw more than once, consists of the following ingredients: 1) The bast of the bark of a tree (called epéna-kési). This bast is dried, toasted and then powdered. 2) The outermost bark of young stems of an Acacia-like plant (called ama-asita). It is dried com- pletely and burnt to ashes (called jupi-uschi). An herb about 30 cm. tall (called mashi-hiri) is dried and powdered and mixed with both of the other ingredients." Subsequent to this letter, Seitz added even more data in several articles (30, 31) and has produced an excellent motion picture showing the preparation and use of epéna-snuff amongst the Waiká of the Rio Maturacá. This more recent information has been supported by botanical specimens that have been submitted to taxonomists for identification and prepared snuff-power that has been studied chemically and toxico- logically (12). In 1965, Seitz (30) published observations on the use and effect of epéna in which he summarized what he then knew concerning the plant ingredients of the snuff. He recognized two different snuff preparations. One was, as he had previously stated, prepared from three ingredients: 1) the bast of a tree; the small leaves of the herbaceous mashi-hiri; and the ashes of the outer- most bark of an Acacia-like species called ama-asita. Botanical material of the first and apparently principal ingredient was identified by Dr. Eberhard Schmidt who *. . . came to the unequivocal conclusion that the specimen was Virola, fam. Myristicaceae. To my great content, I afterwards found the following statement: Engler-Prantl, 130 Rhodora [Vol. 70 Die natürlichen Pflanzenfamilien, 2. Aufl., Band 17 a II, p. 193. The bark from Virola calophylla Warburg and V. calophylloidea Markgr. — gives a narcotic snuff powder (yá-ka, yá-lo) used by various Indian tribes in the Amazon region and in Colombia". Seitz did not offer botanical deter- minations for the second ingredient. The third, he re- marked, resembled an Acacia. We have had an opportunity of examining Seitz' botanical material. Some of it comprises small pieces of bark and twigs that do not lend themselves to easy or exact identi- fication, but some of the specimens are very adequate. Of the principal ingredient, we have one twig (Seitz 1b) and several pieces of bark (Seitz 1c) collected on the Rio Marauiá. They seem definitely to be referable to the Myris- ticaceae. These specimens, together with a photograph of a sterile branch of the epéna plant, allow us to identify the principal ingredient as Virola. Seitz has identified it in his papers as Virola calophylloidea, but the photograph looks to us more like V. theiodora. He submitted no specimens of mashi-hiri, the second ingredient. Of the third plant, ama-asita, we have seen several small bark samples (Seitz 1h) and a small transverse section of a branch (Seitz 1i), both of which are definitely leguminous; and a dried, pressed leaflet and a photograph of a branch, both of which we would refer to the genus Elizabetha. Seitz (31) reported of ama-asita that it “is a tall tree . it seems to be a Trichilia species . . . and seems to be rare." Powder prepared from these three ingredients was analyzed by one of the writers (Holmstedt), using both bioassay and conventional biochemical techniques, with especially significant results from gas chromatography (10). The main component is 5-methoxy-N, N-dimethyl- tryptamine, but smaller amounts of N, N-dimethyltrypta- mine and 5-hydroxy-N, N-dimethyltryptamine (bufotenine) were likewise found in it. Amongst the Waiká in the Marauiá drainage-area, Seitz learned of several other ingredients of snuff, but nothing is 1968] De Plantis Toxicariis — Schultes 131 known of their botanical identification. “There is also used another snuff powder which contains, besides the above mentioned three ingredients, the other vegetables: 1) the leaves of a plant called poschi-have-moschi-hena (hena means ‘leaf’); 2) the leaves of another vegetable called ai-amo-hena. In the villages we visited, the Indians either could not or did not want to show us these two plants. They always said that they only grew in the higher region of the mountains and not nearby. For this reason, the powder compound of the five ingredients was not at hand. In my opinion, it is the same compound whose snuffing we saw in our first expedition and whose effect was described as noxious for health . . . I cannot at the moment say more about this powder. Neither the missionary with whom I am corresponding and who lives in continued contact with several tribes, nor myself, saw in our other expeditions a similar effect again, and in no other visited tribe were we able to get this powder." There are, further, several very interesting and possibly significant Seitz collections from the Cachoeira Guirapajé on the Rio Marauiá. One (Seitz 4b) consists of a small piece of a branch of Anadenanthera peregrina. The other (Seitz 4) is a very adequate herbarium specimen in fruit of this same species. Excellent photographs of the tree from which these specimens came were published by Wassen (36): one, an unmistakable trunk of Anadenan- thera peregrina with the characteristic mammilose pro- jections; the other, leaves and pods of the same tree. Seitz nowhere indicates that the Waika employs this tree in preparing a snuff. His collections and photographs do, however, establish the presence in this part of the Rio Negro of Anadenanthera peregrina. This area is deeply and uninterruptedly forested, and Anadenanthera peregrina does not occur there naturally. It is obvious that these Waika must have imported seeds of this tree from rather distant localities of semi-open savannah, such as parts of southern Venezuela or the Boa Vista area of the Rio Branco of Brazil, where Anadenanthera peregrina grows sponta- 132 Rhodora [Vol. 70 neously. The question remains, nonetheless: why are these Rio Marauiá Waika cultivating Anadenanthera peregrina? Seitz (31) has given further valuable data on the myristi- caceous snuffs. In the town of Tapurucuara, a large village on the main course of the Rio Negro above its confluence with the Rio Branco, Seitz witnessed the preparation of a snuff by a Tukano medicine-man originally from the Río Papurí on the Colombo-Brazilian boundary. This powder was called paricd. The medicine-man still used the snuff in his practice. In its preparation, according to Seitz, he em- ployed: “... the same raw material as the Waiká Indians — the inner layer of the bark from Virola calophylloidea Markgraf, but he prepared the powder in a very different qv? way". Scraping off the inner layer of the bark, he cast the pieces into a pot of water, kneaded them and squeezed until the water was reddish brown and muddy. Then the liquid was evaporated over a slow fire. In several hours, there was a hard crust left; this was scraped off with a knife and ground finely with a smooth stone, This powder, with- out any further admixture, was the snuff. We have examined four herbarium specimens collected by Seitz in Tapurucuara in August, 1965, as representing the species employed by the Tukano medicine-man as his source of paricá. They appear to us to represent, like the material gathered in the Waiká village on the Rio Marauiá, Virola theiodora. During the summer of 1967, we were able to pursue our studies of myristicaceous snuffs in two widely separated Waika villages in the Rio Negro drainage area of Brazil: Maturacá, on the uppermost Rio Cauaburí, an affluent that enters the Negro above Tapurucuara and the rises in the Sierra Neblina land mass on the Venezuelan-Brazilian boundary; and Wayhana-oo-thle, a very small village on the Rio Tototobi, a headwaters tributary of the Rio Demini, a northern affluent that empties into the Negro slightly above the town of Barcellos. 1968] De Plantis Toxicariis — Schultes 133 Plate 1373. Foliage and flowers of Virola theiodora, Manáos, Brazil. The natives at Maturacá are closely related to those con- tacted by Seitz on the Rio Marauiá, a river that parallels slightly to the east, the Cauaburí. Understandably, their method of preparing the snuff is very similar to that de- scribed by Seitz. The natives on the Tototobí, on the con- trary, elaborate their snuff quite differently. The Maturacá Waiká store epéna in a large bamboo tube hanging from the house beams, and it is employed by any adult male singly or in groups at any time as well as during festivals. The tube is kept full, and the snuff, consequently, is always available for use. Every now and then, an Indian 13: Rhodora [Vol. 70 will take the snuff, become intoxicated, dance and sing all alone, with the rest of the village going about its usual chores and not paying any heed to him. As Seitz noted in the neighbouring Karauetarí village on the Rio Marauiá: * . . there is no system for the snuff ceremony ... . there were Indians who took epéna-powder every day at any time in the afternoon; there were others who practiced the cere- mony only once in a fortnight. Seldom did we see any formal motive for taking the snuff, such as curing a sick person, invoking success in the hunt or thanksgiving for a successful hunt". Plate 1374. Waiká Indian collecting resin from heated pieces of bark of Virola theiodora. Rio Tototobí, Brazil. 1968] De Plantis Toxicariis — Schultes 135 The ingredients of the epéna-snuff at Maturacá are three plants: Virola theiodora, Justicia pectoralis var. steno- phylla and Elibabetha princeps. Bark is stripped from trees of Virola theiodora. A young tree (height about forty feet, diameter fifteen inches) is preferred, although older and larger trees may be used. The natives assert that the yonger trees have a "stronger blood", but more probably the choice of the young tree is dictated by easier stripping of the bark. The strips are bundled and brought into the village. The soft, inner layer is carefully scraped off; the shavings are dried — first in the sun, then by gentle toasting over a slow fire — and Stored until needed. When a new batch of snuff is to be made, these crisp shavings are pulverized by grinding them between the palms of the hands, the hands being held firmly between the knees while the grinding motion is in process. The ground shavings drop from the hands onto a banana leaf. The crudely ground shavings are collected from the leaf and placed in a hollow lecithydaceous fruit as a mortar ( Bertholettia excelsa H. & B.) and triturated with a heavy wooden pestle until the material is almost completely pul- verized. This powder is then sifted through a small basket to remove fibres, pieces of wood and other undesired refuse. The resulting powder is very fine, homogeneous, chocolate- brown and highly pungent. Next, a powder of the leaves of a plant called mashi-hiri is prepared. This twelve- or fifteen-inch herb, cultivated in dense patches near the edge of the village, is the acan- thaceous Justicia pectoralis var. stenophylla. It is kept hanging in bunches from the house-beams and is, conse- quently, usually quite dry when needed. The whole plant is crushed between the hands, the powder is sifted to remove bits of the stem and other refuse, and the resulting fine greenish dust is added to an equal amount of the brown Virola-powder. The Justicia plant is pleasingly aromatic as it hangs drying, and the prepared powder is even more highly aromatic. The natives assert that it is added to "improve the smell" of the final epéna-snuff and that it is 136 Rhodora [Vol. 70 Plate 1375. Justicia pectoralis var. sphenophylla growing at the edge of a cultivated plot. Maturacá, Rio Cauaburí, Brazil. not active. While it is true that other groups of Waiká pre- pare a potent Virola-snuff without the Justicia, preliminary chemical investigation of this acanthaceous plant suggests that we may be unwarranted in assuming that it is wholly inert ingredient devoid of pharmacological activity. Seitz found that the Karauetarí did not add mashi-hiri to the snuff, whilst the Indians of the Marauiá did. He remarked of mashi-hiri, which undoubtedly is the same Justicia called mashi-hiri by the Waiká of Maturacá, that in *another Waiká village, near the Maturacá-channel, we saw that a third ingredient was added: the little leaves of a herbaceous plant, called mashi-hiri, like the epéna-scrap- 1968] De Plantis Toxicariis — Schultes 13 Plate 1376. Justicia pectoralis var. sphenophylla hanging to dry, together with a bamboo tube of Virola-snuff and the woven basket used to sift the prepared snuff. Maturaca, Rio Cauaburi, Brazil. ings, dried and powdered. These leaves, however, have no intoxicating effect. The Indians say they are merely aro- matic. I don’t know why the Karauetari didn’t use the plant. Perhaps it was not available at the moment, or the Indians in the Marauria River liked another flavour.” A number of years ago, a missionary working amongst the Waika in the headwater regions of the Orinoco in Vene- zuela handed one of the authors (Schultes) a partially rotted, matted roll of plant material that he claimed to be the source of a narcotic snuff of these Indians (28). The condition of the material was very poor, and it disintegrated 138 Rhodora [Vol. 70 Plate 1377. Waika Indian about to grind to a fine powder crushed and dried leaf material of Justicia pectoralis var. sphenophylla for mixing with the powdered resin of Virola and ashes of Elizabetha. Maturaca, Rio Cauaburi, Brazil. upon study, but it seemed to represent a species of Justicia, an identification that was corroborated by the late Dr. E. C. Leonard, specialist on the Acanthaceae. With the unsatisfactory preservation of the material, Schultes’ in- ability to visit the region personally, and the failure of other botanists who had worked in the general area to report it, Schultes more or less dismissed Justicia as a serious contender for inclusion in our list of hallucinogenic plants. We now realize that it is definitely a common ingre- dient of the myristicaceous snuffs of the Waika. 1968] De Plantis Toxicariis — Schultes 139 There is a further piece of field information that may now have a bearing upon this general topic. Justicia pecto- ralis var. stenophylla has been collected frequently, espe- cially near cataracts, in the Rio Apaporis drainage-area of Colombia. One of the collections (Schultes & Cabrera 15244) from the Rio Pacoa, an affluent of the middle course of the Apaporis, bears an annotation that the name of the plant in Puinave is ya-ka-yoó. Since the Puinave name for Virola-snuff is yá-kee, one might well wonder if this simi- larity of native epithets could indicate some connexion in use of the Virola and the Justicia. Schultes never, with the numerous collections that he made of Justicia pectoralis var. stenophylla, was given any data that might suggest that Indians of the Colombian Amazon employed the plant as an admixture with Virola-resin. Yet the coincidental similarity between the names yá-kee and ya-ka-yoó, viewed in the light of what we know amongst the Waiká, is pro- vocative of conjecture. The third ingredient of the epéna-snuff of Maturacá is the ash of the bark of Elizabetha princeps, known in Matu- racá by the Waiká name a-má. This tree is undoubtedly the same as that reported by Seitz as ama-azita on the Rio Marauiá. Bark is stripped from the trunk of this rather large and majestic tree. The inner, soft part of the bark is scraped off and thrown away. The hard grey outer bark is chopped into small pieces and set in a glowing fire. When they themselves begin to glow, they are removed, set aside and allowed to reduce to ashes. The ashes, which are nearly white, are carefully gathered up when cool, in- spected critically for possible contamination in the form of sticks, bits of leaves or other extraneous material. The ashes that pass this inspection are then mixed, in approxi- mately equal amounts with the Virola-Justicia powder. The resulting epéna snuff is greyish and extremely fine We have every reason to believe that this ash may be an inert ingredient, although it may serve as a means of drying, to free the alkaloids more easily from the resin, to keep the snuff from deteriorating rapidly when stored in the 140 Rhodora [Vol. 70 bamboo tubes, or merely for mechanical purposes. The manufacture of the narcotic snuff amongst the Waika of the Rio Tototobi is most interesting and very different from the methods described for other groups of these Indians. The resin of Virola theiodora is employed alone, without any admixture. The bark of Virola theiodora is stripped first from the base of the tree with the tree standing, then, after felling, the entire trunk is stripped. As at Maturacá, relatively young trees are preferred, If more bark is needed for the batch of snuff required, additional trees are felled. Our visit, fortunately, coincided with the annual endocanniba- listic festival and a large supply of snuff was needed: three trees were felled and completely stripped. The Tototobi Waiká, unlike the Maturacá group, apparently do not usually keep a large quantity of the snuff available at all times, nor is the use of the snuff, it appears, so frequent and casual. The natives assert that the bark on the lower four or five feet of the trunk is richer in resin than higher up on the tree. When the lower bark is stripped and used, and prior to the felling of the tree, an Indian boy climbs to the top of the trunk, to the first main branch, with a circle of stout vines into which he has placed his feet. He then repeatedly slides down the trunk, scraping the bark vigour- ously with the vine-loop each time with his whole weight. This operation, they believe, causes the resin in the bark to “loosen” and, later, to flow more readily when the bark has been stripped from the tree. Occasionally this same scraping operation is carried out with the back of a machete before the lower bark is removed. The pieces of bark that are stripped off measure two feet in length and five or six inches across. All of the prelimi- nary preparation of the snuff is done in the forest where the tree is felled. A fire is made in the path, these strips are laid over the slow fire with the inner surface of the bark facing up. The gentle heating of the bark causes a very copious "bleeding" of the red resin, so copious that 1968] De Plantis Toxicariis — Schultes 141 Plate 1378. Pulverized dried resin of Virola theiodora, ready for use as snuff, together with an unground piece of the resin. Rio Tototobí, Brazil. large drops flow together, and soon the inner surface of the bark is covered with the liquid. Then the Indian in charge of this operation picks up the strips, one by one, holding each piece vertically over a small clay pot. He lets the resin flow down and drop into the pot, running his index finger down the inner surface of the bark to scrape off every drop of the resin. The bark is once again laid over the fire and reheated. More resin appears, This is done several times before the Indian throws the bark away because the resin in it has been exhausted. A large number of these strips are treated in this way, unhurried care and meticulousness 142 Rhodora [Vol. 70 marking the whole operation. The amount of resin ex- tracted when we witnessed this work amounted to about two teacups full and was the result of “bleeding” three trees. One of the trees was about seventy feet in height and measured fourteen inches in diameter; the other two were smaller. When enough resin has been gathered, the Indian gradu- ally heats the pot, and the resin boils slowly and is allowed to boil until it becomes thick enough to carry back to the house without its slopping out of the pot during walking. By this time, most of it has, upon cooling, crystallized into a beautiful, bright amber-red resin. Once the Indians have returned to their house, the crystallized resin is scraped from the sides of the clay pot with a sharp-edged stone and is then pulverized finely by grinding the powder accu- mulated in the bottom of the pot with the same stone as a pestle. Once it is deemed sufficiently fine, the powder is removed and placed on a banana leaf. It is a light coffee colour. The native begins a meticulous scraping of the carbonized resinous material from the sides of the interior of the pot and the grinding, as before, of this very dark blackish brown powder until it, in turn, is deemed fine enough. When the whole operation of grinding had begun, certain lumps of very thick resin that had not crystallized were removed from the pot and put onto a small stick. This stick was put over a fire and hardened, then pressed out on a heated surface over the fire and allowed to carbonize. The carbonized lump was then added to the similar blackish brown powder in the pot and completely reduced to a fine powder. At this point, the first brown powder and the second blacking brown powder were united and fully mixed to prepare the final snuff. Either one of these two powders, "The authors, in self-experimentation, inhaled the second, carbonized , , powder. 1968] De Plantis Toxicariis — Schultes 143 Plate 1379. Waiká Indians picking out stem material from Jus- ticia pectoralis var. sphenophylla prior tc drying the leaves for use in the Virola-snuff. Rio Tototobí, Brazil. however, has the potency and may be used alone as snuff." This snuff — as well as the tree from which the resin is extracted — is known by the Waika of the Tototobi as nya-lcwá-na*. At Tototobí, nyakwana-snuff is made usually without any admixture. Ashes are, according to our information, never "Although we are not anthropologists, we feel constrained to call attention to the curious similarity between the name of a tribe and linguistic group of Karib-speaking Indians on the Río Ventuari in the uppermost Orinoco basin of Venezuela-Yekwana-and one of the Waiká names for the myristicaceous snuffs nyakwdna (the other Waiká name being epéna). This may be coincidental, but it should be pointed out, so that linguistically oriented specialists may con- sider whether or not there may be any significance to this observation. 144 Rhodora [Vol. 70 added. These people know and do cultivate Justicia pecto- ralis var. stenophylla, which they call masha-hára-hanak (hanak means 'leaf"). There is another name for this Justicia at Tototobi: boo-hanák. The Indians know that it can be used as an aromatic mixed with the Virola-resin dust, and they very occasionally do use it in this way, but it is the exceptional, not the usual, method of preparing nya kwána-snuft. Interestingly enough, the Tototobí people do not know the custom of mixing ashes of Elizabetha or of any other plant with the Virola-powder. What may be even more significant is that they do not use Anadenanthera peregrina in preparing a snuff. In fact, several of these Tototobí Waika knew vaguely of the snuff made from seeds ( Anade- nanthera peregrina) amongst the Makiritare far to the north and northeast and affirmed that they themselves had never employed seeds or beans to make snuff. We saw no cultivated trees of Anadenanthera peregrina in or near their village, and they assured us that the tree was un- known in their area. III. There is still an extraordinary facet to this study of Virola amongst the Waiká — its use as the source of an arrow-poison. Salathé (20) reported that the Karimé, culturally and geographically close to the Waiká of the upper Orinoco region, use as a poison for hunting monkeys and birds “. . . jakuana, which is extracted from the bark of a tree”; the term jakuana is, obviously, related to if not the same as nyakwana employed for Virola by the Waiká of the Tototobí. In discussing the Waiká (Surára) of the Rio Arará, Becher (3) reported: “The third kind of arrow- poison is the weakest. It is prepared from the snuff powder epéna by mixing it with water and letting it boil, This method is likewise known to the Pakidái and Shirianá." Becher unfortunately could not witness the preparation of this poison. Since it seemed to be a closely guarded secret, his knowledge of it was only from hearsay. This information is, ethnopharmacologically speaking, 1968] De Plantis Toxicariis — Schultes 145 extraordinary. It is the first time that Virola has been indicated as the basis of an arrow-poison. Since we could not understand what constituent of Virola might be respon- sible for the toxic effects when employed on arrows or darts, and realizing that Becher had not witnessed the actual preparation of the poison, we were constrained to suspect an error in ascribing the source to Virola or to its being the same tree from which the epéna-snuff was made. During our visit with the Tototobi Indians, however, we fully substantiated Becher’s statement. Arrows were poi- soned with the resin from Virola theiodora. Furthermore, they were poisoned with resin from the very same trees from which the snuff was being made. And what is even more surprising, the arrows were poisoned without any prior preparation of the resin. These Indians state that they use no other kind of arrow- poison than this Virola-resin which is called simply nyak- wána-hoó-soo (hoó-soo meaning “latex” or “resin”). They know that other Waiká groups prepare arrow-poison “from a vine" and indicate (with hand motions) a large liana with roundish leaves and a stem diameter of eight or nine inches. The poisoning of arrows in the Tototobi group is inter- esting. During the operation in the forest of heating the pieces of stripped Virola bark and scraping off the warm resin into the clay pot, an Indian occupies himself solely with the poisoning of arrows. Running his index finger up and down the exposed trunk from which the bark has been stripped, he gathers the slime and paints bamboo arrow heads with it repeatedly. He then heats the arrow-heads slowly over the fire to liquefy the resin slightly, spreading it more evenly with his finger. The heads are slowly turned in the smoke of the fire. When they have cooled and the preliminary smearing has completely hardened, the Indian "The Waiká at Maturach pointed out to Schultes several meni- spermaceous vines that they employ as the basis of their arrow-poi- sons. 146 Rhodora [Vol. 70 Plate. 1880. Waiká Indian preparing arrow-points smeared with resin of Virola thetodora. Rio Tototobi, Brazil. begins to apply pure resin to the tips— resin that he scrapes from the heated strips of bark or that he takes from the accumulation in the pot. There are twenty or thirty applications of resin to the arrow heads, each appli- cation followed by a slow heating in the smoke. "Then, finally, our Indian wraps a knot of palm leaves around the top of a stick, inserting the arrow tips in this knot — like pins in a pin cushion — and sets the stick in the ground in the sun to dry slowly. This represents the whole process of poisoning arrows or darts, an exceedingly simple one compared to most of the elaborate curare preparations of South America. 1968] De Plantis Toxicariis — Schultes 147 It is significant to note that during the annual endo- cannibalistic festival, when these Waiká ran out of the snuff that they had prepared for the occasion, they took several bamboo tube cases, in which they had packed the newly prepared arrow tips, scraped the Virola-resin from the arrow points and used the resulting powder as snuff during the dance. It had the same effect as the snuff that they had been using and which had been exhausted, and, inasmuch as both the arrow-poison and the snuff had been prepared in almost identical procedures from the same trees, there is no reason why it should not be interchange- able. We do not yet know which chemical substance in this Virola-resin that, injected into the blood stream, could have properties making it valuable as an arrow-poison. It could also be possible that the hallucinogenic properties of the indoles of the Virola-resin act in the manner of an incapaci- tating substance. It may be of some significance that the Waika refer to this as a slow-acting arrow-poison and that the hunter must follow the animal, after shooting it, through the forest and wait until the poison has had its effect. IV. MYRISTICACEAE Virola, a rather natural genus of trees or rarely shrubs, comprises under forty-five species distributed in the Ameri- can tropics, mainly in the Amazon Valley and adjacent humid tropical regions. It is relatively well understood pri- marily through the monographic works of Warburg and of A. C. Smith, both of which are cited below. Virola calophylla Warburg in Nova Acta Leop. — Carol. 68 (1897) 231; Smith in Brittonia 2 (1987) 474. Distributed in the Amazonia regions of Colombia, Peru and Venezuela and the western half of the Brazilian Ama- zon. One of the most beautiful species of Virola with un- usually large leaves with well spaced lateral veins. [Vol. 70 Rhodora 148 E SRS B. X SEES el 2 + GN » Se 1968] De Plantis Toxicariis — Schultes 149 T D, M fa. Q NA a 3 q | AL "a ÉP 7777 ac? ( P I Q 4 MOE | N i > M 150 Rhodora [Vol. 70 Type of Virola calophylla (Spruce 3207) collected along the Casiquiare in Amazonian Venezuela. COLLECTIONS BASIC TO THIS WORK: COLOMBIA: COMISARÍA DEL AMAZONAS, Río Apaporis, Soratama. *Small tree about 35 feet tall; diameter 8-9 inches. Bark exudes inside a reddish resin when ripped off tree. Externally reddish brown, pebbled. Puinave—yd-kee. Source of narcotic snuff. In flood-forest. June 26, 1951. R. E. Schultes & I. Cabrera 12855. — Same locality. "Large columnar tree. Source of yá-kee-snuff. Flood forest. August 16, 1951, Schultes & Cabrera 13587. Virola calophylloidea Markgraf in Repert. Sp. Nov. 19 (1923) 24; Smith in Brittonia 2 (1937) 475. Distributed in the Amazonian basin of Colombia and the Amazonas of Brazil. Type of Virola calophylloidea (Ule 8846) collected near Manáos in Amazonian Brazil. Smith reported this species as rare in 1937, but since that time it has been shown to be relatively abundant, especially in the Colombian Vaupés and Amazonas. It can be distinguished from Virola calophylla by its shorter and more compact inflorescences and usually smaller leaves, amongst other characters. COLLECTIONS BASIC TO THIS WORK: COLOMBIA: COMISAREA DEL AMAZONAS, Rio Apaporis, Soratama. “Small tree along flood-bank. Flowers brownish. Puinave name— yd-kee. Source of narcotic snuff. July 3, 1951. R. E. Schultes & I. Cabrera 12872. Virola theiodora (Spruce ex Bentham) Warburg in Nova Acta Acad. Leop. — Carol. 68 (1897) 187; Smith in Brittonia 2 (1937) 470. Myristica theiodora Spruce ex Bentham in Jour. Bot. & Kew Misc. 5 (1853) 6. Distributed mainly in the western Amazonia of Brazil and Colombia and probably adjacent Venezuelan territory. Especially abundant in the Rio Negro drainage-area. Type of Myristica theiodora (Spruce 1444) collected near Manaos in Amazonian Brazil. 1968] De Plantis Toxicariis — Schultes 151 VIROLA thetodora ( Spr. ex Bth.) Warburg 152 Rhodora [Vol. 70 In Smith’s monograph of 1937, Virola theiodora is placed in synonymy under V. elongata (Spr. ex Benth.) Warb. There can be no misgiving about its relationship to this species but I have preferred to maintain it as dis- tinct because of the ease by which it can be distinguished in the field. When Bentham published Myristica theiodora, based on a fruiting specimen, and stated that it seemed to be close to M. sebifera, he wrote: ‘I should nevertheless in the absence of flowers, have considered it a mere variety of that species [M. sebifera], had not Mr. Spruce assured me that it is perfectly distinct. The leaves, when drying, are said to emit a strong odour of tea". The crown of Virola theiodora appears to be conspicuously denser and more compact than that of what I take to represent V. elongata, which has a more freely branching looser crown. The bark of the former seems usually to be thinner and less corky than that of the latter. Virola theiodora, furthermore, has, as Smith pointed out, thicker leaves and more prominent lateral veins, and the leaf bases are rounded or subcordate. In addition, the leaf blades in life are lightly but conspicu- ously sinuous near the margins, a character which my field observations indicate may be more or less peculiar to this species-concept. It is a character recognized by the Waiká Indians who more than once have pointed it out to me whilst we were searching the forests for a Virola-tree from which they could extract the resin for preparing snuff. COLLECTIONS BASIC TO THIS WORK: BRAZIL: ESTADO DO AMAZONAS, Rio Cauaburi, Maturacá. “Tree 60 feet tall; diameter small. Bark resin red. Bark ingredient of narcotic epéna-snuff. Waika—epéna. Alkaloid very positive. July 5- August 12, 1967 (R. V. Alpha Helix Expedition Amazon-1967). R. E. Schultes 24574. — Same locality and date. “Small tree, Bark resin red. Ingredient of Waika epéna-snuff. Schultes 24575. TERRITORIO DO RORAIMA, Rio Tototobi, Waika Indian village of Wayhana-oo-thle, “Medium sized tree, 60 ft. tall. Bark resin very abundant, reddish brown. Waika=nya-kwana. Used for preparing snuff and for smearing on arrows". Alkaloid positive. August 6, 1967 (R. V. Alpha Helix Expedition Amazonas-1967). R. E. Schultes 24026. 1968] De Plantis Toxicariis — Schultes 153 Miner ELIZABETHK, vm leg KS princepa ES | pas fl, Schomb. ex Bth. wns l [ T ER? (ab às 15 6 Y 79 18 . P e 4 fef 009%, ve vip Pe DS (Cl " ef (0,9 ^ n! La e = e Pi es * a e. 19 20 s M 21 22 TE dat rhea ig tak? > er e * A q o ees 1 BCE ^, / by 1968] Chromosome Numbers — Coleman 231 in these conflicting reports or that A. brasilianum has naturally occurring tetraploid populations. The count of n = 10 for Alomia fastigata constitutes the second report for that genus. Alomia micro- carpa (Benth.) Rob. has previously been reported also as having » — 10 (Turner and King in Cave, 1965). Three species of Stevia, each having n — 11, are reported. In addition to counts of n — 11, counts of » — 12 and 17 have been reported for North American species of Stevia (Powell and Turner in Cave, 1964). The count of n = 10 for Trichogonia gardneri represents the first report for this South American genus of about 25 species, the majority of which are Brazilian. Symphyopappus cuneatus and S. polystachyus, both n = 10, are the first members of that genus to be reported. Robinson (1919) expressed the opinion that Symphyopappus is a genus of doubtful value as the characters used to separate it from Eupatorium tend to break down. The two species of Symphyopappus reported here are similar cytologically to most species of Eupatorium thus far re- ported in having x = 10. In the present paper ten species of Eupa- torium, all based on x — 10, are reported. Four species of Mikania are reported, each having n = 18 or approximately 18. As the bivalents of these species tended to fall apart rather easily, pre- sumably due to low chiasma frequency, exact counts were difficult to secure. The report of n = c. 18 for Kanimia strobilifera constitutes the first report for that genus. Hoffmann (1894) placed Mikania in the subtribe Ageratinae and Kanimia in the subtribe Adenostylinae. This treatment tends to obscure the evidently quite close affinity of these two genera whose only strong distinguishing character appears to be that the achenes of Mikania are 5-ribbed whereas those of Kanimia are 8-10-ribbed, ASTEREAE — The report of » — 9 for Inulopsis scaposa is the first for this Brazilian genus of perhaps two species. This species was FIGURES 1. Vernonia polyanthes n = 18 2. Orthopappus angustifolius n = 11 3. Adenostemma brasilianum n = 10 4. Alomia fastigata m = 10 5. Stevia veronicae n = 11 6. Eupatorium intermedium n —10 7. Eu- patorium velutinum n = 10 8. Eupatorium ballotaefolium n = 30 9. Symphyopappus cuneatus n — 10 10. Symphyopappus polystachyus n => ]0 11. Kanimia strobilifera n = c. 18 12. Inulopsis scaposa n = 9 13. Solidago chilensis n — 9 14. Podocoma hirsuta n — 9 15. Cony- za chilensis n = 36 16. Baccharidastrum nottobelidiastrum n = 9 17. Baccharis helichrysoides n= 9 18. Baccharis oxyodonta n = 9 19. Pluchea suaveolens n = 10 20. Pterocaulon alopecuroides n = 10 21. Clibadium armani n = 24 22. Jaegeria hirta n= 27 23. Wedelia pilosa n = 28 24. Senecio adamantinus n = 50 25. Trixis pinnatifida n — 11 26. Hieracium commersonii n = 9 232 Rhodora [Vol. 70 included by Baker (1884) in the genus Leucopsis, but served as the type species of the genus Inulopsis which was erected by Hoffman (1894). Since no chromosome numbers have yet been reported for species undisputedly considered to be Leucopsis, no comparison of chromosome numbers between these two genera is possible. The count of n = 9 for Solidago chilensis concurs with a previous report for this species (Beaudry and Chabot, 1959). Aster squammatus has been reported as n = 10 (Solbrig et al, 1964). In the present paper an approximate count of n = 9 or 10 is given. The report of n — 9 for Podocoma hirsuta is the first report for this genus of about 12 species, the majority of which are South American but two of which are Australian. Conyza chilensis has been reported as hav- ing n = 36 (Solbrig et al, 1964) and n = 36-38 (Turner and King in Cave, 1965); the count given in the present paper is n = 36. The genus Baccharidastrum was erected by Cabrera (1937) who distin- guished it from Baccharis on the basis of the hermaphroditic heads of the former. The count of n = 9 for Baccharidastrum notobelli- diastrum (= Gonyza notobellidiastrum Griseb.) is the initial report for the genus. The three species of Baccharis reported brings to about 23 the number of species counted in this genus of about 350 species. All reports to date are based on x — 9 with polyploidy infrequent. INULEAE — The report of » — 10 for Pluchea suaveolens is in agreement with previous reports of other species of this genus. The count of n = 10 for Pterocaulon alopecuroides is the second report for the genus and the first for a South American species, Turner and Flyer (1966) have reported P. undulatum (Walt.) Mohr from Florida also to have » — 10. HELLIANTHEAE — The only previous report for the genus Clibadium is for C. surinamense L. which is given as n — 16 (Turner and King in Cave, 1965). In the present paper C. armani is reported to have n — 24 which would indicate Clibadium to be based on x — 8. Thes haploid complement of C. armani revealed a dimorphism in regard to size with 16 larger and 8 decidedly smaller chromosomes (fig. 21). The possibility that the 8 smaller chromosomes are B chromosomes can not as yet be excluded. Acanthospermum australe has been re- ported to have n — 10 (Carlquist in Cave, 1959) ; however the count given in the present paper is » — 11. Acanthospermum hispidum, the only other member of the genus thus far reported, also has "n —11 (Miége in Cave, 1964). Parthenium hysterophorous has been reported to have n — 18 (Thombre in Cave, 1960), » — 17 (DeJong and Longpre, 1963) and 2» — 34 and 35 (Rollins in Cave, 1963). The count reported in this paper is * — 17. The report of n = 18 for Ambrosia polystachya agrees with a previous renort for this species (Turner and Irwin, 1960). Jaegeria hirta has been re- 1968] Chromosome Numbers — Coleman Zoo ported three times previously, each as having n = 18 (Turner et al, 1962; Turner and King in Cave, 1965; Turner and Flyer, 1966). In the present study individuals from two populations were examined, one clearly showing n — 27 and the other » — c. 27. The conflicting reports for J. hirta suggest either that more than a single taxon is involved or the existence of naturally occurring triploid populations. Species of Wedelia have now been reported to have n — 11, 12, 15, 20, 22, 23, 25, 28, 29 — 1, and c. 36 thus revealing an aneuploid series to exist in the genus. Counts of n = 16 and 19 are reported for two species of Calea. In addition to counts of n = 16 and 19, counts of n — c. 17 and 18 have been reported for North American species (Turner et al, 1961; Turner et al, 1962). SENECIONEAE — The count of n — 10 for Emilia coccinea is in agreement with a previous report for this species (Baldwin, 1946) and the count of n = 20 for Erechtites hieracifolium concurs with previous reports for this species (Arano in Cave, 1963; Turner and King in Cave, 1965). The counts reported for three species of Senecio are consistent with the majority of previous reports for the genus in that they are based on x —10. MUTISIEAE — Counts of n — 24 for Chaptalia nutans and C. in- tegerrima, ccnfirm earlier reports for these species (Turner, 1959; Baldwin and Speese, 1947). Species of Perezia have previously been reported to have n = 8 (Heiser, 1963) and n = 12 (Diers in Cave, 1961). The report of » — 4 for P. cubataensis establishes that x — 4 in Perezia. The report of » — 27 for Trixis divaricata is consistent with previous reports for the genus. However the report of » — 11 for Trixis pinnatifida indicates Trixis to be dibasic with x = 11 and 27. CICHOREAE — Repcrts of n = 4 confirms earlier reports for Hy- pochaeris brasiliensis (Stebbins, 1953) and H. radicata (Heiser, 1963; Turner and King in Cave, 1965). Hypochaeris gardneri, n = 5, is reported for the first time. SUMMARY Chromosome numbers are reported for 68 species of Brazilian Compositae, initial reports being made for six genera: Baccharidastrum, Inulopsis, Kanimia, Podocoma, Symphyopappus and Trichogonia. Adenostemma brasil- ianum, previously reported to have n — 5, is reported to have n = 10 and Jaegeria hirta, previously reported to have n = 18, is reported to have n = 27. It is suggested either that more than a single taxon is involved in these conflicting reports or that naturally occurring polyploid populations 234 Rhodora [Vol. 70 exist. It is established that x = 4 in Perezia and Trixis is indicated to be dibasic with x = 11 and 27. DEPARTMENT OF BOTANY THE UNIVERSITY OF GEORGIA ATHENS, GEORGIA 30601 LITERATURE CITED BAKER, J. G. 1884. Compositae. In Martius, Flora Brasiliensis. 6. BALDWIN, J. T. 1946. Cytogeography of Emilia Cass. in the Amer- icas. Bull. Torr. Bot. Club 73: 18-23. , and B. M. SPEESE. 1947. Chaptalia nutans and C. in- tegrifolia: 'Their chromosomes. Bull. Torr. Bot. Club. 74: 283- 286. BEAUDRY, J. R. and D. L. CHABOT. 1959, Studies on Solidago L. IV. The chromosome numbers of certain taxa of the genus Solidago. Canad. Journ. Bot. 37: 209-228. CABRERA, A. L. 1937. Compuestas Argentinas nuevas o interesantes. Notas de la Museo de la Plata. 2: 171-204. CAVE, M. S. (ed.). 1956-1965. Index to plant chromosome numbers. University of North Carolina Press, Chapel Hill. Davis, P. H. and V. H. HEywoop. 1963. Principles of angiosperm taxonomy. 556 pp. New York. DEJowG, D. C. D. and E. K. LoNGPRE. 1963. Chromosome studies in Mexican Compositae. Rhodora 65: 225-240. GLEASON, H. A. 1906. A revision of the North American Vernonieae. Bull. N. Y. Bot. Gard. 4: 144-243. HEISER, C. B. 1963. Numeracion cromosomica de plantas ecuato- rianos. Cien. y natur. 6: 2-6. HOFFMANN, O. 1894. Compositae. In Engler and Prantl, Die Na- turlichen Pflanzenfamilien. 4(5): 87-391. ROBINSON, B. L. 1919. On tropical American Compositae, chiefly Eupatorieae. Contrib. Gray Herb. 55: 3-41. SoLBRIG, O. T., L. C. ANDERSON, D. W. Kvnos, P. H. RAVEN and L. RUDENBERG. 1964. Chromosome numbers in Compositae V. As- tereae II. Amer. Journ. Bot. 51: 513-519. STEBBINS, G. L., J. A. JENKINS and M. S. WALTERS. 1953. Chromo- somes and phylogeny in the Compositae, tribe Cichorieae, Univ. Calif. Publ. Bot. 26: 401-430. TURNER, B. L. 1959. Meiotic chromosome counts for 12 species of Texas Compositae. Brittonia 11: 173-177. , W. L. ELLISON and R. M. ko 1961. Chromosome num- bers in the Compositae. IV. North American species, with phy- letic interpretations. Amer. Jour. Bot. 48: 216-223. 1968] Chromosome Numbers — Coleman 235 — —-—, and D. FLYER. 1966. Chromosome numbers in the Com- positae. X. North American species. Amer. Jour. Bot. 53: 24- 33. — —-—, and H. S. IRWIN. 1960. Chromosome numbers in the Com- positae 11. Meiotic counts for fourteen species of Brazilian Com- positae. Rhodora 62: 122-126. -—, M. PowELL and R. M. KiNG. 1962. Chromosome numbers in the Compositae. VI. Additional Mexican and Guatemalan species. Rhodora 64: 251-271. Table 1. List of material examined for chromosome number Species m= Locality VERNONIEAE Elephantopus mollis H. B. K. 11 State of São Paulo: Par- que do Estado, São Paulo. 190º Orthopappus angustifolius 11 State of São Paulo; Par- (Sw.) Gleason que do Estado, São Paulo. 191 Vernonia diffusa Less. c. 18 State of Sao Paulo: Par- que do Estado, Sao Paulo. 429 Vernonia polyanthes Less, 18 State of Sao Paulo: Par- que do Estado, Sao Paulo. 416 Vernonia scorpioides Pers. 28+1 State of Sao Paulo: Par- que do Estado, Sao Paulo. 267 Vernonia cognata Less. 26-27 State of Sao Paulo: Par- que do Estado, Sao Paulo. 202 EUPATORIEAE Adenostemma brasilianum 10 State of Rio de Janeiro: Cass. Maromba, Parque Nacion- al de Itatiaia. 369 Adenostemma brasilianum 10 State of São Paulo: 13 km Cass. SW of Ubatuba. 206 Ageratum conyzoides L. 20 State of São Paulo: 25 km NE of São Paulo. 295 Ageratum conyzoides L. 20 State of São Paulo: Par- que do Estado, São Paulo. 204 Alomia fastigiata 10 State of São Paulo: Par- (Gardn.) Benth. que do Estado, São Paulo. 374 *Collection numbers are those of the author. 256 Eupatorium H. B. K. Eupatorium Sch. Bip. Eupatorium H. & A. Eupatorium DC. Eupatorium H. & A. Eupatorium H. & A. Eupatorium Eupatorium Eupatorium Eupatorium Eupatorium Gardn. Eupatorium Gardn. Rhodora ballotaefolium callilepis congestum intermedium laetevirens lanigerum macrophyllum L. orbiculatum pallescens DC. pallescens DC. vellutinum vellutinum Kanimia strobilifera Gardn. Mikania campanulata Gardn. Mikania capricorni Robinson DC. 30 20 20 10 10 30 10 10 10 10 17-18 [Vol. 70 State of São Paulo: Par- que do Estado, São Paulo. 304 State of São Paulo: Par- que do Estado, Sao Paulo. 428 State of Sao Paulo: Par- que do Estado, Sao Paulo. 338 State of Sao Paulo: Par- que do Estado, Sao Paulo. 209, 259 State of Minas Gerais: Brejo de Lapa, Parque Na- cional de Itatiaia. 374 State of Sao Paulo; Par- que do Estado, Sao Paulo. 261 State of Sao Paulo: Mid- way between Bertioga and Maresias. 245 State of Sao Paulo: 5 km E of Sao José dos Campos. 282 State of Rio de Janeiro: Maromba, Parque Nac'onal de Itatiaia. 355 State of São Paulo: 8-10 km ESE of Campos do Jordão. 346 State of São Paulo: 8 km S of Campos do Jordão. 306 State of Sao Paulo: Par- que do Estado, Sao Paulo. 258 State of São Paulo: c. 5 km E of Itirapina. 260 State of Sao Paulo: Par- que do Estado, Sao Paulo. 426 State of Sao Paulo: Par- que do Estado, Sao Paulo. 427 1968] Mikania conferta Gardn. Mikania pachylepis Sch. Bip. Stevia decussata Baker Stevia conmixta Robinson Stevia veronicae DC. Symphyopappus cuneatus Sch. Bip. Symphyopappus polystachyus Baker Trichogonia gardneri Gray ASTEREAE Aster squamatus (Spreng.) Hieron. Baccharidastrum | notobellidi- astrum (griseb.) Herter Baccharis genistelloides (Lam.) Pers. Baccharis helichrysoides DC. helichrysoides DC. Baccharis Baccharis oxyodonta DC. Chromosome Numbers — Coleman 18 18+1 11 11 11 10 10 10 9-10 237 State of Sao Paulo: 8 km S of Campos do Jordao. 181 State of São Paulo: Par- que do Estado, Sào Paulo. 291 State of Rio de Janeiro: Reboucas, Parque Nacional de Itatiaia. 379 State of Sao Paulo: Par- que do Estado, Sao Paulo. 430 State of Sao Paulo: 8-10 km ESE of Campos do Jordào. 334 State of Sao Paulo: Par- que do Estado, Sao Paulo. 192 State of São Paulo: Par- que do Estado, Sao Paulo. 262 State of São Paulo: Par- que do Estado, Sao Paulo. 187 State of Sao Paulo: Par- que do Estado, Sao Paulo. 314 State of Sao Paulo: Par- que do Estado, Sao Paulo. 196-197 State of Sao Paulo: Par- que do Estado, Sao Paulo. 387 State of Sao Paulo: Par- que do Estado, Sao Paulo. 193 State of Minas Gerais: Brejo de Lapa, Parque Na- cional de Itatiaia. 376 State of Sao Paulo: Par- que do Estado, Sao Paulo. 331 238 Conyza chilensis Spreng. Inulopsis scaposa (DC.) O. Hoffm. Inulopsis scaposa (DC.) Hoffm. Podocoma hirsuta Baker Solidago chilensis Meyen. INULEAE Pluchea suaveolens (Vell.) O. Ktze. Pterocaulon alopecuroides (Lam.) DC. HELIANTHEAE Acanthospermum australe (L.) O. Ktze. Ambrosia polystachyus DC. Bidens segetum Mart. ex Colla Calea multipega H. Krasch Calea pinnatifida Banks Calea pinnatifida Banks Clibadium armani (Balbis) Sch. Bip. Eclipta alba (L.) Hassk. Rhodora 36 10 10 11 19 19 24 11 [Vol. 70 State of São Paulo: Par- que do Estado, São Paulo. 188 State of São Paulo: 8-10 km ESE of Campos do Jordão. 312 State of São Paulo: 5 km SE of Sao José dos Cam- pos. 317 State of São Paulo; Par- que do Estado, São Paulo. 230 State of São Paulo: Par- que do Estado, São Paulo. 290 State of São Paulo: Par- que do Estado, São Paulo. 194 State of São Paulo: Par- que do Estado, São Paulo. 254 State of Sao Paulo: Par- que do Estado, Sao Paulo. 201 State of Sao Paulo: 5-8 km E of Amparo. 256 State of Sao Paulo: 8 km S of Campos do Jordao. 390 State of São Paulo: 15 km W of Itirapina. 302 State of São Paulo: Par- que do Estado, São Paulo. 255 State of Sao Paulo: 5-8 km E of Amparo. 263 State of Sào Paulo: Par- que do Estado, Sào Paulo. 265 State of Sao Paulo: Par- que do Estado, Sào Paulo. 210 1968] Chromosome Numbers — Coleman 239 Eclipta alba (L.) Hassk. Jaegeria hirta Less. Jaegeria hirta Less. Parthenium hysterophorous L. Verbesina glabrata H. & A. Verbesina glabrata H. & A. Verbesina glabrata H. & A. Wedelia pilosa Baker Wedelia subvelutina DC. SENECIONEAE Emilea coccinea (Sims.) Sweet. Erechtites hieracifolium Rafin. Senecio adamantinus Bong. Senecio brasiliensis Less. Senecio erisithalifolius Sch. Bip. MUTISIEAE Chaptalia integerrima (Vell.) Burk, 27 17 17 17 10 20 20 20 24 State of Sao Paulo: Cara- guatatuba State Park. 207 State of Sao Paulo: Par- que do Estado, Sao Paulo. 203 State of Sao Paulo: 8 km S of Campos do Jordao. 391 State of Sao Paulo: Par- que do Estado, Sào Paulo. 268 State of Sao Paulo: 13 km SW of Ubatuba. 298 State of Sao Paulo: Par- que do Estado, Sao Paulo. 297 State of Rio de Janeiro: Reboucas, Parque Nacional de Itatiaia. 382 State of Sao Paulo: Par- que do Estado, Sao Paulo. 396 State of Sao Paulo: c. 9 km SW Sao José dos Cam- pos. 211 State of São Paulo: 10 km NE of Pirassununga. 251 State of Sao Paulo: Par- que do Estado, Sao Paulo. 5271 200 State of São Paulo: 8-10 km ESE of Campos do Jordão. 179 State of São Paulo: 8 km S of Campos do Jordão. 179 State of São Paulo: 8-10 km ESE of Campos do Jordão. 305 State of São Paulo: Par- que do Estado, Sao Pauis. 309 240 Chaptalia nutans (L.) Polak. Chaptalia mandoni (Sch. Bip.) Burk. Perezia cubataensis Baker Trixis divaricata (H. B. K.) Spreng. Trixis pinnatifida Less. Hieracium commersonii Monnier Hypochaeris brasiliensis Griseb. Hypochaeris gardneri Baker Hypochaeris radicata L. Rhodora 24 24 27 11 [Vol. 70 State of Sao Paulo: Par- que do Estado, Sao Paulo. 172 State of Sao Paulo: Par- que do Estado, Sao Paulo. 285 State of Sao Paulo: Par- que do Fstado, Sao Pauls. 212 State of Sao Paulo: Par- que do Estado, Sao Paulo. 170 State of Sao Paulo: 5 km E of Paranapiacaba. 389, 397 State of Sao Paulo: 8-10 km ESE of Campos do Jordào. 388 State of Sao Paulo: Par- que do Estado, Sao Paulo. 189 State of São Paulo: 8-10 km ESE of Campos do Jordão. 299, 332 State of São Paulo: Par- que do Estado, São Paulo. 200 STATUS OF THE GENUS CYMOPHORA (COMPOSITAE)! C. E. ANDERSON AND J. H. BEAMAN The genus Cymophora was founded by Robinson (1907) as a monotypic Mexican member of the tribe Heliantheae. He suggested that it shared many characters with, and ap- peared to stand near Eleutheranthera. The latter was placed in the subtribe Verbesininae by Hoffmann, (1894) and considered distinct from Aspilia only in that the asex- ual marginal florets are lacking or aborted-ligulate. While reviewing various Mexican genera of Compositae, we noted a considerable resemblance between Cymophora pringlei and members of the subtribe Galinsoginae, especially cer- tain species of Tridax. A detailed comparison (Table 1) of the species with Eleutheranthera indicates that these taxa differ strikingly in almost all structures utilized for generic classification in the Compositae, such as the in- florescence, receptacle, phyllaries, pales, anthers, style branches, and achenes. On the other hand, it is so similar to Tridax (particularly to T. accedens) that there is no basis for retaining it as a separate genus. The following new name is thus required. Tridax oligantha Anderson & Beaman, nom. nov. Cymophora pringlei Robins, Proc. Amer. Acad. 43: 39. 1907, non Tridax pringlei Robins. & Greenm. Proc. Amer. Acad. 32: 4. 1896. Type: MEXICO. GUERRERO: Iguala Canyon, 2500 ft, 22 Sept. 1905, C. G. Pringle 10068 (GH, holotype, MSC photo 6732!; MICH, MSC, US, isotypes). Erect or suberect annual 1.2-4 dm high; stems rather densely pilose and glandular, branched at each node, the principal internodes 6-9 cm long, branches arcuate or flex- uous; leaves opposite, the lower on petioles to 1.5 cm long, the upper subsessile, blades broadly lanceolate to ovate, slightly acuminate, green above, paler below, hirtellous ‘Supported by NSF grant GB-4592. We appreciate the loan of specimens from the U. S. National Herbarium and the Herbarium of the University of Iowa. 241 [Vol. 70 Rhodora 242 Déeg dei al A[ao1v0s pur popuno. sodepuoedde 'ojv.redos ‘Suo, WU 6' SIleuquy ie[nqni *QI-8 SPOLA ysry ww g-c 'ayenurdureo 94on[OAUIT 'snouredourou SpeoH quaIsIs.tod ‘paAteu-g ‘xaAuod ‘eao 'soorde e3e?e[norde snoaveqaəy QjL4 ‘moq SnoasoevuRiquieul SALA le[npuels-afissas pue aso[rd-Zuo[| A[aye1apout ‘paAtou-g ‘snosvoRq..ey Surdieur “yey ‘ayeao A[Moareu ‘9 SILYA u300uis A[Aey 'snoiqe[d 'snoooeo[ed Apueueuried ‘papuno. Anuzi[s e[oe3do»oqq S[IX& Jeol ou) UI ÁA[durs ou.oq spray 93€A0-pro3[op sasepuadde 'ojeuuoo ‘SUO, UU g Siouqu1uy 07-06 $3940g XSIp ‘G s391og Avy ysiy Wu €-p *o3e[nue dureo ÁA[peoiq exon[oAur 'fsnouredoiojoQ speaH serie[[Aud əy} 193]? snonploep 'po9A.9u-J, ‘xaauoo 'xode oje[notde [eus e UM ojv2unj -qns 'oj€A0qo-Suo[qO *SNOMRUBIQqUIOUL So[eq Jepnpue[s pue eso[rd ÁAjasuap 01 Á[sq&1opour “postou -QI Bo “snolivos A[MOLIeU SULSIRUL X9AUOO “1ejnoiqioqns 0j 93840 A]peoiq ‘G 8d salaelfAyg peqqouy ‘snoaqejs '[eotuoo apoeydavey @SOULAD BDUVOSa.OYUT (eB[NSUBIIY səgepuədde “azeuuoo ‘Suo, ut g SIGUJUY 9jeIqe[IQ. Ap[eoa 07-06 SITU US unu c-p vo *eje[nuedureo 2aon[oAur ‘snoulesouloy speoH sor1e[[4ud au) 9e10goq snonproop “poaltou-G 'xoAuoo 'osnjqo *oje[|osou?[qO A[peO.q *snooo€utgaquiour Sed Je[npue[d pue oso[rd Ajasaeds 07 snoidqe[3 ‘paAlau-g-f 'snorieos suldieut 'XoAuO2 '9jeAOqO ‘OL Bo SƏMLA peqqouy 'snoadqe[3 “[eotuoo o[oe3doooq] 9SOULÃO adUadSa1OYUT le[n suey sosepuodde 'oj1euuoo ‘Suo, WU g'g S1oujuv oy erqeliq ALam “OT $s32X0[d usn unu p eo 'eje[nuedureo A[Moareu AI9A O.I0n[OAUI *snouredouroqg spray sorie][Aud ayy 910jgoq Snonplvep “postou -G-6 fpepeex Aysus oi X9AUOD ‘SUO[GO-9}B[090UB] ‘snosoevuBiquieul SO9[eq ce[npueysd pue eso[td 'peAideu-g-p “snolIvIS Surdieur 'XoAuoo *oj3eAoqo ‘9 SƏLA peqqouy ‘asojid Ayasaeds '[eo1uoo ajorydadoy aSOULAD 92Uu92S9.I0pUT SyDLapnet DLIYJUDLIY INA] T »iqnp sopi, SUIPIIIÐ LOPU I, DYJUDBYO LOPLI, ‘SUDLIPNL DLIYJUDLIYINIQ A pue *miqnp "LL 'suopooon uL “Dyjunbmo rop Jo sainjeoy 3ug31iodui[ “T aATaV 245 Cymophora — Anderson and Beaman 1968] juesqe snddeq SUO WU j'( BO sno[[eo ie[nqn3 'eso[rd-3.10u$ “ULIOJIUUICUL 9 YIM xode 'surdieur pagdura A[MOIIBU əy} u99439q pur uo Suorjoefo.id. Áurea UL ‘asojid AÁp[os.teds ‘paqqia -$-g 'ejeurq.nj ‘Buoj unu g-c ‘UMOIG souoqoy 9je[ooou?[ Á[p9o.q ‘SUO, WU GT Bd Sje1op ysip ay} ur ‘suo, Wut T vo Sjoloy Ári ou) ur “sopros ap QT Jo snddeg pe[sue-g “agriqeis souoyor ysıp *'eso[rd Aja] e1apoul sousyor ABA “eotuoogo A[mo.reu ‘Buol unu g Bd f*xove[q souoyoy 919 [090UB [-.1£ 9 UT] ‘SUO WU [ Bd .iouut ay} ur ‘Suo, ww J/0-c'( $j91op 19jno SU) ur “safras 9jUI[I2 OZ vo Jo snddeg pe[sue-c ‘aseq ay} Ye Asou qusosaqnd 1əuur ayy *sno[[I^ A[ssuep togno ayy '[eoruooqo Ajmorreu “Suo] WU g Bo ‘yaRIq seuoqQoy quasqe snddeg Daun: *ojeiqe[d 07 3uoeoseqnd Ájosuop SSe[ Ieuut au) fsno[[I4 A[esuop Jogno au “[eotuoogo A[MOLIeU ‘Buoj WU CU "ap souoyoy 244 Rhodora [Vol. 70 above and below, 3-nerved from 1-2 mm above the mostly broadly rounded, short-acuminate base, 2-5.5 cm long, 1-3.8 em wide, margins subentire to serrulate, ciliate; inflores- cence a much-branched, many-headed cyme, pedicels 1-2 cm long, filiform, pilose and glandular, the glands reddish on long stalks; heads very narrowly campanulate, homog- amous, 10-flowered, 6 mm high, 3-3.5 mm wide; receptacle conical, pilose, knobby, 1 mm wide, ca 0.8 mm high; in- volucre 2.5-3 mm wide, 4 mm high, phyllaries 6, pilose and glandular, green below, purplish above and near the scari- ous margins, convex, obovate, 4-8-nerved, striate, 2.5- (outer 1 or 2) 4 mm long, 1-2 mm wide, apex obtuse and somewhat lacerate-apiculate, tardily deciduous; pales ca 7, membranaceous, lanceolate-oblong, 3.8-4 mm long, 0.7-1.2 mm wide, rather obtuse to acuminate-apiculate, subentire, glabrous, convex to slightly keeled, 3-5-nerved, deciduous ; ligulate florets absent; outer disk florets 5, corollas min- utely glandular and pilose on the purplish tube, minutely pilose near the tips of the white lobes, papillose on the mar- gins, 4 mm long, tube 0.5 mm long, throat 1.8 mm long, the three larger, outer lobes 1.5 mm long, 0.7 mm wide, broadly lanceolate, the 2 smaller, inner lobes 1 mm long, 0.5 mm wide, triangular; inner disk florets 5, similar to the outer but the corollas 3-3.5 mm long with nearly equal lobes; anthers connate, 1.8 mm long, purplish, appendages triangular, bases sagittate; style branches 0.8-1 mm long, elongate, recurved, acuminate-appendiculate; achenes black, narrowly obconical, 2.1-2.3 mm long, 0.6-0.9 mm wide, apex with a low callous ring 0.05 mm high, the outer achenes slightly laterally compressed, densely villous with whitish trichomes (when dry appearing wing-like on the margins) which extend pappus-like from the outside of the apical callous ring, the inner achenes 4-5-angled with the trichomes concentrated on the margins of the inner sur- face and at the apex, less densely pubescent than the outer. Additional specimens examined. MEXICO, GUERRERO: 10 mi by gravel road from Tixtla to Chilapa in the arid mts E of Chilpancingo, Melchert & Sorensen 6145 (IA); 1968] Cymophora — Anderson and Beaman 245 Iguala Canyon, 3000 ft, 2 Oct. 1906, Pringle 10068 1/2 (MICH, US). The characters by which T. oligantha differs from the closely related T. accedens are enumerated in Table 1. Most of these are minor, but it may be noted that the very nar- row heads of T. oligantha, with 10 florets, in contrast to the more broadly campanulate heads of T. accedens, with 30-40 florets, provide a ready means for distinguishing the two species. Powell (1965) has suggested that T. accedens might pos- sibly link Tridax with Galinsoga. Tridax oligantha might thus appear to provide an even closer approach to that genus. But neither species actually bridges the gap between Tridax and Galinsoga. Among the characters which ally T. oligantha to Tridax rather than to Galinsoga are the corollas longer than 3 mm, anthers 1.8 mm long, the elon- gate, appendiculate style branches, and the narrowly ob- conical, densely pubescent achenes (see Powell, 1965, p. 47 for a tabulation of differences between Tridax and Galin- soga). The narrow heads might seem to relate it to Galin- soga, but the heads are narrow as a result of the few florets. Other species of both Tridax and Galinsoga have more nu- merous florets per head. In all aspects except width, the heads of T. oligantha are much more similar to T. accedens than to any species of Galinsoga. Generic status for T. oligantha might be argued on the basis of its epappose and eligulate features, but there are other epappose forms in both Tridax and Galinsoga, and several eligulate species, including T. accedens, are known in Tridax. Tridax dubia, T. accedens, and T. oligantha, respectively, appear to form an evolutionary reduction series. The largest plants, perennial in duration, occur in T. dubia, while T. accedens and T. oligantha are smaller annuals. The heads of T. dubia and T. accedens have 30-40 florets while those of T. oligantha are reduced to 10. Ray florets are present in T. dubia and absent in T. accedens and T. oligantha. The corollas of the disk florets are longer in T. dubia than in the other two species. The pappus is longest 246 Rhodora [Vol. 70 in T. dubia, shorter in T. accedens, and absent in T. oli- gantha. Although T. dubia is the basic species in this evolu- tionary line, it is probably not a very primitive member of the genus. The morphological similarities of these three species are paralleled by their geographic relationships. Each has a localized distribution on the Pacific slope of southwestern Mexico. Tridax dubia occurs near sea level in Colima, Jalisco, and Nayarit at the northwest end of an axis that extends southeast and east through Michoacán to northern Guerrero. Tridax accedens occupies an intermediate geo- graphic position, being known only from Coalcoman, Mich- oacán at an altitude of ca 1000 m. The easternmost member of the three, T. oligantha, occurs at about the same altitude as T. accedens, and is localized in the Iguala-Chilpancingo region of Guerrero. McVaugh and Rzedowski (1965) have noted that endemism is frequent in southwestern Mexico, and a number of the species of Bursera which they studied have distributions corresponding to these species of Tridax. In contrast to the evident phytogeographie relationship of T. oligantha when associated with T. accedens and T. dubia, it has no geographic connection with Eleutheranthera. The latter is widely distributed in tropical America but has not been reported from Mexico. DEPARTMENT OF BOTANY, UNIVERSITY OF MICHIGAN, ANN ARBOR 48104 DEPARTMENT OF BOTANY AND PLANT PATHOLOGY, MICHIGAN STATE UNIVERSITY, EAST LANSING 48823 LITERATURE CITED HOFFMANN, O. 1894. Compositae in Engler & Prantl, Natürl. Pflan- zenfam, Ai: 87-391. McVAuGH, R. & J. RZEDOWSKI. 1965. Synopsis of the genus Bursera L. in western Mexico, with notes on material of Bursera collected by Sessé & Mocino. Kew Bull. 18: 317-382, 6 pl. PowELL, A. M. 1965. Taxonomy of Tridax (Compositae). Brittonia 17: 47-96. RoBINSON, B. L. 1907. New or otherwise noteworthy spermatophytes, chiefly from Mexico. Proc. Amer. Acad. 43: 21-48. A CYTOTAXONOMIC STUDY OF THE HERBACEOUS SPECIES OF SMILAX: SECTION COPROSMANTHUS* JOSE K. MANGALY (Continued from page 82) TAXONOMY Smilax L. Plants dioecious, with annual or perennial shoots from perennial rhizomes, either smooth or armed and terete or angular, Leaves alternate, palmate and reticulate veined, long or short petiolate, petioles often sheathing at base and bearing a pair of tendrils, Flow- ers in axillary umbels on long or short peduncles, typically trimerous and hypogynous. Tepals 6, usually greenish, with those of staminate flowers larger than those of pistillate flowers. Stamens 6, fixed at the base of the tepals. Anthers oblong, 2-celled, introrse and basi- fixed. Ovary tricarpellary, syncarpous, superior; stigma solitary or 3, sessile; ovules 1-2 in each carpel, orthotropous. Fruit a berry. Seeds 1-6, with horny endosperm. Stem annual, herbaceous and smooth, ovules two in each carpel ............... section Coprosmanthus Stem perennial, woody and spinous, ovules one in each carpel ........................ sections Pleiosmilaz, Smilax, Macranthae, Coilanthus, and China (woody Smilax) Section Coprosmanthus Smilax section Coprosmanthus Torrey, Fl. New York. 2: 303. [1843] 1847. Nemexia Rafinesque, Neogeneton 3. 1825. Coprosmanthus Kunth, Enum. Pl. 263. 1850. Smilax sect. Nemexia (Raf.) A. DC. Monogr. Phaner. 1: 46. 1878. Smilax Subgen. Nemexia (Raf.) Pennell in Bull. Torrey Bot. Club. 43: 409. 1916. Rhizome tuberous and perennial, with long thick roots. Stem an- nual, 1-to many per season, 3 dm to 3 m high, erect or climbing, terete or slightly angular. Leaves few and approximate at top to many and uniformly distributed, alternate; petioles (with sheathing base in some) with a pair of tendrils at the base (absent or with few rudimentary ones in S. ecirrhata complex); blades membraneous, ovate to cordate; tip rounded and cuspidate or acute to acuminate. Inflorescences are umbels, 1-to many, on long and slender peduncles, either in leaf axils or in bract axils below foliar leaves; generally one from each node, occasionally three from the same node in one species. Flowers pedicellate, greenish, malodorous in some, trimerous 247 248 Rhodora [Vol. 70 and hypogynous. Pedicels inserted in small pits on a conical or globose receptacle and slender. Tepals 6, greenish, outer ones gen- erally broader than inner ones, those on staminate flowers larger than those on pistillate flowers, slightly ovate or elliptic to obovate. Stamens 6, reduced and nonfunctional in pistillate flowers, filaments slightly flattened at base, attached at the base of the tepals and free. Anthers oblong, 2-celled, introrse and basifixed. Ovary 3-celled, rudi- mentary or absent in staminate flowers. Style very short or none; stigmas three, ligulate and recurved. Ovules 2 in each carpel, or- thotropous. Fruit a berry, blue to black, with three branches of strengthening tissues running through the pulp connecting the base and apex. Plants flowering generally in early Spring. KEY TO SPECIES 1. Plants tall, 1 to 3 meters high, climbing; tendrils many, long and functional; peduncles usually many, and from leaf axil. 2. Leaves pubescent beneath. 3. Blades dark green, shiny beneath, petiole long; fruits black .... BEE 3. S. pulverulenta 3. Blades light green, not shiny beneath, petioles generally short; berries blue ............... eee 2. S. lasioneuron 2. Leaves glabrous and glaucous beneath. 4. Pedicels nearly 2X or more as long as the fruit; larger leaves on the main stem broad ovate, not hastate, with 7-9 main VEINS BERE 1. S. herbacea 4. Pedicels more or less equal to diameter of the fruit, larger leaves distinctly hastate, with concave margins, and distinctly cordate base, 3-5 main veins reaching the tip, tip rounded and cuspidate ............ sese 4. S. pseudo-china 1. Plants short, up to 1 meter or less in height, erect, with a few short tendrils or none; peduncles few, the lowest one usually from the bract axil. 5. Leaf blades glabrous and glaucous beneath; leaves few, usually 5-7, approximate at top, cordate at base, tip short acuminate or ACULO ARR 8. S. biltmoreana 5. Leaf blades pubescent beneath. 6. Basal leaves broadly ovate, base cordate, tip acuminate, petioles generally equal to or shorter than blade ..................- :——-———————— "— —— 6. S. ecirrhata 6. Basal leaves narrowly ovate or elliptical, base mostly ovate. 7. Leaves approximate at top, tendrils mostly absent, petioles shorter than lamina ..............:4.. ccce 7. S. hugeri 7. Leaves subverticillate, or equally distributed, progressively smaller and narrowly ovate, tendrils usually many, petioles usually longer than or equal to lamina ........ 5. S. illinoensis 1968] Smilax — Mangaly 249 1. Smilax herbacea L. Smilax herbacea L. Sp. Pl. 2. 1030. 1753. Type: in Linnaean Herbarium, London, marked by Linnaeus. This specimen was collected by Clayton in Virginia. Linnaeus obtained this specimen evidently from Gronovius and labelled it “Smilax annua inermis caule fusco purpureo claviculis plurimis teneris vestito. Exalis foliorum oritur pedunculus teres glaber semipedalis gerens flores in capitulos globose, collectors". "This specimen appears to be a duplicate cf Clayton's No. 541 now in the Gronovian Herbarium in the British Museum. Both specimens appear to be a small branch characteristically bearing slightly elliptieal leaves with pointed tips. These leaves differ from the ovate to slightly cordate leaves expected on the main stem. S. peduncularis Muhl. ex Willd. in Linn. Sp. Pl. ed. 4: 786. 1806. Type: Photograph of the type (Willdenow No. 18404, now in Botanisches Museum, Berlin — Dahlem labelled Smilax “pedun- cularis . . . habitat in america boreali”)! Kunth (Enum. Pl. 264) refers to Herb. 18402, but not seen. Nemexia nigra Raf. Neogeneton 3. 1825, Used as a new name for S. herbacea but, as Pennell points out, it was apparently intended for the black-fruited S. pulverulenta Michx. Nemexia cerulea Raf. 1. c. 3. 1825. Used as a new name for S. pedun- cularis Muhl. Coprosmanthus peduncularis (Muhl.) Kunth, Enum. Pl. 5: 264. 1850. C. herbaceus Kunth, 1. c. 265. 1850. S. herbacea var. Simsii A. DC. Monogr. Phan. I: 51. 1878. Type not seen. Muhl. in h. Willd, fol. 8394. «S. herbacea Sims. Bot. Mag. pl. 1920". This apparently refers to a narrow-leaved S. herbacea. S. herbacea var. peduncularis A. DC. 1. c. 51. 1878. N. herbacea Small, Fl. S.E.U.S. 2800. 1903. S. herbacea subsp. crispifolia Pennell, Proc. Acad. Nat. Sc. Phil. 62: 559. 1910. Type examined. *Herb. Acad. Nat. Sc. Phil. Pennsylvania, Dela- ware Co. serpentine, Mineral, Sept. 6, 1908. Pennell.” This ap- parently represents a form with narrow, rigid leaves growing in dry and sunny places. Annual shoots up to 2.5 meters high, climbing and branching; lower cauline leaves narrow and small grading into larger upper leaves (except in younger plants wherein leaves are larger below). Leaves at tip ovate to round, most- ly cordate at base. Leaves in general smooth and glaucous beneath, short or long acuminate at tip on the stem and cuspidate on branches; peduncles many on both stem and 250 Rhodora [Vol. 70 branches, axillary to leaves; basal peduncles up to 30 cm long, upper ones progressively shorter. Flowers many in a globose head; tepals greenish, 3.5-4.5 mm long, 1.5-2 mm broad; anthers 1-1.5 mm long, shorter than filaments; style broadly ligulate; fruit about 1 em in diameter, sub-globose, with 3-6 brownish seeds. Flowering begins in the fourth week of May in the south and extends to the end of June in New England states, Distribution of S. herbacea is mostly along the Appala- chian Mountains extending from Georgia and Alabama to Maine and north-eastern Canada. Specimens have been col- lected from Connecticut, Delaware, Kentucky, Maine, Mary- land, Massachusetts, New Hampshire, New Jersey, New York, North Carolina, Pennsylvania, Rhode Island, Ten- nessee, Vermont, Virginia and West Virginia. Its closest relative apparently is Smilax lasioneuron , a species recognised by many taxonomists as a variety of S. herbacea. However, S. lasioneuron differs from S. herbacea by the pubescence on the leaves and its ecological adaptation to the plains. The smooth glaucous leaf surface of S. her- bacea is very similar to that of S. pseudo-china, which is comparatively smaller in size. Smilax pulverulenta is an- other close relative of this species, but differs from it by the presence of shiny dark green leaf surfaces, pubescence long and mostly on the veins, comparatively fewer flowers on the peduncle, and black fruits. REPRESENTATIVE SPECIMENS." ALABAMA. AUBURN Co: Earle & Baker, 1898 (F). LEE Co: Earle & Baker, May 15, 1898 (RM). CON- NECTICUT. FAIRFIELD co: Stamford, E. M. Eames, June 10, 1895 (GH). NEW LONDON CO: Groton, Janssen, June 16, 1929 (TEX). DIS- TRICT OF COLUMBIA. River flat, Chain Bridge, Pennell 12072 (PH). GEORGIA. DE KALB CO: Base of Stone Mt., Mill, Boyd and Myres 481 (GH). HART CO: Wooded bluffs on Savannah R, Smith McGee Bridge, Harper 1600 (PH). LUMPKIN co: Blood Mt., Leeds 1560 (PH). KENTUCKY. HARLAN co: Near summit of Big Black Mt., McVaugh 8677, (TEX). MAINE. KENNEBEC CO: Sidney, Fernald & Long 13295 (PH). PENOBSCOT CO: Milford, Fernald & Long 13296 (GH, PH). PISCATAQUIS CO: Piscataquis River Valley, Dover, Fernald, June 6, 1895 (GH). MARYLAND, GARRETT co: Kelso Gap, Backbone *Only a partial list of specimens examined is provided here. 1968] Smilax — Mangaly 251 Mt., Core, June 24, 1931 (wva). MASSACHUSETTS, BARNSTABLE CO: vicinity of Woods Hole, J. M. Fogg, Jr. 1745 (MO). DUKES CO: Gay Head, Pennell 3317 (PH). ESSEX CO: Ipswich, Morong, June 23, 1875 (NY). HAMPDEN CO: Seymour 566 (GH, NY). NORFOLK CO: Milton, Churchill 295 (Mo). PLYMOUTH co: Vail, July 1888 (NY). WORCESTER CO: near Lancaster, W. V. Brown, June 9, 1946 (TEX). NEW HAMPSHIRE. coós co: Shelburne, Deane, June 29, 1916 (GH). GRAFTON co: Lebanon, G. Kennedy, Aug. 16, 1890 (GH). ROCKINGHAM Co: Hedge about rich meadows, Rye, Robinson 705 (GH). NEW JERSEY. cAMDEN co: Merchantville, Meredith, June 2, 1921 (GH). CAPE MAY co: Cold Springs, Pennell 2200 (MO). MONMOUTH CO: along Big Brook, Marlboro, Long 37491 (GH). SOMER- SET Co: Watchung, Moldenke 2879 (NY). SUSSEX Co: Britton, Sept. 28, 1887, (NY). WARREN CO: 2.5 mi. s. of Asbury, Godfrey 62142 (FSU). NEW YORK. BROOME co: Binghamton, Millspaugh, 1884 (F). ESSEX co: shore of Lake Harries, New Comb, ‘Howe 10162 (GH). HAMILTON co: Adirondack Mt. along Hudson River, Lambert 22 (NY). New York City: Edmondson, June 3, 1899 (NY). ONTARIO CO: Hall, 828-1834 (F). ST. LAWRENCE CO: Narros Island, Blake Lake, Fernald, Wiegand & Eames 14219 (GH). TOMPKINS CO: South Hills, Ithaca, Thomas 2018 (GH). ULSTER CO: vicinity of Lake Mennewaska, Sowden, July 3, 1922 (PH). WESTCHESTER CO: Bedford Hills, Pennell 7085 (PH). NORTH CAROLINA. ALLEGHENY CO: 1.7 mi. n. of Amelia, Radford 34093 (NCU). ASHE CO: 0.6 mi. s. of Warrensville, Radford, Aug. 1, 1958 (NCU). BURKE CO: Tim's Creek, east of Drowning Creek Mt., Bell 6518 (NCU). DAVIDSON CO: Ellis Creek, about 2 mi. e. of Newsome, Ahles with Haesloop 57412 (NCU). GUILDFORD CO: Greensboro, Causey (NCU). HAYWOOD co: Piney Mt., Coker, July 31, 1926 (NCU). JACK- SON co: 1 1/4 mi. s.w. of Balsam, Radford, June 26, 1946 (NCU). MADISON CO: 7.4 mi. n. of jet. U. S. 19 and 23 on U. S. 23, n. of Mars Hills, Ahles, with Duke 46324 (NCU). MONTGOMERY CO: 1.8 mi. w. of Eldorado on N. C. 109, Ahles with Haesloop 57346 (NCU). ORANGE CO: Chapel Hill, Radford & Steward 650A, (NCU). POLK CO: Tryon, Peattie 1173 (NCU). ROWAN CO: 3 1/2 mi. n.w. of Cleveland, Ahles 12980 (GH). SWAIN co: Andrews Bald Ridge, Great Smoky Mt. Nat. Park, Sharp & Jennison 585 (TENN). TRANSYLVANIA CO: Glauces- ter, Memminger, Sept. 20-30, 1895 (NCU). YANCY CO: 3 mi. n. of Burnsville, Radford, July 9, 1946 (NCU). WILKES CO: 4 1/2 mi. n.w. of Vashti, Radford 13718 (NcU). PENNSYLVANIA. BEDFORD CO: 11/2 mi. n. ne. of Grand View, Berkheimer, July 8, 1939 (MO). BUCKS CO: Hedgerow, 2 1/2 mi. n.w. of Neshaminy, McDowell 267 (GH). CARBON co: 1 mi. s. se. of Little Gap Village, Wherry, June 13, 1952 (PH). CENTRE co: Bear Meadows, Leeds 1405 (PH). CHESTER CO: Nottingham Barrens, Pennell 1421 (MO). CLARION CO: 1 3/4 mi. s. of Sligo, Wahl 18790 (FSU). DELAWARE Co: Pennell, July 5, 1905 (F). ELKS CO: 3 252 Rhodora [Vol. 70 mi. w. of Ridgeway, Westerfield 2159 (GH). JEFFERSON CO: 1 mi. n.e. of Summerville, Wahl 3593 (GH). MONTGOMERY CO: near Sumney- town, Leeds 1956 (MO). NORTHAMPTON CO: Fallowfield, 1 mi. e. of Lower Saucon, Schaeffer, Jr., 5069 (GH). PERRY CO: above Marysville, Small, June 18, 1888 (F). WARREN Co: Heart's Content, Allegheny National Forest, 20 mi. s. of Warren, Pennell 26613 (PH). WESTMORE- LAND CO: Block, June 1881 (LCU). YORK co: Britton, July 2-6, 1904 (NY). RHODE ISLAND. NEWPORT CO: north of Crescent Beach, Block Island, Fernald, Long & Torrey 9249 (GH, PH). PROVIDENCE CO: Providence, Thurber, June 1844 (GH). SOUTH CAROLINA. DAR- LINGTON CO: at Lauthers Lake, B. E. Smith 9, (NCU). OCONEE CO: House 2119 (us). TENNESSEE. COOKE co: along Highway 75, near Cosby, Raper & Jenison 3383 (TENN). HENDERSON CO: near Club Lake, moist slope, Natchez Trace State Park, Sharp, Fairchild, E. Clebsch & A. Clebsch, July 8, 1948 (TENN), MONROE CO: moist slope along Sycamore River Gorge, Gillespie, Laing & Sharp 18846 (TENN). VERMONT. appisoN co: Middlebury, Brainerd, June 1877 (GH). BENNINGTON CO: Manchester, Day, Aug. 30, 1898 (GH). WINDHAM co: banks of Corn River, Westminster, Robinson 4, (GH). VIRGINIA. APPOMATTOX CO: 8 mi. n.w. of Pamplin City, Kral 11215 (Ncw). BLAND CO: Dismal Creek, near Giles Co. line, Potter, July 12, 1936 (NCU). BOTETOURT CO: along Bearwallow Creek, below Gap, Freer 1602 (GH). DINWIDDIE CO: east of Burgess Station, Fernald & Long 9898 (GH). FAIRFAX CO: Great Falls, Pennell 1670 (NY). GILES CO: west side of Mt. Lake, Kral 9176 (FSU). GRAYSON Co: Troutdale, Gleason, June 21, 1940 (NY). MONTGOMERY CO: 1.5 mi. s. of Ellett, Kral 9176 (FSU). PAGE CO: Stony Man Mt. and vicinity in the Blue Ridge, near Luray, E. S. Steele & Mrs. Steele 176 (GH, NY). PULASKI co: along Peak Creek on Peak Mt., Small, July 16, 1892 (GH, MO). Richmond: University Campus, Angus and Haywood 4471 (GH). SHENANDOAH CO: Valcour, Pennell 12119 (PH). SMITHE CO: northwest slope of Mt. Rogers, Ramseur 4402 (NCU). SOUTHAMPTON CO: along Three Creek, n.w. of Carey Bridge, Fernald & Long 10204 (GH). SUSSEX CO: along Nottoway River, s.e. of Homeville, Fernald & Long 10208 (GH). WEST VIRGINIA. BARBOUR CO: 1 mi. n. of Kasson, Rt. 92, Haller, June 5, 1951 (WVA). BRAXTON CO: Sugar Creek, s. of Grassaway, Boggs, June 29, 1953 (WVA). CALHOUN CO: Grantsville, Harris, July 1, 1933 (wvA). MERCER CO: Oakvall, Mtn., McNeill, May 30, 1931 (WVA). MONONGALIA CO: Millspaugh 63 (WVA). OHIO co: Wheeling Girls Scout Camp, woody hillside Middle Wheeling Creek, Bartholomew, July 22, 1934 (WVA). PENDLETON CO: Spruce Knob, Rydberg 9175, (NY). POCAHONTAS CO: along U. S. 250, 2 mi. n. of Durbin, Clarkson 3956 (WVA). PRESTON CO: Cheat River near Kingwood, Core, Aug. 1, 1931 (WVA). TUCKER co: Sharp, July 18, 1940 (wvA). TYLER CO: along Tygarts River, 2 mi. n. of Crafton, Core 1968] Smilax — Mangaly 253 5496 (WVA). WEBSTER CO: Hinkle, Aug. 1957 (WVA). WETZEL CO: thickets at north of Sugar Run, 1 1/2 mi. w. of Littleton, Haught 407, (WVA). WRIT CO: on shady bank of Little Kanawha River, about 3/4 mi. above mouth of Reedy Creek, Bartholomew W-1312 (WVA). 2. Smilax lasioneuron Hook. Smilax lasioneuron Hooker, Fl. Bor.-Am. 2: 173. Pl. 187A. 1840. Type: Royal Bot. Garden, Kew, *Hab. Carlton House, Fort on the Saskatchewan, Dr. Richardson"! Coprosmanthus lasioneuron. (Hook) Kunth, Enum. Pl. 5: 265. 1850. S. herbacea var. lasioneuron A. DC., Monogr. Phan. 1: 52. 1878, synonym. S. herbacea var. inodora M. E. Jones, Bot. Gaz. 4: 248. 1879. Type not seen. S. diversifolia Small, Bull. Torrey Bot. Club. 25: 607. 1898. Type: Herb. Columbia University at NY, collected along the Flint River near Albany, Ga., J. K. Small. Apparently same as S. lasioneuron Hook! S. tenuis Small, Bull. Torr. Bot. Club. 25: 607. 1898. Type seen. Herb. of Columbia University at the New York Bot. Garden. “The original specimens were collected by Dr. Hale, many years ago in Louisiana”. This is taken to be a depauperate young specimen of S. lasioneuron Hook. Nemezia tenuis Small, Fl. S.E. U.S. 281. 1903. N. herbacea var. melica. A. Nelson, Proc. Biol. Soc. Wash. 17: 175. 1904. Type not seen. N. lasioneuron (Hook.) Rydb. Bull. Torr. Bot. Club. 32: 610. 1905. Shoots 2-2.5 meters high and branching, Basal cauline leaves narrow and small; larger leaves above, short petioles, ovate to round, mostly cordate at base, pubescent below and margin convex, tips acuminate to round and cuspidate: younger plants with larger leaves below : the transition be- tween leaves and bracts is abrupt. Peduncles many on stems with a few occasionally on branches, axillary to leaves, generally short, with many flowers in a globose head at tip. Tepals greenish, 3.5-4.5 mm long, 1.5-1.75 mm broad. An- thers 1.5-2 mm long, generally equal to or shorter than the filaments. Fruit a subglobose berry, dark blue, glaucous, with 3-5 seeds. Flowering begins in the second week of May and reaches its peak in the fourth week in the south ; about 2 weeks later in the north. 254 Rhodora [Vol. 70 Distribution is mostly in the central plains and lower ele- vations of the Appalachian and Rocky Mountains. In the northern and western drier areas, plants are shorter than those in the southern humid areas. Plants have been col- lected from Alabama, Arkansas, Colorado, Florida, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Michigan, Min- nesota, Missouri, Montana, Nebraska, New York, North Dakota, Ohio, Oklahoma, Pennsylvania, South Dakota, Ten- nessee, West Virginia, Wisconsin and Wyoming. Its closest relative appears to be Smilax herbacea. Apart from its ecological adaptation to lower elevations and drier conditions, the only easily recognisable morphological dif- ference is the presence of pubescence on the leaves in S. lasioneuron, The much-emphasised shorter peduncle in this Species is a highly fluctuating character. Specimens from central Canada (the type specimen was collected from this region) show greater uniformity in this character, but in the humid southern regions of central United States, the peduncles are generally long. There is considerable morphological variation in Smilax lasioneuron populations in western Florida, especially along Chippola and Appalachicola Rivers. Two other species, S. hugeri and S. biltmoreana, are also found in the same gen- eral area, Intermediate forms of these species are found in herbaria. I feel, therefore, that it is safe to assume that there is some amount of hybridization between all combina- tions of these species in this general area. REPRESENTATIVE SPECIMENS: ARKANSAS, BENTON co: Sulphur Spring along Butler Creek, Pennell 5366 (PH). FULTON Co: Mammoth Springs, Pennell 11563 (PH). LOGAN Co: top of Magazine Mt., Pyle 555 (TEX). COLORADO. BOULDER CO: near Boulder, Ramaley 2612 (RM). LARIMER CO: along Buckhorn Creek, Osterhout, June 24, 1897 (RM). Horsetooth Gulch, Crandall 2438 (RM). FLORIDA, GADSDEN CO: Chattahoochee, Godfrey 54563 (GH). GEORGIA. BURKE CO: woods facing the Shell Bluff on Savannah River, N. Leeds, Harper 2749 (PH). DE KALB CO: Eggert, May 23, 1897 (MO). ILLINOIS. CHAM- PAIGN CO: Gates 2020 (MICH). COOK CO: west shore of Lake Michigan, Glencoe, Churchill, July 26, 1926 (MO). DU PAGE CO: Glen Elbyn, Shaughnessy 109, (WIS). JACKSON CO: Big Muddy and Crab Orchard Creek, McGee Jr., 778 (MO). JO DAVIESS CO: east of Apple River 1968] Smilax — Mangaly 255 Canyon State Park, Koelling 423 (TENN). JOHNSON CO: Reevesville, Palmer 15354 (MO). KANKAKEE CO: near the mouth of Rock Creek, Jones 15888 (MO). LAKE CO: Gates 3115 (MICH). LA SALLE CO: Starved Rock, Greenman, Jansing Jr., & Dixon 118 (GH). MC DANOUGH CO: Kauffman 204, (MICH). MENARD CO: Athens, Hall, 1861 (GH). PEORIA co: Stewart, July 2, 1869 (F). STARK co: along Mud River, Chase 1608 (MICH). ST. CLAIR CO: Eggert, May 13, 1896 (MO). UNION CO: Wolf Lake, Fuller 595 (wis). INDIANA. ALLEN CO: west end of Preglacial Lake Maumee, Friesner 20388 (GH). CLINTON CO: M. Ek, July 4, 1942 (MO). HARRISON CO: along Mosquito Creek, about 3 mi. n. of New Boston, about 4 mi. e. of Laconia, Deam 63685 (FLAS). MONTGOMERY CO: Ward, Aug. 26, 1944 (FLAS). POSEY CO: in a low wood about 2.5 mi. s.w. of Hovey Lake, Deam, 24325 (A). SULLIVAN co: woods near Paxton, McCoy 3791 (FSU). WAYNE CO: Earlham, Pen- nell 9819 (PH). WHITE co: small island in Freeman Lake, Ward, Aug. 22, 1945 (FLAS). IOWA. BOONE CO: Cass Twp. 5 mi. s.w. of Luther, along Des Moines River, Hyden 8508, (NY, MO). DECATUR CO: Anderson, June 25, 1904 (RM). DICKINSON CO: West Okohaji Lake, Shimek, June 12, 1916 (WIS). EMMET CO: Esterville Twp., Hyden 3031 (GH). FAYETTE CO: Fink, June 1894 (GH). JOHNSON CO: quy Fitzpatrick & M. F. L. Fitzpatrick (F). PLATO ALTO CO: eastern shore of Lost Is. Lake in Lost Is. Park, Hyden 8510 (TEX). KANSAS. ATCHINSON CO: 3 mi. n.w. of Atchinson, Sr. Jeanette, July 18, '96 (LCU). DONIPHAN CO: 1 mi. s.e. of Iowa Point, Horr 4477 (GH). LINN co: Bush 15502 (Mo). RILEY co: Norton 515 (GH, MO, NY, RM, US). WILSON CO: near Neodasha, Palmer 21385 (A). WYANDOTTE CO: Rose- dale, Mackenzie, July 12, 1892 (NY). KENTUCKY. NELSON CO: Samuels, Sr. Agnes 1522 (LCU). MICHIGAN. BERRIEN CO: New Buf- falo, Lansing 3290 (F, GH, WIS). MONROE CO: 2 mi. s.e. of Milan, Robertson 384 (MICH). ST. CLAIR CO: near Port Huron, Dodge, June 18, 1893 (MICH). ST. JOSEPH'S CO: Mendon, Umback 7642 (MICH). WASHTENAW co: n. cf Ann Arbor, Ehlers 3792 (MICH). WAYNE CO: Farwell 9080 (MicH). MINNESOTA. HENNIPIN CO: Spring Grove, Rosendahl, June 6, 1902 (GH). OLMSTEAD CO: 2 mi. s. of Morion, Butters, Moore & Schultz 7562 (LCU). ST. LOUIS CO: 1 mi. w. of Cot- ton, Lakela 6907 (WIS). MISSOURI. ADAIR CO: 6 mi. n. of Novinger, Palmer & Steyermark 4114 (MO). CASS CO: Broadhead, June 7, 1864 (MO). CLARK CO: near Dumas, Palmer 21864 (A). CRAWFORD CO: 4 mi. s. of Bourbon, Bauer 225 (F). DALLAS CO: 2.5 mi. n.w. of Junas, Steyermark 24205 (MO). FRANKLIN CO: Pammel, June 25, 1898 (M0). GREEN CO: along U. S. 66, e. of Springfield, Henderson, May 10, 1959 (FSU). HOLT CO: above Missouri River, near Forest City, Palmer 25394 (A). HOWELL CO: 8 mi, w. of Burnham, Steyermark 28540 (MO). JACKSON CO: Independence, Bush 719 (MO). JASPER CO: near Forest Hill, Palmer 18371 (GH). JEFFERSON CO: Eggert, July 15, 1892 (MO). 256 Rhodora [Vol. 70 Ditmar, Kellog 1660 (MO). LAFAYETTE CO: 1 mi. n. of Chapel Hill, Steyermark 24699 (MO). MADISON CO: Steyermark 71206 (F). MARION CO: near Taylor, Palmer & Steyermark 40650 (MO). MERCER CO: 3 mi. s. of Linesville, Palmer & Steyermark 41302 (MO). NEWTON CO: George Washington Carver National Monument near Diamond, Palmer 62804 (wis). OZARK CO: Steyermark 22800 (MO). PIKE CO: Davis 4172 (MO). POLK co: near Gydon Springs, Palmer 29022 (a). sT. LOUIS Co: Allenton, Letterman, June 7, 1898 (MO, NY). TEXAS CO: 3 mi. w. of Houston, Steyermark 15537 (MO). WEBSTER CO: 3 mi. w. of North View, Palmer 60209 (F). MONTANA. ROOSEVELT CO: Calais, Blankinship, July 14, 1900 (MO). NEBRASKA. CHERRY CO: Tolstead 105 (GH). LANCASTER Co: Lincoln, Hedgcock, May 22, 1900 (TEx). NANCE CO: 4 mi. n. of Fullerton, Osborn 1200F (MO). SHERIDAN CO: 13 mi. n. of Haysprings, Nixon 81 (RM). THOMAS CO: Emerson Lake, Kiener 20061 (F). NEW YORK. OSWEGO CO: near Cleveland, House 15937 (GH). NORTH DAKOTA. BOTTINEAU co: Turtle Mt., Lunell, July 23, 1902; Sept. 9, 1902 (RM). cass co: Banks of Red River, Fargo, A. Nelson & A. Nelson 1018 (RM, MO). DUNN Co: Matt Olson's ranch, Killdear, Fancy 235 (TEX). MCHENRY CO: Leeds, Lunell, Sept. 10, 1908 (NY). NELSON CO: Stump Lake, Tufte 240 (RM). OHIO. BUTLER CO: near Oxford, Wetmeyer & Waters 162 (MICH). HAMILTON CO: near Cincinnati, Lloyd, May 23, 1888 (MO). LUCAS Co: Reservation, Sylvania, Wilson 1482 (WIS). PORTAGE CO: Webb 1542 (GH). SCIOTO CO: Shawnee State Forest, Camp Gordon, Demaree 10682 (GH, Mo). OKLAHOMA. Kay co: White 1898 (NY). SOUTH DAKOTA. FALL RIVER CO: Visher 2687 (RM). HARDING CO: 102 (RM). LAWRENCE CO: Piedmont, Palmer 31230 (GH). MELLETTE CO: Mellette, Brenckle 41-16 (MO). PENNINGTON CO: Spring Creek, Over 16208 (RM). SHANNON CO: along wooded River Creek, Over 16210 (PH). SPINK CO: near James River e. of Mellette, Brenckle 41-16 (GH). WASHABAUGH CO: Visher 2280 (ny). TENNESSEE. HARDEMAN CO: slopes near Chickasaw State Park, Adams 12684 (TENN). WEST VIRGINIA. HAMPSHIRE Co: Hanging Rock, Frye 563 (wvA). WIS- CONSIN. aDAMS CO: 2 1/2 mi. n.e. of Jhondshy's, Brown, July 15, 1948 (WIS). DROWN co: Treble, Schuette, May 19, 1884 (GH). BUF- FALO CO: Eagle Bluff, Fountain City, Fasset & Wilson 5099 (MICH, WIS). DANE CO: near Monona Village, Kuhlman 41 (wis). DUNN Co: Oakwoods, s. of Elk Mount, Meyer 128 (wis). IOWA co: sandstone cliff, Kieckhefer, Lindfors & Thompson, June 8, 1958 (WIS). JACKSON co: 2 mi. w. of Hatfield, Grether 6726 (WIS). LAFAYETTE CO: 2.5 mi. w. of Gratiot, Hagens, Melville & Shaughnessy, Sept. 28, 1956 (wis). MARQUETTE CO: Iltis, Trenk & Noamesi 6354 (WIS). OZAUKEE CO: banks of Milwaukee River, about 3 mi. s. of Fredonia, Zedler, June 23, 1963 (WIS). RICHLAND CO: near Sextonville, Fosberg 300 & 380 (WIS). ROCK Co: Croff, Sept. 23, 1961 (WIS). SAUK CO: Reedsburg, 1968] Smilax — Mangaly 257 Hansen 1018 (wis). ST. CROIX CO: edge of bog n. of Wis. Rt. 64, H. H. Iltis, F. G. Iltis & G. Noamesi (WIS). TREMPEALEAU CO: H. H. Iltis 6308 (WIS). VERNON CO: Viroqua, Smith 7248 (wis). Bergan Twp., Hartley 1029 (WIS). WALWORTH CO: Town of Geneva, Wadmond (WIS). WASHBURN CO: Minong, Fassett 8601 (WIS). WAUKESHA CO: Musbego, Lake Huron, Shirler, July 4, 1928 (wis); vicinity of Chen- equa, w. side of Pine Lake, Schenke & Niss, June 18, 1942 (wis). WYOMING. crook co: Sundance Mt., R. Nelson 2139 (Mo, RM). CONVERSE CO: Mill Creek, E. Nelson 703 (RM). SHERIDAN CO: Dayton, Tweedy 2277 (NY). 3. Smilax pulverulenta Michx. Smilax pulverulenta Michaux, Fl. Bor.-Americ. 2:238. 1803. Type: JI. Li IK. Piat ` Ss Ka pc sia re . SS Léo fhe J^ coma A, |t CA due A ETT T E 2 a i tg Moi fi gt t^. Ay gegen, hes A lay C^ thee d ee Mae / sca Cire foto E» wee. fod Mosa. Ar er WE a. E Ae seen ai ny efe Por ne SC — x wire “Ff : EW ja p. : a - Apr Ate POG M A a Ze Á A ^y ec pá e S , e Pago; E ch may Ae tentit, beter) AT Cee RR A ps y) dane on om de Fahad 4a MT Fre ue A a ft Aas ont, ver bps ne A 4 Cen Coffee euer £4 fcc nd tudo Ctm FA enid fig f "C a^ > mE gk, 4 s. yet ! KaD At... 9t A re qs m teresi È Jr oe É * as E Any sret da. Wees apfel e or Ge y PE AM. amm y m Meses Me eh on! langle 7 EA e OF P» a P, m Pu e Lima MO D irast A inde er doces AI. C T. m fp terne agen ah dd. Man Ho acne Mb A ëss A etr ai tn l hyper edant, Th. d flams avn flan nenv OG ED iy ol (wong E Mager ) pope pert tongue? pee th. € ndi AM Lat ymy fiers dar ar & fot SPREE gr 4e dr e à C. eter Z2 eat) boh f Ao nob cnt) mi RM ben Fo Ive me Pn Quest a Als dera; pe 7 Comet ES. LL . * 1 vv Ihr ka 2 d. Dare v Af Ath dan Lon K D D D ii 4 a E: he hibe Indians, O Hame, jur e : ne n^ H cs i sí- J mus t. feu ty hare 5 qund Le KIM finito n pune nsd f ry . ( See d ~s : pola m] 7] Ee Lev Act): Seep ae sey 2 pow A 6 nnm Plate 1385. A page from Spruce's field notebook describing the possibility that Haemadictyon might be added occasionally to the caapi drink made from Banisteriopsis Caapi. Courtesy of the Royal Botanie Gardens, Kew, England. 326 Rhodora [Vol. 70 that it was called caapi-pinima or “painted caapi.” In his field notebook, preserved at Kew, he wrote: “2712 Banis- teria Caapi Mss. From this is prepared an intoxicating drink known to all the natives of the Uaupés by the name of caapi. The lower part of the stem, which is the thickness of the thumb swollen at the joints, is the part used. This is beaten in a mortar with the addition of water and a small quantity of the slender roots of the Apocynac. ( Apparently a Haemadictyon) called caapi-pinima or painted caapi, from its leaves being stained and veined with red ... Query? May not the peculiar effects of the caapi be owing rather to the roots of the Haemadictyon (though in such small quantity) than to the stems of the Banisteria? The Indians, however, consider the latter the prime agent, at the same time admitting that the former is an essential ingredient. The two plants are planted near all mallocas (villages) ..." When these notes were published in 1908, they suffered a slight, but significant, change of emphasis, The terms of qualification disappeared. The published version states that caapi-pinima “is an apocynaceous twiner of the genus Hae- madictyon, of which I saw only young shoots, without any flowers, The leaves are of a shining green, painted with the strong blood-red veins. It is possibly the same species ... distributed by Mr. Bentham under the name Haema- dictyon amazonicum n.sp. It may be the caapi-pinima which gives its nauseous taste to the caapi ... and it is probably poisonous, . . . but it is not essential to the narcotic effect of the Banisteria, which (so far as I could make out) is used without any admixture by the Guahibos, Zaparos and other nations out of the Uaupés." Spruce was one of South America's most meticulous of scientific observers. A less careful and botanically un- trained observer might easily have confounded the young shoots of Haemadictyon amazonicum (now correctly called Prestonia amazonica) with Banisteriopsis, for the leaves of both are opposite, and the leaves of some species of Pres- tonia do resemble remarkably those of Banisteriopsis in shape and texture. But Spruce could never have confused 1968] Spruce's Explorations — Schultes 327 an apocyaceous plant, full of a white latex, with a Banis- teriopsis. He might have erred as to genus, for the genera of the Apocynaceae are often hard to distinguish even with flowers. But even this possibility would seem, in the case of Spruce, to be remote. In my own many years of explora- tion in the Rio Negro basin of Colombia and Brazil, I fol- lowed along Spruce’s itinerary in great part and searched especially for an apocynaceous vine growing around Indian huts, as described by Spruce: although every Indian Mani- hot-plot boasts its several cultivated vines of Banisteriopsis, nothing resembling a Prestonia was ever seen under culti- vation. A careful reading of Spruce’s notes reveals that he never claimed more for Prestonia or caapi-pinima than the role of an admixture. We know from reports of later investi- gators that other plants are sometimes added in small amounts to the beverage prepared from Banisteriopsis, in the belief that they change or fortify the effects of the narcotic drink. For example, I (18) reported the admix- ture of leaves of an apocynaceous tree — possibly Malouetia Tamaquarina —by the Makunas of the Rio Popeyaca of Colombia. Later writers, without supporting herbarium specimens and taking their cue from Spruce whose notes and files they misread or misunderstood, now propose that the narcotic drink in one part of the Amazon, where it is known as yajé, is prepared exclusively from Prestonia ama- zonica. For this assertion, there is absolutely no basis in field work. Prestonia amazonica is known, furthermore, from only one collection — the type collection made by Spruce in 1859 at Trombetas on the lower Amazon. In more than a cen- tury, the species has never been found again. We are forced to conclude, consequently, that Prestonia amazonica is either a very rare species or else a strict endemic, confined to the general region of the type collection. The Rio Trombetas lies more than 1200 miles in a straight line from the eastern slopes of the Andes of Colombia, where we are expected to believe that the natives are using this rare species in rela- 328 Rhodora [Vol. 70 tive abundance as the source of their frequently employed yajé. The chances that Prestonia amazonica is used are, for all practical purposes, non-existent; and there seems to be no evidence that any species of Prestonia is so employed along the eastern slopes of the Andes of Colombia and Ecua- dor — or indeed in the entire Amazon, Even in the area where Spruce reported its possible use a hundred years ago, there is all probability that, if employed at all, it was utilized solely as an admixture with Banisteriopsis. Never- theless, modern phytochemistry would do well to investi- gate this reportedly poisonous apocynaceous genus of some 30 species of tropical America. Little is known of Prestonia chemically, and the recently published analysis of P. ama- zonica, in which the presence of N — N dimethyltryptamine was reported, was undoubtedly based on a misidentified in- fusion made from Banisteriopsis Caapi (22).** Spruce reported that the Indians of the Rio Uaupés dis- tinguished two kinds of caapi, the minor one called caapi- pinima. Much later, Koch-Grünberg (7) found that the Tukanos of the Uaupés know two kinds, but he could iden- tify botanically only one: Banisteriopsis Caapi. In 1948, one hundred years after Spruce, I discovered the Indians on the Rio Tikié, a Brazilian affluent of the Uaupés, pre- paring an intoxicating drink from the malpighiaceous genus **Very recently, one of my graduate students, Mr. Homer V. Pinkley, learned that the Kofán Indians of Amazonian Ecuador sometimes add the leaves of Banisteriopsis Rusbyana to the drink prepared from the bark of B. Caapi or B. inebrians to “fortify” it. Holmstedt (in litt.) informs me that an analysis of these leaves col- lected by Pinkley indicates that the species contains harman alka- loids and N, N-dimethyltryptamine. Banisteriopsis Rusbyana is called óco-yaje by the Kofán; B. Caapi and B. inebrians are known simply as yaje, as is the narcotic drink. One is led to wonder, in view of this recent discovery, whether or not the admixture so often stated to be Prestonia amazonica, from which the tryptamine was erroneously reported, is, in reality, Banisteriopsis Rusbyana. The material an- alysed by Holmstedt has been identified against a voucher collection: Pinkley 449, deposited in the Economie Herbarium of Oakes Ames in the Botanical Museum of Harvard University. (R. E. S. July, 1968). 1968] Spruce's Explorations — Schultes 329 TETRAPIERIS 3 methystica N RE Schultes `` 330 Rhodora [Vol. 70 Tetrapteris: from a species called T. methystica (16). From the bark of this gigantic forest liana, with no admixtures, a bitter beverage is prepared. I experimentally learned that it had strong hallucinogenic properties. The drink was rather yellowish, unlike the usual chocolate-brown of the drink prepared from Banisteriopsis Caapi. One wonders whether or not the term “painted caapi" reported by Spruce could have referred to the kind of caapi that makes this unusual yellowish drink. The confusion resulting from the misinterpretation, mis- representation and misunderstanding of Spruce's very ob- vious suggestions concerning Prestonia as a possible narcotic are, unfortunately, typical of so much that plagues modern phytochemistry and that will continue to do so, until more thorough botanical evidence is demanded by non- botanical writers before categorie statements are made either on carefully qualified suggestions or without voucher Specimens. A century of research work, from Spruce's time to the present, has given us an appreciable body of information about ayahuasca, caapi and yajé, but much still remains to clarify and much still remains to discover. There are numerous problems demanding research, problems not even touched upon by Spruce. One of these concerns the use of admixtures with Banisteriopsis and the identity of the plants employed as admixtures. Another concerns different "kinds" of basic plants employed. This problem has been reviewed recently in the literature (5, 18). Most recently, a Brazilian anthropologist, Alves da Silva (1) has indi- cated, through aboriginal names, the use of five presumably different lianas and several admixtures in the preparation of caapi by the Tukanos in the very region where Spruce's work was done one hundred years ago. The second hallucinogen that Spruce encountered and studied assiduously in the field was the yopo or niopo snuff of the upper Orinoco of Colombia and Venezuela, It was Humboldt's expedition that first identified the source plant as a species of the leguminous genus Pipta- 1968] Spruce’s Explorations — Schultes 331 denia in 1801 (6). There is a very brief account of its narcotic use: “Ex seminibus tritis calci vivae admixtis fit tabacum nobile que Indi Otomacos et Guajibos utuntur." The plant was first described by Humboldt as a species pos- sibly of Acacia: then it was transferred to the genus Inga; finally, it was accommodated in Piptadenia as P. Niopo. Spruce collected material of the plant (Spruce 1786) along the Rio Janauarí. Bentham determined it as Lin- naeus’ Mimosa peregrina, but placed it in the genus Pipta- denia, thus altering its name to P. peregrina. This is the name by which the source of yopo snuff has been known until recently, when modern taxonomic studies have tended to show that the plant may belong to the genus Anadenan- thera, in which case the modern name should be A. pere- grina (11). We owe to Spruce the earliest reliable ethnobotanical studies on yopo, although I strongly suspect that some of the present confusion surrounding the use of this snuff may trace back to his writings. It will be necessary here to diverge for a moment from Spruce's work and consider briefly the state of our knowl- edge concerning South American narcotic snuffs. Most snuffs used by natives in South America are conveniently identified as being made either from tobacco or from Pipta- denia, Cooper’s distribution map, for example, indicates that Anadenanthera peregrina snuff is taken in most of the western Amazon and in many other isolated spots through the Amazon Valley (2). That Cooper himself recognised the unsatisfactory status of our knowledge is evident from his remarks that *some of the attributions may not be correct, since, in some cases, the lack of exact botanical identification makes it doubtful whether we have to do with Piptadenia snuff, tobacco snuff or snuff from some other plant . . .” We have now come to suspect (21, 24) that the snuffs used in the Amazon Valley itself are rarely if ever made from this species. In fact, it is now clear that Anadenanthera peregrina is a rare plant in the Amazon, where it occurs, for the most part, along the upper Rio 332 Rhodora Nas Fletcher —S\ ke ANADENANTHE om K a s d W A M H AA, Ki 4 Ze y ] A pereg Y S, ud rina [Vol. 70 1968] Spruce's Explorations — Schultes 333 Branco or elsewhere rarely. In my long period of botanical exploration in the Amazon, including a number of years in Spruce's territory of the upper Rio Negro basin, I never once met with Anadenanthera peregrina, although, as a specialist on native narcoties, I kept a sharp outlook for it. Eight other species referred to Piptadenia are known to occur in the Amazon, and I suspect that, in the fleld, Spruce attributed some of these— not known to have narcotic properties — to his concept of P. peregrina. He said“... I have traced the tree all the way from the Amazon to the Orinoco, . . . although it bears a different name on the two rivers...” The species apparently does not have this dis- tribution now. Recent ethnobotanical work has uncovered other plants as the sources of psychotomimetic snuffs, One of the most interesting, used in the northwesternmost part of the Ama- zon, in the upper Rio Negro basin, and adjacent areas of the uppermost Río Orinoco and its tributaries, is prepared from the red resin of the bark of certain species of the myristicaceous genus of Virola, especially V. calophylla and V. calophylloidea (17). I suspect that a species of the acanthaceous genus Justicia may be used in the preparation of a snuff of the Waiká Indians of the headwaters of the Orinoco.*** In the central part of the Brazilian Amazon, à psychotomimetic snuff, said to be prepared from the fruits of a jungle tree of the Moraceae, Olmedioperebea sclero- phylla, has recently been found (20). Other snuffs, as yet unidentified, are known from several localities in South America. In view of the brevity of Humboldt's report of the prep- aration of yopo snuff, Spruce's detailed notes acquire much more importance to ethnobotany and phytochemistry. “I did not see the snuff actually prepared from the seeds and ***Very recent investigation has shown that the snuff of the Waiká Indians is prepared basically from Virola theiodora, sometimes with the admixture of pulverized leaves of Justicia pectoralis var. spheno- phylla and ashes of the bark of Elizabetha princeps. This report was published in Rhodora 70 (1968) 113-160. (R. E. S. July, 1968). 334 Rhodora [Vol. 70 in use until June, 1854, at the cataracts of the Orinoco. A wandering horde of Guahibo Indians, from the river Meta of Colombia, was encamped on the savannahs of Maypures and on a visit to their camp, I saw an old man grinding niopo seeds, and purchased of him his apparatus for making and taking the snuff... “The snuff is kept in a mull made of a bit of the legbone of the jaguar, closed at one end with pitch and at the other end stopped with a cork of marima bark, It hangs around the neck... “The Guahibo had a box of caapi hung from his neck, along with the snuff-box, and as he ground his niopo he every now and then tore off a strip of caapi with his teeth and chewed it with evident satisfaction. ‘With a chew of caapi and a pinch of niopo,’ said he, in his broken Spanish, ‘one feels so good! No hunger — no thirst — no tired!’ From the same man, I learnt that caapi and niopo were used by all the nations on the upper tributaries of the Ori- noco. .. ." Much research — ethnobotanical, botanical, phytochem- ical and pharmacological — has been devoted to yopo snuff. It was probably Spruce's writings more than those of any other investigator, however, that have spurred this research on over the past century. One of the interesting characteristics about Spruce's writings is that I have never been able completely to un- derstand the paucity of incidental notes on plants with physiological activity. Such long residence amongst the natives of the Amazon might be expected to have enabled Spruce to see so much more that he could report concerning medicinal plants, fish poisons, arrow poisons and other plants of medicinal notice. It may be that Spruce kept such observations in a separate notebook that has been lost, but it is rather more probable that he never dreamed of follow- ing up seriously any of these aboriginal uses and, therefore, failed to jot more of them down. In the two volumes Notes of a botanist on the Amazon and. Andes, I find no references to any of the arrow or dart 1968] Spruce's Explorations — Schultes 335 poisons. Nor do I see any mention of plants referred to as fish poison, although this aspect of Indian culture could not have escaped an explorer who lived so long amongst the natives. Furthermore, even though some of the tribes of the upper Rio Negro and Uaupés are known as makers of the most powerful curare preparations or arrow poisons, Spruce is strangely silent on this point. A study of the botanical material amassed on his trips indicates that Spruce did, in fact, collect many of the plants employed in the region as fish poisons or in making curare, but they usually bear no notes regarding these uses. How Spruce, for example, could have failed to discover the use of Virola in snuff-making in the uppermost Rio Negro I can never explain, especially when he described as new a number of myristicaceous species. My own botanical explorations in the northwest Amazon from 1941 through 1953 enabled me to re-trace, in part, Spruce's itinerary. Many are the notes on toxic or medicinal uses that the natives have for plants that Spruce himself collected for the first time and introduced to science with not a note. A few examples will indicate why I am puzzled in this regard. Spruce collected several species of the genera Connarus and Rourea in the Rio Negro basin. The roots and stems of these vines are pounded and employed as fish-poisons. Nothing, apparently, is known phytochemically of these uses, nor did Spruce mention them. Spruce collected the type of Caryocar gracile (Spruce 2550) of the Caryo- caraceae at Ipanoré on the Rio Uaupés. In this region to- day the leaves of Caryocar gracile and other species of the genus are crushed and made into a paste that is ad- ministered to dogs for the purpose of killing them when they have once fled with fear in the hunt. Another member of the same family collected by Spruce (Spruce 2146) at San Carlos, Venezuela, is Anthodiscus obovatus, said to be an ingredient of curare, along with Strychnos. Spruce failed to report these relatively important uses of Caryocar and Anthodiscus. The root of the rubiaceous genus Duroia — 336 Rhodora [Vol. 70 especially D. petiolaris collected by Spruce — is reputedly toxic. Another genus of this family, Retiniphyllum, has the reputation of possessing a toxic resinous exudation from the inflorescence. Spruce collected three species of Retiniphyllum in the upper Rio Negro basin — R. truncatum, R. concolor and R. speciosum, all of which are reputedly poisonous, although R. truncatum has the reputation of being the strongest. He apparently did not mention the interesting and widespread native belief that the bones of the bird pajuíl (Nothocrax urumutum) are toxic to dogs if the bird has eaten the ripened fruit of the apocynaceous tree Malouetia (19). There is, however, so much of value for phytochemistry to be gleaned from Spruce and other contemporary ex- plorers that we should not dwell with too much severity on such minor shortcomings — if, indeed, shortcomings they be. We should here decide to utilise efficiently what they leave us. We cannot but admire a man like Spruce who, in spite of oppressive sickness, weariness and frustration, was en- couraged and spurred on to incredible feats by a dynamic love of knowledge for the sake of knowledge. In a letter to Mr. George Bentham he wrote in part: “I have lately been calculating the number of species that yet remain to be discovered in the great Amazonian forest, from the cataracts of the Orinoco to the mountains of Matto Grosso; taking the fact that by moving away a degree of latitude or longitude I found about half the plants different as a basis, and considering what very narrow strips have, up to this day, been actually explored, and that often very inadequately, by Humboldt, Martius and myself, and others, there should still remain some 50,000 or even 80,000 spe- cies undiscovered. To anyone but me and yourself this esti- mation will appear most extravagant, for even Martius (if I recollect rightly) emits an opinion that the forests of the Amazon contain but few specimens. . . “At the highest point I reached on the Uaupés, the Jagua- raté Caxoeira, now the boundary line between Brazil and 1968] Spruce’s Explorations — Schultes 337 Colombia, I spent about a fortnight, in the midst of heavy rains, when (according to my constant experience) very few forest trees open their flowers. But when the time came for my return to Panuré . . . . the weather cleared up, and as we shot down among the rocks which there ob- struct the course of the river, on a sunny morning, I well recollect how the banks of the river had become clad with flowers, at it were by some sudden magic, and how I said to myself, as I scanned the lofty trees with wistful and disappointed eyes, ‘there goes a new Dipteryx — there goes a new Qualea — there goes a new the Lord knows what! until I could no longer bear the sight and covering up my face with my hands, I resigned myself to the sorrowful reflection that I must leave all these fine things 'to waste their sweetness on the desert air” From that point up- wards, one may safely assume that nearly everything was new, and I have no doubt that the tract of country lying eastward from Pasto to Popayán where are the head-waters of the Japuré, Uaupés, and Guaviare —. . . offer as rich a field for a botanist as any in South America. But I have made enquiries as to the possibility of reaching it, and I find that it will be necessary to cross páramos of the most rugged and inhospitable character, and afterwards risk oneself among wild and fierce Indians, so that I fear its exploration must be left to some one younger and more vigorous than myself." It is in great part from men like Spruce, that we may continue to expect — even in this commercialised day of high pressure and efficiency — progress in our study of the numberless unknown chemical compounds resident in the vegetation of the world. BOTANICAL MUSEUM OF HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS 02138 REFERENCES CITED 1. ALVES DA SILVA, ALCIONILIO B. “A civilização indigena do Uaupés" (1962) 228. 2. COOPER, JOHN M. *Stimulants and narcotics" in Handbook of South American 338 Rhodora [Vol. 70 Indians Ted. J. H. Steward] Bur. Am. Ethnol. Bull. No. 143, 5 (1949) 525. 3. FABRE, RENE “Quelques plantes médicinales de l'Amérique Latine: leur utilisation thérapeutique" Rev. Gen. Sci. Pures Appl. 62 (1955) 49. 4. FISCHER C., G. "Estudio sobre el principio activo del yagé [Thesis, ined.] (1923) Universidad Nacional de Colombia, Bogotá, Colombia. FRIEDBERG, CLAUDINE "Des Banisteriopsis utilisés comme drogue en Amérique du Sud" Journ. Agr. Trop. Bot. Appl. 12 (1965) 403-437, 550- 594, 729-780. 6. HUMBOLDT, ALEXANDER & AIME BONPLAND "Voyage aux regions equinoxiales" 2 (1819) 620. KOCH-GRUNBERG, THEODOR "Zwei Jahre unter den Indianern" 1 (1909) 300. 8. MICHAELS, M. and E. CLINQUART "Sur les réactions chimiques d'identification de la yagéine” Bull. Acad. Roy. Belg., S. 5, 6 (1926) 19. 9. MORTON, C. V. "Notes on yagé, a drug plant of southeastern Colombia" Journ. Wash. Acad. Sci. 21 (1931) 485. 10. REINBERG, P. "Contribution à l'étude des boissons toxiques des indiens du Nordouest de l'Amazone, l'ayahuasca, le yajé, huanto" Journ. Soc. Améric. Paris, n.s., 13 (1921) 25-54, 197-216. 11. von REIS, SIRI S. P. “The genus Anadenanthera: a taxonomie and ethnobotanical study” [Ph.D. Thesis ined.] (1961) Radcliffe College, Cam- bridge, Mass. 12. ROCHA, JOAQUÍN "Memorandum de viaje (regiones amazónicas)" El Mercurio (Bogotá) (1905) 43. 13. RoumniER, A[lexandre] "Les plantes divinatoires" Rev. Metapsych. (1926) 325. 14. ScHULTES, RICHARD EVANS "Plantae Austro-Americanae VII. De festo seculari Ricardi Sprucei America Australi adventu commemoratio atque de plantis principaliter vallis amazonicis diversae observationes" Bot. Mus. Leafl, Harvard Univ. 15 (1951) 29-78. Ku N 15. “Richard Spruce still lives” Northern Gard. 7 (1953) 20-27, 55-61, 87-93, 121-125. 1968] 16. Spruce’s Explorations — Schultes 339 DC *Plantae Austro-Americanae IX. Plantarum novarum vel notabilium notae diversae" Bot. Mus. Leafl, Harvard Univ. 16 (1954) 202-205. 18. *A new narcotic snuff from the northwest Amazon" Bot. Mus. Leafl., Harvard Univ. 16 (1954) 241-260. 19. “The identity of the malpighiaceous narcotics of South Amer- ica" Bot. Mus. Leafl., Harvard Univ. 18 (1957) 1-56. 20. *A reputedly toxic Malouetia from the Amazon" Bot. Mus. Leafl. Harvard Univ. 19 (1960) 123-124. 21. “The search for new natural hallucinogens” Lloydia 29 (1966) 293-308. 22. “The botanical origins of South American snuffs" in Ethno- pharmacologic search for psychoactive drugs [ed. D. Efron] U.S. Publ. Health Serv. Publ. No. 1645 (1967) 291-306. and ROBERT F. RAFFAUF “Prestonia: an Amazon narcotic or not?” Bot. Mus. Leafl., Harvard Univ. 19 (1960) 109-122. 23. SPRUCE, RICHARD “A botanist on the Andes and Amazon” [ed. A. R. Wallace] 2 (1908) 414-4380. 24. WASSEN, S. HENRY “The use of some specific kinds of South American Indian snuff and related paraphernalia” Etnolog. Stud. No. 28 (1965) 1-132. THE SPORES OF FOUR SPECIES OF SPINULOSE WOOD FERNS (DRYOPTERIS) IN EASTERN NORTH AMERICA DONALD M, BRITTON In a recent paper Nannfeldt (1966) indicated that the diploid Dryopteris assimilis S. Walker could be readily dis- tinguished from the tetraploid Dryopteris dilatata, (Hoffm.) A. Gray by means of spore morphology and the color of the perispore. By using these characters, he was then able to accurately identify these species, and plot their distribution in Sweden. Earlier, Walker and Jermy (1964) emphasized that the spores of D. assimilis had a thin pale brown peri- spore with “widely spaced acute spinnules up to lu in length." 'The spore characters of various species of Dryop- teris in North America were used by Crane (1963) to construct a key for identification, The data in this paper are presented to compare and contrast the spores of four species of spinulose wood ferns that are found in Eastern North America. Although there is as yet no unanimity in the names used for these species, or indeed if they should be considered as species, workers are familiar with the names used here. The four taxa considered here as separate species consist of two diploids and two tetraploids. They are the diploid D. intermedia. (Muhl.) A. Gray, a diploid segregate species allied to D. dilatata (Hoffm.) A. Gray which may be con- specific with D. assimilis S. Walker, D. spinulosa (O. F. Muell.) Watt (D. carthusiana), and D. campyloptera Clark- son. Dryopteris campyloptera is considered by Wagner (1963) to be an amphidiploid of the first two species, where- as D. intermedia is part parental to D. spinulosa (Walker 1961). On the basis of cytological and morphological studies, all four species are closely related. MATERIALS AND METHODS Sori on mature fronds were gently scraped with a needle dipped in Permount in order to obtain spores, These spores 340 1968] Spores of Dryopteris — Britton 341 were then transferred to a drop of Permount on à slide and a cover glass was added. It was usually necessary to put a small weight on the cover glass in order to have à thin enough mount for observation with oil immersion. Measurements of exospore length were made using an ocu- lar micrometer and a 90X apochromatic oil immersion ob- jective. The calibration of the eye-piece was made with the aid of a micrometer slide. The calibration was carefully checked by an independent observer. One ocular division was equivalent to 1.4 microns. It is considered that due to operator error or idiosyneracy that the error in individual measurements is probably + 1 ocular division or might be considered to be ca. +1.5 microns. Whenever possible, cytologically determined material was used for spore measurements. However, this proved impos- sible for four collections from Mt. Washington, N.H. of D. campyloptera, and for five collections of D. dilatata from near Lake Superior, Ontario. In these cases, it was neces- sary to rely on material that was morphologically similar to cytologically determined material from the same limited area, e.g. 12 collections of D. campyloptera had been studied cytologically from Mt. Washington (Britton 1962). Voucher Specimens are all at OAC except as indicated for five specimens designated by TRT: (Where no collector is given, the collection is by the author) D. intermedia 620, Cone. IV, lot 11, Puslinch Tp., Wellington Co., Ontario, 26 July 1962. 741, 745, 750, 753, Swan Lake, Algonquin Park, 20 July 1968. 825, 826, 843, 853, The Beaver Pond, Algonquin Park, Ontario, 15 August 1964. 861, H. M. Dale, 2 mi N of Jefferson Notch near Mt. Washington, N.H. Alt. 2,300 ft., 8 Sept. 1964. D. dilatata 715, 717, Garden transplants from N end of Jack- fish Lake, Thunder Bay Dist., 25 June 1963. vouchers 23 August 1965. 854, 900, 901. Garden transplants, A. Asselin, Lac Beaudoin, Amos, P.Q. 26 May 1965, vouchers 23 August 1965. (TRT 2, TRT 4), Taylor, Bannan and Harrison, Nos. 129, 130, Port Coldwell, vic. Peninsula, Thunder Bay Dist, 16 August 1939 (TRT). (TRT 6), Hosie, Losee and Bannan, No. 89, Walker Lake, Schreiber, Thun- 342 Rhodora [Vol. 70 der Bay Dist., 11 July 1937 (TRT). (TRT 7), Taylor et al., No. 288, Mamainse Mt., Algoma Dist., 18 July 1935 (TRT). (TRT 8), Hosie, Losee and Bannan, No. 91, Slate Islands, Thunder Bay Dist., 30 July 1937 (TRT). D. spinulosa 286, 290, Conc. VIII, Lot 20, Puslinch Tp., Welling- ton Co., Ontario, 22 August 1960. 299, Conc. III, Lot 15, Puslinch Tp., Wellington Co., 19 August 1960. 305, Conc. IV, Lot 11, Puslinch Tp., Wellington Co., 19 August 1960. 488, Conc. IV, Lot 11, Puslinch Tp. Wellington Co., 28 Sept 1961. 504, Conc. IV, Lot 14, Nelson Tp., Halton Co., 21 Sept 1961. 623, 624 (see 620) 716 (see 715 and 717) 744 (see 741, 745 etc.) D. campyloptera 857-860 (see 861) 971, Garden Transplant, M. Lan- don, Yarmouth, N.S. 14 July 1965. OBSERVATIONS The average lengths of 20 exospores for each collection of ten plants of each of D. intermedia, D. dilatata and D. spinulosa and for five plants of D. campyloptera are given in Table 1. The extreme measurements for individual spores are shown as well as the overall mean for the species and the range of means of collections. In Plate 1386, small sections of the perispore are illus- trated to show the morphology of some of the spines present. DISCUSSION One of the initial decisions, was to decide how many spores of each collection to measure, The measurements of 10, 20, and 30 spores from each of five collections were analyzed by Dr. G. C. Ashton, Dept. of Mathematics. Ac- cording to the analysis of variance made by him, 20 spores were a suitable number to measure. The 30 spore sample did not yield any further information, The measurements, together with the morphology show that D. intermedia has the most distinctive spores of the four species. The spores of this species are smaller and are covered by long, narrow, sharp spines. The other three species cannot be easily identified by spore characters. These results are not in agreement with those of Wagner 1968] Spores of Dryopteris — Britton 343 TABLE 1 Lengths of exospores for four species of Dryopteris Species Collection Mean length Range of of 20 exospores lengths in microns in microns D. intermedia 620 36.3 32-39 741 34.4 34-39 745 32.8 31-38 750 322 29-35 753 35.3 31-39 825 35.2 28-39 826 35.9 31-41 843 36.9 31-42 853 33.9 31-39 861 30.9 28-34 Mean 34.4 (31-37) 28-42 D. dilatata 715 37.4 34-41 717 36.1 32-41 854 37.1 34-42 900 37.4 31-41 901 36.8 34-41 TREZ 39.6 35-48 TRT4 37.8 34-42 TRT6 37.8 35-39 PRA 36.3 34-42 TRT8 Ot 35-41 Mean 37.3 (36-40) 31-48 D. spinulosa 286 40.0 35-43 290 39.8 35-43 299 85.1 34-38 305 39.4 36-43 488 41.0 36-45 504 36.8 32-42 623 37.1 34-41 624 37.0 34-39 716 37.4 35-42 744 39.3 36-43 Mean 38.3 (35-41) 32-45 D. campyloptera 857 41.3 38-43 858 40.6 36-46 859 38.6 35-42 860 39.9 35-45 971 38.4 35-45 Mean 39.8 (38-41) 35-46 344 Rhodor: [Vol. r Plate 1386 A, Spore of Dryopteris intermedia, X 2000. B-E Portions of perispores to show the morphology of the spines, X 4700: B, D. dilatata; C. D. spinulosa; D, D. intermedia; E, D. campyloptera. Note the similarity between B and C and between D and E, 0 1968] Spores of Dryopteris — Britton 345 and Hagenah (1962) and Wagner (1963). My results are contrasted with those of these authors in Table 2. TABLE 2 Mean exospore length and range of means in microns for four species of Dryopteris D. intermedia D. dilatata D. spinulosa D. campyloptera Britton 34.4 (31-37) 37.3 (36-40) | 38.8 (35-41) 39.8 (38-41) Wagner 32 (30-32)* — (83-87) — 38 (37-40) *4 collections As can be seen, Wagner's limited sample of four collec- tions of D. intermedia from Virginia had smaller spores. The data are in good agreement for D. campyloptera, but differ markedly for D. dilatata. It was originally hoped that spore measurements would easily separate the diploid D. dilatata from the tetraploid D. campyloptera. Such is not the case. Although Wagner and Hagenah (1962) sug- gest that the averages for individuals of D. dilatata will be between 33-37 microns and for D. campyloptera from 37-40 microns my data do not support this contention. Eight of the ten collections of D. dilatata that were studied had an average exospore length (20 spores) of 937 or greater than 37 microns and none were smaller than 36 microns. Although the average size of all the collections of D. cam- pyloptera is greater (39.8) than that of D. dilatata (37.3), this does not help in identifying single plants. Crane (1960) has illustrated the spines of D. spinulosa as being blunt. In reality, all are sharp-tipped, but in morphology the sides of the spines are approximately as long as the base is broad, in contrast to the needle-like, narrow-based spines of D. intermedia. (Plate 1386) The spines of D. dilatata are widely set and are more similar to those of D. spinulosa than to those of D. inter- media with its closely set, long, sharp spinnules. The spines of D. campyloptera can be visualized as showing the influence of both putative parents, D. intermedia and D. dilatata. The spines seem slightly longer and more pro- nounced than those of D. dilatata but they are more widely spaced than those of D. intermedia. However, the differ- ences are not clear-cut, and it would be difficult to identify one species from another by perispore morphology alone. 346 Rhodora [Vol. 70 Lovis (1964) has discussed the difficulties in obtaining accurate measurements for spore size in Asplenium. Cali- bration of the microscope, mounting medium, operator error and number of spores measured have been mentioned here. Besides these, there is the orientation of the spore on the slide, the possible unconscious selection of the largest spores, spores from sporangia produced late in the season and hence poorly developed, rather than from early sporangia from which the spores were shed. Also, spores from plants growing in very humid conditions versus those growing in drier locations and of course, genetic variability of the in- dividual plants. However, unlike Asplenium trichomanes (Lovis 1964), it would seem that spore size and ploidy are not closely correlated in Dryopteris. For example, D. fragrans is di- ploid and has large spores (Crane 1960), and in this paper, the diploid D. dilatata has spores which are approximately the size of those of the tetraploids D. spinulosa and D. campyloptera. Brown (1964) showed in Woodsia that the spores of three tetraploid taxa were not larger than those of closely related diploid species. In this respect, Dryop- teris and Woodsia are similar. CONCLUSIONS The size of the spores as measured by the length of 20 exospores together with the ornamentation of the perispore allows one to identify diploid Dryopteris intermedia. D. dilatata, D. spinulosa, and D. campyloptera cannot be easily separated by either spore size or perispore morphology. Accordingly, D. dilatata and D. campyloptera which have a similar leaf morphology cannot be easily identified, Hence, until further cytological studies are made, identification of D. dilatata (2X) and D. campyloptera (4X) for the purposes of plotting their distribution is not possible. The critical cytogenetic evidence is still not available (Britton 1962) to decide the status of the taxon referred to as D. dilatata. However, the description of D. assimilis given by Walker and Jermy (1964) with particular em- phasis on spore characters does not disagree with the spore 1968] Spores of Dryopteris — Britton 347 characters of the diploid segregate species referred to here as D. dilatata. The possibility exists that the diploid segre- gate species is conspecific with D, assimilis. DEPARTMENT OF BOTANY UNIVERSITY OF GUELPH ONTARIO, CANADA REFERENCES BRITTON, D. M. 1962. Dryopteris dilatata (Hoffm.) A. Gray in North America. Rhodora 64: 207-212. Brown, D. F. M. 1964. A monographic study of the fern genus Woodsia. Nova Hedwigia 16: 1-154. CRANE, F. W. 1960. A key to American Dryopteris species based on characters of the perispore. Amer. Fern J. 50: 270-275. Lovis, J. D. 1964. The taxonomy of Asplenium trichomanes in Europe. Brit. Fern Gaz. 9: 147-160. NANNFELDT, J. A. 1966. Dryopteris dilatata och Dr. assimilis i Sverige. Bot. Not. 119: 136-152. WAGNER, W. H., JR. 1963. Pteridophytes of the Mountain Lake Area, Giles County, Virginia, including notes from Whitetop Mountain. Castanea 28: 113-150. WAGNER, W. H. JR, & D. J. HAGENAH. 1962. Dryopteris in the Huron Mountain Club Area of Michigan. Brittonia 14: 90-100. WALKER, S. 1961. Cytogenetic studies in the Dryopteris spinulosa complex. II. Amer. J. Bot. 48: 607-614. WALKER, S. & A. C. JERMY. 1964. Dryopteris assimilis S. Walker in Britain. Brit. Fern Gaz. 9: 137-140. A MONOGRAPH OF THE GENUS RATIBIDA' EDWARD LEON RICHARDS The genus Ratibida, comprised of six perennial species, one variety, and one color form, is native to southern Canada, the United States, and northern Mexico. The plants inhabit prairies, edge of woodlands, coastal plains, foothills, pine forests and meadows. The meaning of the name Ratibida, as far as can be determined, is unknown. Ratibida columnifera, R. pinnata and R. tagetes are sold by nurseries under the name of coneflowers; the other species are of little economical importance. Although In- dians of northern Mexico use R. mexicana as a wash and potion for bites, rheumatism and other ailments, the plants are of little forage value and are frequently classified as weeds because of their ruderal nature. This study is the first monographic treatment of Ratibida and is based on 3703 specimens from 15 herbaria. Extensive field studies of all taxa from Minnesota to northern Mexico and Colorado to Ohio have afforded the author an oppor- tunity to make population studies and to make more than 300 collections throughout the range of the genus. The completion of this work will add to the knowledge of the genus and provide a basis for future biosystematic study. Chromosome numbers for the species of Ratibida are listed in Table 1. Recorded chromosome counts for the genus are n — 14 in R. columnifera and R. peduncularis var. picta (Jackson, 1959), n = 16 in R. tagetes (Jackson, 1959), 2n — 26, 27 in R. columnifera (Perdue, 1959) and 2n — 28 in R. pinnata (Perdue, 1959). These chromosome numbers were verified during this study from root tip material; counts of other taxa of Ratibida not previously recorded are shown in Table 1. 'Submitted in partial fulfillment for the degree of Doctor of Philos- ophy in Botany at the University of Kansas, Lawrence, Kansas. 348 1968] Ratibida — Richards 349 Table 1. Chromosome Numbers in Ratibida Species n Seed Source’ R. columnifera 14 Ford Co., Kans. 3763 f. columnifera R. columnifera 14 Sheridan Co., Kans. 1069 f. pulcherrima R. latipalearis 14 Colonia Garcia, Chi- huahua, Mex. D. Spilsbury R. mexicana 14 50 km sw. of Guerrero, Chihuahua, Mex. Hector and Raul Sea R. peduncularis 14 Willacy Co., Tex. 3743 var. peduncularis R. peduncularis 14 Gonzales Co., Tex. 3728 var. picta R. pinnata 14 Anderson Co., Kans. 3767 R. tagetes 16 Mora Co, N. Mex. 3773 "Unless otherwise indicated, the numbers quoted are those of the author's collections. Pollen of all taxa was examined and the only significant difference in stainability was noted in R. peduncularis var. picta which was 85 percent as compared to 91-97 for the other species. Pollen size for the genus is 17-26 micra as indicated in Table 2. Cotton blue in lactophenol was used and pollen grains taking a full, dark stain were considered viable. Percent of stainability is based on 500 grains for each taxon and measurements of the outside diameter, exclusive of spines, were taken from 50 grains at 450 magnification. Rafinesque (1817) founded Ratibida with a brief descrip- tion based on Rudbeckia columnaris Pursh (1814). In his description, Rafinesque said, “The species with naked seeds such as Rudbeckia pinnata Vent. (1800) must form my genus Obelisteca and those with a simple perianthe, such 350 Rhodora [Vol. 70 Table 2. Size and Stainability of Pollen in Ratibida Percent Size Average Sources Species Stainable Range in Size in of Micra Micra Material R. columnifera 97.2 17.1-21.5 19.01 Richards 8722 f. columnifera R. columnifera 94.5 19.8-26.64 24.37 Richards 3725 f. pulcherrima R. latipalearis 93.9 18.2-23.3 20.20 Richards 3770 R. mexicana 97.4 18.9-26.64 22.11 Richards 3768 R. peduncularis 91.5 19.8-26.9 22.77 Richards 3739 var. peduncularis R. peduncularis 85.0 18.4-23.1 21.14 Richards 562 var. picta R. pinnata 95.2 18.9-25.8 22.87 Richards 3766 R. tagetes 96.2 18.7-23.3 21.68 Richards 8764 as Rudbeckia columnaris, must form my genus Ratibida.” Rafinesque (1819) published a complete description of Ratibida and the same year described the genus Lepachys based on Rudbeckia pinnata Vent. Lepachys, a later name than Ratibida, is not acceptable under the International Rules of Botanical Nomenclature. Cassini (1825) described Obeliscaria based on Rudbeckia pinnata Vent. and deCandolle (1836) treated Ratibida and Lepachys as synonyms of Obeliscaria. Obeliscaria, described later than Ratibida, is also congeneric with Ratibida Raf. Don (1838) recognized Ratibida as being the first validly published name for the genus, even though Lepachys and Obeliscaria were still being used by other authors. Torrey and Gray (1842) used Lepachys for three species and one variety, L. columnaris, L. pinnata, L. peduncularis and L. colummaris var. pulcherrima, Gray (1852) named L. peduncularis var. picta and later transferred Rudbeckia tagetes James to Lepachys. Watson (1888) described Lepachys mexicana, and Barnhart (1897) recognized Rati- bida and rejected Lepachys on a priority basis. 1968] Ratibida — Richards 351 The only recent study of the genus was by Sharp (1935) who did a comparative study on all epappose genera of the Heliantheae. TAXONOMIC CHARACTERS ROOTS. Fibrous and taproots occur within the genus. STEMS. The stems are 1.6 to 12.3 dm tall, herbaceous, and branched or unbranched. The leafy branches terminate upward into a naked peduncle. The stems are sulcate with those of R. mexicana and R. latipalearis averaging 10 ridges, R. tagetes 5-6, and all other species 8 ridges. LEAVES. The leaves of all species are alternate with larger ones at the base and smaller leaves higher up. Petioles on lower leaves are long, becoming short to absent on the uppermost ones. The leaf blades are entire to pinnate, lyrate-pinnate or pinnatifid, linear to lanceolate or obovate. VESTITURE. The leaves of all species are punctate and hirsute to strigose-hirsute. The trichomes on the lower parts of the stem are up to 1 mm long or longer in R. mexicana, R. peduncularis, R. pinnata, and also R. columni- fera f. pulcherrima occurring at high elevations. The re- maining taxa possess trichomes of 0.5 mm or shorter. With few exceptions the leaf pubescence is more dense on the lower surface and limited to the veins and margins on the upper surface; stems and peduncles are hirsute to strigose- hirsute on the ridges with unicellular globular glands in the grooves. PEDUNCLES. Length of the peduncle is important taxo- nomically as it ranges from 6 cm in R. tagetes to 62 cm in R. mexicana. The heads are solitary at the ends of naked peduncles. PHYLLARIES. The herbaceous phyllaries are strigose- hirsute on the outer surface and glabrous to strigose-hirsute only at the tip and along the vein of the inner surface. They are biseriate with the inner 0.1-6.0 mm long, ovate, and the outer series 0.1-14 mm long and linear-lanceolate. Occa- sionally they appear as a single row or poorly defined double series, e.g. R. mexicana. 352 Rhodora [Vol. 70 PALEA. The palea are conduplicate, enclosing the achenes and falling from the receptacle with mature seed in all species except R. mexicana. The adaxial edges are hyaline and may be winged or ciliate. The abaxial edges are velu- tinous-hirsute to about half way down with unicellular glands from the middle to the base in all species except R. mexicana which occasionally lacks glands. The tips are recurved and densely velutinous-hirsute, giving the charac- teristic hoary appearance to the immature heads, A dark resinous linear or oblanceolate gland is along the adaxial edge. RECEPTACLE. The receptacle is narrow with a tapered apex or oblong with a rounded apex and is usually 2.5 cm or longer in all species other than R. pinnata and R. tagetes. RAY COROLLAS. The ray corollas are yellow, yellowish- purple or purple in various species and are all neutral, They are ovate-oblanceolate or lanceolate-oblong with two to three teeth at the apex and reflexed in all species except R. mexicana and R. latipalearis. DISC COROLLAS. The disc corollas are fertile, 0.7-2.0 mm long. Each has a light yellowish-green, tubular corolla with purple in the lobes in all species except R. peduncularis var. peduncularis which has a dark purplish-green apex and R. latipalearis with a dark yellow-green apex. They are glabrous except the outer surface of the lobed apex which is velutinous in immature flowers and has globular glands as maturation proceeds. Ratibida mexicana has a vertical row of ciliate hairs on the tube opposite the adaxial edge of the palea. Petal segments are distinguishable by a verti- cal resinous gland from the base of the lobed apex to the base of the tube. STIGMAS. The stigmas are of taxonomic significance due to variation in the length of the stigmatic surface in various species. At anthesis, the stigmas of R. mexicana are green- ish-purple, but they are reddish-purple in all other taxa. PAPPUS. The pappus is a crown or small ring in R. tagetes, awn-like or lacking in R. mexicana and R. pinnata, 1968] Ratibida — Richards 353 and of one to two tooth-like projections in the remaining taxa. ACHENES. The dise achenes are taxonomically important because of the pubescence on the lateral margins. The adaxial and abaxial margins are glabrous, ciliate or pecti- nate-fimbriate with the rest of the surfaces glabrous. The base is indented and the shapes vary from linear-oblanceo- late to oblong-oblique. Ray achenes are triagonal and strigose-hirsute at the apex. GENERIC RELATIONSHIPS Ratibida is a member of the epappose group of the tribe Heliantheae, family Compositae, and is similar to Echinacea and Rudbeckia. It is closely related to Rudbeckia on the basis of columnar receptacles, phyllaries, palea and pappus development. The achenes of Ratibida are flattened later- ally and are triagonal while in Echinacea and Rudbeckia they are four-sided. The phyllaries of Ratibida are in two series of unequal lengths while Echinacea and Rudbeckia have two to three series of about equal length. Palea are absent in the columnar receptacle types of Rudbeckia, but the phyllaries and palea development of Ratibida is similar to Rudbeckia (Dracopsis) amplexicaulis, which some au- thors consider a distinct genus. Based on distribution, morphology, method of seed dis- persal, and the ruderal nature of the species, Sharp (1935) believed Rudbeckia (Dracopsis) amplexicaulis and Ratibida originated from a common ancestor. Anatomically, Ratibida and Rudbeckia (Dracopsis) amplexicaulis are similar in stem structure in that the resin canals are opposite the vascular bundles instead of interfascicular as in Echinacea. Reported haploid chromosome numbers for Ratibida are n = 16 and n = 14 (Jackson, 1959 and Perdue, 1959) ; Rudbeckia has a reported haploid number of 16, and Echinacea has a basic number of n = 11 (Darlington and Wylie, 1955). SYSTEMATIC TREATMENT RATIBIDA Raf., Flor. Ludovic., p. 79. 1817. Rudbeckia Linn., Sp. Pl. 906. 1753, in Part. 354 Rhodora [Vol. 70 Obelisteca Raf., Flor. Ludovic., p. 73, 1817. Lepachys Raf., J. Phys. 89:100. 1819. Obeliscaria Cass., Dict, Natur. Sci. 35:272. 1825. Plants perennial with herbaceous stems. Sterns simple or branched, mostly erect, strigose-hirsute with vertical light green ridges and dark green grooves containing uni- cellular-globular glands. Leaves alternate, entire to pinnate, lyrate-pinnate or pinnatifid, linear to lanceolate or obovate, strigose-hirsute with scattered punctate glands interspersed on both surfaces. Heads globular to columnar, solitary on naked peduncles, receptacle linear or oblong; phyllaries reflexed, strigose-hirsute and in two series, a short ovate inner and a longer linear-lanceolate outer series. Ray flowers neutral, yellow, purplish-yellow or purple, reflexed or upright, ovate-oblanceolate to lanceolate-oblong, outer reflexed surface glabrous, inner strigose to hirsute and punctate, apex 2-3-toothed. Disc corollas tubular, greenish- yellow to purplish-green with slight purple at apex, 5-lobed, lobes ovate, reflexed, 0.3-1.0 mm long, 0.3-0.8 mm wide, inner lobe surface velutinous in immature flower but with punctate glands at anthesis; styles short, flat or filiform, style branches short and stigmatic surface a fourth to half the length of the style branches; stamens with sagittate anthers basifixed to short filaments. Palea conduplicate, apex densely velutinous and incurved, adaxial margins hyaline with some ciliation, a small linear to oblanceolate resin gland present on both surfaces near the adaxial edge, abaxial edge velutinous-hirsute to near the middle and glandular to the base, falling at maturity with the achenes except in R. mexicana, Achenes glabrous except the edges may be ciliate or pectinate-fimbriate, linear-oblanceolate or oblong-oblique, laterally flattened, with a grooved base. Pappus a crown, or of 1-2 teeth, or lacking. Native to North America north of the 20th parallel. Type species. Ratibida pinnata (Vent.) Barnhart, Bull. Torr. Bot. Club. 24:410. 1897. 1968] Ratibida — Richards 355 KEY TO THE SPECIES Plants with fibrous roots, achenes lacking a crown or tooth-like projections. Heads columnar; stems unbranched ................. 1. R. mexicana. Heads globular; stems branched ..............................— 2. R. pinnata. Plants with a taproot, achenes with a crown or tooth-like projections. Stigmatic surface half the length of the style branch; achenes pectinate-fimbriate on one or two edges. Leaves bipinnatifid; achenes pectinate-fimbriate on the edges .... IA EE IDE. EUER ng 3a. R. peduncularis var. peduncularis. Leaves lyrate-pinnate; achenes pectinate-fimbriate along only one edgen -n PS DRE On ee 3b. R. peduncularis var. picta. Stigmatic surface one-fourth the length of the style branch; achenes ciliate or glabrous on one edge. Heads columnar; pappus of tooth-like projections. Rays reflexed; palea longer than broad. Rays wëllten 4a. R. columnifera f. columnif era. Rays purplish-yellow to purple ............. HH EXT o NEIN 4b. R. colummifera f. pulcherrima. Rays upright; palea broader than long ........ 5. R. latipalearis. Heads globular; pappus a Crown seen 6. R. tagetes. 1. RATIBIDA MEXICANA (Wats.) Sharp, Ann. Mo. Bot. Gard. 22:76,77. 1935. Lepachys mexicana Watson, Proc. Am. Acad. Sci. & Arts 203241. 19888. One to several unbranched stems 4.4-9.2 dm tall, pilose-hirsute, spreading above. Leaves ovate-obovate to lanceolate-oblong, basal ones long-petioled, 0.4-2.3 dm long, 1.3-5.6 em wide, serrate or crenu- late, 3-nerved, hirsute below, less so above; upper leaves 2-5 pinnate- parted or cleft, 4.3-16.1 cm long, terminal lobe lanceolate and two-thirds the length of the entire leaf, lateral lobes narrowly oblong or lanceo- late, serrate, 0.5-3.3 cm long, 0.1-1.6 em wide. Peduncles 21.3-61.9 em long. Heads solitary on naked peduncles, columnar, 1.3-3.2 cm long, 1.0-1.2 em wide, broader at the base; receptacle linear, tapering at apex. Phyllaries 7-13, 0.4-2.1 cm long, 1.0-2.5 mm wide. Ray flowers 8-15, yellow, oblong or ovate-obovate, 1.5-4.0 cm long, 1.5-11.0 mm wide, outer surface densely hirsute with numerous dark punctate glands, inner surface glabrous, tube 0.8-1.4 mm long, apex 2-3 lobed, 0.1-3.5 mm long, 0.2-2.8 mm wide. Disc corollas 1.8-2.3 mm long, lobes ovate, 0.8-1.0 mm wide, 0.6-1.0 mm long, tube ciliate along one margin; anthers 1.4-2.0 mm long, filaments 0.7-1.0 mm long; styles 1.3-2.0 mm long, the branches 1.0-1.2 mm long, stigmatic surface light yellow, a fourth the length of the style branch. Palea 3.3-5.0 mm long, 1.7-2.1 mm wide, non-deciduous, adaxial edge glabrous or with an occasional unicellular gland, abaxial edge ciliate; resin gland 356 Rhodora [Vol. 70 Plate 1387 Fig. 1. Ratibida mexicana. ca. 1/3 actual size. Richards 3768 (KANU). 1968] Ratibida — Richards 357 linear, 2.2-3.1 mm long. Achenes oblong, 2.6-4.0 mm long, 1.2-3.0 mm wide, glabrous except for occasional ciliate hairs on adaxial edge and apex, pappus lacking. (Figs. 1, 9, 11d, 12d, & 13d). Flowering time: July-October. Holotype. Mexico. Chihuahua: cool slopes of the Sierra Madre Occidental, Sept. 19, 1887, Pringle 1305 GH; Iso- type NY !, US!). Habitat and Distribution. Pine forests and meadows of the Sierra Madre Occidental, Chihuahua, Durango, and Sonora, Mexico. Ratibida mexicana, occurring at elevations from 6000- 9500 feet, is scattered along the Sierra Madre Occidental of northwestern Mexico and has been infrequently collected. Much or this area is accessible only by foot or horse travel and is sparsely populated. The plants are characterized by the upright ligules, unbranched stems, and undivided basal leaves. The Tarahumar Indians refer to the plant as “Howinowa” and crush the roots to use on legs for rheu- matism. Other Indian tribes use the roots for a wash and potion and the leaves in a beverage for headaches and colds. Representative specimens selected from 37 sheets: MEXICO: CHIHUAHUA: Gentry 2712 (ARIZ, F, GH, MO, UC, US); Richards 3768 (KANU); Richards 3769 (KANU); Townsend & Barber 101 (F, MO, NY, US). DuRANGO: Nelson 4777 (NY, US). SONORA: Phillips 683 (GH, LL). 2. RATIBIDA PINNATA (Vent.) Barnhart, Bull. Torr. Bot. Club. 24:410. 1897. Rudbeckia pinnata Vent., Desc. Nouv. Pl. Jard. Cels. p. 71. 1800. Lepachys pinnatifida Raf., J. Phys. 89:100. 1819. Obeliscaria pinnata (Vent.) Cass, Dict. Sci. Natur. 35:372. 1825. Lepachys pinnata (Vent.) Torr. & Gray, Fl. N. Am. 2:314. 1842. Stems one or many, branched, 3.4-12.3 dm tall, hirsute pubescence above. Leaves orbicular-oblong to ovate-oblanceolate or obovate, lower leaves long-petioled, 1.2-3.9 dm long, 1.0-2.5 em wide, 3-7 pin- nate segments, lanceolate-ovate, upper leaves short petiolate or sessile, 0.1-3.1 dm long, segments 3-9, narrowly lanceolate to ovate, 0.1-1.3 dm 358 Rhodora [Vol. 70 Plate 1388 Fig. 2. Ratibida pinnata. ca. 1/3 actual size. Richards 1152 (KANU). 1968] Ratibida — Richards 359 long, 0.2-3.2 em wide, 3-nerved, serrate or lobed, Peduncles 3.0-26.5 cm long. Heads solitary, globular, 1.1-2.5 cm long, 1.1-1.8 cm wide; receptacle oblong, rounded at apex. Phyllaries 10-14, outer series 0.3-1.1 em long, 1.0-3.0 mm wide, inner 0.3-0.6 cm long, 1.0-3.0 mm wide. Ray flowers 6-13, yellow, linear to oblong-oblanceolate, 1.6-6.0 cm long, 0.4-1.5 em wide, outer surface puberulous-hirsute and with punctate glands, inner surface puberulous, apex 2-3 lobed, lobes 0.2- 9.0 mm long, 0.2-6.0 mm wide. Disc corollas 1.3-3.0 mm long, lobes 0.6-0.9 mm wide, 0.5-1.0 mm long; anthers 1.7-2.3 mm long, filaments 0.6-1.2 mm long; styles 1.0-2.1 mm long, the branches 1.0-2.0 mm long, stigmatic surface half the length of the style branch. Palea 1.2-5.0 mm long, 1.0-1.8 mm wide, adaxial edge hyaline, entire to ciliate, abaxial edge strigose on upper half and unicellular glands on lower half, resin gland linear, 2.2-3.8 mm long. Achenes linear-oblanceolate, glabrous except for occasional ciliate adaxial edge and apex, 2.4-3.0 mm long, 1.3-2.0 mm wide; pappus lacking. (Figs. 2, 10, 11a, 12a, 13a): Flowering time: May-October. Holotype. Illinois: discovered in the country of Illinois by Michaux, grown in Jardin de Cels from seed collected by Michaux (G? Photograph !). Habitat and Distribution. Edge of woodlands, limestone outcroppings, and prairies. Central and eastern United States and southeastern Canada. Ratibida pinnata is distributed in Canada and the United States east of the 98th meridian and is less frequent east of the 85th meridian. Colonies can be found along fence rows, edge of woods, banks and ditches. Characteristics of the species are the numerous globular heads, long, yellow ligules, pinnate leaves with coarsely serrate to lobed seg- ments and many branched stems. Because of the similarity in ligules and heads, R. pinnata has been confused with Rudbeckia. Representative specimens selected from 789 sheets: UNITED STATES: ALABAMA: Choctaw Co.: Sehuchert, in 1896 (NY, US). Madison Co.: Harper 3683 (F, GH, MO, NY, us). Marengo Co.: Mc- Vaugh 8560 (F, GH, TEX). Montgomery Co.: Howard 12918 (ARIZ, F, GH, KANU, MO, NY, OKL, TEX, UC, us). ARKANSAS: Baxter Co.: Palmer 6003 (Mo, US). Benton Co.: Richards 1149 (KANU). Carroll Co.: Richards 1151 (KANU). Clark Co.: Demaree 17799 (F, MO, OKL). Fulton Co.: Demaree 26864 (OK, TEX). Hempstead Co.: Palmer 8037 (MO). Lawrence Co.: Demaree 30999 (KANU, OKL). Marion Co.: 360 Rhodora [Vol. 70 Moore 450809 (TEX). Sharp Co.: Demaree 26894 (KANU, OKL). Wash- ington Co.: Moore, in 1927 (TEX). FLORIDA: Jackson Co.: Wiegand & Manning 3371 (GH). GEORGIA: Bartow Co.: Duncan 8422 (GH, MO, NY, UC). Catoosa Co.: Cronquist 5406 (GH, MO, NY, UC, US). Sumter Co.: Harper 1051 (GH, MO, NY, US). ILLINOIS: Adams Co.: Evers, Jones & Jones 567 (GH, MO, TEX, UC). Carroll Co.: Lodewyks 391 (MO). Champaign Co.: Jones 15307 (MO, TEX, UC). Cook Co.: Smith 6035 (F, GH, MO, UC). Cumberland Co.: Richards 1696 (KANU). Du Page Co.: Umbach 5657 (NY). Fulton Co.: Chase 10567 (F). Han- cock Co.: Mead, Augusta (Nv). Henderson Co.: Patterson, Oquawka (F, MO, NY). Jackson Co.: Gleason 2802 (GH). Kankakee Co., De Selm 543 (F). Lake Co.: Gates 2891 (F). Macon Co.: Clokey 2406 (F, GH, MO, NY, UC, US). Mason Co.: Gleason 9222 (NY). McHenry Co.: Bennett 7971 (F). McLean Co.: Robinson, in 1886 (GH). Peoria Co.: Chase 13497 (F, LL, UC). Pope Co.: Palmer 17003 (Mo). Rich- land Co.: Ridgway 742 (GH). St. Clair Co.: Tesch (NEB). Sangamon Co.: Fuller 1367 (F). Stark Co.: Chase 709 (F, MO). Whiteside Co.: Reecher 219 (F). Will Co.: Swink 2453 (F). Winnebago Co.: Bebb 3508 (MO). INDIANA: DeKalb Co.: Shoop, in 1933 (F). Delaware Co.: Brady, in 1888 (uc). Elkhart Co.: Ek, in 1942 (F, MO). Greene Co.: Springer 825 (KANU, NEB, TEX). Jasper Co.: Welch 1821 (GH, MO). Lagrange Co.: Perdue 2062 (LL). La Porte Co.: Coulter, in 1878 (F, GH). Lawrence Co.: Kriebel 2314 (GĦ). Marion Co.: Floyd, in 1887 (F). Marshall Co.: Stovell & Walton 823 (us). Newton Co.: Friesner 14707 (Ny, UC). Noble Co.: Brooks 2595 (KANU). Porter Co.: Swink 1772 (F). St. Joseph Co.: Coll. unknown (E.B.U.), in 1891 (F). Union Co.: Rose, in 1886 (F). Wabash Co.: Friesner 9006 (F, TEX). Wells Co.: Deam 1370 (MO, NY, US). Iowa: Adams Co.: Richards 1264 (KANU). Audubon Co.: Richards 1296 (KANU). Black Hawk Co.: Burk 556 (Mo). Buena Vista Co.: Richards 1276 (KANU). Carroll Co.: Richards 1272 (KANU). Cass Co.: Richards 1267 (KANU). Cedar Co.: Pammel 712 (GH). Clay Co.: Richards 1277 (KANU). Clayton Co.: Shimek, in 1923 (ny). Dallas Co.: Allen, in 1867 (GH, MO). Decatur Co.: Anderson, in 1904 (Mo). Dickinson Co.: Richards 1281 (KANU). Emmet Co.: Cratty, in 1897 (mo, US). Fayette Co.: Fink 545 (us). Harrison Co.: Fay 3678 (KANU). Iowa Co.: Easterly 757 (ARIZ). Johnson Co.: Loufek, in 1938 (NY). Madison Co.: Shimek, in 1917 (Mo). Mahaska Co.: Augustine 377 (OKL). Monona Co.: Gleason 9324 (NY). Montgomery Co.: Richards 1265 (KANU). Muscatine Co.: Shimek, in 1910 (F, MO). O’Brien Co.: Shimek, in 1929 (NY). Osceola Co.: Richards 1282 (KANU). Pottawattamie Co.: Bates 437 (NEB). Poweshiek Co.: Rohrbaugh 100 (OKL, TEX). Ringgold Co.: Richards 1261 (KANU). Sac Co.: Richards 1274 (KANU). Story Co.: Pammel & Ball 56 (CS, F, GH, MO, NY, US). Taylor Co.: Richards 1263 (KANU). Union Co.: Richards 1262 (KANU). Washington Co.: Wagenknecht 568 (F). 1968] Ratibida — Richards 361 Webster Co.: Somis 3892 (NEB). Winneshick Co.: Tolstead, in 1933 (Mo). Woodbury Co.: Hayden 10783 (uc). KANSAS: Allen Co.: Richards 1137 (KANU). Anderson Co.: Richards 1134 (KANU). Atchi- son Co.: Horr 4468 (GH, KANU). Bourbon Co.: Thompson, Ham- mond (KANU). Brown Co.: Horr 4486 (GH). Cherokee Co.: McGregor 15802 (KANU). Coffey Co.: Richards 1080 (KANU). Doniphan Co.: Horr 4261 (KANU). Douglas Co.: Richards 3766 (KANU). Franklin Co.: Richards 3752 (KANU). Greenwood Co.: Englemann 1033 (OKL). Jefferson Co.: Richards 1255 (KANU). Johnson Co.: McGregor 4381 (GH, KANU). Labette Co.: Rydberg & Imler 310 (KANU, NY). Leavenworth Co.: Richards 1256 (KANU). Linn Co.: Richards 3754 (KANU). Lyon Co.: McGregor 14071 (KANU). Montgomery Co.: Richards 1139 (KANU). Neosho Co.: Richards 3166 (KANU). Shawnee Co.: Norton & Clothier 266 (GH, MO, NY, US). Wabaunsee Co.: McGregor 14073 (KANU). Wilson Co.: Horr, in 1930 (KANU). Woodson Co: Horr & McGregor, in 1947 (KANU). Wyandotte Co.: Richards 3256 (KANU). KENTUCKY: Bullitt Co.: Bishop 386 (GH). Edmondson Co.: Lix, in 1947 (Us). Jefferson Co.: Gunn J-165, 1655 (MO, NY). Logan Co.: Gleason 8929 (NY). Mercer Co.: King 125 (F). Scott Co.: Singer 364 (US). Warren Co.: Price, in 1900 (Mo). LOUISIANA: La Salle Parish: Shinners 23990 (GH, TEX). MASSACHUSETTS: Berkshire Co.: Britton, in 1909 (ny). MICHIGAN: Calhoun Co.: Parmelee 2901 (TEX). Ingham Co.: Parmelee 700 (TEX). Jackson Co.: Camp & Camp, in 1897 (F, US). Kalamazoo Co.: Burgess 378 (F). Kent Co.: Crozien, in 1886 (Ny). Mackinac Co.: Ehlers 281 (GH). Washtenaw Co.: Hermann 9067 (NY). Wayne Co.: Chandler, in 1917 (US). MINNE- SOTA: Cass Co.: Richards 1406 (F). Cottonwood Co.: Richards 1285 (KANU). Dodge Co.: Jeffs, in 1908 (OKL). Hennepin Co.: Sandberg, in 1889 (F, NY, UC). Mower Co.: Brenckle 47-537 (NY). Murray Co.: Richards 1288 (KANU). Nicollet Co.: Ballard, in 1892 (NY, UC, US). Nobles Co.: Richards 1284 (KANU). Pipestone Co.: Richards 1290 (KANU). Ramsey Co.: Churchill, in 1912 (M0). Rice Co.: Hutchin- son 47 (NEB). Rock Co.: Richards 1295 (KANU). Winona Co.: Holzinger, in 1905 (Ny). Mississrppr: Lowndes Co: Diener 934 (NY). Oktibbeha Co.: Tracy, in 1892 (us). Missouri: Barry Co.: Richards 1148 (KANU). Benton Co.: Richards 3340 (KANU). Butler Co.: Eby, in 1893 (Mo). Camden Co.: Chandler 4731 (MO). Carter Co.: Palmer 6204 (MO, US). Cass Co.: Richards 3312 (KANU). Clay Co.: Mackenzie 262 (Mo, NY). Dallas Co.: Richards 1162 (KANU). Douglas Co.: Richards 3401 (KANU). Franklin Co.: Kel- logg 1283 (MO). Greene Co.: Standley 9622 (Us). Henry Co.: Hitchcock, Windsor Springs (Mo). Hickory Co.: Richards 1163 (KANU). Jackson Co.: Bush 164 (MO, NY, OKL, US). Jasper Co.: Richards 3437 (KANU). Jefferson Co.: Steyermark 1036 (MO). Johnson Co.: Richards 1167 (KANU). Laclede Co.: Richards 1160 362 Rhodora [Vol. 70 (KANU). Lincoln Co.: Davis 2732 (Mo). Livingston Co.: Sparling 111 (ri. Marion Co.: Davis 7260 (Mo). Mercer Co.: Palmer & Steyermark 41291 (GH, MO). Nodaway Co.: Richards 1258 (KANU). Ozark Co.: Richards 1156 (KANU). Pettis Co.: Richards 1166 (KANU). Phelps Co.: Steyermark 72173 (F). Pulaski Co.: Dixon, in 1928 (Mo). Ralls Co.: Davis 4645 (MO, TEX). St. Clair Co.: Bush 12797 (MO). St. Francois Co.: Bower 17 (F). St. Louis Co.: Sherff 68 (GH). Saline Co.: Marsh 1629 (NY). Shannon Co.: Bush 6147 (MO). Stone Co.: Richards 3415 (KANU). Taney Co.: Richards 3387 (KANU). Wright Co.: Richards 3347 (KANU). NEBRASKA: Cass Co.: Richards 1308 (KANU). Dakota Co.: Richards 1306 (KANU). Dodge Co.: Richards 1307 (KANU). Douglas Co.: Cleburne, Omaha (NEB). Nemaha Co.: Webber 6086 (NEB). Otoe Co.: Richards 1309 (KANU). Sarpy Co.: Prind & Saunders 3126 (NEB). Saunders Co.: Williams, Ashland (us). NEW York: Chautaqua Co.: Sartwell (GH). Yates Co.: Sartwell (F). Onto: Adams Co.: Ste- phenson, in 1930 (MO). Delaware Co.: Crane 2447 (NY). Erie Co.: Moseley, in 1895 (GH, US). Franklin Co.: Condit, in 1909 (UC). Greene Co.: Demaree 11477 (MO, US). Ross Co: Bartley & Pontius 49 (NY). OKLAHOMA: Adair Co.: Waterfall 9676 (KANU). Cherokee Co.: Teachers College, in 1924 (OKL). Craig Co.: Stevens 2207 (GH). McCurtain Co.: Clark 2946 (OKL). Muskogee Co.: Waterfall 9514 (KANU). Nowata Co.: Richards 1144 (KANU). Ottawa Co.: Stevens 2283 (GH, MO, NY, OKLA, US). SovTH DAKOTA: Clay Co.: Richards 1805 (KANU). Codington Co.: Johnson 78 (GH, TEX). Minnehaha Co.: Richards 1299 (KANU). Roberts Co.: Johnson (MO, NY). TEN- NESSEE: Davidson Co.: Ward, in 1877 (us). Franklin Co.: Eggert, in 1898 (Mo, NY). Lewis Co.: McDougall 1411 (US). Marshall Co.: Bain, in 1892 (vs). Maury Co.: Shimek, in 1891 (F, MO). Meigs Co.: Sharp & Underwood 2331 (Mo, UC). Rutherford Co.: Demaree 45751 (KANU). Warren Co.: Coll. unknown 2639 (Us). Wilson Co.: Svenson 9588 (GH, UC). VERMONT: Chittenden Co.: Wright 2040 (MO). WISCONSIN: Brown Co.: Schuette, in 1884 (F, GH, NY, US). Buffalo Co.: Fassett & Wilson 4467 (GH). Dane Co.: Seymour 1084 (KANU). Door Co.: Millspauth 3691 (F). La Crosse Co.: Hartley 997 (US). Milwaukee Co.: Hasse (UC). Outagamie Co.: Melchert, in 1960 (TEX). Sauk Co.: Eggert, in 1903 (Mo). Walworth Co.: Shannon 133a (MO). Winnebago Co.: Kellerman (us). CANADA: ONTARIO: Macoun 34829 (GH, MO, NY, US). 3. RATIBIDA PEDUNCULARIS (Torr. & Gray) Barn- hart, Bull. Torr. Bot. Club. 24:411. 1897. Stems one or many, branched, 2.8-11.2 dm tall, thickly hirsute near the base. Leaves oblong-obovate, bipinnatifid to lyrate-pinnate, 1.6- 13.7 em long, 1.0-6.4 em wide, leaf divisions 3-10, 0.3-6.0 em long, lobes 0.8-4.3 mm wide. Peduncles 14.8-45.5 em long. Heads solitary, 1968] Ratibida — Richards 363 Plate 1389 Fig. 3. Ratibida peduncularis var. peduncularis. ca. 1/3 actual size. Kickurds 3742 (KANU). 364 Rhodora [Vol. 70 columnar, 1.7-6.9 cm long, 0.8-1.3 cm wide; receptacle linear, tapering at the apex. Phyllaries 7-14, outer series 0.6-1.2 cm long, 0.8-1.3 mm wide, inner 0.1-0.5 cm long, 0.9-1.2 mm wide. Ray flowers 6-13, linear- oblong to oval-orbicular, yellow, purplish-yellow and purple, 4.0-13.5 min long, 3.0-10.0 mm wide, outer surface hirsute and with unicellular- globular glands, inner glabrous, tube 1.0 mm long, apex 2-3 lobed, lobes 0.2-6.0 mm long, 0.2-3.0 mm wide. Disc corollas 1.5-2.5 mm long, the lobes 0.6-0.8 mm wide, 0.5-0.8 mm long; anthers 1.3-2.0 mm long, filaments 0.5-1.0 mm long; styles 1.0-2.0 mm long, the branches 0.5-2.0 mm long, stigmatic surface half the length of the style branches. Palea 3.0-4.3 mm long, 1.5-3.0 mm wide, adaxial side ciliate, abaxial edge ciliate half way down and with unicellular glands to base, resin gland oval, 1.0-1.8 mm long. Achenes oblique- oblong, 2.3-5.0 mm long, 1.2-3.0 mm wide, margin pectinate-fimbriate, edges winged; pappus of one to two teeth. 3a. R. PEDUNCULARIS (Torr. & Gray) Barnh. var. PEDUNCULARIS. Lepachys peduncularis Torr. & Gray, Flora N. Am. 2:315. 1842. Obeliscaria peduncularis (Torr. E Gray) Walp., Repert. 2:979. 1843. Leaves bipinnatifid, ovate-oblong, 2.1-11.2 em long, 1.1-4.0 cm wide, segments 4-10, linear-lanceolate to ovate-obovate, 0.1-3.2 cm long, 0.1-1.1 em wide; hairs on upper part of stem with raised base. Achenes pectinate-fimbriate on both edges. Ray flowers yellow to purplish- yellow. (Figs. 3, 10, 11g, 12g, 13g). Flowering time. March-November. Lectotype. Texas: S. Felipe de Austin, in 1835, Drum- mond 108 (GH!). Since Torrey and Gray did not select a type from the original material, the above lectotype was selected by the author. Habitat and Distribution. Sandy coastal plains of south- eastern Louisiana, Texas, and northeastern Tamaulipas, Mexico. Ratibida peduncularis var. peduncularis occurs mainly on the coastal plains of Louisiana, Texas and Mexico. The coastal forms are usually characterized by yellow or pur- plish-yellow ligules while those more inland have purplish- yellow ligules. Typically the plants have dissected leaves, long peduncles, anc achenes with pectinate-fimbriate mar- gins. 1968] Ratibida — Richards 365 —9 | a Plate 1390 Fig. 4. Ratibida peduncularis var. picta. ca. 1/3 actual size. Rich- ards 3732 (KANU). 366 Rhodora [Vol. 70 Representative specimens selected from 69 sheets: UNITED STATES: LOUISIANA: Cameron Parish: Tracy 8548 (F, MO, NEB, NY, TAES). TEXAS: Aransas Co.: Richards 3739 (KANU). Bexar Co.: Schulz 422 (TEX, US). Brooks Co.: Richards 3744 (KANU). Calhoun Co.: Parks & Cory 11644 (TAES). Cameron Co.: Runyon 441 (US). Dimmit Co.: Hamby 950 (LL). Galveston Co.: Wright (Us). Goliad Co.: Beamer, in 1953 (KANU). Hidalgo Co.: McKelvey 1751 (GH). Jim Hogg Co.: Lehmann & Davis 38 (F). Kenedy Co.: Richards 3741 (KANU). Kleberg Co.: Johnston 54422 (TAES, TEX). Nueces Co.: Heller 1789 (GH, MO, NY, UC, US). Orange Co.: Small & Wherry 11802 (NY, US). San Patricio Co.: Cory 49140 (GH). Willacy Co.: Richards 3743 (KANU). MEXICO: TAMAULIPAS: LeSueur 509 (F, TEX). ob. R. PEDUNCULARIS (Torr. & Gray) Barnh, var. PICTA (Gray) Sharp, Ann. Mo. Bot. Gard. 22:74. 1935. Lepachys peduncularis Torr. & Gray var. picta Gray, Smiths. Contr. Knowl. (Pl. Wright.) 3:107. 1852. Lepachys serratus Buckl., Acad. Natur. Sci. Phila. 13:457. 1861. Ratibida picta (Gray) Small, Flora S. E. United States, ed. 1, 1260 p. 1903. Leaves lyrate-pinnate, ovate-obovate or oblong, 1.6-13.7 em long, 1.0-6.4 em wide, segments 3-7, terminal lanceolate-ovate, toothed to parted, 0.6-7.2 em long, 0.3-5.0 cm wide, lateral segments linear-oblong to obovate, 0.1-4.5 em long, 0.1-1.4 em wide; long pilose and hirsute hairs on lower stem becoming shorter and more numerous on upper stem with the longer hairs strigose, punctate glands mostly few or absent. Achenes pectinate-fimbriate on adaxial edges, Ray flowers purplish-yellow to purple. (Figs. 4, 10, 11e, 12e, 13e). Flowering time: April-October. Lectotype. Texas: Gulf Coast near Galveston, in 1849, Wright (GH!). Since Gray did not select a type from the original material, the above lectotype was selected by the author. Habitat and Distribution. Sandy seacoast and wooded areas of the Texas coastal region. Ratibida peduncularis var. picta is limited in distribution to the seacoast and inland coastal woods of Texas. This variety is recognized because of the lyrate-pinnate leaves and the pectinate-fimbriate adaxial margin of the achenes. 1968] Ratibida — Richards 367 The achenes are usually more narrow and the unicellular globular glands on the stems and peduncles are less frequent than in R. peduncularis var. peduncularis. The range over- laps with that of R. columnifera and the two species readily hybridize (Jackson, 1965). Ratibida peduncularis var. picta probably hybridizes with R. peduncularis var. peduncularis where their ranges over- lap. Collections seem to indicate this as plants intermediate in leaf and achene characteristics and usually more robust than either of the above have been found (Richards 3745, KANU). Representative specimens selected from 52 sheets: UNITED STATES: Texas: Aransas Co.: Richards 3737 (KANU). Bexar Co.: Richards 3735 (KANU). Calhoun Co.: Tharp 5639 (TEX, US). Frio Co.: Higdon s.n. (TEX). Gonzales Co.: Richards 3730 (KANU). Guadalupe Co.: Richards 3731 (KANU). Harris Co.: Thurrow, in 1890 (F). Medina Co.: Richards 3745 (KANU). Victoria Co.: Tracy 8941 (F, GH, MO, NEB, NY, TAES, TEX). Waller Co.: Hall 336 (F, MO, NY, US). Wilson Co.: Richards 3734 (KANU). 4. RATIBIDA COLUMNIFERA (Nutt.) Woot. & Standl. Contrib. U. S. Nat. Herb. 19:706. 1915. Stems one or many, branched, 2.2-10.4 dm high. Leaves ovate- obovate to orbicular-oblanceolate, 1.8-15.1 cm long, 0.8-6.0 cm wide, bipinnatifid, segments 3-14, narrowly linear-lanceolate to ovate-oblong, 0.1-6.0 em long, 0.1-1.6 em wide. Peduncles 1.4-47.8 cm long. Heads solitary, columnar, 1.0-4.7 em long, 0.7-1.2 cm wide. Phyllaries 5-14, the inner row lanceolate-ovate, 0.8-2.0 mm long, 0.8-1.8 mm wide, outer linear, 0.4-1.4 em long, 0.5-2.0 mm wide. Ray flowers 4-11, yellow, purplish-yellow to purple, ovate-oblanceolate to oblong-ellipti- eal, 0.7-3.5 em long, 0.4-1.7 cm wide, inner surface glabrous to puberu- lent, outer surfaces hirsute, punctate, apex 2-3 lobed, lobes 0.1-5.3 mm long, 0.2-7.1 mm wide. Disc corollas 1.6-2.8 mm long, lobes 0.5-1.0 mm wide, 0.3-0.8 mm long; anthers 1.2-1.9 mm long, filaments 0.4-1.0 mm long; styles 0.8-1.5 mm long, the branches 0.5-1.4 mm long, stigmatic surface a fourth the length of the branch. Palea 2.3-3.5 mm long, 0.5-3.0 mm wide, adaxial edge hyaline, ciliate, abaxial edge strigose from apex to middle, unicellular glands from middle to base; resin gland oval-oblanceolate, 0.6-1.5 mm long. Achenes oblong, 1.2-3.0 mm long, 1.2-2.0 mm wide, glabrous except for the ciliate adaxial edge and ciliate or squamellous apex; pappus of 1-2 teeth or absent. 368 Rhodora [Vol. 70 Plate 1391 Fig. 5. Ratibida columnifera f. columnifera. ca. 1/3 actual size. Richards 1019 (KANU). 1968] Ratibida — Richards 369 4a. RATIBIDA COLUMNIFERA (Nutt.) Woot. & Standl. forma COLUMNIFERA. Rudbeckia columnifera Nutt., Fraser's Catal. 1815. Rudbeckia columnaris Pursh, Flora Am. Sept. 2:575. 1814. Ratibida sulcata Raf. J. Phys. 89:100, 1819. Obeliscaria columnaris DeCandolle Prodr. 5:559. 1836. Ratibidia columnaris (Sims) D. Don, Sweet's Brit. FI. Gard. 4:361. 1838. Lepachys columnaris Torr. & Gray, Flora N. Am. 2:315. 1842. Lepachys columnifera (Nutt.) MacBride, Contrib. Gray Herb. New Ser. 3(65):45, 1922. Ratibida columnifera (Nutt.) Woot. & Standl. var. bre- viradiata Cockerell, Am. Natur. 49(586) :620. 1915. Ratibida columnifera (Nutt.) Woot. & Standl. var. incisa Cockerell, Am. Natur. 49(586) :620. 1915. Ratibida columnifera (Nutt.) Woot. & Standl. var. tubu- laris Cockerell, Am. Natur. 49(586) :620. 1915. Ratibida columnifera (Nutt.) Woot. & Standl. var. ap- pendiculata Cockerell, J. Hered. 7:428. 1916. Ratibida columnaris (Pursh) Raf. forma denudata Boi- vin, Natur. Can. 87:49. 1960. Flowering time. March-November. Holotype. As far as can be ascertained, there are no known types. Evidence indicates that the plants were first collected from the “country of the Missourie", 1813. Nut- tall. Habitat and Distribution. Prairies and foothills. South- ern Canada, United States, and northern Mexico. Rudbeckia columnifera Nutt. was first published in Fraser’s Catalogue in 1813. While there is some dispute as to the authorship and legality of the Catalogue, I am accepting Nuttall as the author, at least of the plants ac- companied with descriptions. This opinion is based on my studies of several of Nuttall’s works plus additional papers favoring him as the author. These are as follows: Dress (1961), Cronquist et al (1956), Fernald (1938), Britten 370 Rhodora [Vol. 70 (1899) and Greene (1890). Britten (1899) states that the list of names in the Catalogue was written by Nuttall who added his name in ink to the copy from which Professor Greene’s reprint was made. Greene (1890) says: “Nut- tall wrote his name in ink as the author and even made a few corrections on the copy of the catalogue which was used for the reprint in Pittonia." Nuttall (1818) refers 22 times to his names in “Fraser’s Catalogue", Nuttall (1841), writing on Aplopappus spinulosus, states: “to which I applied the name of Sideranthus in “Fraser's Cata- logue.” Sims (1814) used plants grown at Fraser's nursery and credits Nuttall as introducing them from the “country of the Missourie". Pursh (1814) cites Rudbeckia columni- fera Nutt. as a synonym for Rudbeckia columnaris. While the description of Rudbeckia columnifera in the catalogue is meager, Fernald (1938) sums it up by saying: "certainly this is a far more intelligible diagnosis of a species than was the original characterization of Ratibida for a genus." Cronquist et al (1956) recognized the names with descrip- tions in “Fraser's Catalogue" as being validly published. The author accepts Ratibida columnifera (Nutt.) Woot. and Standl. as the legitimate name. Cockerell (1915) described three varieties of Ratibida columnifera from Boulder, Colorado, as follows: var. breviradiata (US) with short rays, 10 mm long; var. incisa (US) with emarginate and cleft rays; var. tubularis with rays 25 mm long and 3.5 mm wide, completely quilled, being hollow cylinders. The three varieties named by Cockerell have all been observed over the range of the species by me and are not recognized as varieties, although their occur- rence may be uncommon. Aberrant and teratological forms of many kinds have been observed and it is my opinion that they should not be afforded taxonomic rank. Cockerell (1916) described Ratibida columnifera (Nutt.) Woot. & Standl. var. appendiculata (GH) with the rays possessing long appendages (usually paired) arising from the throat. Similar aberrant forms have been observed by the author and this variety is not recognized in this study. 1968] Ratibida — Richards SAL Boivin (1960) described Ratibida columnaris (Pursh) Raf. forma denudata (Boivin & Alex 9886 DAO) where the head was devoid of ray flowers. Even though this condi- tion apparently is not a common occurrence in the above species, it is not recognized in this study. Ratibida columnifera is the most widespread species of the genus and exhibits many morphological forms over its range. Plants of the prairies are characterized by being tall and robust with long, columnar heads while those of drier areas as well as those growing later in the season tend to be shorter and have shorter heads. In sandy areas the plants are more annual, shorter, and less branched, In the southern part of the range, the ligules tend to be some- what shorter as are those growing later in the season and in dry areas. In ruderal areas, the plants may assume a “bushy” appearance with much branching and short heads. Typically, the species is characterized by the columnar head, yellow ligules, and bipinnatifid leaves. Representative specimens selected from 1,489 sheets: UNITED STATES: ARIZONA: Apache Co.: Palmer 601 (GH, NY, US). Coco- nino Co.: MacDougal 416 (GH, KANU, NY, US). Navajo Co.: Mayer- hoff 59(F). ARKANSAS: Crittenden Co.: Demaree 11380 (MO). Washington Co.: Palmer 8216 (MO, NY). COLORADO: Adams Co: Johnston 217A (MO). Arapahoe Co.: Clokey 2819 (MO). Baca Co.: Harrington 3311 (cS). Boulder Co.: ‘Hanson C139 (ARIZ, MO). Chaffee Co.: Sheldon 472 (us). Cheyenne Co.: Christ 983 (cs). Denver Co.: Clokey 2819 (ARIZ, F, GH, UC). Douglas Co.: Christ 1145 (cs). Elbert Co.: Woodward, in 1882 (GH). El Paso Co.: Ehlers 7717 (ARIZ, GH). Fremont Co.: Brandegee 757 (NY). Jefferson Co.: Cleburne, in 1914 (NEB). Kiowa Co.: Harrington 4263 (cs). Kit Carson Co.: Preston, Vona (cs). Larimer Co.: Marshall 2985 (cs, NEB, US). Las Animas Co.: Rogers 4428 (us). Lincoln Co.: Har- rington 792 (cs). Phillips Co.: Harrington 2255 (CS). Prowers Co.: Harrington 3480 (cs). Pueblo Co.: Mathias 526 (MO). Washington Co.: Shantz 1133 (Us). Weld Co.: Johnston 217 (MO). Yuma Co.: Harrington 4242 (cS). IDAHO: Bannock Co.: Sharpe 17204 (NY). Power Co.: Christ & Ward 15030 (Ny). ILLINOIS: Cook Co.: Gates 1221a (F). Du Page Co.: Umbach (F, US). Madison Co.: Eggert, in 1891 (Mo). Menard Co.: Hall, in 1861 (F). Peoria Co.: Chase 8923 (KANU, OKL). INDIANA: Clark Co.: Lansing 981 (F). Iowa: Clay Co.: Hayden 10784 (NY). Dickinson Co.: Shimek, Milford (NEB). Fremont Co.: Harms 616 (KANU). Linn Co.: Smith, in 1871 372 Rhodora [Vol. 70 (NY). Lyon Co.: Gleason 9370 (NY). Muscatine Co.: Shimek, in 1915 (Mo). Polk Co.: Hayden 10546 (MO, NEB, NY, TEX). Potta- watomie Co.: Shimek, in 1909 (F, NY). Sioux Co.: Hayden 10547 (GH, MO, NY, UC). Webster Co: Somis 3893 (NEB). Woodbury Co.: Thorne 13366 (US). KANSAS: Anderson Co.: Richards 1133 (KANU). Barber Co.: Richards 792 (KANU). Barton Co.: Richards 876 (KANU). Bourbon Co.: Thompson (KANU). Butler Co.: Richards 1086 (KANU). Chase Co.: Richards 1082 (KANU). Chautauqua Co.: Hulbert 3162 (KANU). Cheyenne Co.: Horr 4692 (GH, KANU). Clark Co.: Richards 835 (KANU). Clay Co.: McGregor 5015 (KANU). Cloud Co.: Kellerman, in 1888 (MO). Coffey Co.: Richards 1081 (KANU). Comanche Co.: Richards 822 (KANU). Cowley Co.: Richards 1087 (KANU). Decatur Co.: Richards 2835A (KANU). Dickinson Co.: Hulbert 3164 (KANU). Douglas Co.: Richards 3765 (KANU). Edwards Co.: McGregor 10556 (KANU). Elk Co.: Richards 687 (KANU). Ellis Co.: Richards 1074 (KANU). Ellsworth Co.: Richards 1076 (KANU). Finney Co.: Coville 25 (us). Ford Co.: Richards 3763 (KANU). Franklin Co.: Richards 3753 (KANU). Geary Co.: Richards 1077 (KANU). Gove Co.: Horr & McGregor 3262 (GH, KANU). Graham Co.: Richards 1071 (KANU). Grant Co.: Thompson 15 (F, GH, MO, NY, UC, US). Greeley Co.: Rose & Fitch 17076 (NY, US). Greenwood Co.: Richards 654 (KANU). Hamilton Co.: Richards 865 (KANU). Harper Co.: Richards 772 (KANU). Harvey Co.: McGregor 12515 (KANU). Hodgeman Co.: McGregor 10597 (KANU). Jefferson Co.: Marsh 1615 (KANU, US). Jewell Co.: Horr 4363 (KANU). Kearney Co.: Richards 871 (KANU). Kingman Co.: MeGregor 10708 (KANU). Kiowa Co.: Horr 3989 (KANU). La- bette Co.: Richards 1140 (KANU). Lincoln Co.: McGregor 7444 (KANU). Linn Co.: Richards 3755 (KANU). Logan Co.: McGregor 9405 (KANU). Lyon Co.: McGregor 1868 (KANU). Marion Co.: Richards 1084 (KANU). Marshall Co.: Heller 13975 (Mmo). Mc- Pherson Co.: Richards 879 (KANU). Meade Co.: Richards 844 (KANU). Miami Co.: Oyster 4149 (F, MO, NY). Montgomery Co.: Rydberg & Imler 366 (KANU, MO). Morris Co.: Horr, in 1930 (KANU). Morton Co.: Richards 2966 (KANU). Neosho Co.: Richards 8182 (KANU). Norton Co.: Richards 2793 (KANU). Osborne Co.: Shear 121 (GH, NEB, US). Pawnee Co.: Richards 874 (KANU). Phillips Co.: Richards 2740 (KANU). Rawlins Co.: MeGregor 9470 (KANU). Republie Co.: Morley 241 (KANU, UC). Rice Co.: Richards 877 (KANU). Riley Co.: Gayle 526 (NY). Rooks Co.: Horr 4987 (KANU). Rush Co.: Lathrop 3845 (F, KANU, NY). Russell Co.: Richards 1075 (KANU). Scott Co.: Gates 16692 (MO). Sedgwick Co.: Poole 217 (GĦ). Seward Co.: Rydberg & Imler 866 (KANU, NY). Sheridan Co.: Richards 1069 (KANU). Smith Co.: Richards 2888 (KANU). Stafford Co.: Richards 875 (KANU). Stanton Co.: Richards 861 (KANU). Trego Co.: Richards 1073 (KANU). Wallace 1968] Ratibida — Richards 373 Co.: Letterman (Mo). Washington Co.: Horr 4620 (GH, KANU). Wichita Co.: Agrelius & Agrelius, in 1912 (KANU). Wilson Co.: McGregor 14374 (KANU). Woodson Co.: Richards 603 (KANU). Wyandotte Co.: Isely 2604 (NY, uc). MASSACHUSETTS: Middlesex Co.: Jenks, in 1922 (GH). MICHIGAN: Wayne Co.: dePaul (KANU). MINNESOTA: Big Stone Co.: Gleason 9405 (NY). Chippewa Co.: Moyer, in 1897 (ARIZ, UC). Clay Co.: Dewart (Mo), Goodhue Co.: Sandberg 566 (F, US). Hennepin Co.: Aiton, in 1892 (F, GH, NY, uc). Murray Co.: Richards 1287 (KANU). Otter Tail Co.: Chan- donnet, in 1911 (GH). Pipestone Co.: Anderson, in 1987 (uc). Polk Co.: Johnson 520 (GH, NY). Pope Co.: Taylor, in 1891 (uc). Rock Co.: Richards 1296 (KANU). Traverse Co.: Sheldon, in 1893 (GH, MO, UC, US). Washington Co.: Churchill (NEB). MISSOURI: Greene Co.: Standley 1533 (us). Jackson Co.: Bush 163 (MO, OKL, NY, UC, us). Jasper Co.: Palmer 23389 (GH, Ny). Lincoln Co.: Davis 1390 (mo). Taney Co.: Steyermark, in 1938 (F). MONTANA: Broadwater Co.: Hermann 12323 (us). Cascade Co.: Cantrell, in 1947 (OKL). Dawson Co.: Ward, in 1883 (us). Gallatin Co.: Blankinship 283 (F, KANU, MO, US). Jefferson Co.: Cronquist 6778 (NY). Lewis & Clark Co.: Kelsey, in 1888 (F, US). Madison Co.: Booth, in 1941 (KANU). Mineral Co.: Cronquist 6739 (KANU, NY, TAES, TEX). Mis- soula Co.: Barkley 1686 (Mo, vC). Park Co.: Scheuber 56 (NY). Rosebud Co.: Bennett, in 1957 (F, UC). Sheridan Co.: Larsen 112 (GH, MO). Stillwater Co.: Swingle (CS). Sweetgrass Co.: Hitchcock & Muhlick 13310 (GH, MO, NY, UC). Wheatland Co.: Spragg 498 (cs). Yellowstone Co.: House 4904 (US). NEBRASKA: Adams Co.: Pammel 232 (Mo). Antelope Co.: Webber 6087 (NEB). Blaine Co.: Richards 1019 (KANU). Box Butte Co.: Williamson (NEB). Brown Co.: Weinard (NEB). Buffalo Co.: Richards 1014 (KANU). Butler Co.: Cleburne (NEB). Cass Co.: Morrison 1187 (MO, NEB). Cedar Co.: Clements 2a44 (US). Cherry Co.: Tolstead 450 (GH, NEB). Custer Co.: Richards 1016 (KANU). Dawes Co.: Webber 6201 (NEB). Dodge Co.: Engberg (NEB). Douglas Co.: Benke 4314 (F, US). Franklin Co.: Bates 2955 (NEB). Frontier Co.: Richards 1028 (KANU). Furnas Co.: Davidson, Dixit & Romberg 4678 (NEB, OKL). Gage Co.: Washburn, in 1899 (vs). Hall Co.: Richards 1012 (KANU). Hamilton Co.: Richards 1010 (KANU). Jefferson Co.: Rohbraugh 11 (0KL). Kearney Co.: Hapeman, in 1893 (vs). Keith Co.: Werner (NEB). Keya Paha Co.: Lawton 35 (F). Kimball Co.: Moldenke 7228 (NY). Knox Co.: Wilcox, in 1888 (Nv). Lancaster Co.: Webber 6090 (MO, NEB). Lincoln Co.: Richards 1026 (KANU). Logan Co.: Richards 1022 (KANU). Morrill Co.: Bates (NEB). Nance Co.: Osborn 1136R (Mo). Pierce Co.: Harms 582 (KANU). Polk Co.: Turrell (Ariz). Red Willow Co.: Richards 1029 (KANU). Saunders Co.: ‘Williams 296 (US). Scotts Bluff Co.: Rydberg 182 (NEB, NY). Sheridan Co.: Sandoz 172 (NEB). Sioux Co: Woods 374 Rhodora [Vol. 70 354 (NEB). Thomas Co.: Richards 1021 (KANU). Webster Co.: Bates (NEB). York Co.: Sperry (TAES). NEW JERSEY: Hudson Co.: VanSickle, in 1895 (us). NEW MEXICO: Catron Co.: Eggleston 20242 (US). Colfax Co.: Lucas 194 (TEX, UC). Curry Co.: Clark 15241 (UNM). Eddy Co.: Grassi 88 (F). Guadalupe Co.: Richards 1110 (KANU). Harding Co.: Sands (UNM). Lincoln Co.: Wooton 261 (MO). Mora Co.: Richards 3774 (KANU). Otero Co.: Richards 1121 (KANU). Quay Co.: Richards 1108 (KANU). Sandoval Co.: Ellis 207 (us). San Miguel Co.: Richards 1114 (KANU). Santa Fe Co.: Standley 5279 (MO). Taos Co.: Marcelline 2864 (F). Union Co.: Clark 16179 (UNM). NORTH DAKOTA: Barnes Co.: Barber 154 (GH). Benson Co.: Rider 452 (F). Cass Co.: Alexander (uc). Grand Forks Co.: Brannon 213 (MO). Grant Co.: Bell 604 (vc). Kidder Co.: Rollins & Munoz 2811 (GH, US). McLean Co.: Rudd (OKL). Morton Co.: Sarvis 99 (us). Richland Co.: Metcalf 176 (GH). Stutsman Co.: Mearns 167 (US). Ward Co.: Larsen 187 (MO, NY). Williams Co.: Bell 112 (UC). OKLAHOMA: Bryan Co.: Taylor, Silver & Fox 1230 (OKL). Caddo Co.: Graber 63 (OKL). Canadian Co.: Taylor 263 (OKL). Carter Co.: Richards 543 (KANU). Cherokee Co.: Wallis 1402 (UC). Cimarron Co.: Rogers 4664 (US). Cleveland Co.: Eskew 1007 (MO, OKL). Comanche Co.: Demaree 13136 (MO, NY, OKL). Cotton Co.: Waterfall 7277 (OKL). Craig Co.: Richards 1145 (KANU). Creek Co.: Bush 307 (Mo). Custer Co.: Mericle 460 (OKL). Dewey Co.: Richards 3751 (KANU). Ellis Co.: Richards 1096 (KANU). Garfield Co.: Richards 1091 (KANU). Garvin Co.: Richards 536 (KANU). Grady Co.: Warner 30 (OKL). Grant Co.: Richards 1090 (KANU). Greer Co.: Bull, in 1928 (OKL). Har- mon Co.: Stevens 1079 (GH). Harper Co.: Stratton 399 (MO). Hughes Co.: McGregor 13897 (KANU). Johnston Co.: Alder, in 1950 (OKL). Kay Co.: Richards 1088 (KANU). Kingfisher Co.: Blankinship (Gu, MO, US). Latimer Co.: McGregor 18894 (KANU). Logan Co.: Keyser 6014 (NY, OKL). Love Co.: Richards 545 (KANU). Major Co.: Richards 1094 (KANU). Marshall Co.: Duff 93 (ARIZ). McClain Co.: Hopkins 13355 (OKL). Mc- Curtain Co.: Poole 15 (KANU). McIntosh Co.: McGregor 14346 (KANU). Murray Co.: Richards 540 (KANU). Muskogee Co.: Bebb 4437 (GH, OKL). Noble Co.: McGregor 13914 (KANU). Nowata Co.: Richards 1143 (KANU). Oklahoma Co.: Waterfall 1397 (NY, OKL). Osage Co.: Stevens 1952 (GH). Payne Co.: McLean 132 (ARIZ). Pittsburg Co.: McGregor 13896 (KANU). Pontotoc Co.: Richards 535 (KANU). Pottawatomie Co.: Faulkner 131 (OKL). Roger Mills Co.: Engleman 1630 (OKL). Rogers Co.: Hubricht, Shoop & Heinze 1485 (M0). Seminole Co.: Robbins 3082 (OKL, TAES, UC). Texas Co.: Harms 1971 (KANU). Tillman Co.: Fleming 120 (oKz). Tulsa Co.: Richards 3723 (KANU). Washington Co.: Richards 3722 (KANU). Woods Co.: Stevens 764 (GH, MO, OKL, US). Woodward 1968] Ratibida — Richards 375 Co.: Locke 32 (us). SOUTH DAKOTA: Brookings Co.: Thornber, in 1893 (ARIZ, MO). Custer Co.: Rydberg 802 (NEB, US). Day Go Brenckle 39-78 (NY, US). Fall River Co.: Hayward 517 (F). Harding Co.: Visher 374 (F). Hutchinson Co.: Richards 1302 (KANU). Jack- son Co.: Palmer 37623 (GH, MO, NY). Kingsbury Co.: Thornber, in 1894 (GH, Mo). Lawrence Co.: Carr 93 (F, NEB, NY, US). McCook Co.: Richards 1301 (KANU). Meade Co.: Forwood 218 (Us). Min- nehaha Co.: Richards 1298 (KANU). Pennington Co.: Visher, in 1908 (F). Roberts Co.: Powell, in 1903 (GH). Spink Co.: Ricksecker 44 (uc). Turner Co.: Johnson 77 (GH, MO, NY). Washabaugh Co.: Visher 2033 (F). Yankton Co.: Richards 1303 (KANU). TEXAS: Aransas Co.: Uzzell 37 (LL). Armstrong Co.: Reverchon 3339 (MO). Bell Co.: Wolff 892 (TAES, US). Bexar Co.: Groth 95 (F, GH). Brazoria Co.: Correll & Edwin 16450 (LL). Brewster Co.: Mueller 8216 (GH, MO, NY, TEX). Cameron Co.: Runyon 2638 (F, US). Childress Co.: Blake 7399 (TEX). Collinsworth Co.: Correll & John- ston 16901 (LL). Crockett Co.: Warnock 15176 (LL). Culberson Co.: Moore & Steyermark 3618 (GH, MO, NY, UC, US). Dallam Co.: Gorder 116 (LL, TEX). Dallas Co.: Reverchon 493 (F). Eastland Co.: Ruth 1314 (NY). Erath Co.: Gould 5615 (ARIZ, TAES). Galveston Co.: Traverse 1651 (TEX). Gillespie Co.: Jermy (mo). Gray Co.: Richards 1104 (KANU). Hall Co.: Fairview School 6515 (TEX). Harris Co.: Reverchon 3987 (Mo). Hartley Co.: Cory & Parks 16466 (TAES). Hemphill Co.: Richards 1100 (KANU). Hutchinson Co.: Thornton 52-430 (TEX). Jeff Davis Co.: Moore & Steyermark 3133 (GH, MO, NY). Karnes Co.: Richards 561 (KANU). Kaufman Co.: Reverchon 3337 (Mo). Kenedy Co.: Tharp 49090 (F). Kent Co.: Tracy & Earle 369 (MO, NEB). Kerr Co.: Heller 1850 (ARIZ, GH, MO, NY, UC, US). Lipscomb Co.: Richards 1097 (KANU). Lubbock Co.: Demaree 7706 (GH, MO). Matagorda Co.: Cory 20290 (GH). McLennan Co.: Smith 865 (TEX). Medina Co.: Richards 3746 (KANU). Menard Co.: Cory (TAES). Moore Co.: Jespersen 2704 (F, MO, NY, UC, US). Nueces Co.: Heller 1789 (TEX). Ochiltree Co.: Wallis 7198 (TEX). Oldham Co.: Richards 1106 (KANU). Parker Co.: Ruth 1314 (us). Pecos Co.: Tharp 43-938 (OKL, TEX, UC). Potter Co.: Palmer 12535 (MO, UC). Presidio Co.: Hinckley 656 (F). Randall Co.: Palmer 13878 (MO, US). Reeves Co.: Gillespie 5253 (vc). Roberts Co.: Richards 1102 (KANU). Robertson Co.: Marsh & Calley 74 (TEX). Sherman Co.: Jackson 5 (KANU). Smith Co.: Reverchon 2229 (MoO). Stephens Co.: Denton, in 1931 (TEX.) Sutton Co.: Gould 5729 (TAES, TEX, UC). Tarrant Co.: Reverchon 1435 (F, GH, KANU, NEB, NY). Taylor Co.: Tracy 7891 (F, GH, MO, NEB, NY, TEX, us). Terrell Co.: Palmer 33538 (NY). Tom Green Co.: Tweedy 297 (us). Travis Co.: Barkley 13321 (TEX). Uvalde Co.: Kincaid & Johnston 54637 (TEX). Val Verde Co.: Jones 25903 (MO). Victoria Co.: Richards 3736 (KANU). Williamson 376 Rhodora [Vol. 70 Co.: York 46218 (NEB, TAES, TEX). Wise Co.: Shinners 7930 (uc). Young Co.: Gentry 1254 (TEX). UTAH: Cache Co.: Holmgren 3752 (GH, NY, UC, US). WISCONSIN: Milwaukee Co.: Shinners 2342 (GH, uc). Sauk Co.: Steele 72 (vs). Walworth Co.: Wadmond L.9 16211 (GH, LL). WvoMING: Albany Co.: Goodman 776 (MO, OKL). Camp- bell Co.: Porter & Porter 7568 (UC). Carbon Co.: Hanna 1053 (MO). Converse Co.: Porter 3151 (GH, MO, UC). Crook Co.: Hay- ward 210 (F). Hot Springs Co.: Mulford, in 1892 (GH, Mo). John- son Co.: Tweedy 3143 (NY). Laramie Co.: Pammel, in 1918 (GH). Niobrara Co.: Harms 2131 (KANU). Park Co.: Witt 1385 (NY). Platte Co.: Nelson 386 (GH, MO, NY, US). Sheridan Co.: Rollins 556 (NY). CANADA: ALBERTA: Moodie 1181 (F, GH, NY, UC, US). BRITISH COLUMBIA: Shaw 1143 (GH, MO, NEB, NY, US). MANITOBA: Dawson, in 1873 (GH). ONTARIO: Dodge, in 1911 (Mo, TEX). SAs- KATCHEWAN: Macoun 5068 (GH, US). MEXICO: CoanuILA: Palmer 345 (F, GH, MO, NY, US). HIDALGO: Chase 7117 (F, GH, MO, NY). NUEVO LEON: Taylor 27 (F, MO, TEX). SAN Luis Porosi: Lundell 5103 (ARIZ, MO, US). TAMAULIPAS: Tracy 8947a (NEB). VERA CRUZ: Det. by Hoffman 213 (GH). 4b. R. COLUMNIFERA (Nutt.) Woot. & Standl. forma PULCHERRIMA (DC.) Fernald, Rhodora 40:353. 1938. Obeliscaria pulcherrima DC., Prodr. 5:559. 1836. Obeliscaria columnaris DC. var. pulcherrima (DC.) D. Don, Sweet's Brit. Fl. Gard. Ser. 2, 4:361, 1838. Lepachys columnaris Torr. & Gray var. pulcherrima (DC.) Torr. & Gray, Flora N. Am. 2:315. 1842. Ratibida columnifera (Nutt.) Woot. & Standl. var. pul- cherrima (DC.) Woot. & Standl., Contrib. U. S. Nat. Herb, 19:706. 1915. Lepachys columnifera (Nutt.) Rydb. var. pulcherrima (DC.) Rycb., Flora Prairies and Plains of Central North America, 838 p. 1932. Ratibida columnaris (Sims) D. Don forma pulcherrima (DC.) Sharp, Ann. Mo, Bot. Gard. 22:70-71. 1935. Differs from R. columnifera only by the purplish-yellow to purple ray flowers. (Figs. 6 and 9). Flowering time, March-November. Lectotype. Texas: Mexican province. San Fernando to Bejar, June, 1832, Berlandier 1960 (G-DC?, Microfiche!; Isotype GH!). Since deCandolle did not select a type from 1968] Ratibida — Richards 377 Plate 1392 Fig. 6. Ratibida columnifera f. pulcherrima. ca. 1/3 actual size. Richards 567 (KANU). 378 Rhodora [Vol. 70 the original material, the above lectotype was selected by the author. Habitat and Distribution. Prairies, foothills, and moun- tains of Canada, United States, and Mexico. Ratibida columnifera f. pulcherrima is characterized by having showy, purplish-yellow to purple ligules. Since de- Candolle (1836) first described this as a species, its has attracted attention in nurseries and gardens, and it is in- variably collected in the field as being something different. It is recognized as a color form in this work because of its obvious color difference and its distinct distribution pattern. It occurs abundantly in northern Mexico, Texas, and New Mexico and is infrequently found farther north in the prairie and central states as compared to the yellow-liguled plants. Pure populations of the purple forms are common above 7,000 ft. in the Sacramento Mountains of New Mexico with isolated pure purple plants found infrequently over the range of the form. Representative specimens selected from 842 sheets: UNITED STATES: ARIZONA: Apache Co.: Griffiths 5382 (us). Gila Co.: Johnson 1697 (KANU). Navajo Co.: Thornber 4576 (ARIZ). ARKAN- SAS: Hempstead Co.: Palmer 8024 (MO, NY). Lawrence Co.: Demaree 30991 (KANU, MO). Miller Co.: Eggert, in 1898 (Mo, NY). COLORADO: Archuleta Co.: Baker 698 (F, MO, NY, US). Baca Co.: Harrington & Smith 296 (ARIZ). Denver Co.: Clokey 2819 (TEX). El Paso Co.: Durshel 4191 (MO). Jefferson Co.: Letterman, in 1913 (MO). Larimer Co.: Crandall 2980 (ARIZ). Las Animas Co.: Richards 1130 (KANU). Weld Co.: Osterhout 2310 (NY). KANSAS: Barber Co.: Richards 773 (KANU). Cheyenne Co.: Horr 4685 (GH, KANU). Coffey Co.: Horr 3467 (KANU). Edwards Co.: Carleton 308 (us). Ellis Co.: Crockett 182 (OKL). Decatur Co.: Richards 2835 (KANU). Grant Co.: Thompson 15 (GH, MO). Gove Co.: Kellerman, in 1893 (KANU). Meade Co.: McGregor & Horr 4954 (KANU). Mor- ton Co.: Richards 2967 (KANU). Sheridan Co.: Richards 1059 (KANU). Sherman Co.: McGregor 5212 (KANU). MISSOURI: Jackson Co.: Bush 4019 (NY). MONTANA: Cascade Co.: Safford 591 (us). Dawson Co.: Ward, in 1883 (Us). Meagher Co.: Anderson 232 (MO). Rosebud Co.: Bennett, in 1957 (us). NEBRASKA: Banner Co.: Rydberg (NEB). Box Butte Co.: Williamson (NEB). Buffalo Co.: Richards 1013 (KANU). Cedar Co.: Clements 2644 (NEB). Custer Co.: Richards 1018 (KANU). Dawes Co.: Webber 6200 (MO, NEB). Dawson Co.: Bates (NEB). Franklin Co.: Laybourn 79 (MO). 1968] Ratibida — Richards 379 Hooker Co.: Rydberg 1445 (NEB, US). Kearney Co.: Hapeman, in 1893 (ARIZ, US). Lancaster Co.: Hedgcock, in 1899 (Mo). Lincoln Co.: Richards 1024 (KANU). Morrill Co.: Bates (NEB). Rock Co.: Bates, in 1896 (GH). Scotts Bluff Co.: Rydberg 183 (NEB, NY). Sheridan Co.: Hatcher, in 1886 (NY). Sioux Co.: Woods 353 (NEB). Thomas Co.: Richards 1020 (KANU). Webster Co.: Bates (NEB). NEW Mexico: Bernalillo Co.: Jackson 2085 (UNM). Catron Co.: Eggleston 20388 (NY). Chaves Co.: Earle & Earle, in 1900 (Ny). Colfax Co.: Richards 1128 (KANU). Curry Co.: Reynolds 243 (UNM). Eddy Co.: Grassl 88 (GH). Grant Co.: Jones (UNM). Guadalupe Co.: Richards 1109 (KANU). Harding Co.: Sands (UNM). Lea Co.: Castetter 9003 (UNM). Lincoln Co.: Wooton 247 (GH, MO, NY, UC, US). Otero Co.: Richards 1122 (KANU). Quay Co.: Fisher 56 (Mo, US). Rio Arriba Co.: Standley 6762 (Us). Sandoval Co.: Hershey (UNM). San Migtel Co.: Richards 1125 (KANU). Sante Fe Co.: Arsene 15909 (UC). Socorro Co.: Metcalfe 415 (ARIZ, GH, MO, NY, UC, us). Taos Co.: Dixon A-370 (UNM). Torrance Co.: Jackson 2166 (UNM). Valencia Co.: Jones 122 (UNM). NORTH CAROLINA: Moore Co.: Pinehurst Nurseries, in 1902 (Mo). NoRTH DAKOTA: Benson Co.: Lunell, in 1917 (Mo, US). Grand Forks Co.: Churchill (NEB). McKenzie Co.: Stevens 727 (Mo, UC). Stark Co.: Mearns 168 (US). Steele Co.: Stevens, in 1949 (US). Ward Co.: Larsen 194 (MO, NY, US). Wells Co.: Bolley, in 1891 (NY). OKLAHOMA: Beckham Co.: Mericle 378 (OKL). Cimarron Co.: Waterfall 7441 (0KL). Cleveland Co.: Myers 49 (OKL). Comanche Co.: McCurry 1190 (OKL). Mayes Co.: Wallis 7676 (KANU, OKL, TEX). McCurtain Co.: Wehling 24 (OKL). Payne Co.: Waugh 260 (Us). Tillman Co.: Duncan 28 (Mo). SovurmH DAKOTA: Brookings Co.: Thornber (ARIZ). Custer Co.: Murdoch (F). Hutchinson Co.: Harms 554 (KANU). Kingsbury Co.: Thornber, in 1892 (Mo, UC). Meade Co.: Forwood 218a (vs). Moody Co.: Ball & Ball 882 (us). Spink Co.: Ricksecker 45 (UC). Washabaugh Co.: Visher 2033 (NY). TENNESSEE: Davidson Co.: Gattinger, in 1886 (F, Mo, NY). TEXAS: Anderson Co.: Eggert, in 1899 (Mo). Andrews Co.: Scudday (LL). Aransas Co.: Richards 3738 (KANU). Armstrong Co.: Reverchon 494 (MO). Bailey Co.: Correll 13102 (LL). Bastrop Co.: Richards 3725 (KANU). Bee Co.: Richards 559 (KANU). Bell Co.: York & York B44 (TEX). Bexar Co.: Richards 568 (KANU). Brazoria Co.: Correll & Edwin 16451 (LL). Brazos Co.: Palmer 7792 (MO, NY). Brewster Co.: Tharp 8833 (TEX). Caldwell Co.: Richards 550 (KANU). Cameron Co.: Traverse 1015 (F, LL, MO, TEX, US). Childress Co.: Childress Bio- logical Class, in 1930 (Mo). Comal Co.: Lindheimer 642 (M0). Crane Co.: Warnock 14697 (LL, TEX). Crockett Co.: Degener 5093 (NY). Culberson Co.: Shreve 8121 (ARIZ). Dewitt Co.: Richards 554 (KANU). Dickens Co.: Graham 8 (TEX). Eastland Co.: Hodge Oak Park School 50 (TEX). Edwards Co.: Gould 6810 (TAES). Erath 380 Rhodora [Vol. 70 Co.: Hoisington 88 (OKL, TEX, UC). Galveston Co.: Reverchon 3987 (NY). Garza Co.: Ruth 1313 (us). Gillespie Co.: Richards 3747 (KANU). Goliad Co.: Richards 558 (KANU). Gonzales Co.: Richards 551 (KANU). Gray Co.: Stemen 122 (OKL). Grayson Co.: Richards 3724 (KANU). Gregg Co.: Eggert, in 1899 (Mo). Guadalupe Co.: Richards 567 (KANU). Hall Co.: Reverchon, in 1904 (mo). Harris Co.: Traverse 1492 (F, LL). Hays Co.: Walker 44 (TEX). Hemphill Co.: Eggert, in 1900 (Mo). Hidalgo Co.: Tracy 8923 (F, GH, MO, NEB, NY, TAES, TEX, US). Howard Co.: Correll 16340 (LL). Jackson Co.: Drushel 9564 (US). Jeff Davis Co.: Waterfall 4730 (GH, MO, NY). Karnes Co.: Richards 560 (KANU). Kenedy Co.: Johnston 53.280.85 (TAES, TEX). Kent Co.: Tracy & Earle 369 (TEX). Kerr Co.: Perdue 1597 (LL, UC). Kinney Co.: Mearns 1467 (US). Kleberg Co.: Cory 51294 (GH, NY, TAES, US). Lamar Co.: Shinners 7864 (GH). La Salle Co.: Mauermann 22 (TEX). Llano Co.: Parks & Cory 6310 (TAES). Lubbock Co.: Demaree 7675 (GH, LL, MO, Us). Martin Co.: Eggert, in 1900 (Mo). Mason Co.: Taylor (TAES). Matagorda Co.: Beamer, in 1953 (KANU). Maverick Co.: Bray (TEX). MeMullen Co.: Thompson & Graham 67 (TEX). Medina Co.: Shinners 7302 (GH). Midland Co.: Tracy 7881 (F, GH, MO, NEB, NY, TEX). Nueces Co.: Pennell 10341 (NY). Oldham Co.: Hubricht A1435 (MO). Parmer Co.: Shinners 8332 (GH). Pecos Co.: Warnock 11845 (LL). Potter Co.: Richards 1105 (KANU). Presidio Co.: York 48101 (KANU, TAES, TEX). Randall Co.: Eggert, in 1901 (MO). San Patricio Co.: Richards 3740 (KANU). Scurry Co.: Harris, in 1927 (TEX). Smith Co.: Moore 828 (GH). Starr Co.: Tharp & Johnston 541899A (TEX). Sutton Co.: Aldwell Bros. 17499 (TAES). Tarrant Co.: Ruth 1054 (GH). Taylor Co.: Tolstead 7274 (GH, MO, NEB, NY, TEX). Terrell Co.: Webster 255 (TEX). Tom Green Co.: Smith 16 (GH, OKL). Travis Co.: Richards 549 (KANU). Val Verde Co.: Jones 25904 (MO, UC). Victoria Co.: Richards 556 (KANU). Wash- ington Co.: Brackett 285 (GH, TEX). Webb Co.: Orcutt 5768 (MO). Wharton Co.: Drushel 10723 (NY). Willacy Co.: Johnston 53.280.87 (TEX). Williamson Co.: York 46218 (0KL). Wilson Co.: Richards 563 (KANU). Zapata Co.: Clover 699 (Us). UTAH: Grand Co.: Holmgren & Hansen 3496 (GH, NY, US). WYOMING: Converse Co.: Nelson 8416 (ARIZ). Laramie Co.: Shantz (NEB). Natrona Co.: Nelson 570 (MO, NY, UC, US). CANADA: ALBERTA: Moodie 1188 (NY, US). MANITOBA: Macoun 12243 (NY). MEXICO: CHIHUAHUA: Shreve 7979 (ARIZ). COAHUILA: Nelson 6152 (F, US). NUEVO LEON: Frye & Frye 2358 (ARIZ, GH, MO, NY, UC). TAMAULIPAS: Tracy 8947a (F, GH, MO, NY, TAES, TEX). 5. RATIBIDA LATIPALEARIS Richards, Rhodora 66 (767) :267, 268. 1964. 1968] Ratibida — Richards 381 Plate 1393 Fig. 7. Ratibida latipalearis. ca. 1/3 actual size. Richards 3770 (KANU). 382 Rhodora [Vol. 70 Stems one or more, branched, sulcate, 5.5-11.7 dm high, strigose- hirsute, unicellular globular glands interspersed in the grooves, average number of ridges 10 just below inflorescence. Leaves strigose- hirsute, punctate on both surfaces, basal leaves long petioled, 3.2-26.5 cm long, ovate-obovate, crenate to 3-7 pinnately cleft, 3-nerved, lateral segments entire to serrate-crenate, oblong-oblanceolate, 0.2-4.0 cm long, 0.2-2.4 em wide, terminal lobe ovate, serrate-crenate to cleft-parted, 2.0-7.0 em long, 0.2-3.6 em wide, upper leaves ovate-obovate or oblong, 2.7-14.3 em long, 1.9-8.5 em wide, 3-9 pinnate or pinnately cleft, terminal segment lanceolate, parted or cleft, 2.0-7.0 em long, 0.6-4.5 em wide, lateral segments oblanceolate, entire or cleft, 0.2-4.7 cm long, 0.1-1.5 em wide. Peduncles 11.8-33.5 em long. Heads solitary, colum- nar, 1.7-3.5 em long, 0.9-1.3 em wide, on naked peduncles; receptacles linear, 1.7-3.3 em long, 1.2-4.0 mm wide, narrowing to apex. Phyl- laries reflexed, 10-12, linear-lanceolate, black punctate glands present, inner suface glabrous except for strigose-hirsute marginal tips, bi- seriate, outer row 0.6-1.1 cm long, 1.0-1.2 mm wide, inner row 0.3-0.8 em long, 0.8-1.2 mm wide. Ray flowers 8-17, yellow, lanceolate- oblanceolate, or oval-oblong, 0.8-3.1 cm long, 0.4-1.2 cm wide, outer surface hirsute, black resinous and punctate glands interspersed on outer surface, inner surface glabrous, tube short, 1.0-2.0 mm long, apex 2-3 lobed, lobes 0.2-8.2 mm long, 0.2-3.5 mm wide. Disc corollas dark greenish-yellow, 1.6-2.0 mm long, lobe segments 0.6-1.0 mm wide, apex lobes reflexed, ovate, 0.5-1.0 mm long, black resinous glands present on outer lobe surface; anthers sagittate, 1.2-2.0 mm long, tips with black resinous glands on outer surface, filaments 0.5-1.2 mm long, attached to base of anther; styles 1.0-1.8 mm long, filiform, the branches 2, 0.8-1.2 mm long, stigmatic surface restricted to tip, a fourth the length of the style branch. Palea 2.8-3.0 mm long, 4.0-5.0 mm wide, adaxial side hyaline, winged, ciliate, abaxial side hirsute half way down to base, few unicellular, black glands near base; vertical oblanceolate resin gland present in center of palea 1.0-1.2 mm long. Achenes oblong-oblanceolate, 2.0-3.0 mm long, 1.1-1.4 mm wide, adaxial margin glabrous, occasionally ciliate or with black, stalked glands; pappus 1-2 minute teeth or absent. (Figs. 7, 9, 11f, 12f, 13f). Flowering time: August-October. Holotype. Mexico. Chihuahua: Colonia Garcia, 37 miles southwest of Colonia Juarez, pine forest meadow, August 29, 1962, Richards 3770 (KANU!; Isotypes F!, GH!, KANU!, MO!, NY!, UC!, US!). Habitat and Distribution. Pine forest meadows, valleys and rocky pine slopes of Sierra Madre Occidental, Chihua- hua, Mexico. 1968] Ratibida — Richards 383 Ratibida latipalearis, as presently known, is limited in distribution to Chihuahua, Mexico. It was recognized as a new species because of the non-dissected, triple-nerved, first basal leaves, the greenish-yellow heads with 8-17 up- right ligules, black glands on the ligules, phyllaries, petal lobes of the disc corollas and the apex of the anthers, palea mostly broader than long with unicellular black glands from the middle to the base of the adaxial margin, and the branching habit of the stem. Representative specimens selected from 19 sheets: MEXICO: CuiHUAHUA: Dobie 2 (TEX); LeSueur 988 (F, MO, TEX, UC) ; Muller 3547 (LL, UC); Richards 3770 (KANU); Townsend & Barber 230 (F, MO, NY, UC, US); Tucker 2573 (ARIZ). 6. RATIBIDA TAGETES (James) Barnh. Bull. Torr. Bot. Club 24:410. 1897. Rudbeckia tagetes James, Long's Exped. 2:68. 1825. Rudbeckia globosa Nutt., J. Acad. Natur. Sci. Phila. 7:79. 1834. Obeliscaria tagetes DC., Prodr. 5:559. 1836. Lepachys columnaris Torr. & Gray var. tagetes Gray, Smiths. Contrib. Knowl. (Pl. Wright.) 3:106. 1852. Lepachys tagetes (James) Gray, Pac. Railw. Report. 4:103. 1856. Ratibida tagetes (James) Barnh. var. cinera Standley, Muhlenb. 5:30. 1909. Stems one or many, branched, 1.6-5.4 dm tall. Leaves ovate-obovate or linear-oblong, 0.6-8.6 cm long, 0.2-6.6 cm wide, pinnately or bipin- nately cleft into 2-7 lobes, narrowly lanceolate, 0.3-4.3 em long, 0.8- 4.0 mm wide, lower leaves are lanceolate and may be entire, a few upper leaves cauline and entire. Peduncles 0.3-6.4 cm long. Heads solitary, globular, 0.6-1.4 cm long, 0.6-1.2 em wide. Phyllaries 10-12, the inner series ovate, 1.2-3.8 mm long, 0.4-1.5 mm wide, the outer lanceolate, 1.2-6.0 mm long, 0.6-2.4 mm wide. Ray flowers 5-10, yellow, purplish-yellow to purple, oval-oblong, 2.8-8.5 mm long, 2.0-6.0 mm wide, hirsute on outer surface, punctate glands numerous on outer surface, inner surmace glabrous, 2-3 lobed apex, lobes 0.1-2.1 mm long, 0.3-3.0 mm wide. Disc corollas 1.2-1.9 mm long, lobes 0.5-0.8 mm wide, apex punctate, 0.5-1.0 mm long; anthers 1.0-1.5 mm long, punctate glands abundant on outer apex, filaments 0.4-1.0 mm long; styles 1.0- 1.9 mm long, the branches 0.8-1.2 mm long, stigmatic portion a fourth the length of the style branch. Palea 2.1-4.0 mm long, 1.5-2.5 mm 384 Rhodora Fig. 8. Ratibida (KANU). tagetes. ca. 1/3 actual size. Plate 1394 Richards 1111 1968] Ratibida — Richards 385 wide, punctate glands at apex, adaxial edge hyaline, ciliate, abaxial edge strigose to below the center, unicellular glands from below cen- ter to the base, resin gland linear-oblanceolate, 1.1-2.1 mm long. Achenes oblong-oblique, 1.9-2.8 mm long, 1.2-2.0 mm wide, glabrous except some ciliation at apex, winged adaxial edge; pappus a thick- crowned corona. (Figs. 8, 10, 11c, 12c, 13c). Flowering time: May-October. Holotype. Colorado: Otero County, south of La Junta, in prairie near stagnant pool, July 24, 1820, James (NY!). Habitat and Distribution. Prairies, high plains, and rocky hillsides. Colorado, Kansas, Oklahoma, southwestern United States, and Chihuahua, Mexico. Standley (1909) described Ratibida tagetes var. cinera (Holotype Standley 5156 NY!; Isotype MO!, US!) based on the stem and peduncles being cinerous and densely tomen- tose-hirsute with gray leaves also very densely strigose- hirsute. This is the only known occurrence of Standley’s variety but with transplants from Seward County, Kansas (Richards 3777 KANU), a pubescence similar to the above named variety was observed on greenhouse plants. Further observation led to the discovery of a minute gall or bud mite (Eriophyidae) feeding in the tomentose-hirsute areas. Use of insecticides stopped further outbreak of this phe- nomenon and the plants recovered, with the exception of small dead portions, to appear normal and flower, Careful study of Standley’s type 5156 revealed the remains of simi- lar mite exoskeletons. One other collection of Ratibida tagetes, Colfax County, New Mexico, Lucas 193 (OKL), exhibited this cinerous type of pubescence and again similar mite remains were seen in the areas showing the pubescence of Standley's variety. In addition to the bud mite, hyphal filaments of a Deu- teromycete were seen on a slide made by scraping the leaf surface in the noticeable areas of pubescence from Stand- ley's specimens and staining with lacto-phenol cotton blue. Standley (1909) mentions the possibility that his variety might be the typical form of R. tagetes affected by some fungus. On another specimen, Lucas 193 (OKL), conidia and hyphal filaments of Alternaria were seen on a mounted 386 Rhodora [Vol. 70 Plate 1395 Fig. 11. Disc Corollas of Ratibida; ca. X 9. a. R. pinnata; b. R. columnifera; c. R. tagetes; d. R. mexicana; e. R. peduncularis var. picta; f. R. latipalearis; g. R. peduncularis var. peduncularis. slide. Conidia and hyphal filaments were also taken from Richards 3777 (KANU). The above leads the author to believe that either mites or a fungus or both are the cause of the cinerous pubescence of Standley’s variety and it is, therefore, not recognized in this study. Ratibida tagetes is a short species characterized by small globular heads, narrowly linear or dissected leaves, vellow to purple ligules, and a well-formed tap root which produces adventitious buds from horizontal branches, leading to clonal development. The purple and purplish-yellow ligules have never been recognized as a color form probably be- cause they usually occur together, infrequently being found in pure colonies. The rays are small and inconspicuous com- 1968] Ratibida — Richards 387 Plate 1396 Fig. 12. Paleae of Ratibida; ca. X 9. a. R. pinnata; b. R. columni- fera; c. R. tagetes; d. R. mexicana; e. R. peduncularis var. picta; f. R. latipalearis; g. R. peduncularis var. peduncularis. pared to the larger showy ones of the color form of R. columnifera. Wooton and Standley (1915) mention the small rays of R. tagetes and the various forms as reasons for not naming a new variety. Representative specimens selected from 398 sheets: UNITED STATES: ARIZONA: Apache Co.: Griffiths 5092 (US). Coconino Co.: Deaver (ARIZ). COLORADO: Archuleta Co.: Bethel & Payson, in 1917 (NY). Baca Co.: Richards 1132 (KANU). Bent Co.: Coll. unknown 572 (cs). Cheyenne Co.: Robertson 1474 (cS). Denver Co.: Engelmann, in 1874 (Mo). Fremont Co.: Shear 3769 (NY, UC, US). Gunnison Co.: Baker 667 (GH, MO, NY, UC, US). Huerfano Co.: Parry 110 (MO). Kit Carson Co.: Preston (cs). Las Animas Co.: Richards 1131 (KANU). Lincoln Co.: Ownbey 1348 (GH, MO, NY, UC). Otero Co.: Richards 3776 (KANU). Prowers Co.: Christ 583 (cS). Pueblo Co.: Woodward 156 (GH). KANSAS: Barber Co.: McGregor 14790 (KANU). Edwards Co.: Richards 3764 (KANU). Finney Co.: Wilson 388 Rhodorz [Vol. 70 Plate 1397 Fig. 13. Achenes of Ratibida; ca. X 10. a. R. pinnata; b. R. columnifera; c. R. tagetes; d. R. mexicana; e. R. peduncularis var. picta; f. R. latipalearis; œ. R. peduncularis var. peduncularis. & Miller, in 1912 (KANU). Ford Co.: Richards 3762 (KANU). Grant Co.: McGregor 13993 (KANU). Gray Co.: MeGregor 15773 (KANU). Greeley Co.: Reed, in 1892 (Mo, vc, US). Hamilton Co.: Rydberg & Imler 1017 (KANU, NY). Kearney Co.: Rydberg & Imler 938 (KANU, NEB, NY). Logan Co.: McGregor 17245 (KANU). Meade Co.: Richards 3759 (KANU). Morton Co.: Richards 3570 (KANU). Russell Co.: Hitchcock 268 (GH, KANU, MO, NY, US). Scott Co.: Agrelius, Wilson & Agrelius, in 1912 (KANU). Seward Co.: Richards 3761 (KANU). Sherman Co.: Harshbarger 3801 (US). Stanton Co.: Thompson 68 (F, MO, NY, UC, US). Stevens Co.: Harms 752 (KANU). Trego Co.: Bodin 20 (F). Wallace Co.: Letterman, in 1884 (MO, NY). NEW Mexico: Bernalillo Co.: Jones 4118 (F, US). Colfax Co.: Richards 1129 (KANU). Dona Ana Co.: Wooton 5 (GH, MO, NY, US). Eddy Co.: Standley 40347 (US). Grant Co.: Bigelow 5571 (NY). Guadalupe Co.: Richards 1111 (KANU). Hidalgo Co.: Wolf 2591 (GH). Lea Co.: Harms 1786 (KANU). Lincoln Co.: Skehan 32 (F, GH, MO, NY, UC, US). McKinley Co.: Dittmer 9020 (UNM). Mora Co.: 1968] Ratibida — Richards 389 R. columnifera f columnifera E. columnifera f. pulcherrima R. mexicana e u O +. latipalearis A X. Fig. 9. Distribution of Ratibida. Richards 3773 (KANU). Quay Co.: Fisher 157 (us). Rio Arriba Co.: Jackson 2511 (UNM). Roosevelt Co.: Dittmer 9018 (UNM). Sandoval Co.: Ellis 208 (us). San Miguel Co.: Richards 3772 (KANU). Santa Fe Co.: Heller & Heller 3829 (ARIZ, GH, KANU, MO, NEB, NY, US). Sierra Co.: Jackson 2814 (KANU). Socorro Co.: Dunn 5281 (UNM). Torrance Co.: Clark 10328 (UNM). Union Co.: Harms 1958 (KANU). Valencia Co.: Dittmer 9019 (UNM). OKLAHOMA: Cimarron Co.: Demaree 13301 (GH, NY, OKL). Harmon Co.: Stevens 1080 (GH, MO, NY, OKL). Texas Co.: Williams 95 (uc). TEXAS: Brewster Go.: Warnock W919 (GH, TEX). Crane Co.: Tharp, in 1941 (TEX). Crockett Co.: Parks & Cory 29672 (TAES). Culberson Co.: Tharp & Janszen 49-1002 (TEX). Dallam Co.: Griffiths 5641 (Mo, US). Deaf Smith Co.: Shinners 8334 (TEX). El Paso Co.: Waterfall 3956 (ARIZ, GH, MO). Hartley Co.: Palmer 14142 (Mo, US). Hudspeth Co.: Mulford 260 (Mo, NY). Lamb Co.: Shinners 8354 (GH, UC). Lub- 390 Rhodora [Vol. 70 h. pinnata ae? kh. peduncularis var. peduncularis R. peduncularis var. ricta R. tagetes Fig. 10. Distribution of Ratibida. bock Co.: Demaree 7728 (GH, MO, US). Nolan Co.: Tharp 47466 (TAES). Ochiltree Co.: Cory 32228 (TEX). Oldham Co.: Reverchon 3340 (MO). Pecos Co.: Tharp 43939 (F, MO, NY, TEX, UC). Potter Co.: Palmer 12534 (Mo, UC). Randall Co.: Cory 16505 (GH). Reagan Co.: Merrill (TAES). Reeves Co.: Warnock 6229 (TEX). Sherman Co.: Weaver 17800 (TAES). Sutton Co.: Cory 15414 (XL). MEXICO: CHIHUAHUA: Pringle 1061 (F, MO, NY). ACKNOWLEDGMENTS The writer expresses appreciation to Dr. R. L. McGregor for the suggestion of this problem and for valuable help throughout. Thanks are due to The Society of The Sigma Xi and Sigma Xi-RESA for a grant enabling travel to Texas and Mexico for study of Ratibida. The writer wishes to 1968] Ratibida — Richards 391 express his appreciation also to Dr. R. C. Jackson for sug- gestions and criticism and to Dr. R. W. Baxter and Dr. R. H. Thompson for help with the photographs. Acknowl- edgment is due to the other staff members and graduate students of the Department of Botany at the University of Kansas. To the curators of the herbaria of the Univer- sity of Arizona (ARIZ), University of California (UC), Chicago National History Museum (F), Colorado State University (cs), Gray Herbarium (GH), Missouri Botani- cal Garden (MO), University of Nebraska (NEB), Univer- sity of New Mexico (UNM), New York Botanical Garden (NY), University of Oklahoma (OKL), Texas A & M College (TAES), Texas Research Foundation (LL) and the Univer- sity of Texas (TEX) the author is indebted for the loan of specimens. BIBLIOGRAPH Y BARNHART, J. H. 1897. Nomenclatural notes. Bull. Torr. Bot. Club 24:409-411. BorviN, B. 1960. Centurie de plantes canadiennes III. Natur. Can. 87:25-49. BRITTEN, J. 1899. Bibliographical notes. J. Bot. 37 :481-487. BUCKLEY, S. B. 1861. Description of new plants from Texas. Acad. Natur. Sci. Phil. 13:448-463. CANDOLLE, A. P. DE. 1836. Prodromus systematis naturalis regni vegetabilis. Paris. 5:1-706. CassINI, H. 1825. Obeliscaire. Dict. Sci. Natur. 85 :272-278. COCKERELL, T. D. A. 1915. Specific and varietal characters in annual sunflowers. Am. Natur. 49 (586) :609-622. 1916. Collarette flowers. J. Hered 7:428-431. CRONQUIST, A., D. D. Keck, and B. MAGUIRE. 1956. Validity of Nettal's names in Fraser’s Catalogue. Rhodora 58:23,24. DARLINGTON, C. D. and A. P. WYLIE. 1955. Chromosome atlas of flowering plants. George Allen & Umwin Ltd. London. 519 p. Don, D. 1838. Ratibida colummaris. 4:361, Fig. 361. In: Sweet's Brit. Fl. Gard. (Ser. 2) James Ridgeway, London. Dress, W. J. 1961. Notes on the cultivated Compositae, 6. The coneflowers: Dracopsis, Echinacea, Ratibida, and Rudbeckia. Baileya 9:67-83. FERNALD, M. L. 1938. New species, varieties and transfers. Rho- dora 40:353-356. Gray, A. 1852. Plantae Wrightianae — Texano — Neo Mexicanae. Smiths. Contrib. Knowl. 3:1-145. 392 Rhodora [Vol. 70 1856. Reports of exploration and Surveys to ascertain ... the route... from the Mississippi River to the Pacific Ocean. Pacif. Rail. Report 4:103. GREENE, E. L. 1889-1892. Reprint of Fraser's Catalogue. Pittonia 2:114-119. JACKSON, R. C. 1959. Documented chromosome numbers of plants. Madrono 15:52. JACKSON, S. W. 1963. Hybridization among three species of Rati- bida, Univ. Kans. Sci. Bull. XLIV (1):27 p. JAMES, E. 1823. S. H. Long's expedition: account of an expedition from Pittsburg to the Rocky Mountains, 1819-1820. H. C. Carey & K. Lea, Philadelphia. 2:1-442. MACBRIDE, J. F. 1922. Various North American spermatophytes, new or transferred. Contrib. Gray Herb. Harvard Univ., New Ser. 3(65) :39-46, NUTTALL, T. 1818. Genera of North American plants. D. Heart, Philadelphia. 2:254 p. 1854. A description of some of the rarer or little known plants indigenous to the United States. J. Acad. Natur. Sci. Phila. 7:61-115. 1841. Descriptions of new species and genera of plants in the natural order of the Compositae. Trans. Am. Philos. Soc. Ser. 2. 7:284-453. PERDUE, R. E. JR. 1959. The somatic chromosomes of Rudbeckia and related venera of the Compositae. Contrib. Gray Herb. 185 :129-162. PursH, F. T. 1814. Flora Americae septentrionalis. White, Coch- rane & Co., London. 1:1-751. RAFINESQUE, C. S. 1817. Florula Ludoviciana. C. Wiley & Co., New York. 178 p. 1819. Prodrome: des nouveaux genres de plantes observés en 1817 et 1818 dans l'intérieur des États-Unis d'Amérique. J. Phys. Chim. Hist. Natur. et Arts. 89:96-107. RicHArps, E. L. 1964. A new species of Ratibida (Compositae) from Mexico. Rhodora 66(767) :267,268. RvpBERG, P. A. 1932. Flora of the prairies and plains of central North America. New York Bot. Gard. 969 p. SHARP, W. M. 1935. A critical study of certain epappose genera of the Heliantheae — Verbesininae of the natural family Compo- sitae. Ann. Mo. Bot. Gard. 22:51-148. SIMS, J. 1814. Rudbeckia columnaris, High-erowned Rudbeckia. Curtis’s Bot. Mag. or Fl. Gard. Displayed. 39:1601; Fig. 1601. SMALL, J. K. 1903. Flora of the southeastern United States. Era Printing Co., Lancaster, Pennsylvania. 1370 p. STANDLEY, P. C. 1909. Notes on the flora of the Pecos River Na- tional Forest. Muhlenb. 5:17-30. 1968] Ratibida — Richards 393 TORREY, J., and A. GRAY. 1842. A flora of North America. Wiley & Putnam, New York and London. 2:1-504. VENTENAT, E. P. 1800. Description des plantes nouvelles et peuc- connues cultivées dans le jardin de J. M. Cels. Crapelet, Paris. 100 plates and descriptions. WALPERS, G. G. 1843. Repertorium botanices systematicae. S. F. Hofmeister, Lipsiae. 2:1-1029. WaTSON, S. 1888. Some new species of Mexican plants, chiefly of Mr. C. G. Pringle's collection in the mountains of Chihuahua, in 1887. Proc. Am. Acad. Arts & Sci. 23:267-283. WooTON, E. O. and P. C. STANDLEY. 1915. Flora of New Mexico. Contrib. U. S. Nat. Herb. 19:1-794. DEPARTMENT OF BIOLOGY ARKANSAS STATE UNIVERSITY STATE UNIVERSITY, ARKANSAS 72467 PLANT SPECIATION ASSOCIATED WITH GRANITE OUTCROP COMMUNITIES OF THE SOUTHEASTERN PIEDMONT! W. H. MURDY Granite outcrop communities of the southeastern Pied- mont have long been of interest to the systematic botanist because of their large assemblage of endemic taxa. Crystal- line rock exposures, called “flat-rocks” or "granite out- crops," occur sporadically from North Carolina to Alabama, and their combined extent is estimated to be about 12,000 acres. Ten of the best-documented species reported to be Table 1. ENDEMIC SPECIES GEOGRAPHIC ZONES (FIG. 1) 10 9 8 7 6 3 Isoetes melanospora Rhynchospora saxicola Viguiera porteri Quercus georgiana Amphianthus pusillus Phacelia maculata Sedum pusillum Portulaca smallii Cyperus granitophilus Juncus georgianus x x x Occurrence of 10 plant species endemic to granite outcrop commu- nites in the Piedmont of southeastern United States. X XXXXXXXXx X XXXXXXXX X XXXXXXX XX xXx x x xxx xX xX x enedmic to granite outcrop communities are listed in Table 1. When their collective ranges are plotted on a single map, all ten species are found to occur within a small region of the upper Piedmont of Georgia, east of Atlanta (the inner- most circle of Fig. 1). Toward the southwest and northeast the number of endemic taxa gradually decreases until only three of the species listed in Table 1 are found to be associ- ated with granite outcrops in eastern North Carolina. This “center of endemism” in Georgia is coincident with the geographic center of the greatest concentration of exposed rock. "This work was supported by National Science Foundation grant NSF-GB2340. 394 1968] Plant Speciation — Murdy 395 D 1 Fig. 1. Diagram to show how the number of species of granite outcrop endemics listed in Table 1 diminishes along a northeast- southwest axis from a “center of endemism" located in the central Piedmont of Georgia. The number of species is also taken as the number for the geographic zone (Table 1). A variety of habitats are afforded by the granite outcrops. Two of the most important types available to higher plants are: 1) depression pits in the rock surface, which can be classified into a number of different community types by the correlates of soil depth and vegetation cover (Burbanck and Platt 1964); and 2) glade-like, border areas of thin soil situated between rock exposures or between rock and adjacent forest or field. 396 Rhodora [Vol. 70 The granite outcrops have been open to plant habitation for a long period of time although the longevity of any single exposure may not be great. Sufficient time has elapsed, for example, for the rare endemic Amphianthus pusillus Torr. to become adapted to an exceedingly special- ized and infrequent habitat. It presently occupies depression pits of a critical size, topography, soil depth, and water retentive capacity. Furthermore, it has had time to diverge from close relatives to the extent that it now constitutes a monotypic genus of uncertain taxonomic affinities. Pennell (1935), in his monograph of the Schrophulariaceae, notes that the Amphianthus habit of bearing flowers both on elongated floating stems and also at the base of the plant, where they are immersed and cleistogamous, has no counter- part among the Scrophulariaceae. The strikingly high degree of endemism associated with the outcrops suggests that they have long served as active sites for speciation in the southeast. A few of the endemic species may be remnants of once more widespread species which had their origin elsewhere. However, most endemic taxa appear to have had their origin in adaptation to out- crop habitats; most by gradual, ecogeographical processes and perhaps a few by an abrupt, saltation process. The Piedmont Region marks the geographic limit of distribution for many Coastal Plain and Appalachian species and the granite outcrops provide a variety of ecologically-marginal habitats to which geographically-peripheral populations may adapt. The processes associated with speciation may actually undergo acceleration in the region of granite outcrops. Outcrops frequently occur in clusters with exposures sepa- rated by a few miles or less, and the clusters, in turn, separated from one another by greater distances. Species characteristic of, or endemic to, outcrop communities are consequently subdivided into a number of disjunct popula- tions. Thus, semi-isolated populations are able to diverge from nearby populations as well as occasionally cross with them. Wright (1931) considered this type of population to be capable of rapid evolution. 1968] Plant Speciation — Murdy 397 * R meculete OR dubla georgiano = Fig. 2-5. Distribution maps. Fig. 2, Rhynchospora saxicola and R. globularis var. typica. Fig. 3, Phacelia dubia (open circles) and P. dubia var. georgiana (solid black circles). Fig. 4, Phacelia macu- lata and P. dubia var. georgiana. Fig. 5, Cyperus granitophilus. ECOGEOGRAPHICAL SPECIATION The fact that several endemic species are situated at the geographic periphery of their nearest relative is taken as indirect evidence of past ecogeographical speciation in association with granite outcrop communities. For example, Rhynchospora saxicola Small and Portulaca smallii P. Wilson occur at the boundary of the ranges of their puta- tive, Coastal Plain progenitors, R. globularis var. typica (Gale 1944), as illustrated in Fig. 2, and P. pilosa (Wilson 1932, Steiner 1944, Cotter and Platt 1958) respectively. These taxa are in need of modern systematic treatment in order to more accurately assess their specific status and 398 Rhodora [Vol. 70 evolutionary history, Another endemic species, Phacelia maculata Wood, which will be discussed below, is presently distributed at the southeastern boundary of the range of its closest relative, P. dubia of the Appalachians. Recent ecogeographical speciation is indicated by the number of sub-specific taxa endemic to the outcrops. The systematics of one of these, Phacelia dubia var. georgiana, has been described in a recent report (Murdy, 1966). This variety was named by McVaugh (1943) as a race of P. dubia which could be distinguished from the latter on the basis of leaf characters. Constance (1949), in his revision of Phacelia species in the subgenus Cosmanthus, referred to it as a distinct regional phase of the Appalachian P. dubia restricted to granite rocks in the Piedmont of Ala- bama and Georgia (Fig. 3). Population studies have revealed that the ecotype P. dubia var. georgiana can be separated from the main body of the species on the basis of any one of a number of quantitative traits, such as length of calyx, petals, and sepal hairs. The ecotype has the same chromosome number as P. dubia (n — 5) and artificial crosses between them have produced vigorous hybrids. The ecotype occurs in abundance on out- crops of the lower Piedmont (Fig. 3), and in this region, it has become an integral component of communities both in depression pits and in border areas. Its origin appears to have been recent and in the lower Piedmont. It has not yet migrated to the "center of endemism," in the upper Piedmont of Georgia, where a niche appears to be open to it. Phacelia maculata is a species endemic to granite outcrops of the upper Piedmont from North Carolina to Alabama (Fig. 4). It is a poor competitor when compared with P. dubia var. georgiana and when both occur at the same outcrop site, the former is largely confined to deciduous glades bordering the exposed rocks, whereas P. dubia var. georgiana occupies a variety of open habitats. All of the systematic data to date suggests that P. maculata and P. dubia formerly diverged from a common, 5-chromosome 1968] Plant Speciation — Murdy 399 stock. Today, they are reproductively isolated by means of a postfertilization, incompatibility factor (Murdy 1966). The tentative hypothesis is advanced that P. dubia is presently in the act of speciation and within recent times has produced an ecological race, P. dubia var. georgiana, at the southernmost boundary of its range and in adapta- tion to an ecological niche afforded by the granite outcrops. P. maculata may have had a similar origin in the past, but now exhibits the properties of a relict endemic because of its inability to compete successfully outside of a narrowly- restricted habitat which is gradually diminishing in extent. SALTATIONAL SPECIATION The occurrence of “species pairs" within outcrop com- munities, strikingly similar in life-form and ecology, but radically different in other important respects, suggests that saltational speciation may have been involved in their origin. The widely-fluctuating, outcrop environment would be a factor of significance in such a process, Three pairs of species will be discussed below: Sedum pusillum Michx. and S. smallii (Britt.) Ahles (Diamorpha cymosa (Nutt.) Britt.) ; Talinum mengesii Wolf and T. teretifolium Pursh; and Cyperus granitophilus McVaugh and C. aristatus Rottb. In none of these examples has a saltational mode of origin been clearly established. Instead, such a process is merely postulated on the basis of evidence presented below. Sedum pusillum and S. smallii Sedum pusillum is a small, succulent annual wholly con- fined to granite outcrop communities (Fig. 6, B + Fig. 8). It will grow in a variety of micro-habitats during “good” years, but is typically found growing in shallow, depression pits under the partial shade of pine or, more commonly, Juniperus virginiana. lt is of systematic interest within the genus because of its low chromosome number, 2» — 8 (Baldwin 1940), the lowest of any North American Sedum. It is of general evolutionary interest because of its endem- ism and remarkable similarity to S. smallii (Fig. 6, A). The latter is one of the most characteristic members of outcrop communities, but also occurs on sandstone outcrops 400 Rhodora [Vol. 70 Fig. 6-8. Fig. 6. Habit of: A, Sedum smallii and B, S. pusillum. Fig. 7, 8. Distribution maps. Fig. 7. S. smallii. Fig. 8. S. pusillum. 1968] Plant Speciation — Murdy 401 in Alabama, Tennessee, and Georgia (Fig. 7). It grows in abundance in the shallowest soils of depression pits where it has no competitors. Fróderstróm (1935), in his mono- graph on Sedum, considered the two species to be closely related. However, S. smallii has a very different chromo- some number, 2n — 18 (Baldwin 1940, O'Connell 1949, McCormick and Platt 1964). Common base chromosome numbers in the Crassulaceae are 4 and 5, with most species being polyploid. The lowest numbers in each series are found in annuals (Sedum pusil- lum, n = 4, and S. stellatum, n = 5) within a family mostly composed of perennials. Sedum smallii is unusual in having a base chromosome number of 9 and Baldwin (1940) suggested that it may have originated as an amphi- diploid. If this is taken as a working assumption, one of its parents could very well have been S. pusillum, the only extant member of the family with 4 pairs of chromosomes. Attempts should be made to produce interspecific hybrids which can be used for cytogenetic verification of the above hypothesis. For example, if 4 of the 9 chromosomes of S. smallii are homologous with the chromosomes of S. pusil- lum, 4 bivalents and 5 univalents should constitute the paired, meiotie condition in the hybrid. Positive results would warrant a search for the other parental species which should have 5 pairs of chromosomes, Strong barriers to natural hybridization seem to be operative between the species. Populational analysis of a large outcrop in Rockdale County, Georgia, where both spe- cies are abundant, has yielded neither hybrids nor any indi- cation of introgression. However, McCormick and Platt (1964) reported that: “Apparent intergeneric hybrids be- tween Sedum pusillum X Diamorpha cymosa were observed on one outcrop out of more than 100 visited”. Several hun- dred artificial pollinations between the two species in the spring of 1967 failed to yield any seed, while control pollina- tions within each species resulted in a good yield. Talinumm engesii and T. teretifolium The habit, habitat and distribution patterns of these two 402 Rhodora [Vol. 70 Talinum species suggest that their separation into distinct, morphological entities might also have been brought about by a saltational process of speciation, which also involved polyploidy. Talinum teretifolium (2n = 48) ranges from Pennsyl- vania to Georgia and grows in sandy or dry, rocky sites. In Georgia it is almost exclusively associated with granite exposures where it represents the only perennial in many of the shallow soils. T. mengesii (2n — 24) is found in western Georgia and it is the predominant species in Ala- bama, where it grows on the sandstone outcrops ın Appa- lachian counties as well as on the granites of the Piedmont (Wolf 1939). Their ranges appear to have an allopatric pattern with a considerable overlap in Georgia. In the central Piedmont of Georgia, the predominant form is T. teretifolium, bui pockets of T. mengesii populations have been located in the following Georgia counties: Douglas, Heard, Paulding, Rockdale, and Walton. Outcrops in the region of overlap support almost exclusively either one or the other species. At three sites in Georgia where plants of both species are contiguous, triploid hybrids have been found which are vigorous perennials, but apparently sterile. The close morphological and distributional similarities between the two prompted Wolf (1939), in his evaluation of Talinum in Alabama, to postulate a common ancestry for them. Both species have a similar life-form and appear to occupy a very similar outcrop niche. Nevertheless, they are morphologicaly distinct species and can be readily separated on the basis of flower structure. T. teretifolium has small, rounded petals, fewer than 20 stamens, a short style, and flowers that open late in the afternoon, whereas T. men- gesti has long, acute to mucronate petals, more than 40 stamens, a long, exserted style, and flowers that open early in the afternoon. Plants of eighteen populations of T. teretifolium, from Georgia and South Carolina, were found to be tetraploid (2n — 48). Steiner (1944) gave similar counts for this 1968] Plant Speciation — Murdy 408 species but also reported a plant from Bibb Co., Alabama to be diploid (2n — 24). This is likely a case of mis- identification since T. teretifolium has not otherwise been reported from this location which is in the heart of T. mengesii (2n = 24) territory. It is suggested that the widespread, Talinum tereti- folium is an amphidiploid species and that T. mengesii was one of its diploid progenitors. The proposed relation- ship for the Talinum species parallels that postulated for the Sedum species. In both pairs the polyploid member has a greater geographical range and is better adapted to exposed outcrop sites than its diploid partner, which in the case of Sedum pusillum is restricted to several granite outerops in Georgia, South Carolina and North Carolina and in the case of T. mengesii is confined to the sandstone outcrops of Alabama and a few granite outcrops of Ala- bama and Georgia. Cyperus granitophilus and C. aristatus Cyperus granitophilus was first named by McVaugh in 1957 who considered it to be a form sufficiently distinct from the related C. aristatus (C. inflexus) to rank as a good species endemic to granite outcrop locations from North Carolina to Alabama. This classification was upheld by O'Neil (1942), in his taxonomic study of North and South American species of Cyperus, who added characters by which the two similar-looking sedges could be separated. A recent systematic study (Wynne 1964) suggests that the endemic, Cyperus granitophilus, arose from the widely- distributed C. aristatus through autopolyploidy and subse- quent selection for high ploidy levels in outcrop habitats. Data gathered from several natural populations and from herbarium specimens show the endemic to be distributed throughout the outcrop region (Fig. 5). It constitutes the predominant form in outcrop communities and grows in depression pits of little soil and scant competition. C. aris- tatus, on the other hand, is mostly found in border areas of deep soil where competition is more intense. Plants of both species have been found to grow side by 404 Rhodora [Vol. 70 side and both have more than once been mounted on the same herbarium sheet. Furthermore, character differences that separate them are almost all quantitative. When compared on the basis of spikelet, scale, and achene dimen- sions, the endemic is larger in every respect. Chromosome counts showed high, but variable, numbers within popula- tions and C. granitophilus had a consistently higher chromosome number (2n = ca, 80, 88, 96) than C. aristatus (Zn = ca. 48, 56, 64). Many interesting questions have been raised by this pre- liminary study. One of the most widespread effects of poly- ploidy is to increase the water content of the cell relative to the amount of protoplasm, which in turn is believed to be a result of increase in cell size (Noggle 1949, Stebbins 1950). Could this provide increased drought resistance necessary for C. granitophilus to become the prevailing form in xeric, outcrop habitats? Futhermore, does this example represent a mechanism whereby a monotypic, wide- spread species, itself probably polyploid, could produce an ecotype eble to exploit an available, but marginal, habitat by further increase in the ploidy level? Further experi- mental work is needed before definite answers can be given to the above questions. CONCLUSION The granite outcrop region constitutes an area of high endemic frequency in the southeast and the greatest num- ber of endemic taxa occur in the central Piedmont of Georgia. Several different types of endemics are repre- sented in outerop communities including at least one mono- typic genus, endemic species with and without related species nearby, and infraspecific taxa which appear to be ecotypes of extant species. The implication derived from this array of endemic types is that the granite outcrops have served as active sites for plant speciation for a long period of time. The scattered distribution of rock exposures increases the probability that their populations will diverge from one another with time. An ideal situation for ecogeographical 1968] Plant Speciation — Murdy 405 speciation is established when such unusual and disjunct habitats are located at the boundary of a species range. Semi-isolated, geographically-peripheral populations are afforded an ecologically-marginal habitat to which they may adapt. Subsequent migration, immigration and geneti- eal interaction among such populations could eventually produce an endemie ecotype, which, in time, could ulti- mately become an endemic species. This appears to be the manner by which the ecotype Phacelia dubia var. georgiana originated in the recent past from the Appalachian P. dubia, and may, in turn, closely parallel the manner by which the endemie P. maculata, originated from an Appalachian progenitor in the remote past. Species which also appear to have had a gradual, ecogeographical origin, but from Coastal Plain progenitors, include Portulaca smallii and Rhynchospora saxicola. The disjunct distribution of outcrop communities, to- gether with their variable size and extreme environmental conditions, might all be factors contributing to the salta- tional origin of various taxa. Drastic fluctuations in popu- lation size could on the one hand lead to local extinction (endemic species may be totally lacking at one outcrop and abundant at another outcrop a few miles away), but could also be conducive to “catastrophic selection" (Lewis 1966), whereby a few individuals, possessing favorable genotypes, would be the sole survivors at a particular outcrop site and provide progeny for repopulation. Three sympatric “species pairs," similar (but not identi- cal) in vegetative morphology and in ecological adaptation, but different in other important respects, are presented as possible cases of saltational speciation. Sedum smallii (2n = 18) may have originated through allopolploidy in- volving the rare endemic S. pusillum (2n = 8) as one pa- rental species and an unknown, 5-chromosome species as the other parent. Cyperus granitophilus may be an autopoly- ploid form of C. aristatus and appears to be better adapted to outcrop habitats than the latter. Finally, T. teretrifolium (2n — 48) may have originated as an amphidiploid with 406 Rhodora [Vol. 70 T. mengesii (2n = 24) as one of its parental species. In all three examples, the polyploid member of each pair is more widely distributed than its diploid, or, in the case of Cyperus, its lower polyploid relative. DEPARTMENT OF BIOLOGY EMORY UNIVERSITY ATLANTA, GEORGIA 30322 LITERATURE CITED BALDWIN, J. T. 1940. Cytophyletic analysis of certain annual and biennial Crassulaceae. Madroño 5: 184-187. BURBANCK, M. P. and R. B. PLATT. 1964. Granite outcrop commu- nities of the Piedmont Plateau in Georgia. Ecology 45: 292-306. CONSTANCE, L. 1949. A revision of Phacelia Subgenus Cosmanthus (Hydrophyllaceae). Contr. Gray Herb., Harv. Univ. 168: 1-48. Correr, D. J. and R. B. PLATT. 1959. Studies on the ecological life history of Portulaca smallii. Ecology 40: 651-668. FRODERSTROM, H. 1935. The genus Sedwm L. Part IV. Act. Hort. Gotab. 10: 135-138, GALE, S. 1944. Rhynchospora, section Eurynchospora, in Canada, the United States and the West Indies. Rhodora 46: 207-249. LEWIS, H. 1966. Speciation in flowering plants. Science 152: 167- 172. McCormick, F. and R. B. PLATT. 1964. Ecotypic differentiation in Diamorpha cymosa. Bot. Gaz. 125: 271-279. McVaucH, R. 1937. A new species of Cyperus from the granite region of central Georgia. Castanea 2: 100-104. 1943. The vegetation of the granite flatrocks of the southeastern United States. Ecol. Monog. 13: 119-165. Murpy, W. H. 1966. The systematics of Phacelia maculata and P. dubia var. georgiana, both endemic to granite outcrop communi- ties. Amer. Jour. Bot. 53: 1028-1036. NoGGLE, G. R. 1946. The physiology of polyploidy in plants. I. Review of the literature. Lloydia 9: 153-174. O’CONNELL, J. F. 1949. The cytology, morphology and taxonomy of Diamorpha cymosa. J. Elisha Mitch. Sci. Soc. 65: 194. O'NEILL, H. 1942. The status and distribution of some Cyperaceae in North and South America, Rhodora 44: 43-64; 77-89. PENNELL, F. W. 1935. Scrophulariaceae of eastern temperate North America. Acad. Nat. Sci., Phila. Monog. 1. STEBBINS, G. L. 1950. Variation and evolution in plants. Columbia Univ. Press, New York. STEINER, E. 1944. Cytogenetic studies on Talinum and Portulaca. Bot. Gaz. 105: 374-379. 1968] Plant Speciation — Murdy 407 WILSON, P. 1932. Portulaca. North American Flora 21: 328-336. WoLF, W. 1939. The status of Talinum in Alabama. Amer. Midl. Nat. 22: 315-332. WRIGHT, S. 1931. Evolution in Mendelian populations. Genetics 16: 97-159. WYNNE, L. L. 1964. Systematic relationship of the granite outerop endemic Cyperus granitophilus (McVaugh) to Cyperus inflexus (Muhl.). M. S. Thesis, Emory Univ. Library, Atlanta, Ga. A NEW SPECIES OF PHYSALIS FROM THE GALAPAGOS ISLANDS U. T. WATERFALL While preparing a treatment of Physalis for A Flora of the Galapagos Islands by Ira L. Wiggins of Stanford Uni- versity, it became evident that some of the material under consideration was not readily referable to a known species. The following species is therefore described. Physalis galapagoensis Waterfall, sp. nov. Planta her- bacea, annua, 15-90 cm alta; ramis, petiolis et pedicellis plus minusve antrorse hispidulis; foliis ovatis, marginibus inaequaliter crasse dentatis, principalibus 3.5-8 em longis et 3-6 cm latis, petiolis 1.5-5 em longis; calycibus floriferis 2.5-5 mm longis et 1.5-2 mm latis ad basim loborum ; ealycis lobis lanceolatis vel anguste lanceolato-attenuatis, (1-) 2-3 mm longis; pedicellis floriferis 4-6 mm longis; corollis luteolis, immaculatis, 4-6 mm longis; antheris luteis vel marginibus violaceis, oblongis vel ovato-oblongis, 1-1.5 mm longis; filamentis 1.5-3 mm longis; calycibus fructiferis pentangulatis, glabris, 30-45 mm longis et 37-40 mm latis; calycis lobis lanceolatis vel anguste lanceolato-attenuatis, 4-8 mm longis, interdum abrupte porrectis; pedicellis fructiferis 15-30 mm longis; baccis 12-15 mm latis. TYPE: Ira L. Wiggins 18724, Feb. 15, 1964, on nearly bare lava ridge at dormitories, vicinity of Charles Darwin Research Station, Acadamy Bay, Isla Santa Cruz, Galápagos Islands, Ecuador (Ds; Isotype: OKLA). OTHER COLLECTIONS SEEN: ALBEMARLE ISLAND: Villamil, Apr. 27, 1932, John Thomas Howell 8939 (CAS) ; abundant near beach in shady places, Tagus Cove, R. E. Snodgrass & E. Heller 185, Mar. 1899 (ps); in shady places around 900 ft., Tagus Cove, Alban Stewart 3393, Mar. 24, 1905-1906 (CAS); CHARLES ISLAND: Post Office Bay, John Thomas Howell 8814, Apr. 23, 1932 (cas); in shady places near shore, Albam Stewart 3397, May 24, 1905-1906; CHATHAM ISLAND: corolla yellow, concolorous, Wreck Bay, John Thomas Howell 8524, Apr. 15, 1932 (CAS) ; GARDEN ISLAND: near Hood Island, John Thomas Howell 8772, Apr. 22, 1932 (CAS); HOOD ISLAND:R. E. Snodgrass & E. Heller 740, May 1899 (DS); INDEFATIGABLE ISLAND: corolla greenish-yellow, Acad- 408 1968] New Physalis — Waterfall 409 emy Bay, John Thomas Howell 9020, May 2, 1932; JARVIS ISLAND: John Thomas Howell 9759B, June 6, 1932 (CAS); NORTH SEYMOUR ISLAND: John Thomas ‘Howell 9962B, June 11, 1932 (CAS); SANTA CRUZ ISLAND: near Charles Darwin Research Station in Bursera for- est, alt. 30 meters, Syuzo Itow 267, Feb. 23, 1964, (ps); vicinity of Charles Darwin Research Station, Academy Bay, in partial shade among Opuntia echios trees, sea level to 30 meters, Ira Wiggins 18314, Jan. 23, 1964 (Ds). Physalis galapagoénsis is characterized by its small, immaculate corollas, small anthers, yellow all over or violet- margined, and its usually large, glabrous, 5-angled fruiting calyces, much-inflated about the berry. In the latter characteristic it is reminiscent of P. cordata Miller, a common species of southern Mexico, Central America and the West Indies (Waterfall, 1967), a species with much larger, dark-maculate corollas. Or it might be compared with the less common P. porrecta Waterfall (1. c. 237, 238) of Oaxaca, Costa Rica and Guatemala, a species with abruptly-beaked fruiting calyces and large corollas, immaculate, or inconspicuously spotted. The length of the corollas is similar to that of P. clarionensis Waterfall (1. c. 326, 327) from Clarion Island off the coast of Baja Cali- fornia, but the latter species has shorter, and proportionally narrower, fruiting calyces. The chromosome count, according to the label is: n=12 (by Kyhos) based on 10 cells. DEPARTMENT OF BOTANY AND RESEARCH FOUNDATION OKLAHOMA STATE UNIVERSITY STILLWATER 74074 LITERATURE CITED WATERFALL, U. T. 1967. Physalis in Mexico, Central America and the West Indies. Rhodora 69: 82-120; 203-239; 319-329. A FREEZE-DRY TECHNIQUE FOR THE PLANT COLLECTOR RICHARD A. HOWARD During a visit to the United States several years ago Dr. Hans-Christian Friedrich, Oberkonservator at the Munich Botanic Garden, described his work with the extremely succulent plants of Southwest Africa. In reply- ing to a question on the preparation of herbarium specimens from such materials he reported on the success they had had in freezing the plants from a greenhouse collection in Munich and then drying the plant specimens in a plant press in the fashion familiar to collectors. The implication was made that the liquid from the plant tissues nearly “flowed out” of the defrosted specimens in the drying process. It was my feeling at that time that the technique might be useful also in meeting two distinct problems I had en- countered in collecting plants in the American tropics. The first concerned the preparation of some of the fleshy or latex-possessing tropical plants encountered in field work in the Caribbean Islands. Specimens of the Cactaceae, Euphorbiaceae, Polygonaceae and Vitaceae usually dry slowly, fragment at nodes or articulations, discolor, become distorted in shape, or adhere to the drying papers. Sug- gestions have been made previously to overcome these problems involving treatment of specimens with solutions of formaldehyde, alcohol or chemical fixatives, or boiling in water. All of these techniques have been used at times with varying degrees of difficulty in the field and of success in the preparation of dried specimens. The second problem occurs in the course of field work when the amount of rare or unusually high quality material to be pressed and dried exceeds the capacity of the drying apparatus. In tropical areas this material cannot be stored fresh in plastic bags or held loosely in buckets of water or even lightly contained in improvised presses without some deterioration of the specimens. In their Manual for Tropical Herbaria (Regnum Vegetabile 39: 45-48. 1965) Fosberg and Sachet summarize the various techniques that have been 410 1968] Freeze-Dry Technique — Howard 411 developed in the tropics of both hemispheres for collecting plants in quantity and sending the material back to a dis- tant base for arrangement later into individual specimens and for drying. These techniques call for arranging the specimens in papers alone or in presses and placing such bundles in plastic sheets, tubes or containers, or in metal cans to be sealed after alcohol, formaldehyde or other chemicals have been used to soak, paint, dust or immerse the specimens. Both of these problems were encountered again during the summer of 1967 in the course of field work in St. Kitts and in Puerto Rico. Mr. Robert Wadsworth, who was col- lecting representative specimens on the former island as part of a study of the dwarfed elfin vegetation in the Antilles, worked with very limited drying facilities. At one stage of the operation a chest-type household freezer was made available through the courtesy of Mr. Colin Napier of St. Kitts. The plant specimens were prepared in the usual manner and the portion of the day’s work which could not be accommodated on the small heater frame were placed in the freezer. Eventually, when space became available, the frozen press was removed from the freezer and placed directly on the heater. Now that all the specimens are mounted for insertion in the herbarium the frozen, then dried, specimens are indistinguishable in most cases from those dried directly from the field. In Puerto Rico we had the opportunity of using the same technique while also making some comparative tests of the freezer-heater combination. Plants known to be difficult to prepare as specimens were deliberately chosen for trial. Mr. Joseph B. Martinson, who has supported many peculiar requests in relation to our study of the vegetation on Pico del Oeste, tolerated the use of his chest type of freezer. The cooperation of Dr. and Mrs. Richard J. Wagner and of Mr. Luis Maldonado helped complete the studies. Our attempts to hold general collections for several days by freezing them and then completing the drying process were successful. There was some indication that the drying 412 Rhodora [Vol. 70 process was slower than with freshly pressed specimens, but for the majority of collections the specimens prepared by the two methods were indistinguishable. Specimens of certain native plants from the Luquillo Mountains and of certain species under cultivation by Mr. Martinson and at the Pennock Nurseries in Hato Rey were prepared in equal sets for comparative processing studies. One complete set was placed in a press and dried imme- diately, using electric space heaters with low-speed blowers. Other sets were prepared in presses and frozen for 48 hours. One of these was allowed to defrost to air temperature before it was placed on the heater, while a comparable set was placed on the heater directly from the freezer. The fresh material which was placed directly over the heat without freezing dried from the lower side of the press, as expected, unless the specimens were rotated. The loss of liquid from the specimens was gradual and the sepa- rators rarely became very wet. The frozen material, how- ever, tended to lose liquid quickly when defrosted so that the papers, blotters, or separators became saturated. Unless these were changed frequently the specimens often de- veloped a growth of mold and the presses ultimately required 10-25% more time on the heater than did the fresh material. Recent correspondence with Dr. Friedrich reveals that he, too, experienced the problems of rot and fungus attack, which he controlled by soaking the press with alcohol after the first 24 hours of drying. Only after several unsatisfactory attempts to use the freezer as an aid to plant collecting did we finally develop the technique and knowledge necessary to recommend it to those collectors fortunate enough to have a freezing com- partment handy. For leathery or succulent specimens or those with thickened portions such as the large inflores- cences of the Araceae or Musaceae, the stems or rhizomes of the Zingiberaceae, the rosette bases of the Bromeliaceae, or the pseudobulbs of the Orchidaceae, the freezing tech- nique has real advantages. The specimens can be prepared whole or cut longitudinally and frozen either in newspaper 1968] Freeze-Dry Technique — Howard 413 in the freezing compartment of a refrigerator, with dry ice in a chest or box, or in a loose press in a larger freezing unit. The frozen material is then returned to air tempera- ture with the press loosely arranged. The liquid will often drain out of the specimens by gravity and the wet papers of the press must be changed. They now can be placed in a press and the pressure increased slightly or gradually for several hours to squeeze out more liquid before the speci- mens are finally placed on dry papers and arranged in a press for drying over heat. These specimens will dry more rapidly than fresh material since the liquid has been re- leased from all of the tissues simultaneously. With this preliminary treatment the actual drying time is reduced 25 to 50%. The benefits of this method of handling certain kinds of plant materials are evident on a comparative basis. There is a noticeable reduction in the fragmentation of stems at nodes, the disjunction of leaves from stems, the separation of parts of compound leaves, and the fragmentation of in- florescences. The parts of inflorescences of plants of Heli- conia, and Musa or various members of the Araceae or Zingiberaceae remain distinct, and single flowers or fruits can be lifted free when dry. This result contrasts dramatic- ally with the agglutinated mass that is usually formed when the specimens are dried directly over heat. There is also less tendency for the specimens to adhere to the pressing papers. Succulent stems or woody ones which contain large amounts of liquid also dry more quickly after being frozen and tend to retain a more natural shape and color. There is much less distortion of fleshy parts of the plants if the pressure of the press is not excessive. Tracings were made of leaves of Alloplectus and of Begonia before and after drying and freeze-drying. Although both species showed the translucent characteristic of frozen material, the final leaf shape was less distorted and more truly representative of the fresh condition than was the material dried directly. The translucent condition of frozen-dried material even has an advantage in revealing more easily the vascular patterns. 414 Rhodora [Vol. 70 In general, there was also a lessening of the darkening in those species which tend to darken on drying if the speci- mens had been frozen. There were, however, some notable exceptions. The following materials have been treated by freeze-dry technique. Specimens cited bear the collecting numbers of the author and a representative set is deposited in the herbarium of the Arnold Arboretum, Harvard University. Alocasia sp. Araceae (17041). A stout herb under cultiva- tion with very thick inflorescence and rhizome. The stems, inflorescence, and petiole sections retained a more charac- teristic shape after freezing. Alloplectus ambiguus Urb. Gesneriaceae (16641, 16815). An epiphytic herb with succulent stems and foliage. The frozen specimens have a translucent appearance, while heat- dried specimens tend to darken. The fleshy fruit is much less distorted after freezing. Anthurium dominicense Schott. Araceae (16642). An epiphytic rosette-forming plant with succulent and leathery leaves. Heat-dried specimens turn a pale brown color, while freezing causes a blackening of the leaves. The rhizomes retained more natural shape after freezing. All specimens proved to be very susceptible to fungus attack during drying. Begonia decandra Par. Begoniaceae (16939). A ter- restrial herb with normally pink-colored translucent succu- lent stems and leaves. Heat-dried specimens tended to darken, while those frozen before heat drying became more translucent and dried a lighter color. Fragmentation of the leaves from the stems or of the inflorescense was con- spicuously less in the material which had been frozen. Caryota sp. Palmae (16912). A cultivated stout cane- type fishtail palm. Both inflorescences and leaves were prepared. The frozen leaves retained a green color while the fresh material darkened. The flowers and young fruits are retained on the rachi much more securely in the frozen material. The mature flowers, however, fell from the rachus in both methods of preparation. 1968] Freeze-Dry Technique — Howard 415 Cissus trifoliata L. Vitaceae. An herbaceous vine which has succulent leaves and stems. The fresh dried specimens showed the fragmentation botanists have come to expect with most tropical species of Cissus. A considerable reduc- tion in fragmentation was evident in the specimens which had been frozen. The leaflets were retained intact in most cases and the stem fragmented only at the very young nodes. Clusia grisebachiana (Pl. & Tr.) Alain. Guttiferae (16632). Small tree with thick coriaceous leaves and copi- ous latex in tissues of leaf and stem cortex. Fruits thick and inflorescence commonly fragile. Specimens dried notice- ably faster after freezing and retained a clearer yellow-tan color. The fragility was not reduced in any noticeable manner. Epidendrum | ramosum Jacq. Orchidaceae (16640). Epiphytic. This species of orchids tends to break apart in normal drying. The frozen specimens retained their leaves in more satisfactory numbers, but the specimens were abnormally translucent when dried. Forsteronia portoricensis Woodson. Asclepiadaceae (16637). This is a rampant vine with heavy semi-fleshy leaves, abundant latex and deep carmine red flowers. The fresh-dried specimens showed the dark brown or black dis- coloration often obtained with plants having a latex. The specimens which had been frozen first became an unexpected greenish-brown color. Guzmannia berteroniana (R. & S.) Mez Bromeliaceae (16638), An epiphytic rosette plant with conspicuously bracted inflorescence. Following the usual practice the rosette was reduced to one quarter the original size before processing. In some specimens the inflorescence was split down the middle before drying. This plant does not produce beautiful herbarium specimens with any technique. The fresh-dried material showed the straw-brown color charac- teristic of most herbarium specimens. The frozen material retained a greenish cast. The colorful bracts and flowers discolored in both processes but the parts were more easily separated in the frozen specimens. 416 Rhodora [Vol. 70 Hedyosmum arborescens Sw. Chloranthaceae (16633). A tree with fragile branches, abundant liquid in the stems, succulent leaves, and white fleshy perianth parts to the fruit. All Caribbean material of Hedyosmum shrinks notice- ably in drying and turns a characteristic chocolate brown. In this test the frozen material showed a slight translucence not noticeable in heater-dried material. Heliconia bihai L. Musaceae (16643). A clump-forming herb with succulent petioles and a thick distichous inflores- cence. Material was prepared of the leaf blade and the petiole base, and of the inflorescence, which was split longitudinally. Heliconia specimens are notoriously difficult to dry. Several collectors have attempted to boil the speci- mens before drying, with very unsatisfactory specimens resulting. In this test all frozen material dried much more rapidly and made much better looking specimens. The frozen leaves retained a better color. The frozen inflores- cense was flatter, but the individual flowers could be sepa- rated easily and the fruits remained distinct, very much in contrast to the heater-dried material. Hillia parasitica Jacq. Rubiaceae (16634). An epiphytic or climbing succulent woody plant with pendant branches. The long white tubular flowers are fleshy and the leaves are thick. The specimens dried by heat alone showed the characteristic black color of most herbarium specimens, while the flowers are an ivory white color. Following freezing, however, the leafy specimens were less discolored and a lighter green color suggesting a translucence, while all of the flowers were either pale tan or brown with a definite translucent aspect. Musa sapientum L. Musaceae (16938). A succulent herb with massive inflorescence. Herbarium specimens of banana inflorescences are infrequent in herbaria. Very young inflorescences were split lengthwise for this test, while older ones were separated into bract and flower components for both pistillate and staminate flowers, The heater-dried material discolored and became distorted in shape. The frozen material dried more quickly, showed similar dis- coloration but much less distortion. 1968] Freeze-Dry Technique — Howard 417 Philodendron aff. bipinnatifidum Araceae (16928). A heading type of Philodendron. Inflorescences were split and frozen, then defrosted and placed in press for drying. Regular drying methods produced much-distorted specimens in which the flowers and fruits were tightly agglutinated. Freezing produced equally darkened specimens, slightly less distorted but with separate flowers and young fruits. Leaves dried and darkened in both techniques. Philodendron krebsii Schott. Araceae (16644). A high- climbing herbaceous vine with succulent stems and leaves and a fleshy inflorescence. Normal heater drying produced the usual blackened specimens and agglutinated inflores- ences. Material which had been frozen before drying was tan or light brown in color and the spathe and spadix sepa- rated more readily for dissection. Pitcairnia angustifolia L. Bromeliaceae (16636). A rosette-forming terrestrial plant with stout inflorescence axil and red succulent fruits. Heater-dried material turned a straw-brown color, while the frozen and dried material retained the natural glaucous-gray color. In flower and fruit, however, the plant parts showed better natural pig- mentation when dried directly. Plumeria alba L. Apocynaceae. The stout stems of Plumeria contain copious amounts of a low viscosity latex which causes specimens dried normally to adhere to paper. The flowers are fleshy and in normal drying techniques discolor slightly and usually fall free from the inflorescence. The Puerto Rican specimens tend to have the leaves more tightly curled or involute than those of other West Indian populations. The frozen dried material produced noticeably flatter leaves with a slight degree of translucence. The frozen flowers were conspicuously translucent but were not held on the inflorescence to any greater degree. Frozen material generally did not adhere to the drying papers. Vriesia sintenisii (Baker) Sm. & Pittend. Bromeliaceae (16639). An epiphytic plant with a rosette of leaves and a stout erect rhizome. This species under normal prepara- tion does not make attractive specimens. Those dried from 418 Rhodora [Vol. 70 fresh condition showed considerable shriveling of the leaves, contraction of the inflorescence, and general dark- ening of the specimen. The frozen material, in contrast, retained a more natural form of rosette and leaves, and all parts showed more natural pigmentation. Not every expedition to a tropical area will be fortunate enough to have a freezer with space available at a base camp. When such circumstances do occur or when it is necessary to prepare unusually succulent materials in other surroundings, the freezer may aid the collector in holding massed collections and in speeding the drying while im- proving the technique for processing difficult materials. The support of a grant from the National Science Founda- tion (GB-3975) for the basic studies of the vegetation of Pico del Oeste, Puerto Rico, is gratefully acknowledged. The present observations were made during the course of the ecological study. ARNOLD ARBORETUM JAMAICA PLAIN, MASSACHUSETTS NOTES ON ROCKY MOUNTAIN CARICES F. J. HERMANN In the preparation of a Manual of the Carices of the Rocky Mountains and Colorado Basin (to be published in the Agriculture Handbook series of the U. S. Department of Agriculture) the need for a change in status for several taxa and for proposing two new varieties became apparent. These are as follows: Carex foenea Wiild. var. tuberculata F. J. Herm., var. nov. A varietate typica recedit perigyniis ventraliter tuber- culatis. This variety differs strikingly from the typical form of the species in having the perigynia ventrally granular- tuberculate, much as in C. potosina Hemsl. of Mexico. It is a variable characteristic in the present taxon, being sometimes very prominent but in other collections barely perceptible. Since it appears to be geographically segre- gated, apparently never occurring in eastern United States where the species is not uncommon from Maine to New Jersey, varietal status seems most appropriate for it. SPECIMENS EXAMINED: WASHINGTON: Mission Peak head- waters, alt. 6000 ft., Wenatchee National Forest, Chelan County, G. Garrison, D. Wooldridge & R. Gibbons 80-56, July 24, 1956 (TYPE — US; isotype — USFS); COLORADO: Rio Grande National Forest, alt. 10,800 ft., D. C. La Font 18, (USFS) ; north shore of Trappers' Lake, Garfield County, alt. 9500 ft, F. J. Hermann 5433, (US); NEW MEXICO: West Fork of Tusas River, Carson National Forest, alt. 9500 ft, R. F. Copple 240, (USFS); ALBERTA: in sand under Jack pines, 15 mi. N. of Fort Saskatchewan, G. H. Turner, July 2, 1937 (US). Carex vallicola vor. rusbyi (Mack.) F. J. Herm., comb. nov. C. rusbyi Mack., Smithson. Misc. Coll. 657: 2. 1915. Intermediates between this plant of the mountains of Arizona and New Mexico and the more wide-ranging C. vallicola are too frequent to warrant its retention in specific rank, 419 420 Rhodora [Vol. 70 Carex microptera Mack., var. crassinervia F. J. Herm., var. nov. A forma typica differt perigyniis ventraliter valde 5-7- nerviis. In typical C. microptera the perigynia are lightly few- nerved ventrally, the nerves often being evident only at the base. The present plant is in striking contrast to this with its very stout 5-7 ventral nerves. SPECIMENS EXAMINED: COLORADO: open parks, basin below Engineer Pass, alt. 11,000 ft., Ouray County, Colorado, W. M. John- son 594, Aug. 14, 1967 (TYPE — Us); Wyoming: moist meadow (Bear- trap Meadows) at edge of forest, west of Circle Drive road, summit of Casper Mountain, alt. 7900 ft., Natrona County, F. X. Jozwik 262 (USFS). Carex luzulina Olney var. ablata (Bailey) F. J. Herm., comb. nov. C. ablata Bailey, Bot. Gaz, 13: 82. 1888, not Boott apud Miq. 1867. The great frequency of transitional forms between this and typical C. luzulina make it impractical to attempt to maintain it in specific rank as was done by Mackenzie. Carex parryana Dewey, var. brevisquama F. J. Herm., var. nov. "C. aboriginum M. E. Jones", Mack., N. Am. FI, 18: 364. 1935, not M. E. Jones, 1910. A varietate typica recedit spica terminali staminata, spicis femineis 4-5 mm. latis, squamis femineis perigyniis conspicue brevioribus, perigyniis 2.75-3.5 mm. longis, 1.75- 2.5 mm. latis. Dr. Arthur Cronquist, in his studies of the flora of the Great Basin, first noted the discrepancy between Macken- zie's "Carex aboriginum" and Jones’ type of that species, which is still the only collection known, and called it to the attention of the writer. Mackenzie's concept may be sepa- rated from typical C. parryana by the following key: 1968] Carices — Hermann 421 Terminal spike usually gynaecandrous; pistillate spikes 2-3 mm. wide; pistillate scales mostly equaling to exceeding and concealing the perigynia; perigynia 1-1.5 mm. wide, 2-2.5 mm. long ad ut audi POA O NS ano terc tail Eesen C. parryana Terminal spike staminate; pistillate spikes 4-5 mm. wide; pistillate scales 1/2-2/3 the length of the perigynia; perigynia 1.75-2.5 mm. wide, 2.75-3.5 mm. long .................. C. parryana var. brevisquama Carex heteroneura W. Boott var. epapillosa (Mack.) F. J. Herm., comb. nov. C. epapillosa Mack. in Rydb., Fl. Rocky Mts. 138. 1917. This and the following two sedges are amply distinct from C. heteroneura in their extreme forms, but inter- grading forms between all four taxa occur with such fre- quency that attempts to maintain them in specific rank have been impractical. Carex heteroneura var. chalciolepis (Holm) F. J. Herm., comb. nov. C. chalciolepis Holm, Am. Jour. Sci. IV. 16: 28. 1908. Carex heteroneura var. brevisquama F. J. Herm., var. nov. C. atrata auct. Amer. (including Mackenzie, N. Am. Fl. 18: 371. 1935), not L., Sp. Pl. 976. 1753. True Carex atrata L. is Eurasian and is not known to occur in America. It belongs to the group in Section Atra- tae having subinflated, scarcely flattened perigynia. The American plant that has been passing as C. atrata, although it often merges more or less into C. heteroneura and the above two varieties, is sufficiently different from all of these in its typical form to merit varietal recognition. FOREST SERVICE HERBARIUM RANGE AND WILDLIFE HABITAT RESEARCH U. S. DEPARTMENT OF AGRICULTURE WASHINGTON, D. C. BOOK REVIEW The Evolution and Classification of Flowering Plants by Arthur Cronquist, Senior Curator, The New York Botanical Garden. Hough- ton Mifflin Company, Boston. 1968. i-xii, 1-396, with glossary. $6.95. As the title implies, this book is conveniently divided into two sections. The first, which amounts to approxi- mately one-third of the work, is concerned with taxonomic principles, the origin of the angiosperms, and the evolution of characters. The second and by far the longer section is devoted to the author’s classification of the angiosperms (Magnoliophyta) and to explaining in some detail his rea- soning in erecting the proposed system. The first chapter deals with certain taxonomic concepts such as the natural system, character values, monophylesis, parallelism, survival value (adaptive significance) of certain characters, and the grouping and ranking of taxa. This chapter, as the other two in the first section, ends with a short list of selected references. In general, the discussion is interesting, although the topics are highly selective. One might disagree that the best definition of taxonomy is “a study aimed at producing a system of classification of organisms which best reflects the totality of their simi- larities and differences", even though the author readily admits that "taxonomy might be variously defined, accord- ing to the definer ...”. A biochemist might also disagree with the statement that we know perhaps more about man than about any other species in light of the recent statement of one molec- ular biologist that we now know at least one-fifth and maybe more than one-third of all the chemical reactions in Escherichia coli. On the other hand a palynologist might take the author to task for the statement on page six that identifications of fossil pollen “cannot provide new or independent infor- mation on the evolutionary diversification of a group, or on the transitions between groups; they merely document the existence of a modern group at a particular time in the past”. 422 1968] Book Review — Walker 423 Among the major conclusions of this chapter are the following: taxonomy is based on multiple correlations, character value is determined a posteriori, the presence of a character is likely to be more important than its ab- sence, and a proper taxonomic system must reflect evolu- tionary relationships. The second chapter very briefly characterizes the angio- sperms, reviews the fossil record (or lack of it), and surveys the groups that have been suggested as possible angiosperm precursors: conifers, the Gnetales, and various cycadophytes — cycads, the Bennettitales, the Caytoniales, and the seed ferns. After eliminating most of the above groups as potential ancestors, the author concludes, as so many others have in the past, that although there is a vast abyss between the seed ferns and the angiosperms, the former are the best candidate for the ancestors of the angiosperms, The third and final chapter of the first section deals with the evolution of characters. The topics covered are the determination of primitive characters and evolutionary trends, the characters of the primitive angiosperms, and finally a relatively lengthy discussion of the evolutionary morphology of the angiosperms beginning with habitat and growth habit, proceeding through leaf structure, vascular structure, inflorescences, and flower structure, and ending with fruits and seed dispersal. The chapter ends with another discussion of the non- adaptive value of the majority of the characters of the higher angiospermous categories. A previous dicussion of the same topic is to be found under the heading “Ecologic Niches, Survival Value, and Taxonomic Groups” in the first chapter. The author is well known for his somewhat unorthodox views concerning this subject and for various papers on orthogenesis in the angiosperms. He himself admits that this viewpoint is “not in favor among most modern students of evolutionary mechanisms, who feel that any evolutionary trend must ultimately be related to survival value". Perhaps this is the reason that the term 424 Rhodora [Vol. 70 “orthogenesis is not to be found even once in the book, such phrases as “evolutionary momentum” or “predisposed evolution” being used in its place. Evidently the author has not recognized the possibility that while the higher angiosperm categories per se may not have evolved in response to a certain adaptive zone (fide Simpson), the majority of the individual characters of the higher categories may themselves be adaptational and have their origin in natural selection. Be that as it may, the discussion in the first part of the book (and indeed in the whole work) airs a number of good ideas which are expressed in a refreshing style. The following excerpts are taken as examples: Evolutionary trends have all degrees of stability, from those which are so vague and subject to reversal that they can scarcely be recognized as trends, to those which fasten an inescapable grip on the destiny of the group. p. 62. My nose tells me that the ethereal oils in the Magnoliidae are all chemically allied, and different from the ethereal oils found in other groups of flowering plants, but I have not found an account of the chemistry. p. 135. To borrow an analogy which Stebbins has used in another con- text, evolutionary increase in parts of relatively small and definite number follows a goatpath, whereas reduction follows a paved highway. p. 313. Anyone who has pushed a lawnmower should recognize the significance of the intercalary meristem in permitting a plant to withstand grazing. p. 344. The last two-thirds of the book is a rather lengthy expo- sition of the author's system of angiosperms. Each recog- nized order is discussed, followed by a very useful synoptical key to the families of the order. This is followed by selected references of predominantly recent vintage. Six subclasses of dicots and four monocots are recognized. The dicot subclasses and roughly their composition are as follows: 1968] Book Review — Walker 425 1) the Magnoliidae, which is a slightly restricted concept of the Englerian order Ranales, 2) the Hamamelidae, which is basically the Amentiferae. 3) the Caryophyllidae or Centrospermae. 4) the Dilleniidae, which is primarily the Englerian order Parietales with the ericaceous and ebenaceous Sympetalae thrown in along with a few other groups such as the Malvales, the Lecythi- daceae, and the Capparales. 5) the Rosidae, which includes among others the sapindalean- geranialean complex, the Myrtales, the Euphorbiales, and the Umbellales. 6) the Asteridae, which includes the epigynous Sympetalae (Ru- biaceae, Compositae, etc.) and the remainder of the hypogynous Sympetalae other than the Ericales and Ebenales (i.e. the Contortae and the Tubiflorae). A few comments on certain specific points of the system will now be made. It is evident that the author has made use of a number of personal communications and pre-publication manu- scripts. It is unfortunate, however, that he was not able to discuss the Hamamelidaceae with Dr. A. L. Bogle, who for the past several years has been investigating the floral morphology and vascular anatomy of the family. Unpub- lished evidence available from this research strongly sug- gests that several genera of the family (e.g., Distylium, Sycopsis) possess bisexual pseudanthia (thus contradicting the key to the subclasses of Magnoliatae, p. 131), that the petals of the Hamamelidaceae are staminal in origin rather than of the probably more primitive tepalar origin (result- ing from the differentiation of a primitively undifferentiated perianth of tepals into a petaloid upper part and a sepaloid lower part — as in the Magnoliaceae), that the flowers of the family were primitively apetalous, and that the sup- posedly primitive position of the genus Disanthus is at best questionable. In light of this knowledge one might justifiably disagree with the author's statement that the Trochodendrales “have already progressed farther in floral reduction than the Hamamelidaceae, which have unmistakable (though some- what reduced) petals, in contrast to the complete absence 426 Rhodora [Vol. 70 of petals in the Trochodendrales". Conceivably the sepaloid perianth in Tetracentron may represent the vestiges of a once undifferentiated perianth while both the Trochoden- drales and the more primitive members of the Hamame- lidaceae are apetalous. Thus the problem of how to account for the petals in some members of the otherwise more advanced Hamamelidaceae vanishes when one considers that these petals may have arisen secondarily from apetalous ancestors by sterilization of stamens and do not represent petaloid vestiges of a once undifferentiated perianth, The author’s concept of the Rafflesiales as having “rather large to very large flowers” leads him to conclude among other things that a notable difference between that order and the Santalales is the size of the flowers. However, genera with very small flowers such as Apodanthes and Pilostyles immediately come to mind, even though the largest flower known is in the genus Rafflesia. Also this group is not composed solely of root parasites as the author implies, since certain members occur on the stem and minor branches of their host. The author’s treatment of the monocots is somewhat novel, but I am not in a position to comment on the Com- melinidae and the Liliidae. The placement of the Juncales warrants some attention and is an interesting treatment of the group. Unfortunately, the book is too short to treat familial relationships in any detail. It is to be regretted that a further treatment at the family level as is done in the latest book by the Russian phylogenist Takhtajan was not possible. A chart showing the over-all classification at the ordinal level would be greatly desirable. Attention should also be called to the fact that families which the author does not accept are often not mentioned in the body of the discussion and their placement in the system can only be determined by reference to the list of families and orders at the end of the book. 1968] Book Review — Walker 427 The great number of personal communications and pre- publication manuscripts alluded to will certainly help to keep the book up-to-date longer than usual. There is a good coverage of new chemical data as they relate to phylogeny and the reader is especially referred to the excellent dis- cussion of betalains in the Caryophyllidae (Centrospermae). One of the major accomplishments of both Cronquist and Takhtajan is the erection of six rather large and for the most part probably natural subclasses within the dicots. The erection of these major divisions greatly aids one in grasping the prominent evolutionary trends within the dicotyledons. At last the major patterns are emerging and the work of the future phylogenist will consist mainly of deciding where to place a certain family or order here and there, i.e., it will be a matter of filling in the details and confirming previous ideas. Although obviously opinionative, the author’s frankness on certain morphological questions is quite refreshing. This includes his ideas on such topics as the following: 1) that the staminal fascicles in many of the Dilleniidae sensu Cronquist represent fusion and reduction rather than splitting and multiplication. 2) that the tendency in the Apocynaceae-Asclepiadaceae for a separation of the pistils from the base upwards is secondary rather than primary. 3) that the centrifugal maturation of the stamens in the Winter- aceae (or in any family for that matter) is not phylogenetically significant; rather it is the centrifugal versus centripetal direc- tion of the initiation of the staminal primordia which has phylogenetic significance. In conclusion, the book should be read by every botanist who is interested in the latest thoughts concerning the phylogenetic treatment of the flowering plants. It is the most detailed treatment of a modern “Besseyan-based”’ system of angiosperm classification in English, being second only to the more detailed work of Takhtajan (unfortunately in Russian). The book should also serve admirably for plant taxonomy courses as a supplement to the usual 428 Rhodora [Vol. 70 morphology and taxonomy texts. It will serve to put the orders and families of angiosperms into better perspective in the light of modern phylogenetic thought, which is something that is often sorely neglected or more or less omitted entirely in many introductory taxonomy courses. As the editor’s introduction states, it serves to fill a place that was previously empty for the English reader. In the final analysis, it combines the remnants of the out-worn system of Engler and Prantl with a modern re- interpretation of the Besseyan system. Like the phoenix arising from its ashes, it recreates in the light of modern knowledge another stage in the ever continuing search for the best system to reflect the evolutionary history of the angiosperms. JAMES W. WALKER GRAY HERBARIUM HARVARD UNIVERSITY 02138 BIOGRAPHY OF THOMAS NUTTALL; A REVIEW WITH BIBLIOGRAPHY: 2? RONALD L. STUCKEY Thomas Nuttall (1786-1859) is frequently regarded as one of North America’s most important, widely traveled, pioneer naturalists. Miss Graustein’s full-length biography presents for the first time in scholarly detail the career of this brilliant naturalist, born in Yorkshire, England, who at the age of 22 emigrated to Philadelphia to begin his scientific explorations. These travels through the Old North- west, up the Missouri River, down the Mississippi River, through southeastern United States, down the Ohio River, up the Arkansas River, across the Rocky Mountains, down the Columbia River, over to Hawaii, and into southern California are discussed at length and mapped. His physi- cal hardships in the American wilderness, vivid descriptions of frontier scenes and living conditions of the peoples, personalities of the companions with whom he sometimes traveled, information about many of the plants he collected and described, and the geological and archaeological phe- nomena he observed are all presented in great detail. Although it may be inferred from the title that this book treats only Nuttall’s travels in America, the contrary is the case, for the author adds considerable information on his family affairs, his several visits to England, and his horticultural pursuits in England during his later life. In addition, major portions of five chapters deal with his 11 years at Harvard University, a chapter in his life previously little known. For a man who left scattered on two conti- nents bits and pieces of information concerning his personal life in letters, itineraries, travel narratives, or scientific papers and books, Miss Graustein has woven these facts remarkably effectively among the reports of his exploring ‘Thomas Nuttall Naturalist Explorations in America 1808-1841. Jeannette E. Graustein. Harvard University Press, Cambridge. 481 pp. 1967. $11.95. “Paper No. 746 from the Department of Botany, The Ohio State University, Columbus 43210. 429 430 Rhodora [Vol. 70 expeditions. Through the book she frequently lets Nuttall speak for himself in many quotations from unpublished letters. This biography also presents another look at the story of American botany during the period 1808-1841. Not only does the reader learn in detail about Nuttall’s botanical pursuits in exploration and publication, but also about the many botanical friends and associates he had in America and England, perhaps the most important being William Baldwin, Benjamin Smith Barton, John Bradbury, Zaccheus Collins, William Gambel, Henry Muhlenberg, Charles Pickering, and Frederick Pursh. These naturalists provide much of the background against which the story of Nuttall’s career is enacted. For American botanical historiography, this book is comparable to other biographies of American plant taxonomists, such as McAllister on Amos Eaton; Rodgers on William Starling Sullivant, John Torrey, John Merle Coulter, and Liberty Hyde Bailey; Dupree on Asa Gray; and Berkeley on John Clayton. Of Nuttall’s botanical associates, probably more new in- formation is given on Zaccheus Collins than any of the others. As indicated in Miss Graustein’s extensive foot- notes, this information apears to be gleaned primarily from the numerous letters written to Collins now preserved in the library of The Academy of Natural Sciences, Phila- delphia. At times, however, the source of the facts regard- ing Zaccheus Collins is not always presented. For example, on page 24, there is a quotation saying that Collins is “A retentive gulph ...” without source, and on pages 272-273, the sale of Collins’ Herbarium including dates, bidders, purchasers, and prices is given but not documented. The former item was probably quoted from Rev. L. D. von Schweinitz’s April 19th, 1821, letter to John Torrey (Mem. Torrey Bot. Club 16: 142. 1921) and the latter data were probably taken from Daniel Steinhauer’s and C. S. Rafi- nesque’s unpublished letters to Rev. L. D. von Schweinitz now in The Academy of Natural Sciences Library, Phila- delphia. 1968] Nuttall Biography — Stuckey 431 There appear to be some inconsistencies in the use of scientific names of plants, especially in cases where Nut- tall’s names are no longer in use because of either subse- quent taxonomic and/or nomenclatural changes. Citing later new combinations instead of the names of new species that Nuttall published is confusing. I point out the follow- ing items (in chapters 5 and 7) which pertain to Nuttall’s plants from the 1810-1811 expedition in the Great Lakes region and Missouri River valley and from his 1816 trip in the Ohio River valley. Among a list of five of Nuttall’s new species from the Mackinac Straits region mentioned on page 51, Miss Graustein includes Cirsium undulatum (Nutt.) Spreng. Actually this species was published by Nuttall as Carduus undulatus (Gen. 2: 130), but later transferred to Cirsium by Sprengel. Other instances of this kind of re- porting occur on page 71 for Opuntia fragilis (Nutt.) Haw., actually Cactus fragilis (Gen. 1: 296), Mamillaria vivipara (Nutt.) Haw., actually Cactus viviparus (Gen 1: 295), Lactuca ludoviviana (Nutt.) Riddell, actually Sonchus ludo- vicianus Nutt. (Gen. 2: 125). In other instances, the author follows an accepted procedure by giving the recog- nized new combination or new name in square brackets following Nuttall’s published name, for example, p. 37, Hedeoma glabra [Satureja arkansana (Nutt.) Briq.]; p. 98, Enslenia albida [Ampelamus albidus (Nutt.) Britt.] ; p. 106, Dracocephalum cordatum [Meehania cordata (Nutt.) Britt.]; and p. 102 Liatris corymbosa [Carphephorus corymbosus (Nutt.) T. & G.]. Concerning some other records of plants mentioned in chapters 5 and 7, the following four cases are noteworthy: (1) The author believes apparently without evidence that Nuttall omitted the Euonymus [E. obovatus Nutt.] from his 1810 diary (p. 45, footnote 8). Recent evidence suggests that Nuttall first found the Euonymus on his 1816 foray into western Pennsylvania (Mich. Bot. 6: 92- 93. 1967). (2) Miss Graustein states that Nuttall obtained his Dalea enneandra at Prairie du Chien (p. 54). This species 432 Rhodora [Vol. 70 is described in Fraser's Catalogue without locality, but one would assume from the title of the catalogue that the plant came from somewhere in “Upper Louisiana" or on the "River Missourie.” Pursh (Fl. Am. Sept. 2: 741. “1814” [1813]) described a D. laxiflora from “Upper Louisiana" based on a collection by Bradbury and seen in his herbarium. Pursh placed D. enneandra of Fraser's Catalogue in syno- nymy. In his Genera (2: 101), Nuttall recognized Pursh's D. laxiflora but not the Fraser Catalogue D. enneandra. Nuttall gave the localities as “On the high hills and naked grassy plains of the Missouri, also on the banks of the Mississippi near the Praire du Chien.” Without further evidence one can only infer that Nuttall found the D. enne- andra at the latter location. Nuttall’s manner of treating the locality data suggests that this is another plant that he originally collected at Prairie du Chien, lost it, and recol- lected another one later in the Missouri region. This sug- gested interpretation of these data is consistent with my theory that Nuttall lost most of his plants on the 1810 trip, but still cited their localities in his Genera (Stuckey, Mich. Bot. 6: 81-94. 1967). (3) The author says (p. 58) that “the species of poppy [Stylophorum diphyllum (Michx.) Nutt.] on which he [Nuttall] based his genus *Stylophorum” came from “Near St. Louis." Although the locality may have been near St. Louis, Nuttall wrote, “In the shady woods of Kentucky and Tennessee, also on the banks of the Missouri;" (Gen. 2: 7). Unless the author has other uncited documentation, one should not read added information into the locality data Nuttall gave in his Genera. (4) Concerning Nuttall's 1816 travels in the Ohio River valley (p. 106), Miss Graustein notes: “At Portsmouth he made a foray up the Scioto River where he saw Juglans nigra L." The reference given is Nuttall’s Sylva I, p. 41, but I fail to find this plant mentioned on that page or any comment concerning a trip up the Scioto River. Perhaps this note appears in another reference unknown to me. Miss Graustein believes that Nuttall's work in the Genera 1968] Nuttall Biography — Stuckey 433 was impaired because of his deficiency in taxonomic train- ing (p. 121-122). Although that may be true, she supports her viewpoint by saying that Nuttall “remained oblivious to simple conventions such as the retention of the original specific name when transfer is made to a different genus.” To state that Nuttall was ignorant of so simple a rule is completely unjustified. In Nuttall’s day there was no Code of Botanical Nomenclature, but only the guidelines provided by Linnaeus’ Philosophia Botanica and Critica Botanica. It was not until 1866, when the Botanical Congress com- missioned Alphonse DeCandolle to draw up some “Laws . ," which were adopted by the Congress in 1867, that there was any written “rule” regarding retention of name on transfer (see C. A. Weatherby, Am. J. Bot. 36: 5-T. 1949). Furthermore, it has not been until the present century that this procedure has been widely practiced. On the whole, the book is well documented by notes oc- cupying 54 pages, all placed at the back of the book. There is no concise alphabetically (by author) arranged list of con- sulted references, although a list of 15 papers concerned mostly with biographical sketches of Nuttall do appear on page 401. On pages 473-474 of the index there is a partial list (20 entries) of the titles only of Nuttall's own publica- tions. A more complete list of 51 titles usually with pages, dates, and place of publication can be found in Smith and Thieret (Leafl. West. Bot. 9: 33-42. 1959), an important reference not cited by Graustein. I believe Miss Graustein's biography, like some others in recent years concerned with botanists and perhaps with scientists in other fields, is defi- cient in its presentation or listing of the scientist's own published works. It seems to me that one of the minimum requirements of a biography would be the preparation of a complete chronologically arranged bibliography of the man or woman about whom an author is writing. In the botani- cal field, and especially in the area of plant taxonomy, where exact dates of publication are often extremely im- portant in settling questions of priority of plant names, a bibliography including full title, publication date, volume 434 Rhodora [Vol. 70 number, and page numbers accompanied by pertinent annotations, is essential. Usually this information can best be supplied by the biographer who is familiar with the many aspects and sequence of events of the scientist's life. As part of my research on Thomas Nuttall's contributions to plant taxonomy, I have been accumulating biographical references, as well as papers that discuss taxonomic and nomenclatural problems of his plants, his itineraries, corre- spondence, and/or relationships with other botanists. Many of these references are useful to the plant taxonomist, but some of them are not cited by Graustein for various rea- sons. It seems desirable to include below a bibliography concerned with Nuttall, his plants, and plant exploration. This list is not intended to be complete, but represents those references gleaned primarily from journals on taxo- nomic botany. Aside from the above criticisms of certain specific details, the student interested in biography, travel, and/or natural history in pioneer North America will find that this im- portant book adds another rewarding chapter in those fields. Written for both the scholar and the layman, this book should appeal to a wide range of interests. PUBLICATIONS ABOUT THOMAS NUTTALL AND His BOTANICAL WORK (Excluding Nuttall's Own Publications in Smith and Thieret, 1959) ANONYMOUS. 1860. Biographical sketch of the late Thomas Nuttall. The Gardeners Monthly 2: 21-23. (Graustein treats this reference with Thomas Meehan as the author; Pennell, 1936, believed Charles Pickering was the author.) 1937. Thomas Nuttall. Science 85(2201): Sci. Suppl. 18; Travels of early naturalist are traced and mapped. Sci. News Let. 31: 185, (Although not stated, each of these notes is a review of Pennell, 1936.) ALDEN, ROLAND H., and JOHN D. IFFT. 1943. Early naturalists in the Far West. Occ. Pap. Calif. Acad. Sci. 20: 1-59. (Nuttall, p. 42-46.) BARNHART, JOHN HENDLEY. 1909. Some American botanists of former days. Jour. N.Y. Bot. Gard. 10: 177-190. (Nuttall, p. 182.) 1935. The botanical correspondents of Schweinitz. Bartonia 16: 19-36. (Nuttall, p. 31-32.) 1968] Nuttall Biography — Stuckey 485 BEIDLEMAN, RICHARD G. 1956. The 1818-20 Arkansas journey of Thomas Nuttall. Arkansas Hist. Quart. 15: 249-259. 1960. Some biographical sidelights on Thomas Nuttall, 1786-1859. Proc. Am. Philos. Soc. 104: 86-100. (Two of Nuttall’s published obituaries are reprinted here, trans- lated from the French (Société Botanique de France 4: 759. 1860), and translated from the German (Flora 43: 16. 1860.) 1967. Thomas Nuttall Botanist of the American wilderness. Holticulture 45: 36-37, 45. [BIGELow, JACOB]. 1823. Review of Nuttall’s Journal of Travels into the Arkansa Territory. North Am. Rev. 16: 59-75. BooTrH, THOMAS JONAS. 1861. Mr. Thomas Nuttall, the Naturalist. The Settle Chronicle and North Ribblesdale Advertiser. Janu- ary 1 and February 1. (Reprinted by Pennell, 1938.) BRITTEN, JAMES. 1899. Bibliographic notes. XXI — Frasers’ Cata- logues. J. Bot. 37: 481-487. (Including reprint of Fraser's Cata- logue, p. 485-487.) CHICHESTER, H. MANNERS. 1895. Thomas Nuttall. Dict. Nat. Biog. 41: 277-278. COVILLE, FREDERICK V. 1899. The botanical explorations of Thomas Nuttall in California. Proc. Biol. Soc. Wash. 13: 109-121. (In- cludes list of new species based on Nuttall’s Californian collec- tions. ) CRONQUIST, ARTHUR. 1953. Notes on specimens of American plants in European Herbaria. Leafl. West. Bot. 17: 17-31. (Nomen- clatural notes on some of Nuttall's species in Artemisia, Chryso- thamus, Cirsium, Grindelia, and Senecio.) , Davin D. KECK, and BASSETT MAGUIRE. 1956. Validity of Nuttall’s names in Fraser's Catalogue. Rhodora 58: 23-24. [CUSHING, CALEB]. 1821. [Botany of the United States]. North Am. Rev. 13: 100-134. (Review of Nuttall's Genera, p. 116-118.) DuPREE, A. HUNTER. 1952. Thomas Nuttall's controversy with Asa Gray. Rhodora 54: 293-303. DURAND, ELIAS. 1860. Biographical notice of the late Thomas Nut- tall. Proc. Am. Philos. Soc. 7: 297-315. EIFERT, VIRGINIA S. 1965. Chapter 7, Nuttall, p. 101-122. In Tall Trees and Far Horizons. Dodd, Mead & Co., New York. 301 pp. EWAN, JOSEPH. 1950. Thomas Nuttall, 1786-1859, p. 273. In Rocky Mountain Naturalists. Univ. Denver Press, [Denver]. 358 pp. (Includes some references not cited here.) 1952, Nuttall’s Diary of 1810 and some inquirendae. Rhodora 54: 234-236. FERNALD, M. L. 1913. Nuttall's white sassafras. Rhodora 15: 14-18. 1942. Some early botanists of the American Philo- 436 Rhodora [Vol. 70 sophical Society. Proc. Am. Philos. Soc. 86: 63-71. (Nuttall, p. 65-67.) 1944. The confused publication of Monarda russe- liana. Rhodora 46: 491-493. FOSTER, ROBERT C. 1944. The publication-date of Nuttall's *Arkan- sas Flora." Rhodora 46: 156-157. GEISER, S. W. 1956. Thomas Nuttall’s botanical collecting trip to the Red River, 1819. Field & Lab. 24: 43-60. GRAUSTEIN, JEANNETTE E. 1951. Nuttall’s travels into the Old Northwest. An unpublished 1810 diary. Chron. Bot. 14: 1-88, + pl. 68-77. 1952. Audubon and Nuttall. Sci. Month. 74: 84-90. 1954. Nuttall in 1815. Rhodora 56: 253-257. 1956a. Nuttall’s quarrel with Pursh. Rhodora 58: 20-22. 1956b. Manuel Lisa and Thomas Nut- tall. Missouri Hist. Soc. Bull. 12: 249-252. 1961. The eminent Benjamin Smith Barton. Pennsylvania Mag. Hist. Biog. 85: 423-438. (Nuttall, p. 437-438.) 1964. Early scientists in the White Mountains. Appalachia n. ser. 30: 44-63. (Nuttall, p. 55-59.) GREENE, EDWARD L. 1890. Reprint of Fraser's Catalogue. Pittonia 2: 114-119. (Catalogue appears on p. 116-119.) HARPER, ROLAND M. 1904. The type-locality of Arenaria brevifolia. Torreya 4: 138-141. (Ohoopee River, about four miles from Reids- ville, Tattnall County, southeast Georgia.) HARSHBERGER, J. W. 1899. The botanists of Philadelphia and their work. Philadelphia. 457 pp. (Nuttall, p. 151-159, including a bibliography of 18 titles.) HUMPHREY, HARRY BAKER. 1961. Thomas Nuttall 1786-1859, p. 190-192. In Makers of North American Botany. The Ronald Press Co., New York. 265 pp. JACKSON, B. DAYDON. 1922. Thomas Nuttall (1786-1859). J. Bot. 60: 57. JEPSON, WILLIS LINN. 1934. The overland journey of Thomas Nut- tall. Madroño 2: 143-147. (Concerns Nuttall’s trip from Inde- pendence, Missouri, April 28, 1834, to Fort Vancouver, September 16, 1834; journey in California in 1836.) JONES, GEORGE NEVILLE. 1937. Thomas Nuttall. Botanist, orni- thologist, geologist. Little Gardens (Seattle) 8: 6-25. LEONARD, W. E. 1886. Some early Philadelphia botanists: Schweinitz, Nuttall, Rafinesque and Darlington. Minn. Acad. Nat. Sci. Bull. 13: 29-37, 1968] Nuttall Biography — Stuckey 437 LITTLE, ELBERT, L., JR. 1943. Notes on the nomenclature of Carya Nutt. Am. Midl. Nat. 29: 493-508. MCATEE, W. L. 1954. Thomas Nuttall. Nature Magazine 47: 541- 542, 550. McKELVEY, SUSAN DELANO. 1955. Botanical Exploration of the Trans-Mississippi West 1790-1850. The Arnold Arboretum of Harvard University, Jamaica Plain. 1144 pp. (Nuttall, chapter VI, p. 139-149; chapter VIII, p. 164-187; chapter XXIX, p. 586- 626; portions of chapter XXXV, p. 731-752.) OsrERHOUT, GEORGE E. 1907. Nuttall and Pursh and some early spring flowers of Colorado. Plant World 10: 80-84, OWNBEY, GERALD. 1952. Nuttall’s Great Plains species of Cirsium: C. undulatum and C. canescens. Rhodora 54: 27-35. PALMER, ERNEST J. 1927. On Nuttall’s trail through Arkansas. J. Arnold Arb. 8: 24-55. Parry, C. C. 1883. Early botanical explorers of the Pacific coast. Overland Monthly ser. 2, 2: 409-416. (Nuttall, p. 414-415.) PEASE, ARTHUR STANLEY. 1918. Nuttall and Pickering in the White Mountains. Rhodora 20: 39. PEATTIE, D. C. 1927. Nuttall, botanist and ornithologist. Nature Magazine 9: 37-98. PENNELL, FRANCIS W. 1936. Travels and scientific collections of Thomas Nuttall. Bartonia 18: 1-51. . 1938. An English obituary account of Thomes Nuttall. Bartonia 19: 50-53. . 1942. Botanical collectors of the Philadel- phia local area. Bartonia 21: 38-57; 22: 10-31. (Nuttall, 21: 44.) Powers, W. H. 1925. Some facts in the life of Thomas Nuttall. Science 62: 389-391. RAFINESQUE, C. S. 1819. Review of Nuttall’s Genera. Am. Month. Mag. & Crit. Rev. 4: 184-196. REVEAL, JAMES L. 1967. Notes on Eriogonum — III On the status of Eriogonum pauciflorum Pursh. Great Basin Nat. 27: 102-117. (Reference to Nuttall’s and Bradbury’s botanical work on the Missouri.) 1968. On the names in Fraser’s 1813 Catalogue. Rhodora 70: 25-54. , and VIRGINIA S. SPEVAK. 1967. Publication dates and current names of 144 names proposed in two 1848 Thomas Nuttall articles. Taxon 16: 407-414. RODGERS, ANDREW DENNY, III. 1942. John Torrey. A Story of North American Botany. Princeton Univ. Press, Princeton. 352 pp. (Contains numerous references to Nuttall and his various journeys.) SHEAR, C. L., and NEIL E. STEVENS, [ed]. 1921. The correspondence 438 Rhodora [Vol. 70 of Schweinitz and Torrey. Mem. Torrey Bot. Club 16: 119-300. (Many references to Nuttall in various letters.) SHINNERS, LLOYD H. 1955. Non-validity of Nuttallian names in Fraser's Catalogue. Rhodora 57: 290-293. 1956. Nuttall not the author of Fraser’s Catalogue. Rhodora 58: 281-289. SMALL, JOHN K. 1923. The needle palm — Rhapidophyllum hystrix. Jour. N.Y. Bot. Gard. 24: 105-114. (Note on Nuttall’s observation of this plant, with a brief sketch of Nuttall’s life in footnote, p. 111-112, by J. H. B[arnhart]. SMITH, C. EARLE, JR. 1957. A century of botany in America. Bar- tonia 28: 1-30 + 6 pl. (Nuttall, p. 10-14.) , and JOHN W. THIERET. 1959. Thomas Nuttall (1786-1859): An evaluation and bibliography. Leafl. West. Bot. 9: 33-42. SMITH, EDGAR Fans. 1927. Early science in Philadelphia. Pennsyl- vania Mag. Hist. & Biog. 51: 15-26. (Nuttall, p. 17, 18.) SPAULDING, PERLEY. 1909. A biographical history of botany at St. Louis, Missouri. II. Pop. Sci. Monthly 74: 48-57, 124-133, 240-258. (Nuttall, p. 52-57.) STEVENS, O. A. 1959. Bradbury and Nuttall, pioneer Dakota botan- ists. North Dakota Hist. Quart. 26: 159-169. STONE, WHITMER, 1934. Thomas Nuttall. Dict. Am. Biog. 13: 596, 597. STUCKEY, RONALD L. 1966. Thomas Nuttall's 1816 Ohio valley plant. collections described in his “Genera” of 1818. Castanea 21º 187-198. 1967. The “lost” plants of Thomas Nuttall’s 1810 expediton into the Old Northwest. Mich. Bot. 6: 81-94. 1967. Thlaspi tuberosum Nutt., a taxonomic synonym of Cardamine douglassii Britt. (Cruciferae). Rhodora 69: 460-465. TATNALL, ROBERT R. 1940. Nuttall’s plant collections in southern Delaware. Bartonia 20: 1-6. THIERET, J. W., and C. E. SMITH, JR. 1959. An English obituary account of Thomas Nuttall (concluding part). Bartonia 29: 10. [YouMANS, WILLIAM Jav]. 1895. Sketch of Thomas Nuttall. Pop. Sci. Monthly 46: 689-696. YOUMANS, WILLIAM JAY. 1896. Thomas Nuttall, p. 205-214. In Pioneers of Science in America. Sketches of their Lives and Scientific Work. D. Appleton & Co., New York. 508 pp. ZAHNISER, HOWARD. 1937. In January — Thomas Nuttall. Nature Magazine 29: 7. DEPARTMENT OF BOTANY, THE OHIO STATE UNIVERSITY, COLUMBUS 43210. THE LINUM BAHAMENSE COMPLEX C. M. ROGERS! Populations of flax, disjunct about 1000 miles from their nearest relatives in Mexico and western Texas, have been found on five of the Bahama Islands: Abaco, Andros, Eleu- thera, Grand Bahama and New Providence. Small (1907) recognized four species. Recently (1963) I reduced these to one species, L. bahamense, with three varieties. One of these, var. corallicola, is conspicuously different in being densely pubescent throughout. The other two, var. baha- mense and var. bracei are glabrous and were believed to be distinguished from one another by a series of more or less minor features. It was recognized, however, that, particu- larly on Grand Bahama, where the two varieties grow to- gether, numerous intermediates could be found. I have now had the opportunity to collect var. bracei from the type locality on Grand Bahama, var. corallicola from both Grand Bahama and Andros, and var. bahamense from Grand Bahama, Andros and New Providence; to grow them in the greenhouse; and to make a number of crosses. All of the populations have been found to have a haploid number of 34 chromosomes. The basic number for the segment of the genus to which L. bahamense belongs is n — 18, so that the island populations are interpreted as aneuploid derivatives from an unknown tetraploid ancestor. The differences in habit between var. bracei and var. bahamense, which appeared fairly marked on the compara- tively few herbarium specimens previously available, dis- appear completely when these plants are grown together and, therefore, must be due to local environmental differ- ences or season of collection. Crosses between the two are completely fertile and pairing at meiosis in the offspring is entirely normal. Not only are remaining minor differences between the two populations quite lost among the various greenhouse-grown plants and their offspring, but now à fairly large series of collections from Andros and New Contribution No. 199 from the Department of Biology, Wayne State University, Detroit, Michigan. 439 440 Rhodora [Vol. 70 Providence Islands, where var. bahamense was thought to be the only entity, reveals “var, bracei" features inter- mixed. It seems unnecessary, if not impossible, to try to distinguish the two varieties and I propose that they be combined. Until recently I knew of but three collections of var. corallicola, but, in 1965, I found that near San Andros, Andros Island, it was intermixed with the glabrous plant (var. bahamense) as a very common roadside weed. Crosses between a pubescent plant and several glabrous plants uni- formly yielded only pubescent offspring. The hybrids are completely fertile and pairing of chromosomes at meiosis is entirely normal. Self-pollination of these F, plants give F, generation ratios of approximately 3 pubescent: 1 gla- brous (74 plants examined). No plants intermediate in hairiness have appeared and it is quite clear that a very simple inheritance, probably a single gene controls this character. Therefore, in spite of the fact that pubescence is a very conspicuous feature, var. corallicola should be treated only as a form of L. bahamense. When this is done, the complete synonymy for L. baha- mense is then as follows: Linum bahamense Northrop, Mem. Torr. Club 12: 42. 1902. forma bahamense. Cathartolinum | bahamense (Northrop) Small, North Amer. Flora 25: 75. 1907. Cathartolinum bracei Small, loc. cit. 1907. Cathartolinum lignosum Small, loc. cit. 1907. Linum lignosum (Small) Winkl. in Engl. & Prantl, Natürl. Pflanzenfam. ed. 2, 19a: 116. 1931. Linum bracei (Small) Winkl. loc. cit. 1931. Linum bahamense var. bracei (Small) Rogers, Brittonia 15: 107. 1963. Linum bahamense forma corallicola stat. nov. Cathartolinum corallicola Small, North Amer. Flora 25: 74. 1907. 1968] Linum bahamense — Rogers 441 Linum corallicola (Small) Winkl. in Engl. & Prantl, Natiirl. Pflanzenfam. ed 2, 19a: 116. 1931. Linum bahamense var. corallicola (Small) Rogers, Brit- tonia 15: 107. 1963. DEPARTMENT OF BIOLOGY WAYNE STATE UNIVERSITY DETROIT, MICHIGAN 48202 LITERATURE CITED Rocers, C. M. 1963. Yellow flowered species of Linum in eastern North America. Brittonia 15: 97-122. SMALL, J. K. 1907. Linaceae in North Amer. Flora 25: 67-87. SEASONAL OCCURRENCE AND ECOLOGY OF SALT MARSH PHANEROGAMS AT IPSWICH, MASSACHUSETTS' E. E. WEBBER For the past several years, research into the ecology and systematics of benthic algae has been in progress at a salt marsh adjacent to the Castle Neck River, Ipswich, Massa- chusetts. Because of the close association between many of these algae and the salt marsh seed plants, it was essential to characterize this salt marsh area as to its phanerogam vegetation. The present paper reports on the presence and distribution of the dominant seed plants characteristic of the Ipswich salt marsh. Principal references used for species determinations were Gleason and Cronquist (1936), Mason (1957), and Fernald (1950). Specimens of the plants collected are deposited in the Herbarium, University of Massachusetts, Amherst. The phanerogams collected as indicative of the Ipswich salt marsh and its immediate surroundings are as follows: MONOCOTYLEDONAE Ruppia maritima L. Distichlis spicata L. Triglochin maritima L. Phragmites communis Trin. Eleocharis halophila Agropyron pungens Fern. & Brack. (Pers.) R. & S. Scirpus americanus Pers. Spartina pectinata Link Juncus gerardi Loisel. Spartina alterniflora var. Bromus tectorum L. glabra (Muhl.) Fern. Glyceria melicaria Spartina patens (Michx.) Hubb. (Ait.) Muhl. Panicum virgatum L. "Modified from a section of a Ph.D. thesis, “Systematics and Ecology of Benthic Salt Marsh Algae at Ipswich, Massachusetts”, submitted to the Graduate Faculty, University of Massachusetts, Amherst. 442 1968] Salt Marsh — Webber 443 DICOTYLEDONAE Myrica pensylvanica Loisel. Suaeda linearis (Ell.) Moq. Polygonum aviculare L. Spergularia marina var. littorale (Link) Koch (L.) Griseb. Bassia hirsuta (L.) Aschers. Trifolium arvense L. Atriplex patula var. Hudsonia tomentosa Nutt. hastata (G) Gray Glaux maritima L. Salicornia bigelovii Torr. Limonium nashii Small Salicornia europaea L. Gerardia maritima Raf. Suaeda maritima (L.) Dum. Plantago maritima L. Suaeda richii Fern. Solidago sempervirens L. Aster sp. SEASONAL DEVELOPMENT. Two of the dominant species, Juncus gerardi and Spartina patens, began their growth period far earlier than the re- mainder of the marsh phanerogams. Young green shoots of these two species a few centimeters in height were col- lected from beneath the ice and snow covering the marsh in February, 1963 and 1964. By early May Juncus gerardi shoots had increased to 15 cm in height, while those of Spartina patens were 7 cm tall. In contrast, emergent shoots (1-3 cm tall) of the follow- ing plants became apparent in early May throughout the marsh: Glaux maritima Suaeda richii Eleocharis halophila Suaeda linearis Plantago maritima Limonium nashii Atriplex patula var. hastata The first flowering of salt marsh phanerogams began in late May with Glaux maritima, Eleocharis halopila, and Plantago maritima, These were followed in June by Juncus gerardi, Scripus americanus, Phragmites communis, and Ruppia maritima. By July, Spartina patens and Limonium nashii joined the species already in bloom. The majority of seed plants flowered during August, among them the following: 444 Rhodora [Vol. 70 Suaeda linearis Salicornia bigelovii Suaeda richii Salicornia europaea Spergularia marina Bassia hirsuta In September, the marsh appearance changed from a lush summer green to an autumn brown with splotches of yellow, purple, and red belonging to the blooms of Solidago sem- pervirens, Aster sp., and the stems of Salicornia sp. respec- tively. Also in flower during September, but much less evident than the aforementioned autumn species, was Gerardia maritima, growing among the taller culmns of Spartina patens and the scapes of Plantago maritima. DISCUSSION OF PHANEROGAM DISTRIBUTION, Throughout tne Ipswich marsh the phanerogam vegeta- tion follows a recognizable pattern of distribution. In a landward direction the dominant species succeed each other in the following sequence: Spartina alterniflora var. glabra ; Spartina patens ; a mixture of Spartina patens and Distichlis spicata or, where the topography was sloping, a Spartina patens — Scirpus americanus mixture; and Juncus gerardi. Historically, explanations of seed plant distribution in the salt marsh habitat have emphasized the importance of tide levels in conjunetion with the ability of a particular species to tolerate varying concentrations of salt in the soil solu- tion, as well as periods of tidal immersion. Johnson & York (1915) related tide levels on Long Island marshes to seed plant zonation, and suggest that the upper and lower distributional limits for a particular plant are controlled by the ability of that plant to withstand specific periods of submergence and emergence, These authors also indicate that light availability may be involved in the maxi- mum downshore extension of S. alterniflora var. glabra, a point recently reemphasized by Ranwell, et al. (1964). On the Saugus marsh (Massachusetts), Chapman (1940) found the tidal factor to be ultimately the most important ecological variable in controlling phanerogam distribution. 1968] Salt Marsh — Webber 445 Additional investigations of phanerogam distribution on other North American salt marshes indicate the importance of tide levels as they influence soil salinities and soil drain- age (Penfound & Hathaway, 1939; Uphof, 1941; Purer, 1942; Reed, 1947; Hinde, 1954). Results of culture experiments reported to date suggest that, in general, different salinity tolerance levels (Taylor, 1939) and specific nutrient requirements (Adams, 1963) of the various species are responsible for their zonation in salt marshes. An excellent review of pertinent physio- logical data attempting to explain, at least in part, the distribution of salt marsh seed plants has been presented by Chapman (1960). At the Ipswich marsh I have recognized the importance of several factors related to phanerogam distribution, e.g., proximity to saline or fresh water influences, marsh surface topography, and character of the substrate. The pioneer phanerogam to colonize the marsh was Spar- tina alterniflora var. glabra, forming a band of vegetation along marsh edges and creek margins, Its landward extent coincides with high water neap tide levels. Maximum vege- tative and reproductive growth was restricted to those plants on the marsh near this level. At its lower limit, this species was only several centimeters tall and typically vegetative. The distribution of S. alterniflora var. glabra at Ipswich is confined to marsh areas with muddy substrate, poor drain- age, repeated tidal inundation, and continual mud and silt deposition. Above high water neap tide levels an extensive Spartina patens - Distichlis spicata mixture occurred on flat sections of the marsh which lacked even the slightest slope. In Connecticut, Nichols (1920) found that poorly drained marsh areas above high water neap tide levels were popu- lated by Distichlis spicata, and with improved soil drainage S. patens became abundant. Taylor (1939) noted that D. spicata was more vigorous in growth and flowering at higher salinities in culture solutions. While Purer (1942) 446 Rhodora [Vol. 70 has reported D. spicata to be typical of marsh sections having reduced salinities and improved aeration, Adams (1963) using nutrient solution experiments, has shown that Distichlis grows best in a saline medium. It may be that the flat and more fully exposed marsh areas at Ipswich had a higher soil salinity than sloping, better drained situations, thereby accounting for the characteristic distribution of Distichlis spicata, The presence of a zone lacking vegetation, a “barren Zone", was evident immediately landward to the Juncus gerardi consocies on the marsh adjacent to the Castle Neck River. The gently sloping character of the marsh here per- mitted the retention of tidal debris in this “barren zone". In their study of Connecticut tidal marshes, Miller and Egler (1950) noted that tidal trash in the upper Juncus gerardi consocies commonly killed the plants which it covered. At Ipswich the mass of debris, with its accompanying high temperatures and poor air circulation, very likely had a similar effect on a potential vegetation becoming established in this zone. Perhaps high temperatures beneath this debris resulted in excessive salinity, both factors acting to prevent the establishment of seedlings. A tidal creek at the head end of the marsh was the sole habitat for Ruppia maritima, where the plants, anchored in the soft mud of the creek bed, undoubtedly aid in stabi- lizing this substrate. The plants were never exposed during periods of low tides; indeed, Ruppia is unable to withstand exposure to the desiccating effects of sun and air (Chap- man, 1960). Citations are common indicating the restricted distribution of R. maritima to “brackish water" areas, without specifying salinity readings. An early account of the response of R. maritima to varying salinities was made by Graves (in Harshburger, 1909) who found cells of Ruppia to tolerate salinities to 259/00, but to plasmolyze in salinities of 30º/00. Maximum growth and development of Ruppia maritima at Ipswich was related to a summer salinity range of 20-25º/00 (once to 279/00). 1968] Salt Marsh — Webber 447 Equally distinctive and more characteristic as a salt marsh phanerogam is the genus Salicornia. Its brilliant searlet autumn coloration together with its clonal distribu- tion on the marsh signify “salt marsh" to the observer. Species of Salicornia have long been recognized as true halophytes. Halket (1915) and Taylor (1939) have shown experimentally that this plant grows best in 2-3% NaCl solutions. Numerous additional reports pertaining to the halophytic character of the various species of Salicornia are contained in Chapman (1960). For the Bay of Fundy marshes (Ganong, 1903), S. stricta is typically one of the first phanerogams to colonize eroded segments of the marsh, and for New England marshes, a Salicornetum character- istically develops on bare mud, where high salinities would be encountered (Chapman, 1960). At Ipswich, Salicornia bigelovii was abundant in the more poorly drained sections of the marsh, and where the plants were entirely submerged during high water spring tide levels. In contrast, Salicornia europaea was always limited to the better drained areas, e.g., as scattered plants along the lower edge of the Spartine- tum patentis. However, in several marsh depressions at the head of the marsh S. europaea was often the sole com- ponent. While soil salinities were not determined for those marsh areas with S. bigelovii and S. europaea, the prox- imity to fresh water influence and the greater retention of rain water in the depressions at the head of the marsh, may have produced a less saline environment more condu- cive to the development of S. europaea. Seed dissemination and germination for the salt marsh annuals is an area of investigation which has progressed little beyond a general consideration of but a few species. During his study of the Saugus marsh, Chapman (1940) noted that seed germination of the annual species occurred in the spring, when the salinity level of the soil surface layers was low. In laboratory experiments he showed that maximum germination of several species occurred in fresh water, although optimum germination for Salicornia bige- lovii from Saugus (Massachusetts) required a 1% NaCl 448 Rhodora [Vol. 70 solution. He also demonstrated that freezing temperatures interact with salinity levels in affecting seed germination of S. bigelovit, for while the per cent germination decreased with increasing salinities, this was partially offset by lower temperatures. In contrast to Salicornia, seeds of Plantago maritima germinate best in fresh water (Chapman, 1940). At Ips- wich Plantago maritima and Glaux maritima appeared in early May, their seeds apparently germinating when the salinity of the marsh surface soil would be expected to be lowest, Another consideration relating to salinity and seed plant growth pertains to changing salinity levels as the plants age. Turner (in Chapman, 1960) has shown that Suaeda novae-zelandiae grows best in a 1% NaCl solution during its early stages, but that when the plants are six months old, a 3% NaCl solution yields maximum growth and de- velopment. It is clear that while salinity levels are im- portant in the germination of some salt marsh annuals, one must also consider salt concentrations of the soil solution. throughout the growing season as the plants develop. The distribution of Swaeda (maritima, linearis, richii) at Ipswich occurred in the upper marsh levels, as scattered plants among Spartina patens or, occasionally, at the upper boundary of Juncus gerardi. Seed germination of Suaeda maritima is best in fresh water, although the per cent germination is very small (4%); however, another species, S. novae-zelandiae, has an optimum of 41% in fresh water (Chapman, 1960). Germination of Swaeda seeds at the higher marsh levels at Ipswich may be related to low soil salinities which probably exist here during the early spring. Atriplex patula var. hastata also occurs with Suaeda at these higher marsh levels. Seed germination in Atriplex is inhibited by the high chloride content of sheathing brac- teoles, for removal of these bracteoles results in immediate germination (Beadle, 1952). At the higher marsh levels, 1968] Salt Marsh — Webber 449 the chloride concentation of Atriplex bracteoles could be reduced readily by spring rains, resulting in early spring germination which was observed for this species at Ipswich. It is immediately evident that continued research is neces- sary with the salt marsh phanerogams to define their specific growth requirements, and therefore more accurately inter- pret their distribution in salt marshes. Additional and detailed field studies coupled with laboratory experiments should be initiated with this view in mind. DEPARTMENT OF BIOLOGY KEUKA COLLEGE KEUKA PARK, NEW YORK LITERATURE CITED ADAMS, D. A. 1963. Factors influencing vascular plant zonation in North Carolina salt marshes. Ecol. 44(3): 445-456. BEADLE, N. C. W. 1952. Studies in halophytes. I. The germination of the seed and establishment of the seedlings in five species of Atriplex in Australia. Ibid. 33(1): 49-62. CHAPMAN, V. J. 1940. Studies in salt marsh ecology. VI and VII. J. Ecol. 28: 118-152. 1960. Salt Marshes and Salt Deserts of the World. Interscience Pub. N.Y. 392 pp. GLEASON, H. A., and A. CRONQUIST. 1963. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. D. Van Nostrand Co. Princeton, N.J. 810 pp. FERNALD, M. L. 1950. Gray’s Manual of Botany. 8th ed. American Book Co., N.Y. 1632 pp. GANONG, W. F. 1903. The vegetation of the Bay of Fundy salt and diked marshes: an ecological study. Bot. Gaz. 36: 161-186; 280- 302; 340-367; 429-455. HALKET, A. C. 1915. The effect of salt on the growth of Salicornia. Ann. Bot. 29: 143-154. HARSHBURGER, J. W. 1909. The vegetation of the salt marshes and of the salt and fresh water ponds of northern coastal New Jersey. Proc. Acad. Nat. Sci. Phil. 61: 373-400. HINDE, H. P. 1954. The vertical distribution of salt marsh phane- rogams in relation to tide levels. Ecol. Monogr. 24(2): 209-225. Jounson, D. W., and H. H. York. 1915. The relation of plants to tide levels. Carnegie Inst. Wash. Publ. No. 206. 162 pp. Mason, H. L. 1957. A Flora of the Marshes of California, Univ. of Calif. Press. Berkeley. 878 pp. 450 Rhodora [Vol. 70 MILLER, W. R., and F. E. EGLER. 1950. Vegetation of the Wequete- quock-Pawcatuck tidal marshes, Connecticut. Ecol. Monogr. 20: 143-172. NICHOLS, G. E. 1920. The vegetation of Connecticut. The associa- tion of depositing areas along the sea coast. Bull. Tor. Bot. Club. 47(1): 511-548. PENFOUND, W. T., and E. S. HATHAWAY. 1938. Plant communities in the marshlands of southeastern Louisiana. Ecol. Monogr. 8: 1-56. Purer, E. A. 1942. Plant ecology of the coastal marshes of San Diego City, California. Ibid. 12: 81-111. RANWELL, D. S., E. C. F. BIRD, J. C. E. HUBBARD, and R. E. STEBBINGS. 1964. Spartina salt marshes in Southern England. V. Tidal sub- mergence and chlorinity in Poole Harbour. J. Ecol. 52: 627-641. REED, J. F. 1947. The relation of the Spartinetum glabrae near Beaufort, N.C. to certain edaphic factors. Amer. Midl. Natur. 38: 605-614. TAYLOR, N. 1939. Salt tolerance of Long Island salt marsh plants. N.Y. State Mus. Circ. No. 23. 42 pp. UrHor, J. C. 1941. Halophytes. Bot. Rev. 7: 1-58. SPOROBOLUS TEXANUS VASEY IN LINCOLN, NEBRASKA Sporobolus texanus was collected in saline soils in Lincoln, Nebraska on July 7, 1965, Ungar 1190 (KANU), and August 28, 1967, Ungar 1231 (Ohio University Herbarium). This is the northeasternmost extension of the range of this species. It is about 70 miles east and 80 miles north of any previously reported collection of S. texanus. Prior to this collection, the most northern and eastern collection had been reported by Gates (1940) for Cloud County, Kansas, which is in northcentral Kansas. Accord- ing to Hitchcock (1950) S. texanus is primarily a south- western species, occurring in Kansas, Oklahoma, Texas, Colorado, Arizona, and New Mexico. It occurs in non-saline prairie soils, prairie bordering saline marshes, and is occa- sionally an invader of open salt flats from Oklahoma north to Lincoln, Nebraska (Ungar, 1965, 1966, 1967). S. texanus occurs in two communities in the Lincoln marsh area on the south side of Highway 6-2, opposite 720 1968] Sporobolus texanus — Ungar 451 west 0 street. In one area which is slightly saline, 0.45% to 0.70% salt on a dry weight basis, it is associated with Hordeum jubatum L., Iva annua L., Distichlis stricta (Torr.) Rydb., Poa arida Vasey, Agropyron smithii Rydb., Melilotus officinale Desr., Atriplex patula L. var. hastata (L.) Gray, and Rumex crispus L. It also occurs rarely in highly saline soils, 0.61% to 2.2%, invading a salt pan occupied by the highly salt tolerant Suaeda depressa (Pursh) Wats., Salicornia rubra Nels., and Distichlis stricta. Its range of salt tolerance appears to be very broad and it has been found in soils ranging from 0.003% to 2.5% salinity (Ungar, 1965, 1966). The possibility of S. texanus occurring in saline soils which are still farther north in South Dakota and North Dakota are being investigated. Field work in 1967 in saline areas of Codington County and Day County, South Dakota indicates that S. texanus is not found in these areas though the remaining vegetation complement is similar to the Lincoln, Nebraska marshes. IRWIN A. UNGAR! BOTANY DEPARTMENT OHIO UNIVERSITY, ATHENS 45701 LITERATURE CITED GATES, F. C. 1940. Flora of Kansas. Contribution #291. Botany Department, Kansas State University. 226 p. HITCHCOCK, A. S. 1950. Manual of the Grasses of the United States. U. S. Government Printing Office, Washington. Misc. Publication 220. 1051 p. Unear, I. A. 1965. An Ecological Study of the Vegetation of the Big Salt Marsh, Stafford County, Kansas, University of Kansas Bull. 46: 1-98. —, 1966. Salt Tolerance of Inland Halophytes of Kansas and Oklahoma. Ecology 47: 154-155. , 1967. Vegetation-Soil Relations on Saline Soils in Northern Kansas. American Midland Naturalist 78: 98-120. 'T would like to thank Ohio University for research grant #211 and the National Science Foundation for grant #GB6009. Thanks to Mrs. Warren A. Wistendahl for corroborating my identification. RANGE EXTENSION OF MUHLENBERGIA SOBOLIFERA IN NEW HAMPSHIRE The collection of Muhlenbergia sobolifera (Muhl.) Trin. proved to be the most significant record of a two year flor- istic study conducted on Rattlesnake Island, East Alton, New Hampshire. This was the first record for Belknap County and the second record for the state. The previous record was by A. A. Eaton in Nottingham, September 1899. The collection was obtained below a shaded ledge on the south facing talus slope. Significant species occurring on the same slope include: Asplenium platyneuron, Hystrix patula var. Bigeloviana, Silene antirrhina, Arabis Drum- mondii, A. missouriensis, Ceanothus americanus, Prunus susquehanae, Specularia perfoliata, and Carya ovata. This talus slope rising some 400 vertical feet from the lake’s surface, resembled closely in species composition a similar talus slope in the Pawtuckaway Mountains, Not- tingham, New Hampshire. The slope at Rattlesnake Island is believed to be of the same geologic origin as the Ossipee Mountains some 20 miles to the north and west. Collections are to be placed in the herbarium of the New England Botanical Club, Cambridge, and the University of New Hampshire Herbarium in Durham. WENDELL H. BERRY, JR.' LAKE FOREST ACADEMY LAKE FOREST, ILLINOIS 60045 "Thesis research conducted from 1964-66 under the direction of Dr. A. R. Hodgdon. A NEW FORM OF MYRICA PENSYLVANICA While collecting in the Ocean Park section of Old Orchard Beach, York County, Maine, in the summer of 1967, I noticed that certain bayberry plants were conspicuous for their extremely small leaves. The fruit were also im- pressively smaller than on typical plants. Such aberrant plants were often found beside typical plants, but no plants with intermediate characters were found, About 5% of the Ocean Park bayberry plants are of this new form, All the bayberry plants in Ocean Park are found in open sandy soil within 14-mile of the beach. Myrica pensylvanica Loisel. forma parvifolia T. W. Wells, forma nova. Type Specimen: T. W. Wells, Old Orchard Beach, York Co., Maine, 5 Sept., 1967 (N.E.B.C.). Differt a forma typica foliis multo minoribus maturisque baccis minoribus: foliis ad 314 cm longis, ad 142 cm latis ; maturis baccis ad 3.5 mm latis. The difference between this form and forma pensylvanica is in the much smaller size of the leaves and fruit: the larger leaves up to 314 cm long and 114 cm wide, but aver- aging under 3 cm long and under 1 cm wide; and the mature fruit up to 3.5 mm in diameter. Voucher specimens are deposited in the herbarium of the New England Botanical Club. THEODORE W. WELLS MILTON ACADEMY MILTON, MASSACHUSETTS 02186 453 A YELLOW-CENTERED FORM OF HELENIUM NUDIFLORUM This form was found growing on the western shore of Hinckley Lake near the mouth of Beaver Meadow Creek, about one mile north of Hinckley, New York. It was grow- ing in close association with the typical form (with brown or purplish disk) and was widely, though sparingly, dis- tributed through the community. It was observed by my- self and by Stanley J. Smith, during the 1966 season and again in 1967. It was collected in August, just as the dise flowers were beginning to open (No. 3231 in the Oneida County collection at Utica College of Syracuse University, Utica, New York) and in September, at which time the ovules were plump and well-developed (No. 3246). Mr. Smith also collected this form and added it to the collection in the state herbarium in Albany. (No. 40666, S. J. Smith and T. C. Baim) It is suggested that this form be named Helenium nudi- florum. forma homochromum Howe, forma nova: Differt forma typica floribus tubulosis luteis. Type in the Herbarium of Utica College of Syracuse Uni- versity, Oneida County Collection, Howe 3246. M. DORISSE HOWE DEPARTMENT OF BIOLOGY UTICA COLLEGE OF SYRACUSE UNIVERSITY UTICA, NEW YORK 13502 454 TWO EUPATORIUM SPECIES NEW TO CONNECTI- CUT. Although usually warned by signs against trespass- ing, it may be found by one who does so that railroad rights-of-way, yards, and terminals are good places to botanize with the hope of finding plants from distant parts which are new to one's area, That conclusion is supported by the following incidents. In 1963 while riding down a sandy lane toward sand-pits in North Haven, Connecticut, I caught sight of a beautiful tall whitish-flowered plant with strongly 3-veined leaves, which proved to be the Tall Thoroughwort, Eupatorium altissimum L. Upon botanizing the area, I discovered that on both sides of the railroad tracks there were many plants of this species intermingled with Hupatorium hyssopi- folium L. Since that time, although one large area has been destroyed by the parking of trucks and tractors of the Connecticut Highway Dept., the species is rapidly spread- ing over a large area. Gray's Manual gives the range of the species as Pennsylvania, westward and south ward. Recently, while driving through the Brewery St. yards at the New Haven Railroad repair shops, I noticed another tall whitish-flowered composite. This one turns out to be Eupatorium. serotinum Michx., the range of which, accord- ing to Gray's Manual, is north to New Jersey and adventive in Massachusetts. It is described in the Manual as "some: what aggressive and weedy," and in this locality it certainly deserves that reputation, as there are scores of plants singly and in clumps scattered over several acres. 455 456 Rhodora [Vol. 70 It may be expected that both of these species are spread by association with the railroad, either "hitch-hiking" on equipment or by suction from passing trains. They should be watched for in other areas. Specimens of each are deposited in the Herbaria of Yale University, the University of Connecticut, Pringle Her- barium of the University of Vermont, and the herbarium of the New England Botanical Club at Harvard University. JAMES J. NEALE 39 FAIRMOUNT AVENUE NEW HAVEN, CONNECTICUT 06513 Volume 70, No. 783 including pages 313-456, was issued September 30, 1968. Hodova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL ` STUART KIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER > Associate Editors ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. Vol. 70 October-December No. 784 CONTENTS: John Goldie and the Republication of his Diary. ALa A OA Taa EE RS dg EE 457 A Case of Mistaken Identity, or the Mysterious C. E. Lloyd. Irving W. Knobloch wcccccssrressrccrssccccccsscsvesscscneccenscsccccsosssceescce 462 Hybridization within the Triticeae of Alaska: A new Ely- hordeum and Comments. William W. Mitchell and H. J. OIE E E 467 Generic Considerations Concerning Carphephorus and Trilisa (Compositae). Henry J.-C. Hebert ........................ 474 Spiranthes ovalis New for Kansas. Kermit L. Johnson .... 485 An Example of a Vernonia Hybrid in a Disturbed Habitat. MIL IE nM. ee EE 486 Wolffia in New Hampshire. William D. Countryman ........ 491 (Continued on Inside Cover) The Nem England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge, Mass, 02138 CONTENTS: — continued Leguminosae of Mexico — Faboideae 1. Sophoreae and Podalyrieae. Velva E. Rudd ....... eene 492 Evidence for the Hybrid Origin of Petasites warrenii and P. vitifolius. Alfred L. Bogle ienes 533 Biometric and Taxonomic Uses of Cellulose Acetate Plastic. Edward E. Terrell |... eee emere 552 A Reversion in Polymnia canadensis (Compositae) and Notes on the Pedicellate Disk Florets in the Genus. Edwin B. Smith eene enter races 559 A “Pillar” Type Black Spruce. Henry I. Baldwin ............ 562 Notes on Festuca arundinacea and F. prateneis in the United States. Edward E. Terrell eee 564 Yellow-Fruited Cornus florida — Cultivar or Form? Robert D. MacDonald. .......... eee 568 Occurrence of Murdannia spirata in Florida. Olga Lakela osses. 571 Emendations in United States Physalis. U. T. Waterfall |... essere 574 Alisma gramineum in Vermont. William D, Countryman 577 Errata for Volume 70 .......... 17... eee 580 Rbodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 70 October-December No. 784 JOHN GOLDIE AND THE REPUBLICATION OF HIS DIARY JOSEPH EWAN John Goldie (1793-1886) is remembered by users of Gray’s Manual through the tallest and largest of Wood ferns, Dryopteris goldiana (Hook.) Gray ( Aspidium goldi- anum), but few know his Diary published in 1897 in Toron- to in a small privately printed edition. In the course of restoring materials in the Toronto Public Library, the Philadelphia artisan-bibliographer, Willman Spawn, discov- ered that the manuscript from which the 1897 printing was made differed from the published version and urged Goldie’s granddaughter, Mrs. Theresa Goldie Falkner, to reissue the Diary, verbatim et literatim. It appeared December 6, 1967: with an introduction and thirty-six superscriptoral notes. To report factually, the notes which Mr. Spawn had so carefully prepared were drastically abbreviated and he did not see proofs before publication. Yet the Diary is now available for the first time as written by John Goldie, attrac- tively bound in stiff-paper cover with a frontispiece por- trait, and two pages reproduced from manuscript. Apart from its intrinsic botanical interest it will be a sought-for Americana item. We are grateful to Mrs. Falkner and Willman Spawn for their contribution to American botany. “Dairy of a Journal through Upper Canada and some of the New England States, 1819. John Goldie. [ Toronto, 1967] x + 65 pp. portr. frontis. May be purchased from University of Toronto Book- store, $1.75 net. 457 458 Rhodora [Vol. 70 The plate of “Aspidium goldianum” and the “original description” (pp. 64-65) are not satisfactorily identified. The plate reproduced here is Faxon’s made for Daniel Cady Eaton’s Ferns of North America (pl. XL) drawn from a Vermont specimen. The “original description” of Aspidium goldianum Hooker was published by Goldie in the Edin- burgh Philosophical Journal in 1822. A map locating Gold- ie's place names would have been welcome, but the two examples of his handwriting will be useful in verifying labels encountered in herbaria. Goldie was not mentioned in Index Herbariorum, Collectors Part II (2) 1957. In addition to specimens at Kew about twenty Goldie speci- mens are preserved in the original Fielding Collection in the Oxford University Herbarium according to Hermia Clokie, An Account of the Herbaria of the Department of Botany in the University of Oxford (1964). Fate dealt harshly with Goldie who came to America in 1817 on the advice of William Jackson Hooker of Glasgow whose enthusiasms for field collecting are well known. On three separate occasions Goldie lost his plant specimens in transit between Canada or New York and England. The loss of his botanical journal describing his collections made in Newfoundland, Quebec, Ontario, New York, and New Jersey, was an immediate tragedy to Goldie for it wiped out recognition of possible discoveries that might have been credited to him, and a permanent loss to us were the records of native vegetation which covered what today are, as Spawn observes, “housing developments, industrial parks, shopping centers and all the manifestations of urban sprawl.” This surviving Diary is a fortunate reminder “of a land and a time we have left behind.” Biographical notices of Goldie have been liberally pub- lished and are listed below. Most of these are based on the autobiographical sketch which appeared in 1822. Almost wholly neglected is Goldie’s own list of his American intro- ductions and “observations” thereon (1827), brought to print by the zealot in horticultural publishing, John Claud- ius Loudon. But none of these accounts, largely repetitive, quote the two letters from Goldie to his professor extraordi- 1968] John Goldie — Ewan 459 nary, William Jackson Hooker, preserved at Kew. These are here published through the genial courtesy of Sir George Taylor. The first letter records his association with James McNab and his botanizing in Russia; his second was written from “Canada West” whither he immigrated to a farm “about a mile from Ayr, in the county of Waterloo,” his new Scotland. “Wrightfield, June 16, 1837 Dear Sir I have been called upon by a medical student, Mr. Paton from Air, who stated that if he obtained a recommendation from any of your acquaintances he would be permitted to attend your Botanical Class: with regard to which, I beg leave to say, that although not very intimately acquainted with him, yet I believe him to be very deserving of encouragement in his studies. He has always sustained an excellent character; and his Mother who is a Widow, having lost her husband at sea, has by industry and economy, been able to support herself and family respectably: yet not able, without a great sacrifice, to afford her son the means requisite for the com- pletion of his studies in the profession which he intends to pursue. He seems exceedingly anxious to attend the Botanical Class, and should you be as kind as [to] give him a ticket, I feel assured that it will be received with much gratitude. My Friend, Mr. James McNab’ visited me the other day, and I expect to see him again to night on his way home. I am going to send by him a sketch of an Iris which I brought from St. Petersburgh in 1830. It flowered last year but I am sorry to say that I have nearly lost the plant. The flower was very singular and very beautiful. The drawing was taken by a Lady; but it is not a good likeness of the flower. Mr. McNab will give more minute information. I would feel much gratified by having the name of the species. Should the plant live and flower with me again I will send vou a specimen — I am Dear Sir Your Honoura' Obedt Sert. John Goldie “James McNab (1810-1878) had visited New Jersey, Pennsylvania, and Richmond, Virginia, in 1834, and was to succeed his father William MeNab (1780-1848) as Curator of the Edinburgh Botanic Garden in 1849 460 Rhodora [Vol. 70 II Ayr, Canada West Sept 2nd 1851 Dear Sir William, I am afraid that you will have forgotten me altogether. A very excellent young man from this quarter is going out to Jamaica as a missionary, and ac he wishes to see through the Kew Gardens requested an introduction to you. My Friend, Mr. Scott, is very fond of plants, and knows all that is interesting in this vicinity. I have now been located, as the Yankees say, in this place for the last seven years and am likely to remain here during life. There are many interesting plants in the neighbourhood, several of which I cannot make out from Pursh, and I have no other scientific work except Nuttall’s Genera. I am sorry that I had no knowledge of Mr. Scott's going to London untill very lately else I would have had some specimens prepared to send you of such things as I could not make out satisfactorily. I have bought a piece of land which I occupy as a farm, and have lately erected a Flour Mill, which has taken both my time and money and has prevented me from travelling so much as I would like to do. I have never been many miles from home except at N. York. I found lately some of the Gerardia quercifolia, a beautiful plant. Batschia canescens® is also fine, but very local, it is only in a few places that it is to be found. Mr. Scott found the Batschia longiflora? near Toronto. You must now have a splendid collection in the Kew Gardens, and I feel glad that the public are admitted to admire them." There is one of my sons in N. Jersey, who has taken a fancy for entomology, and who has made a good collection; but is greatly at a loss for a deseriptive work on American Insects. If you will have the kindness to write to me, and state what may be the best work on the subject, I would feel greatly obliged — Would American Insects *Frederick Pursh, Flora Americae Septentrionalis (London, 1814), 2 vols. ‘Gerardia flava L. “Lithospermum canescens (Michx.) Lehm. “Lithospermum. incisum. Lehm. "P. H. Gosses Wanderings through the Conservatories at Kew (London, [1856]) offers an excellent commentary on Hooker’s achieve- ment. ‘Judging from entomological rosters and history consulted, Walter Horn and Ilse Kahle (1935-36), Mathilde M. Carpenter (1945), and Harry B. Weiss (1986), the son’s insect specimens did not survive. 1968] John Goldie — Ewan 461 sell in London? A letter from you would afford much gratification. I shall never forget your kindness. I feel confident that you will show my Friend all your rarities, and remain, Dear Sir W. yours sincerely John Goldie REFERENCES [BRITTEN, JAMES]. John Goldie. Jour. Bot. 26: 299-301. 1888. CLUTE, WILLARD N. John Goldie. Fern Bull. 8: 73-75. portr. 1900. [CouLTER, J. M.?] John Goldie, gardener and botanist. Bot. Gaz. 11: 272-274. 1886. [GoLDIE, JAMES] [introduction to] Diary of a Journey... Toronto, Wm. Tyrrell & Co., 1897. pp. 3-5. GoLDIE, Jonn. Rare Canadian plants. Edinburgh Philos. Jour. 6: 333. 1822. Hooker wrote to John Torrey, 29 May 1823, regarding Torrey’s opinions on Goldie’s paper of the previous year (A. D. Rodgers III, John Torrey, 1942, 58) GoLDIE, JOHN. On the culture of North American Plants, including Ferns; founded on Observations made during a Journey through Canada, and some of the Northern States of the Union, in the Years 1817, 1818, and 1819. Gardener’s Mag. 2: 129-135. 1827. Gray, Asa. John Goldie. Amer. Jour. Sci. ser. III. 35: 260-261. 1888. Hay, G. U. John Goldie, Botanist. Trans. Roy. Soc. Canada ser. II. 3 (sect. 4): 125-130. 1897. PENHALLOW, D. P. Review of Canadian Botany from 1800-1895. Part II. Trans. Roy. Soc. Canada. ser. II. 3 (sect. 4) : 3-56. 1897. John Goldie noticed on p. 8. 1819. A CASE OF MISTAKEN IDENTITY, OR THE MYSTERIOUS C. E. LLOYD IRVING W. KNOBLOCH One of the most interesting travelogues dealing with the natural history of the New World is “Unknown Mexico", a two volume work by Carl Lumholtz (1902), In this account will be found those discoveries noted by him on a series of expeditions mostly in western Mexico during the years 1890 to 1898, a span of time interrupted from time to time for the purpose of raising money for further exploration. About one and a half years were spent by members of the expedition among the unique Tarahumare Indians of Chi- huahua, a group of cave-dwellers. Incidentally, Carl Lum- holtz was born at Faberg, Norway in 1851 and died at Saranac Lake, New York on May 5, 1922. In this offering we are only concerned with the first Lumholtz Expedition. From reading “Unknown Mexico" one gathers the impression that this lasted from 1890 to 1893 but others believe that the dates were 1890 to 1892. For example, J. A. Allen (1893) published a list of mam- mals and birds collected in northeastern Sonora and north- western Chihuahua, Mexico on the Lumholtz Archaeological Expedition, 1890-1892. It is difficult to see how this article could have gotten into print in 1893 if the expedition did not end sometime in 1892. According to Allen, the first expedition left Bisbee, Ari- zona in September 1890. It traveled 225 miles south to Bacadehuachy and Nacory on the Río Yaqui, passing San Pedro (32 miles south of Bisbee), Los Trincheras, Santa Barbera, Fronteras, Cachuta (102 miles south of Bisbee), Los Trinitas (127 miles south of Bisbee), Los Pinitos, Los Cuevas, Oputo, Granados, Bacadehuachyl (the last three on the Río Yaqui), thence eastward toward and across the Sierra Madre, passing the following — Nacory (3400 ft.), Heurachi (4000 ft.), Napolera, El Puerto (6300 ft.) Bavispe River, Rancho de Los Apaches (6620 ft.), Chuhua- chupa, Tachico (2000 ft.) to San Diego (4000 ft.) on the eastern slope. From February 15th to March 1st trips were 462 1968] C. E. Lloyd — Knobloch 463 made from San Diego to Guanopa, Rio Chico and Tatuara. This itinerary evidently only encompasses the immediate time span during which F. Robinette of Washington, D.C. and A. D. Meeds of Minneapolis, Minnesota collected mam- mals and birds. It should be mentioned that the personnel of the expeditions changed often and that as many as thirty men were involved with the various Lumholtz expeditions. Of course, botanists also accompanied Lumholtz. Several years ago, in doing the biography of all collectors of plants in northern Mexico (unpubl.), we became aware that the names of two botanists, C. E. Lloyd and C. V. Hartman were printed on several hundred (?) labels in various her- baria, as being the collectors of plants on the first Lumholtz expedition. No problem presented itself in the case of Carl Vilhelm Hartman who was born in Orebro, Sweden on August 19, 1862 and who died in Stockholm, Sweden on June 19, 1941. We will pass over his illustrious history because it is not germane to our thesis. When we came to the collection of data on C. E. Lloyd, we located absolutely nothing regard- ing his background. The only major clue we found attesting to his existence is the printed herbarium labels, prepared at Harvard University by B. L. Robinson and M. L. Fernald. We frequently did encounter the name “Francis E. Lloyd." The latter, an eminent botanist, led a life which, in part, is difficult to put down upon an accurate time table. He was born in 1868 in England and entered Princeton University in 1889 as a junior, presumably having attended Lafayette College before this event. He terminated his undergradu- ate work at Princeton with graduation in 1891. Certain plant collections by him on the Lumholtz expedition are dated “1890”, and it is possible, therefore, that he went, to Mexico while still an undergraduate. This matter will be dealt with in more detail later. Lloyd was appointed a Fellow at Princeton 1891-2 but resigned presumably to take a position as assistant professor at Williams College. In 1895 he received an A.M. degree from Princeton. This year coincides with his tenure at Pacific University (1892 to 1897) but his A.M. degree may have been conferred in 464 Rhodora [Vol. 70 absentia. Another overlap is found in the dates 1896 to 1906 when he was associated with the Teachers College at Columbia University. Lloyd received honorary Doctor of Science degrees from University of Wales and from the University of Masaryk. After 1906, his other employers were the Desert Laboratory of Carnegie Institute, the Con- tinental Mexican Rubber Company, Alabama Polytechnic Institute and, lastly, McGill University in Canada. The above information was gleaned from various sources (Ridge 1968, Gibbs 1968, Scarth 1948, Wulf 1967, Thomas 1967, Mason 1958, Tryon 1959, Norwegian Embassy 1958, Schwarten 1959, Rollins 1958). At this point, it seemed obvious to us that we had to choose between C. E. Lloyd and F. E. Lloyd. Accordingly we set down the following facts both pro and con. Pro C. E. Lloyd As stated above, the herbarium labels of the first expedi- tion include a “C. E. Lloyd”. In the Preface to volume 1, page X of *Unknown Mexico" by Carl Lumholtz (1902), we find the statement “Messrs. C. V. Hartman and C. E. Lloyd were the botanists,". A research article by B. L. Robinson and M. L. Fernald (1895) is entitled “New Plants Collected by Messrs. C. V. Hartman and C. E. Lloyd upon an archaeological expedition to northwestern Mexico under the direction of Dr. Carl Lumholtz”. One of F. E. Lloyd's closest friends was C. E. Bradley, research associate in chemistry at California Institute of Technology. Bradley correctly pointed out that F. E. Lloyd taught at Pacific University, Forest Grove, Oregon from 1892 to 1897. He does not believe that Lloyd went to Mexico until 1907. Since the first Lumholtz expedition was from 1890 to 1892, it would appear that either Francis E. Lloyd did not go on the trip, or, he left the expedition early to take the position at Pacific University in 1892. Pro F. E. Lloyd Paul Standley (1920) mentioned the name of F. E. Lloyd as being on the Lumholtz expedition. 1968] C. E. Lloyd — Knobloch 465 The first Lumholtz trip was from 1890 to 1892, Dr. Francis E. Lloyd, having been born in 1868, was 22 years old at the time of the expedition and thus could very well have been on the trip. Both the Barnhart File at the New York Botanical Gar- den and American Men of Science, 1944, indicate that F. E. Lloyd was on the Lumholtz expedition. Dr. F. Shreve and Dr. F. E. Lloyd were well-known to each other, both having been on the staff of the Desert Laboratory in Arizona. In the Foreword to Dr. Howard Gentry's *Rio Mayo Plants" (1942), Dr. Shreve mentioned that C. V. Hartman and F. E. Lloyd accompanied Dr. Lum- holtz on his 1890 expedition, an endorsement which can hardly be in error. In M. L. Fernald's *Synopsis of the Mexican and Central American Species of Alnus” (1904) we note that Alnus oblongifolia Torr. was collected at Huehuerachi, Sonora at an altitude of 1230 meters in December 1890 by C. V. Hart- man (no. 322) and by F. E. Lloyd (no. 464). In the same year F. E. Lloyd collected Salvia, elegans sonoriensis at Rin- conardo, Sonora, twenty-one days after Hartman collected the same taxon at Huehuerachi, Sonara (M. L. Fernald 1900). Notholaena lumholtzii was collected by Hartman, number 298 (GH, US) on the Lumholtz expedition and is the type. Dr. F. E. Lloyd evidently collected the same species (no. 489) on the same trip. This specimen is also at the Gray Herbarium. Dr. Carl Lumholtz (1891) himself mentioned Francis E. Lloyd as one of his collectors on page 389 of "Report on Explorations in Northern Mexico". This contradicts what he wrote in *Unknown Mexico" in 1902 but can be attribut- ed to an excusable lapsus mentis. When all the evidence given above is evaluated, it would seem logical to assume that one of the botanists on the first Lumholtz expedition was Francis E. Lloyd and not the mysterious “C. E. Lloyd". Since Francis E. Lloyd did not graduate from Princeton until 1891, and since some of his collections were dated 1890, we shall be forced to conclude 466 Rhodora [Vol. 70 that Lloyd was still an undergraduate at the time of his participation in the Lumholtz adventure. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY MICHIGAN STATE UNIVERSITY, LANSING LITERATURE CITED ALLEN, J. A. 1893. List of Mammals and Birds Collected in North- western Sonora and Northwestern Chihuahua, Mexico, on the Lumholtz Archaeological Expedition, 1890-1892. FERNALD, M. L. 1900. A Synopsis of the Mexican and Central American Species of Salvia. Proc. Amer. Acad. Arts and Sci. 35: 489-556. 1904. Synopsis of the Mexican and Central American Species of Alnus. Proc. Amer. Acad. Arts and Sci. 40: 24-28. GENTRY, H. S. 1942. Rio Mayo Plants. Carnegie Institution of Washington, Publ. 527. GiBBS, R. D. 1968. (Pers. Com.) LuMHOLTZ, C. 1891. Report of Explorations in Northern Mexico. Jour. Amer. Gco. Soc. N.Y. 23: 386-402. 1992. Unknown Mexico. N.Y. Charles Scribner's Sons. 2 vols. MASON, C.T. 1958. (Pers. Com.) NORWEGIAN EMBassY 1958. (Pers. Com.) RIDGE, A. D. 1968. (Pers. Com.) ROBINSON, B. L. and FERNALD, M. L. 1895. New Plants Collected by Messrs. C. V. Hartman and C, E. Lloyd upon an Archaeologi- cal Expedition to Northwestern Mexico under the Direction of Dr. Carl Lumholtz. Proc. Amer. Acad. Arts and Sci. 30: 114-123. RoLLINS, R. C. 1958. (Pers. Com.) ScARTH, G. W. 1948. In Memoriam, Francis Ernest Lloyd, 1868- 1947. Plant Physiology 23 (1): 1-4. 1948. Francis E. Lloyd. Proc. & Trans. Roy. Soc. Canada, 3rd Ser., 42: 99-102. SCHWARTEN, LAZELLA 1959. (Pers. Com.) STANDLEY, P. C. 1920. Trees and Shrubs of Mexico. Contr. U.S. Nat. Herb. 23(1): 140, footnote. THOMAS, M. H. 1967. (Pers. Com.) TRYON, R. 1959. (Pers. Com.) WULF, GLORIA 1967. (Pers. Com.) HYBRIDIZATION WITHIN THE TRITICEAE OF ALASKA: A NEW X ELYHORDEUM AND COMMENTS. WM. W. MITCHELL! AND H. J. HODGSON? Frequent intergeneric hybridization between members of the Triticeae has led some taxonomists to reconsider generic affiliations in this group. Lóve and Lóve (1965) recently included in the genus Elymus the members of Agropyron referred to Roegneria by some authors, at the same time transferring Elymus arenarius L. and its relatives from Elymus to Leymus. A number of these grasses occur in Alaska, and their crossing relations are pertinent to the proposed taxonomic realignment. This paper reports on an additional intergeneric cross in the Triticeae, summar- izes the recorded crosses involving Alaskan Triticeae, and comments on the above realignment. Plants found at two locations in the vicinity of Palmer, Alaska, are apparent hybrids of Elymus sibiricus L. and Hordeum jubatum L. The inflorescence of the hybrid has the general appearance of H. jubatum but is more nodding with shorter awns that become curved and divergent upon drying. The nodding habit of the inflorescence is not as extreme, however, as in E. sibiricus (Fig. 1). Under favor- able growing conditions the plant is broadly tussocked and leafy and produces numerous flowering culms. The characteristics of Hordeum are expressed in the possession of (i) 3 spikelets at a node, the laterals being subsessile, (ii) awnlike glumes, and (iii) a tendency to disarticulation of the main rachis. The Elymus influence is evident in a plurality of well-developed florets in all 3 spikelets on some of the plants and the occurrence of 3 spikelets at a rode (instead of 1 or 2, which is character- istic of the X Agrohordeum hybrids). "University of Alaska Agricultural Experiment Station, Palmer. "Cooperative Stat» Research Service, U. S. Department of Agricul- ture, Washington, D.C. 467 468 Rhodora [Vol. 70 Plate 1398. | Fig. 1. Elymus sibiricus left, X Elyhordeum arcuatum center, Hordeum jubatum right. The only wild ryes occurring in the area of hybridization are Elymus sibiricus and E. mollis Trin. Elymus mollis is essentially precluded from consideration because of its stiffly erect inflorescence, densely pilose glumes and lem- mas, pilose rachis, large awnless spikelets, long anthers and strongly rhizomatous habit. The tufted E. sibiricus with its pendulous spike, long curved awns, scabrid spike- lets, and spikelet measurements appropriate to the hybrid, quite obviously is one of the parents. Both the long-awned Hordeum jubatum and short-awned H. brachyantherum Nevski occur in the area and possess characters pertinent to the hybrid. Some characteristics of the taxa under consideration are compared in Table 1. Conceivably, either species of Hordeum could have com- bined with E. sibiricus to produce the hybrid. However, the long-barbed prickes and pilose leaf surface of the hybrid definitely favor H. jubatum over H. brachyanthe- rum. Moreover, anther length of the hybrid tends to be intermediate between those of H. jubatum and E. sibiri- cus, CYTOLOGICAL OBSERVATIONS. The results of meiotic studies are summarized in Table 2. Both parents and the hybrid are tetraploids (2n—28). 1968] Triticeae — Mitchell and Hodgson 469 TABLE 1. Some morphological comparisons of X Elyhordeum arcuatum and its possible parents Hordeum x brachy- Hordeum Elyhordeum Elymus antherum jubatum arcuatum sibiricus lemma (mm) 6.5-10 5-7.5 9-11 9.5-14.2 lemma-awn 6-15 85-75 17-25 11.5-31 (mm) anther (mm) 1.3-1.8 0.8-1.4 1.0-1.5 1.4-1.9 abaxial glabrous often pilose glabrous surface of pilose to pilose flag leaf abaxial with scattered, densely scabrid smooth to surface of short-barbed scabrid with with short scabrid with flag leaf prickles long-barbed to long- short-barbed prickles barbed prickles prickles All cells viewed bore univalents varying in number up to 14. Generally 5 to 8 bivalents were present in a cell, and about one-half of the cells contained multivalents. Ana- phase I figures contained from 7 to 16 laggards, and all 249 quartets observed bore micronuclei. The frequency of pairing and multivalent formation indicates some homology between one set of genomes as well as the occurrence of translocations in the differentiation of the genomes. Pollen TABLE 2. Observations of meiosis in X Elyhordeum arcuatum Stage of No. Cells Meiosis Observed Univalents Bivalents Multivalents Laggards Metaphase I 28 245 126 11 Anaphase I 24 311 Average per cell 10.7 5.5 0.5 TAS 470 Rhodora [Vol. 70 grains accepted a light stain in aniline blue but not the deep granular stain of normal pollen. The hybrid appears completely sterile. DESCRIPTION OF HYBRID. X Elyhordeum arcuatum Mitchell & Hodgson, hybr. nov. Fig. 1. (Putative parents: Elymus sibiricus L. XX Hordeum jubatum L.) Gramen perenne, caespitosum; culmi 40-85 cm alti 2-4 nodis; folia pilosa supra, scabra infra, 3-10 mm lata, 7-22 cm longa; spica arcuata, nutans, 6-15 cm longa; spiculae 3 ad nodum; spicula cen- tralis 1-3 flosculis, sessilis; spiculae laterales 0-2 flosculis, subsessiles; floseuli steriles; glumae scabrae, aristiformes, usque ad 25 mm longae; lemmata scabra, 9-11 mm longa, aristis usque ad 25 mm longis; aristae curvantes ubi siccae; antherae 1.0-1.4 mm longae; 2n=28. HOLOTYPE: Roadside at junction of Fireweed Avenue and the Alaska railroad in Palmer, Alaska, 27 Jul 63, Mitchell & Hodgson 980, (ALA); isotynes (AES), (US), (1SC). Waste area at east end of east-west runway, Palmer, Alaska airfield, 23 Sep 64, Mitchell & Hodgson 923A1, paratype (AES). Experimental garden, Alaska Agri- cultural Experiment Station, Palmer, 4 Jul 67, Mitchell 74D1, Mitchell 74D2, paratypes (AES). DISCUSSION. An interesting pattern of intergeneric hybrid links has emerged from recorded crosses within the Triticeae of Alaska (Lepage 1952; Bowden 1958; Hodgson 1964; Hodg- son and Mitchell 1965; Mitchell and Hodgson 1965) as illus- trated in Figure 2. A number of reported crosses, that as yet have not been observed in Alaska, are also designated for species that are nevertheless represented in the state (Stebbins, et al. 1946a, 1946b; Lepage 1952, 1957; Bowden 1958, 1959). These are denoted by dashed lines in Figure 2. See Bowden (1967) for an excellent review of the inter- generic hybrids of the Triticeae of North America. In the proposed reclassification of the taxa of Figure 2 (Lóve and Lóve 1965) Elymus mollis and E. innovatus Beal would be placed in Leymus and the wheat-grasses Agropy- ron sericeum Hitche., A. trachycaulum (Link) Malte, and A. violaceum (Hornemann) Lange (= A. latiglume (Scribn. & Smith) Rydb.) combined with the wild ryes Elymus sibiricus, E. glaucus Buckley and E. hirsutus Presl. 1968] Triticeae — Mitchell and Hodgson 471 Elymus sibiricus E. glaucus E. hirsutus d n \ N À N > 2 pu jubatum pr brachyantherum D » Agropyron sericeum "SA. et A.violaceurm ^ N hes Bé Ee A Elymus innovatus E. mollis Fig. 2. Summary of intergeneric hybridization recorded for the Triticeae of Alaska. Heavy lines denote frequent occurrences, thin lines infrequent occurrences, and dashed lines recorded crosses not yet observed in Alaska. Five of the crosses, indicated by arrows, were observed in the vicinity of Palmer, Alaska. under Elymus. Relationships expressed in Figure 2 indi- cate, however, an affinity between Hordeum jubatum and members of Agropyron which is as close as that between Agropyron and Elymus. Indeed, Hordeum jubatum and its close relative H. brachyantherum, which hybridize (Bowden 1962; Mitchell and Wilton 1964), serve as a hub relating all the taxa. In none of the instances observed has the ability to cross appeared to have led to intergeneric introgression, this apparently being blocked by hybrid sterility. But the possi- bility of its occurring is not precluded, since colchicine treatment of X Agroelymus palmerensis Lepage (Agropy- ron sericeum X Elymus sibiricus) has demonstrated that fertility may be established in the allopolyploids. A wide array of recombinants has been produced through the C; 472 Rhodora [Vol. 70 generation from open pollination of fertile octoploids (Hodgson, unpublished data). The incompleteness of the reproductive barriers between members of this generic complex, embracing as it does such morphologically disparate groups, seemingly confounds any attempt at a wholly satisfactory classification of the com- plex. Nevertheless, we consider it noteworthy that the character difference employed to discriminate Agropyron and Elymus is sufficiently strong to be reflected in crosses involving members of both genera with the same species, Hordeum jubatum. X Agrohordeum hybrids of this cross bear 1 to 2 spikelets at most nodes, rarely 3, while X Ely- hordeum hybrids bear 3 well developed spikelets at a node.' Acknowledgement. We thank Mrs. Zorka D. Mitchell for her assistance in composing the Latin description. LITERATURE CITED BowDEN, W. M. 1958. Natural and artificial X Elymordeum hy- brids. Canad. Jour. Bot. 36: 101-123. 1959. Chromosome numbers and taxonomic notes on northern grasses. I. Tribe Triticeae. Canad. Jour. Bot. 37: 1141-1151. 1962. Cytotaxonomy of the native and adventive species of Hordeum, Eremopyrum, Secale, Sitanion, and Triticum in Canada. Canad. Jour. Bot. 40: 1675-1711. 1967. Taxonomy of intergeneric hybrids of the tribe Triticeae from North America. Canad. Jour. Bot. 45: 711-724. HopGsoN, H. J. 1964. Cytology, morphology, and amino acid characterization of the putative intergeneric hybrid, Agroelymus palmerensis, and its presumed parents. Crop Science 4: 199-203. Hopason, H. J. and W. W. MITCHELL. 1965. A new Elymordeum hybrid from Alaska. Canad. Jour. Bot. 43: 1355-1358. LEPAGE, E. 1952. Etudes sur quelques plantes Américaines II. Hybrides intergénérique. Naturaliste Canad. 79: 241-266. 1957. Etudes sur quelques plants Américaines. VI. Naturaliste Canad. 84: 89-103. Love, A. and D. Lôve. 1965. Taxonomic remarks on some Ameri- can alpine plants. Univ. Colo. Studies, Biol. Ser. 17. 'See Bowden (1958) and Pohl (1966) for other Elymus spp. X Hordeum jubatum crosses not cited in this paper. 1968] Triticeae — Mitchell and Hodgson 473 MITCHELL, W. W. and H. J. HopGson. 1965. A new X Agrohordeum from Alaska. Bull. Torrey Bot. Club. 2: 403-407. MITCHELL, W. W. and A. C. WILTON. 1964. The Hordeum jubatum- caespitosum-brachyantherum complex in Alaska. Madrono 17: 269-280. Pont, R. W. 1966. X Elyhordeum iowense, a new intergeneric hybrid in the Triticeae. Brittonia 18: 250-255. STEBBINS, G. L. JR, J. I. VALENCIA and R. M. VALENCIA. 19462. Artificial and natural hybrids in the Gramineae, tribe Hordeae. I. Elymus, Sitanion and Agropyron. Am. Jour. Bot. 33: 338-351. 1946b. Artificial and natural hybrids in the Gramineae, tribe Hordeae. II. Agropyron, Elymus, and Hordeum. Am. Jour. Bot. 33: 579- 586. GENERIC CONSIDERATIONS CONCERNING CARPHEPHORUS AND TRILISA (COMPOSITAE) HENRY J.-C. HEBERT The genus Carphephorus (Greek: karphe: chaff, and phoros: bearing) was established by Henri Cassini (1816). In describing this genus, Cassini stated that it differed from Liatris “par le clinanthe muni de grandes squamelles comme les calea, et par l'aigrette nonplumeuse” and conse- quently named the species Carphephorus pseudoliatris; in 1817, he placed Carphephorus near Liatris, in the natural tribe Eupatorieae. In 1818, Cassini also erected the subgenus Trilisa which he elevated to generic rank in 1820 when he placed it with 15 other genera in the Eupatorieae. In establishing Trilisa, he stressed that it had the greatest affinity with Carphe- phorus from which it differed by the naked receptacle, but he was also careful to point out that Trilisa odoratissima "portoit accidentellement quelques squamelles". This last remark has been a central point of controversy over the generic limits of Carphephorus and Trilisa. No species were actually transferred to Trilisa until 1828, when Cassini cited Liatris odoratissima and Liatris paniculata of Willde- now (1803) as species of Trilisa. However, the genus Trilisa was not accepted at once, De Candolle (1836) and Torrey and Gray (1841) used it as a sectional name of the genus Liatris, but Bentham and Hooker (1873) referred to the incorrect use of Cassini’s generic name for a section. Carphephorus was thought to be of Siberian origin until Torrey and Gray (1841), prompted by a remark of Chap- man, found that the chaffy receptacle had been overlooked by some botanists in describing other species and rightly recognized that Carphephorus referred to American plants. Consequently they proceeded to transfer Liatris bellidifolia, Liatris tomentosa and Liatris corymbosa to Carphephorus. In 1824, the monotypic genus Litrisa (anagram of Tri- lisa) was described by Small for an endemic plant of Central 474 1968] Carphephorus and Trilisa — Hebert 475 Florida. He concluded his description with the following remark: “Technically it is most closely related on the one hand to Trilisa, by its involucre, and on the other, to Car- phephorus, by its chaffy receptacle.” Apparently Small contradicted himself, because in his flora (page 1304, 1933) he keyed out Litrisa as having a naked receptacle. Robinson (1934) noticing the contradiction, examined the type as well as other material and apparently not finding any pales, he decided to transfer Litrisa to Trilisa rather than Car- phephorus, implying as it had been customary, that Trilisa has no pales. So in addition to T. odoratissima and T. pani- culata, a third species carnosa was added to the small genus Trilisa. C. W. James (1958), upon examination of the three species of Trilisa, found that pales may be borne in any one of them and decided to merge at least Trilisa carnosa with the genus Carphephorus, using as primary generic distinction, the imbrication of the phyllaries. As the other two species had generally non-imbricate phyllaries, he con- sidered that it was better “at the presnt time" to keep them in the segregate genus T'rilisa. The taxonomic discrepancies encountered in the Carphe- phorus-Litrisa-Trilisa complex cannot be attributed to thoughtless "splitting" or “lumping” but to incomplete in- formation and inconsistent primary generic characters. Already in 1913, Robinson had noted in his studies of the Eupatorieae, that it is probable that “when these genera are more satisfactorily represented in herbaria some new and more convincingly natural re-adjustment of generic lines will become possible." Bearing in mind that “fundamental” or “key” character can lead to misplacement of taxa, and that “every taxonomic character is potentially important" (Cronquist, 1957), a careful analysis of all available data was made. As the degree of relation may be determined by the “scientific” use of morphological similarity, a statistical evaluation of various characters that are “repeatable and verifiable” (Davidson,1963) was attempted to corroborate the results 476 Rhodora [Vol. 70 obtained. The validity of this approach is based upon the premise that “similar or identical genetic systems operate to produce similar or identical phenotypes, always of course, allowing for the effects of the environment. If two plants have similar structures throughout, they are related.” (Rollins, 1953) The survey revealed a very close relationship between all seven species under consideration. a) Karyotype: The cytological studies have yielded no evidence for the maintenance of the two species of Trilisa as distinct from Carphephorus. All seven species have 2n = 20. The karyotypes of Trilisa odoratissima and Trilisa paniculata have been found “to be indistinguishable from that of Carphephorus." (Gaiser, 1954). The seven species apparently have: 1 pair of long chromosomes with a median constriction 1 pair of long chromosomes with a submedian constriction 1 pair of long chromosomes with a subterminal constriction 3 pairs of chromosomes of median length with a median constriction 1 pair of chromosomes of median length with a submedian constriction 2 pairs of chromosomes of median length with a subterminal constriction 1 pair of short chromosomes with a median constriction. These chromosomal similarities are particularly signifi- cant, because most of the other members of the Kuhniinae have chromosomes which differ either in shape and/or in number. So far as known the species are perfectly distinctive, present relatively little variation and do not intergrade at all with others; interspecific hybridization has not been reported, in spite of overlapping ranges, similarities of habitats and coinciding flowering times. It is to be noted 1968] Carphephorus and Trilisa — Hebert 477 that Carphephorus populations have generally a limited number of individuals and do not cover a wide area. b) Distribution: The geographical distribution is a natural one; the seven species are confined to the south- eastern coastal plain of the United States, generally on dry pine barrens. c) Habit: All are perennial caulescent herbs with a non- deciduous rosette of basal leaves and alternate cauline leaves decreasing in size upwards. The blades are entire or shal- lowy toothed. d) Basal leaves: Except for Carphephorus pseudoliatris, the basic pattern of the basal leaves is oblanceolate or spatu- late. The ratio: maximal width of distal part/width of proximal part (measured at the quarter of the total length of the leaf) shows that the range of variation is approxi- mately the same; only C. pseudoliatris and C. carnosus have less variation, the first having acicular leaves, the second, leaves tending to be linear-lanceolate. The ratio: maximal length/maximal width is not very significant due to the great variation of the length of the leaves. e) Pubescence: The pubescence has been found to be of use as a character for specific differentiation between C. tomentosus and C. bellidifolius and between T. paniculata and T. odoratissima, The trichomes are generally non- glandular, multicellular, uniseriate, cylindrical, tapering towards the end and variable in length; in 7. paniculata, the trichomes are biseriate. The glandular trichomes are generally present in de- pressions below the epidermal surface and cause the characteristic punctation. These trichomes are biseriate in C. carnosus and C. tomentosus, and in the latter they are accompanied by laterally placed uniseriate filaments. They are uniseriate in the five other species (and occasion- ally also in C. carnosus) and usually occur singly, but in C. pseudoliatris and C. corymbosus they can also be found in groups of 2, 3 or 4. As noted by Gaiser (1954), a very reticulate pattern is evident in the distribution of the glandular trichomes. 478 Rhodora [Vol. 70 f) Inflorescence: All species have the same basic cymose type of inflorescence, the heads being arranged in a corym- bose fashion in six species, and in a racemose fashion in T. paniculata. The cymes are contracted in C. pseudoliatris and C. carnosus, and spreading in the other species. g) Flowers: The number of flowers per head varies from 4 in T. paniculata to 46 in C. pseudoliatris, In the related genus Liatris, it varies from 3 to 70 (Gaiser, 1946). The floral parts are identical in all species. Their size varies but the range is not abnormal for a genus. The corolla for example, varies from 3.4 mm to 11 mm in this complex, while in Liatris, it varies from ca. 5 mm to ca. 20 mm (Gaiser, 1946). The corolla lobes show also a normal grada- tion in length, being longest in C, bellidifolius (Mean: 2.286 mm) and shortest in C. carnosus (Mean: 0.723). The anthers are definitely shorter in C. carnosus, T. odoratissima and T. paniculata than in the four other species. The apical end of the anthers is scarious in the seven species, but notched in only five of them: C. pseudoliatris, C. tomento- sus, C. bellidifolius, C. corymbosus and C. carnosus. h) Pappus: The bristles being of unequal length on each fruit, the overall length of the pappus was obtained by measuring the longest bristles. There is a variation from 3 to 10 mm, while in the genus Liatris it varies from 2.5 to 10 mm (Gaiser, 1946). The bristles generally number 30-55; in T. odoratissima they number 20-35. i) Achenes: They are basically alike, being 10-ribbed, angular and narrowed towards the base. The mature achenes are of brown color, but in some cases they approach black. The length varies from 2 to 7 mm; in Liatris it varies from 3 to 10 mm (Gaiser, 1946). j) Phyllaries and pales: (1) Shape and size: The margin of phyllaries has proved to be a useful specific character; the margin can be pectinate-ciliate, erose-ciliate, eciliate or scarious; the apex also, varying from acute, mucronate to obtuse or rounded, is characteristic; on the other hand, characters like shape, width, length or pubescence vary much within 1968] Carphephorus and Trilisa — Hebert 479 each species and can rarely be used as distinctive characters. It is to be noted that the pales generally resemble the inner phyllaries in shape, pubescence and size; occasionally, they are much narrower than the phyllaries. (2) Number: The pales and the phyllaries which have been used as primary generic characters, show apparently a great range of numerical variation. But as the number of flowers per head varies much also, that range presents nothing unusual. For example, C. pseudoliatris which may have up to 46 flowers in one head has a mean number of pales of 23 and a mean number of phyllaries of 20; T. pani- culata which often has only 4 flowers in one head has a corresponding low mean number of pales, 0.15 per head, and a mean number of phyllaries of 6.2 per head. It seems that the number of phyllaries and pales is related to the number of flowers. If one considers that the phyllaries and the pales are homologous — this is in accord with the theory of formation of the capitulum from an ancestral racemose umbel (Small, 1919) — a correlation can be seen between the number of phyllaries and pales (considered together) and the number of flowers. An increase in one set is matched by an increase in the other set; 7. paniculata and C. carnosus which have the smallest number of flowers have also the smallest number of pales and phyllaries while C. pseudoliatris which has the greatest number of flowers has also the greatest number of pales and phyllaries. On the other hand, a statistical attempt to relate the number of flowers and either the number of pales or the number of phyllaries gave negative results, especially in the case of C. pseudoliatris which seems to possess too many pales and too few phyllaries. This stems from the difficulty of delimiting the pales and the phyllaries ; in fact, no sharp line can be drawn between them, their position being often very indefinite and their shape identical. As a control, one species of Liatris has also been plotted and apparently does not contradict the theory that there is a correlation between flower-number and pale-phyllay-num- ber. So, in the complex under consideration, it appears 480 Rhodora [Vol. 70 sound to consider phyllaries and pales as a whole, because their structure is generally indistinguishable and seems to be controlled by the same genetic factor. From these considerations, it is normal to expect fewer pales in a small head or even none, the flowers being almost all peripheral in position; thus, it would be an inadvisable decision to segregate T. odoratissima and T. paniculata simply because one fails to see pales in some heads. In the case of T. odoratissima, the pales were constantly seen in the plants I collected in September 1965, as well as in most herbarium specimens examined. Unfortunately no fresh material of T. paniculata could be examined and the her- barium specimens showed a great paucity of pales. How- ever, James who had more opportunities of observing fresh material wrote in 1958: “Upon examination of all the species of the genera under consideration, I conclude that pales may be borne in any one of the species." As the pales are deciduous and tend to fall off when they are dry, it is possible that the “odoratissima” specimen examined by Cas- sini (1820) had lost most of its pales; that would have prompted Cassini to establish a distinction between Car- phephorus and Trilisa and to write that “Les Trilisa ont la plus grande affinité avec le Carphephorus, qui n'en différe que par le clinanthe squamellifére; et j'ai observé sur la Trilisa odoratissima que le clinanthe portoit accidentelle- ment quelques squamelles." It is also possible that, finding only one or two pales in each head, or sometimes none, he concluded that the pales are not a permanent feature and occur only occasionally. At any rate, his word “accidentelle- ment" is definitely too strong and conveys an erroneous idea about the presence of pales in T. odoratissima, Pre- sumably, if Cassini had examined some fresh material, he would have merged Trilisa with Carphephorus, because he was not impressed by other differences. These considera- tions show how the exclusive use of the “‘presence of pales” is questionable as a primary generic criterion. As E. E. Sherff has pointed out: “There should not be separation of genera solely upon the presence or absence of one or 1968] Carphephorus and Trilisa — Hebert 481 more supposedly diagnostic characters." (Reported by Turrill, 1942). This principle is particularly valuable in this case where the state of maturity of the plant can affect the character under consideration. (3) Imbrication of phyllaries: The imbrication of phyllaries used by James to segregate “much more natu- rally” Trilisa odoratissima and Trilisa paniculata from Carphephorus is considered to be on a very weak basis (Shinners, 1946), especially when it is not supported by additional characters. Besides, it presupposes a sharp line where none can be drawn. In this complex, a gradation from 1 to 5 series of phyllaries is found. As the number of series is dependent on the number of phyllaries, it is evident that when the phyllaries number only 4 or 5 they can hardly overlap and be imbricated. So the imbrication of phyllaries is simply a qualitative factor based on quanti- tative data and cannot be taken as a distinctive character in itself. Upon examination of numerous specimens, I found that the number of series is highest in C. tomentosus and lowest in T. paniculata. But, for each species, the number varies and no line of demarcation can be set between them. Robinson (1934) had come to the same conclusion when he merged Litrisa with Trilisa: “The alleged involucral differ- ence appears in fact a rather vague one of degree only." Too much emphasis and exclusive reliance placed upon either the amount of pales or the arrangement of phyllaries has led to different results and varied segregation, but when the plants are considered as a whole and not as an aggregate of uncorrelated features, I find no solid ground to keep Trilisa as a separate genus. To maintain Trilisa distinct from Carphephorus, "because of tradition and inertia" (McVaugh, 1945) would impede a better understanding of their fundamental relationship and conserve an unwar- ranted segregation. The generic transfer of Trilisa I propose is somewhat similar to the case of Hymenopappus. That genus had been placed in the tribe Helenieae, because the heads have no chaff; recent investigations showed that Hymenopappus 482 Rhodora [Vol. 70 was not only more naturally placed in the Anthemideae, close to its obviously related prototype, the chaffy Leu- campyx, but had to be united into a single genus, on the basis of the total data (Turner, 1956) ; since then the genus Hymenopappus includes “everything from perennial plants with chaffy receptacle and rayed heads to biennials with neither chaff nor rays." (Turner, 1956). With the transfer of both species of Trilisa, Carphephorus remains a homo- geneous genus, much more than Hymenopappus, Liatris and many other composite genera. Only T. paniculata with its few pales, its rarely imbricate phyllaries and its cyme arranged in a paniculate fashion, would be slightly ano- malous but would hardly alter the generic limits. Thus, if one compares the range of variation of the Carphephorus- Trilisa complex with the one of Liatris, one finds that there is no justification to keep Trilisa as a distinct genus. In fact, it is significant that Liatris which has been studied thoroughly by Robinson and Gaiser has not been split in spite of the numerous variations observed. There is no reason to consider Carphephorus a fundamentally different case. As pointed out by Blackwelder "group which cannot be distinguished at any particular level by the characters used for their neighbors must be combined at that level." (1963) As I find that the biological status of all the species concerned may be best expressed by the grouping of all into one genus, I must follow the only alternative of uniting Trilisa to Carphephorus. This more com- prehensive genus conforms to the definition of a genus given by Buxbaum (1951), as *the sum total of species belonging to a phylogenetic unit recognized as such by the unity of its morphological type." It is also in accord with Recommendation 3 of McVaugh (1945): “The most im- portant criterion of any supposed genus is not the width of the gap between it and another, but its own biological unity. Homogeneity in many characters, regardless of the degree of overlapping of these characters with those of other genera, is the best indicator of this unity." No 1968] Carphephorus and Trilisa — Hebert 488 doubt that those who like to make far too fine generic segregations will find that such a group is a return to the broad concept of genus in folk botany; these forget that "large genera may be quite as truly 'entities of nature' as small ones, . . . and from a practical and linguistic stand- point may be a far more useful concept." (Bartlett, 1940). Carphephorus odoratissimus (J. F. Gmel) Hebert, comb. nov. based on Chrysocoma odoratissima J. F. Gmel, Syst. 2: 1204. 1792. Type: preserved in John Fraser's collec- tion, in the British Museum, London. Place of origin: Carolina. Carphephorus paniculatus (J. F. Gmel) Hebert, comb. nov. based on Chrysocoma paniculata J. M. Grael, Syst. 2: 1204. 1792. Type: preserved in John Fraser's collec- tion, in the British Museum, London. Place of origin: Carolina. Isotype: Lamarck Herbarium, Paris. KEY TO THE SPECIES. a) Heads with 12-43 flowers; pales 4-33 per head; phyllaries 12-33 per head; involucres 6-11 mm high. b) Basal leaves acerose, involute. neen 1. C. pseudoliatris. b) Basal leaves oblanceolate or spatulate, not involute. c) Phyllaries pubescent, their margins pectinate-ciliate. ........ DM —————— —— 2. C. tomentosus. c) Phyllarics glabrous or minutely hairy, their margins erose- or suberose-ciliate. d) Inflorescence an open slenderly-branched cyme; corolla lobes ca. 2-3 mm long. ................ 3. C. bellidifolius. d) Inflorescence a dense compact corymbose cyme; corolla lobes ca. 1 mm long. ............ eene 4. C. corymbosus. a) Heads with 3-12 flowers; pales (0-) 1-3 per head; phyllaries 4-14 per head; involucres 3-6 mm high. e) Inner phyllaries pubescent, with pectinate-ciliate margins. .... e) Inner phyllaries glabrous, with eciliate margins. f) Stems glabrous; inflorescence an open and spreading corym- DOSE cume. ..eeesee emere 6. C. odoratissimus. f) Stems sparingly hirsute, viscid; inflorescence a thyrsoid 11:033 19 [NR —— 7. C. paniculatus. BIOLOGY DEPARTMENT, SEYCHELLES COLLEGE, SEYCHELLES ISLANDS, 484 Rhodora [Vol. 70 LITERATURE CITED BARTLETT, H. H. 1940, History of the generic concept in botany. Bull. Torr. Bot. Club. 67: 349-362. BENTHAM, G. & J. D. HOOKER. 1873. Genera Plantarum. BLACKWELDER, R. E. 1963. Classification of the Animal Kingdom. CASSINI, H. 1816. Bulletin Société Philomatique de Paris. 198-200. - 1817. Carphephore. Dictionnaire Sciences Naturelles. 7: 148-150. 1818. Bulletin Société Philomatique de Paris. 139-142. . 1820. Eupatorieae, Dictionnaire Sciences Naturelles 16: 9-10. — 1828. Trilisa. Dictionnaire Science Naturelles 55: 310. CRONQUIST, A. 1957. Outline of a New System of Families and Orders of Dicotyledons. Bull. Jard. Bot. Bruxelles 27: 13-40. Davipson, R. A. 1963. Initial biometric survey of morphological variation in the Cirsium altissimum-C. discolor complex. Brit- tonia 15: 222-241, DE CANDOLLE, A. 1836. Prodromus systematis naturalis regni vege- tabilis 5: 131-133. GAISER, L. O. 1946. The Genus Liatris. Rhodora 48: 165-183, 216- 263, 273-326, 331-382, 393-412. . 1954. Studies in the Kuhniinae (Eupatorieae). Jour- nal Arnold Arboretum 35: 87-133. JAMES, C. W. 1958. Generic considerations concerning Carphe- phorus, Trilisa and Litrisa (Compositae). Rhodora 60: 117-122. McVavGH, R. 1945. The genus Triodanis Rafinesque, and its rela- tionships to Specularia and Campanula. Wrightia 1: 13-52. ROBINSON, B. L. 1913. A generic key to the Compositae-Eupatorieae. Proc. Amer, Acad. Arts & Sci. 49: 429.437. 1934. Records preliminary to a general treatment of the Eupatorieae. XI. Contrib. Gray Herbarium 104: 3-49. RoLLINs, R. C. 1953. Cytogenetical approaches to the study of Genera. Chronica Botanica 14 (3): 91-159. SHINNERS, L. H. 1946. Revision of the genus Kuhnia L. Wrightia 1: 122-144, SMALL, J. K. 1919. The origin and development of the Compositae. New Phytologist. Vol. 16-18. 1924. Plant Novelties from Florida. Bull. Torr. Bot. Club 51: 379-393. ~~~ 1933. Manual of the Southeastern Flora. TORREY, J. & A. GRAY. 1841. Flora of North America. 1968] Spiranthes ovalis — Johnson 485 TURNER, B. L. 1956. A cytotaxonomic study of the genus Hymeno- pappus (Compositae). Rhodora 58: 163-186, 208-242, 250-269, 295-308. TURRILL, W. B. 1942. The concept of the Genus. Chronica Botanica 7: 188-190. WILLDENOW, C. L. 1803. Species Plantarum. 3(3): 1637. SPIRANTHES OVALIS, NEW FOR KANSAS While collecting in an oak-hickory woodland 2 miles north and ?4 mile east of Baldwin City, Douglas County, Kansas, a specimen of Spiranthes ovalis Lindl. was taken on October 13, 1967 (Johnson 1184, KANU). The plant was associated with Botrychium dissectum Spreng. var. obli- quum (Muhl.) Clute, Corallorhiza odontorhiza (Willd.) Nutt., and Monotropa uniflora L. This orchid has previous- ly been reportec from Florida to Texas, north to Virginia Kentucky, Indiana, and Missouri. In Missouri the species is known only from Mississippi, St. Louis, and Jackson Counties. My collection is an extension westward of the range of the species. KERMIT L. JOHNSON UNIVERSITY OF KANSAS, LAWRENCE, 66044. AN EXAMPLE OF A VERNONIA HYBRID IN A DISTURBED HABITAT SAMUEL B. JONES, JR. The identification of ironweeds has been a persistent source of difficulty, resulting in large part from the blurring of species lines by hybridization. Cain (1944) suggested that hybridization in Vernonia increased with the arrival of European man who cut timber, built roads, drained land, and created numerous ecological niches not previously present. He presented the hypothesis that Vernonia X illinoensis Gleas. was derived from hybridization and intro- gression of three species, V. missurica Raf., V. fasciculata Michx and V. altissima Nutt., and this hybridization was augmented by the disturbance of the habitat by man, A relationship between natural hybridization and disturbed habitats has been recognized for some time. Anderson's classic paper (1948) summarizes our present day concept of the habitat and its significance in natural hybridization. Anderson points out that hybrids often survive where man has disturbed or “hybridized” the habitat. Hybrid swarms are most often found in disturbed habitats, particularly those which are intermediate between the habitats of the parents and to which neither parent is as well adapted as are the hybrids. In 1963, I found an interesting example of Vernonia in a disturbed habitat near Leakesville in Greene County, Mississippi. In the area concerned, Vernonia altissima. grows along the edge of the wet, bottomland forest near the Chickasawhay River with many of the plants of the population growing in standing water. This population of V. altissima is not pure but apparently contains genes of V. missurica reflecting the introgression mentioned above. This does not, however, have any bearing on the problem at hand and to avoid confusion I will call this taxon V. altissima. Southeast of the river is a high, very dry, sandy turkey oak and longleaf pine ridge with V. angustifolia 486 1968] Vernonia — Jones 487 abundant. Some years ago, a road with a long fill was con- structed from the wet bottomland to the dry ridge (Fig. 1) and a hybrid swarm of several hundred clones representing both F, and backcross types can now be found on this fill. The ironweeds of this swarm are intermediate in morpho- logical features and habitat requirements between the nar- row leaved V. angustifolia and the wide leaved V. altissima. The roadfill evidently provides a number of ecological niches between the contrasting wet and dry sites resulting in a typical example of a “hybridized” habitat. | | V. ALTISSIMA FILL V. ANGUSTIFOLIA DRY WET Fig. 1. Diagrammatic scheme of the habitats and leaf tracings of the parents and hybrids. To assess the degree of interbreeding, local population samples of about 25 specimens each were collected from the two extreme habitats and along the roadfil in 1963, 1964, and 1966. In addition, individuals of the present species and selected hybrids were transplanted to the 488 Rhodora [Vol. 70 experimental garden and greenhouse. That natural hybridi- zation could be the cause of the intermediates was demon- strated by crossing Vernonia altissima and V. angustifolia. The artificial F, hybrids appeared to be vigorous and simi- lar to the hybrids found on the fill. Backcrosses were also made between the field collected hybrids and the two parental species. Pollen stainability tests with anilin blue in lacto-phenol indicated that there was no appreciable difference in stainability between the two parental species, the natural hybrids, the backcross types, or the artificial F, hybrids. The average stainability ranged from 87 to 99 percent. The local population samples were measured and scored; the important morphological features are compared in Table 1. A scatter diagram is presented in Figure 2. IO 9 V. altissima E E 8 e. E o of ee © > . © Y Es res O 2º ° ° o hybrids D SE ZP 5 ^A? ln > » V. angustifolia Eat? ^ > 4 5 6 7T 8 9 10 I INVOLUCRE HEIGHT mm. Fig. 2. Scatter diagram of Vernonia altissima, V. angustifolia, and their hybrids from local population sample. 489 i Jones nonla Ver xo) vas 'so[dures uorje[ndod [woo] JO Serpnjs uo postas 9]€UO.I2nUI o9je[nuedureo ÁAjpeoiq unu g'L— T'S wut (r6 — Un OF — €6 uut [6 — Ie unu gcT — sor queseid u3993 os03u9ulo? q1oys Ajesieds 91e [oo2uv[-ondi[[9 0} audio IJN 0} 9j?uOJOnul eje[nuedureo un 28 — L9 uu 0°8 — 8r 66 9T uu 9'0r — L'E uiu LET — SL pou3003 Á[930urod 9so1ueulo Aposieds 0} snoiqeos 91e[ooou?[ ÁJMOLIBU ojeurumoe 0j IL eje[nuedureo Á[ao.rreu uu g9 — Lt unu 9 — OP 9T — OT uu peo — 0I unu 86 — OP iaa) sn[[eo snoiqvos Ivaul] Sdt peg edeys exon[oAu] SH ƏN OAU UIPIM exon[oau peoy/sia MoO] A UIPIM u13ue'] juo] SULSIEW Jeo (198107) S92JINS JLI edeug pat DiS$1]]D "A SpHqÁH pyofysnbun "A 91)S11939€.10U) vsSpriqAuq [vinjeu 1oy} pur DWASS1D “A ‘nyofysnbun DUOULIA FO soin eof ]eo:do[oud.rour urej1ioo JO uosrreduio?) "TI ATAVIL 490 Rhodora [Vol. 70 Pollen size was determined by staining mature pollen grains with anilin blue in lacto-phenol and measuring them with an eyepiece micrometer. The sizes of the pollen grains are given in Table 2. The pollen grains of the hybrids were intermediate in size between those of the parents and had a greater standard deviation than those of the parents. Both taxa have a chromosome number of n — 17. TABLE 2. Diameter of the pollen grainsa Taxon x Range S V. angustifolia — mn 314—408 134 Hybrids 41.2 39.1 — 44.2 2.51 V. altissima 43.0 40.8 — 44.2 1.53 ^n — 125, from 5 plants of each taxon; measurements in microns. The ranges of Vernonia altissima and V. angustifolia are partially sympatric. They are, however, usually eco- logically isolated. The former grows in moist or even wet areas at the edge of low woods or in open fields, while the latter is found in well drained upland habitats such as scrub oak sand hills or dry longleaf pine woods. Studies based on the Greene County, Mississippi site clearly indicate that ecological isolation is not always effective and that the hybrids are intermediate between their parents in both morphological characteristics and in their habitat require- ments. Introgression between V. altissima and V. angusti- folia is certainly conceivable and may serve to enlarge the ecological parameters of those two taxa. I have observed this natural hybridization in scattered locations in both Alabama and Mississippi where the ranges of these two taxa are sympatric. At other sites, however, only one or two hybrids were noted. This is probably due to the lack of suitable ecological niches for the hybrids. Since hybridi- zation has ben observed at other locations, it is possible that introgression is occurring where the Species are sympatric. DEPARTMENT OF BOTANY, UNIVERSITY OF GEORGIA, ATHENS 30601 1968] Vernonia — Jones 491 LITERATURE CITED ANDERSON, E. 1948. Hybridization of the Habitat. Evolution 2: 1-9. CAIN, S. A. 1944. Foundations of Plant Geography. Harper, N. Y. 556 p. WOLFFIA IN NEW HAMPSHIRE. Two small additions to the flora of New Hampshire were discovered on 8 October 1967 when I stopped to investigate an unusual greenish coloration in the shallow water of a small roadside pond. I had visited this pond about a year previously but found no noteworthy plants there at that time. To my surprise I now discovered, growing in great profusion, two of our most diminutive flowering plants: WOLFFIA COLUMBIANA Karsten. Profusely abundant. Forming, with Wolffia punctata Griseb. and Lemna minor L., a dense floating layer about the margins of a small pond near Cohas Brook; on the east side of Interstate Highway 93 approximately 2 miles north of the Manchester City Boundary. Manchester, Hillsborough County, New Hampshire. 8 October 1967, Country- man 1718. WOLFFIA PUNCTATA Grisebach. Data as above except less abun- dant. Countryman 1719. The pond was revisited on 4 November 1967 in company with Albion R. Hodgdon. The Wolffia was still plentiful at that date. We were able to collect nearly pure samples of both species of Wolffia by pouring a pail of the pond water through two graduated sieves (kitchen tea strainers). The coarser sieve held back Lemna minor and miscellaneous debris while the finer sieve was of such a size as to contain the Wolffia. In the samples so collected about 95% of the material was W. columbiana, the balance W. punctata. Specimens of both species have been preserved in FAA in sealed glass ampoules (see Dore, W. G., Canadian. Field Naturalist 1957, 71: 10-16) and deposited in the herbaria of the New England Botanical Club and the University of New Hampshire. WM. D. COUNTRYMAN NORWICH UNIVERSITY, NORTHFIELD, VERMONT 05663 LEGUMINOSAE OF MEXICO — FABOIDEAE I. SOPHOREAE AND PODALYRIEAE VELVA E. RUDD INTRODUCTION The Leguminosae, or Fabaceae, is one of the largest and most important families of flowering plants. In Mexico, where corn and beans, tortillas and frijoles, have long been staples of diet, the legume family probably ranks second only after the grass family in importance to the basic econ- omy. Of a worldwide total of about 600 genera of Leguminosae, slightly over 100 are native to Mexico. Approximately an- other 20 to 30 genera have been introduced, many of which have been naturalized. The figures are somewhat larger if one recognizes the many segregate genera such as have been ascribed to Acacia, Pithecellobium, and Astragalus. The leguminous genera known to occur in Mexico include not only beans but other edible seeds, fruits, and vegetative parts utilized as food for man and animals. Many trees provide economic products such as lumber, resins, and pharmaceuticals; some are cultivated as ornamentals. Early reports of the natural history of Mexico mention leguminous plants utilized by the Indians, especially for medicinal purposes. Oviedo", writing in 1526, mentioned that many kinds of beans, “fesoles”, were gathered in great abundance, especially in “Nueva España é Nicaragua”, some with yellow seeds, others spotted, “pintados de pecas.” The Badianus manuscript, an Aztec herbal of 1552º, in- cludes colored illustrations of several legumes appearing to be species of Calliandra, Mimosa, Haematoxylon, and Phaseolus. White incense, possibly from Myroxylon, is mentioned in the text. Hernandez’ report on the plants of 'Fernández de Oviedo y Valdéz, Gonzalo. História general y natu- ral de las Indias. Madrid, 1851-1855. "Cruz, Martín de la, and Juan Badianus. English translation by Emily Walcott Emmart. The Badianus manuscript — an Aztec herbal. 1940. 492 1968] Leguminosae of Mexico — Rudd 493 Mexico, written in 1570-1575*, includes a great many leg- umes, some clearly recognizable, others dubiously so. From an illustration, as well as the text, “hoitziloxitl”, the “bál- samo de indias", is without doubt identifiable as Myroxylon. Taxonomic works, such as those of Linnaeus, Willdenow, Miller, Kunth, DeCandolle, and Bentham, included descrip- tions of new Mexican Leguminosae based on early collec- tions, especially those of Houstoun, Humboldt and Bonpland, Sessé and Mocifio. They will be cited in the following taxonomic text. In 1880 Hemsley* listed the legumes of Mexico and Cen- tral America as completely as then known from the herbarium material at Kew. Unfortunately, although col- lections were cited, descriptions were given only for new species; and there are no keys. The woody species of leg- umes were included in Standley's Trees and Shrubs of Mexico’, but this treatment is long out of date. There never has been a complete treatment of the herbaceous species. For the *North American Flora" the Mimosoideae, as Mimosaceae’, and the Caesalpinioideae, as Caesalpiniaceae’, were revised by Britton and Rose, and the Faboideae, as Fabaceae*, were presented, in part, by various authors, chiefly Rydberg. Perusal of Langman's tremendous bibliography? will in- dicate how much has been published on the Mexican flora and vegetation. Nevertheless, there is no adequate treat- *Hernández, Francisco. Spanish translation by José Rojo Navarro. História de las plantas de Nueva Espana. Mexico, 1942-1946. ‘Hemsley, W. B. in Godman and Salvin. Biologia Centrali-ameri- cana. Botany 1: 225-364. 1880. ‘Standley, P. C. Trees and Shrubs of Mexico, Contributions from the U. S. National Herbarium 23: 348-515. 1922; appendix 1657-1667. 1926. “Britton, N. L. and J. N. Rose. North American Flora 23: 1-194. 1928. "Britton, N. L. and J. N. Rose. North American Flora 23: 201-349. 1930. *Rydberg, P. A. North American Flora 24: 1-462. 1919-1929. *Langman, I. K. A selected guide to the literature on the flowering plants of Mexico. 1964. 494 Rhodora [Vol. 70 ment for general determination of Mexican legumes, es- pecially the faboid genera. In the preparation of this synopsis of the Leguminosae of Mexico, many general works have been consulted, includ- ing the Genera Plantarum of Bentham and Hooker”, Tau- bert’s treatment of the Leguminosae for Die Natiirlichen Pflanzenfamilien" and the supplement by Harms", The Genera of Flowering Plants by Hutchinson^, and the In- dex Nominum Genericorum"™, For economic uses, the works of Martínez, Las Plantas Medicinales de México" and Plantas Utiles de la Flora Mexicana's, and Timbers of the New World by Record and Hess" have been important sources of information. Burkart's Las Leguminosas Argen- tinas'* has been especially valuable. Also helpful have been such floristic works as the Flora of Yucatan by Standley”, the Flora of Guatemala by Standley and Steyermark^', Leg- umes of Texas by Turner*!, Arizona Flora by Kearney and Peebles?, and the Vegetation and Flora of the Sonoran Desert, by Shreve and Wiggins”, “Bentham, G. and J. D. Hooker. Genera Plantarum 1: 434-600. 1865. "Taubert, P. in Engler and Prantl. Die Natürlichen Pflanzenfam- ilien III, 3: 70-388. 1891-1894. "Harms, H. in Engler and Prantl. Die Natürlichen Pflanzenfamilien III, 3. Nachtrige 2: 30-34, 190-203. 1900. "Hutchinson, J. The genera of flowering plants 1: 221-489. 1964. "Index Nominum Genericorum, International Association for Plant Taxonomy. Utrecht. 1954 — "Martínez, M. Las plantas medicinales de México, ed. 4. 1959. "Martínez, M. Plantas ütiles de la flora mexicana. 1959. "Record, S. J. and R. W. Hess. Timbers of the New World. 1943. “Burkart, A. Las leguminosas argentinas, ed. 2. 1952. "Standley, P. C. Flora of Yucatan. Field Museum of Natural His- tory, botanical series 3: 157-492, 1930. "Standley, P. C. and J. A. Steyermark. Flora of Guatemala. Field- iana: Botany 24(5): 1-368. 1946. "Turner, B. L. The legumes of Texas. 1959. “Kearney, T. H. and R. H. Peebles. Arizona flora, 395-484. 1960. "Shreve, F. and I. L. Wiggins. Vegetation and flora of the Sonoran desert 1: 585-731. 1964. 1968] Leguminosae of Mexico — Rudd 495 As indicated in the key, I am treating the Leguminosae as one large family subdivided into three subfamilies, the Mimosoideae, the Caesalpinioideae, and the Faboideae, the latter also known as the Papilionoideae and the Lotoideae. In this study the Faboideae will be presented first for more or less arbitrary reasons, principally, because this subfam- ily was only partly treated in “The North American Flora" mentioned above. My interpretation is generally conservative, with the genera arranged in the Dalla Torre and Harms** sequence, and ten tribes, essentially as delimited by Bentham, and followed by Taubert and others. I believe that Hutchin- son's separation of the “Fabaceae” into 50 tribes is unduly divisive. Perhaps, a system somewhat intermediate be- tween Bentham's and Hutchinson's can be developed, but, for purposes of an introductory key to tribes in this paper, the former will be followed. At the suggestion of the late Dr. Faustino Miranda I am attempting to provide fairly full descriptions, synonymy, and citations of material studied. To avoid unnecessary bulk, however, I am excluding citations of Old World syn- onymy unless pertinent to Mexican taxa. Relevant mono- graphic works will be cited under the genera, but the authors' interpretations will not necessarily be followed. This work is based chiefly on herbarium material, as in- dicated by the citations. The abbreviations of herbarium names are those of Lanjouw and Stafleu?. I am grateful to the curators of those many institutions for making their specimens available to me. The citations of *F.M.Neg." refer to Field Museum neg- atives of a series of photographs taken in European her- baria by J. F. Macbride during 1929 to 1939. “Dalla Torre, C. G. and H. Harms. Genera siphonogamarum. 1900- 1907. Lanjouw, J. and F. A. Stafleu. Index herbariorum I. The her- baria of the world, ed. 5. 1964, 496 Rhodora [Vol. 70 SYSTEMATIC TREATMENT Leguminosae: Legume or bean family Leguminosae Juss. Gen. 345. 1789, nom. cons. TYPE: Faba Mill. (= Vicia L. p.p.) Fabaceae Lindl. Nat. Syst. ed. 2, 148. 1836, nom. alt. TYPE: Faba Mill. (= Vicia L. p.p.) Trees, shrubs, or herbs, the stems sometimes twining; leaves com- monly alternate, rarely opposite, compound, sometimes simple, usually stipulate, sometimes stipellate; flowers commonly bisexual, 5-merous, solitary or in compound inflorescences, axillary or terminal; sepals free or united, 4 or 5, rarely 1, then spathaceous; petals free or united, 5, sometimes fewer or lacking; stamens commonly 5-96, sometimes reduced to only 1 fertile member, the sterile members may be present or sometimes reduced to staminodes, the filaments free or united, the anthers 2-celled, dehiscing lengthwise or by terminal pores; pistil usually 1, the ovary superior, 1-locular, 1-% ovulate; fruit a legume, dehiscent or indehiscent, 1-many-seeded, 2-valved, but may be modified as a drupe, samara, follicle, or loment; seed common- ly with a coriaceous testa, reniform, lenticular, or spherical, some- times alate, sometimes arillate, the hilum orbicular to linear, sometimes circumcinct, the endosperm little or none. Distribution: Worldwide. KEY TO SUBFAMILIES OF LEGUMINOSAE Flowers actinomorphic, radiate and regular; corolla and calyx valvate in bud (except calyx imbricate in Parkieae); stamens 4-96; leaves commonly bipinnate, sometimes pinnate or reduced to phyl- lodes MM Mimosoideae Flowers generally zygomorphie, sometimes subactinomorphic; corolla and calyx imbricate in bud or, sometimes, valvate. Uppermost (adaxial) petal enveloped by the other petals in bud; stamens (fertile) 1-0, commonly 10 or fewer; leaves usually pinnate, sometimes bipinnate, rarely simple .... Caesalpinioideae Uppermost (adaxial) petal exterior in bud, enveloping the other petals; stamens 10, rarely fewer; leaves simple or pinnate, never bipinnate .............ssssssseseeeeee eene eene Faboideae Leguminosae subfamily FABOIDEAE Leguminosae subfam. Faboideae.” TYPE: Faba Mill. (= Vicia L. p-p.) Papilionaceae Giseke, Praelect. Ord. Nat. Pl. 415. 1792. Fabaceae H. G. L. Reichenbach, Consp. Regni Veg. 149. 1828. "The use of “Faboideae,” based on Fabaceae, agrees with article 19 of the International Code of Botanical Nomenclature, 1966. 1968] Leguminosae of Mexico — Rudd 497 Leguminosae fam. Papilionatae A. Braun in Ascherson, Fl. Prov. Brandenb. Einleitung 67. 1864. Leguminosae subfam. Papilionoideae Robinson & Fernald, Gray's New Man. Bot. ed. 7, 500. 1908. Papilionaceae subfam. Lotoideae Luerssen, Grundzüge Bot. 379. 1877. Type: Lotus L. Legwminosae subfam. Lotoideae (Luerssen) Rehder, Journ. Arn. Arb. 26: 477. 1945. Characters those of the family except as noted in the key to sub- families, viz., leaves never bipinnate; flowers with adaxial petal exterior in the bud, enveloping the other petals, if present; stamens commonly 10, rarely fewer. The taxa not typified above presumably were based on Faba Mill. KEY TO TRIBES OF SUBFAMILY FABOIDEAE IN MEXICO a. Flowers with stamens distinct, the filaments separate to the base. b. b. Leaves pinnately 5- many-foliolate. seess ]. SOPHOREAE b. Leaves digitately 1-3-foliolate. ........... e 2. PODALYRIEAE a. Flowers with stamens united, all or most of the filaments joined, at least in part, monadelphous or diadelphous. c. c. Legumes commonly indehiscent, drupes, samaras, or loments, sometimes compressed, nonalate, usually more than 1 em. long. d. d. Fruits articulated, the joints usually separating at maturity; stamens monadelphous or diadelphous 5:5 or 9:1. eme T. HEDYSAREAE pes eeEnencsrrncsecebessaadeesencovecorenececeaesccscnsneeeeeeseeHeeereceeRseadeeeee ment sometimes separating from the others .... 8. DALBERGIEAE c. Legumes 2-valved, usually dehiscent or, if indehiscent, small, but a few mm. long. e. e. Leaves digitate. f. f. Leaflets serrate or denticulate; leaves commonly 3-foliolate, rarely more or less; stamens diadelphous 9 :1. .. 4. TRIFOLIEAE f. Leaflets entire; leaves 1-many-foliolate; stamens monadel- phous or diadelphous. g. g. Stamens monadelphous, the filaments uniform in width, forming a closed tube or open sheath, the anthers di- morphic, alternately basifixed and dorsifixed. seenen 3. GENISTEAE $660060008800000b50088 00699096 0000000 000 BO aS 005050078869 FOTO NER EEEE apex, the anthers uniform. .......- een 5. LOTEAE e. Leaves pinnate. h. h. Leaf rachis terminating in a tendril or bristle; leaves even- pinnate; stamens 9, monadelphous, or 10, diadelphous Ores 9. VICIEAE Ee des Lbs cle gilt Sone sew EE EE CARE NM A STURM 498 Rhodora [Vol. 70 h. Leaf rachis usually with a terminal leaflet, without tendrils or bristles; leaves mostly odd-pinnate; stamens mostly diadelphous 9:1, sometimes pseudomonadelphous with the vexillar filaments partially adherent to the adjacent fila- ments (Galegeae). i. i. Leaflets serrate or denticulate. ss 4. TRIFOLIEAE i. Leaflets entire or sometimes dentate or lobate (Phaseo- leae). j. j. Flowers in umbels, loosely capitate, or solitary in the axils; stamens with filaments commonly broadened at the apex; anthers uniform, 5. LOTEAE j. Flowers in racemes, panicles, or spikes; stamens with filaments essentially uniform in width; anthers uni- form or sometimes dimorphic. k. k. Leaves commonly with 5 or more leaflets, estipellate. — RR 6. GALEGEAE k. Leaves commonly 3-foliolate or sometimes with 1, 5, or 7 leaflets, usually stipellate. ........ 10. PHASEOLEAE Tribe 1. SoPHOREAE Sophoreae Spreng. Anleit. 2, 2: 741. 1818. Type: Sophora L. Trees, shrubs, or herbs, sometimes climbing; leaves pinnately 1- many-foliolate; stipules present or absent; stipels rarely present; inflorescences racemose or paniculate; flowers with the corolla zygo- morphic or subactinomorphic; calyx valvate or imbricate in bud; petals free or with the keel petals joined; stamens 10 or fewer, the filaments separate to the base, equal or alternately subequal in length, the anthers essentially uniform, dorsifixed, oblong to ellipsoid; ovary l-many-ovulate, the style glabrous, at least toward the apex; fruit 2-valved, dehiscent or indehiscent, sometimes samaroid; seed with hilum apical, lateral, or, rarely, circumcinct. KEY TO MEXICAN GENERA OF SOPHOREAE a. Leaflets with resinous, pellucid lines and dots; fruit elongate- samaroid, I-seeded at apex. b. b. Anthers oblong, about 4-4.5 mm. long; leaves 7-11-foliolate; leaflets mostly ovate, acute, or acuminate; ovary glabrous, ........ HMM 2. MYROXYLON b. Anthers elliptie, about 1 mm. long; leaves 11-23-foliolate; leaf- lets mostly oblong, obtuse; ovary pubescent, at least at the UM 3. MYROSPERMUM a. Leaflets lacking pellucid lines but sometimes with minute punctae; fruit 1-15-seeded, not elongate-samaroid as above. c. c. Fruit about 5 mm. thick or less, the valves thinly coriaceous, usually somewhat marginate or alate; flowers not papiliona- ceous. d. 1968] Leguminosae of Mexico — Rudd 499 d. Flowers with 5 subequal petals; ovary with filiform style, the stigma minutely capitate-penicillate; fruit about 5-9 em. long. NEU E PC Exe oC 1. SWEETIA d. Flowers with 1 petal; ovary with sessile, peltate stigma; fruit semiorbicular, about 4 cm. long or less. ........ 4. ATELEIA c. Fruit about 5-20 mm. thick, the valves coriaceous or lignous, not alate; flowers papilionaceous with 5 petals. e. e. Standard petal pubescent on the outer face; fruit orange- or fulvo-velutinous, usually brownish when dry, the valves en- rolled after dehiscence; seeds elliptic or cylindrical, red, darkening when dry, the hilum subapical; leaflets with ter- tiary veins scalariform; lateral leaflets alternate to subop- posite. eese nent nene enne nnne nnne nnne entente enne 5. DUSSIA e. Standard petal glabrous on the outer face; fruit glabrous or pubescent, the valves not distorted after dehiscence; seeds subglobose, the hilum apical; leaflets with tertiary veins mostly reticulate, the lateral leaflets essentially opposite. f. f. Flowers with curved style, the stigma bilobed, lateral; fruit compressed, somewhat constricted between the seeds but not moniliform; seeds red or bicolored red and black. ........ 6. ORMOSIA seosesssdevitetcqdenssassecnaueveesovesccopsaqesseoos0estesssesosascevasdótatepes enun "MERC vL LL ibit 1. SwEETIA Spreng. Sweetia Spreng. in L. Syst. Veg. 2: 171, 213. 1825, nom. conserv., non DC. 1825. TYPE: S. fruticosa Spreng. Brazil. Acosmium Schott in Spreng. in L. Syst. Veg. 4, Append.: 406. 1827. Type: A. lentiscifolium Schott in Spreng. Brazil. Leptolobium Vog. Linnaea 11: 388. [Feb.] 1837, non Benth. [June] 1837, 1838. Lectotype: L. dasycarpa Vog. (Mohlenbrock, Webbia 17: 233. 1963). Brazil. Thalesia Mart. ex Pfeiffer, Nom. 2(2): 1384. 1874, nom. nud. based on Leptolobium Vog.; non Raf. 1818, nom. nud. Reference; R. H. Mohlenbrock, Webbia 17: 223-263. 1963. Trees, unarmed; leaves alternate, imparipinnate, commonly 5-21- foliolate, the leaflets alternate; stipules and stipels minute or lacking; inflorescences axillary or terminal, racemose or paniculate; flowers small; calyx turbinate-campanulate with 5 subequal lobes, valvate or subimbricate in bud; corolla with 5 free, subequal, whitish petals; stamens 10, rarely 5, the filaments free, subequal, the anthers uni- form, ellipsoid, dorsifixed; stigma terminal; fruit indehiscent com- 500 Rhodora [Vol. 70 pressed 1-4-seeded; seeds brown or reddish brown, elliptic to ovate, compressed, the hilum apical, elliptic. About 20 species, chiefly in South America. l. Sweetia panamensis Benth. Jour. Linn. Soc. Bot. 8: 263. 1865. TYPE: S. Hays 267, Panamá [Canal Zone]. Figure 1. Daloergia laevigata Standl. Trop. Woods 12: 5. 1927. TYPE: S. J. Record s.n., British Honduras. Tree, to about 40 m. tall; leaves 5-15-foliolate, the axis about 8-20 em. long, puberulent or subglabrous, the lateral leaflets alternate, the blades coriaceous or subcoriaceous, ovate to ovate-oblong, about 2-9.5 em. long, 1-4 cm. broad, the apex obtuse or acute, retuse, the base rounded, the upper surface nitid, glabrous, the lower surface moderately pubescent, glabrate, the secondary venation inconspicuous, the tertiary venation reticulate, the terminal leaflet sometimes larger than the laterals; inflorescences with axes moderately pubescent, the bracts and bracteoles linear, 1-3 mm. long; flowers about 6 mm. long, fragrant; calyx moderately pubescent to subglabrous, 3-4 mm. long, the lobes and tube about equal in length; corolla white or cream- colored, the petals glabrous, subequal, spatulate, emarginate, clawed; stamens 10, with the filaments about as long as the petals or slightly longer, the anthers less than 1 mm. long; ovary sparsely pubescent to glabrous, the style glabrous or subglabrous, the stigma terminal, minutely capitate-penicillate; fruit indehiscent, coriaceous, light brown or stramineous, glabrous at maturity, reticulate, compressed, 1-3-seeded, oblong or elliptic, the ends acute to obtuse, about 5-9 cm. long, (1.5-) 2-2.5 cm. broad, about 5 mm. thick at the seeds, otherwise less than 1 mm. thick; seeds lustrous, subovate, about 8 mm. long, 6 mm. broad, and 2-3 mm. thick, the hilum about 1 mm. long. Distribution: In moist forest up to about 800 meters elevation, southern México to northern Colombia and north- western Venezuela. MEXICO: vERACRUZ: Campo Experimental de Hule, El Palmar, Zongolica, Vera-Santos 3264 (MICH). Minatitlan, Saenz s.m. (MICH). Near Cordoba, Froehling s.m. (MICH). GUERRERO: La Parota, Langlassé 383 (GH, K, MEXU, P, US). Agua de Correa, Jimalcota, Langlassé 685 (GH, K, P, US). OAXACA: Tuxtepec, Miranda 4153 (MEXU, US); Gómez-Pompa 1264 (MEXU); Sousa 251 (MEXU). Tuxtepec, El Naranjal, Miranda 4297 (MEXU, US). Tuxtepec, Chiltepec, Martínez-Calderón 62 (GH, MEXU, UC, US), 564 (MEXU, UC, US). Santa María de Chimalapa, Mell s.n. (US). "Matilde, Valle Usmacín," Reko 4095 (US). Ubero, Ll. Wil- 1968] Leguminosae of Mexico — Rudd 501 liams 9165 (DS, F, P, UC) [same collection? but locality as Fortuno, Veracruz (A)]. CHIAPAS: Cacaluta, Escuintla, Matuda 16587 (MEXU), 16663 I Tar, i N MYROXYLON BALSAMUM > VAR. PERE IRAE wow oa, MYROSPERMUM FRUTESCENS —— DUSS 1A MEXICANA Fig. 1. Geographic distribution of Sweetia, Myroxylon, Myrosper- mum, Ormosia, and Dussia in Mexico, 502 Rhodora [Vol. 70 (MEXU), 16984 (MEXU). La Arena, Palenque, Enriquez s.n. [Miranda 7604] (MEXU, US). "Estación Lacandon a Tabasco Hacienda, Palen- que, Miranda 7285 (MEXU). CAMPECHE: Champotón, Miranda 8046 (MEXU, US). Local names: Chacté, chakté, chile, palo chile, huasillo, huesillo, huesito (México) ; guayacán (México, Costa Rica) ; chichipate (Guatemala, El Salvador, Honduras) ; quina sil- vestre (Guatemala); Bill Webb (British Honduras); carboncillo (Costa Rica); malvecino (Panamá); rejo (Colombia) ; vera de agua (Venezuela). The wood of this species is strong and durable. The tim- ber is used for heavy construction such as bridges, for wagons and carts, and for various implements. 2. MYROXYLON L. f. Myroxylon L. f., Suppl. 34, 233. 1781, nom. conserv., non J. R. et G. Forster, 1776. TYPE: M. peruiferum L. f. Colombia. Toluifera L. Sp. Pl. 384. 1753. TYPE: T. balsamum L. Colombia. Myrospermum sect. Myroxylon (L. f.) DC. Prodr. 2: 95. 1825, Reference: H. Harms, Notzbl. Bot. Gart. Berlin 5: 85-98. 1908. Trees, unarmed; leaves alternate, imparipinnate, 5-15-foliolate; leaflets alternate, with pellucid lines and dots; stipules and stipels minute or lacking; inflorescences axillary or terminal, racemose; flowers small; calyx turbinate-campanulate with 5 subequal lobes, valvate in bud; corolla with 5 free petals, subequal except for a broader standard; stamens 10, the filaments free, essentially equal, the anthers uniform, oblong, acuminate, sagittate, dorsifixed; fruit indehiscent, samaroid, commonly with 1 apical seed; seeds subreni- form, resinous, the hilum elliptie, subapical. Two or three species, in Mexico, Central and South America, sometimes cultivated. l. Myroxylon balsamum (L.) Harms, Notzbl. Bot. Gart. Berlin 5: 94. Figure 1. Tree, to about 40 m. high; bark gray, resiniferous; leaves 5-11- foliolate, the axis about 6-15 cm. long, puberulent to subglabrous, the lateral leaflets alternate, the blades moderately coriaceous, with pellucid lines and dots, predominantly ovate, sometimes elliptic, 3-14 cm. long, 1-7 cm. broad, the apex obtuse to acuminate, sometimes retuse, the base rounded to subcordate, the surfaces essentially glab- rous at maturity, the upper surface nitid or subnitid, the secondary venation relatively inconspicuous; inflorescences with axes puberulent, 1968] Leguminosae of Mexico — Rudd 503 the bracts and bracteoles deltoid, the bracts 1-3 mm. long, the brac- teoles about 1 mm. long; flowers 12-15 mm. long; calyx minutely pubescent, suffarinose, 5-8 mm. long, the lobes about 1 mm. long, the tube 4-7 mm. long; corolla white, the petals glabrous, spatulate, clawed, subequal except the standard 2-3 times as broad as the others; stamens 10, the filaments essentially equal in length, slightly longer than the calyx, the anthers oblong, resinous, acuminate, sagittate, about 4-4.5 mm. long; ovary and style glabrous, the stigma terminal, truncate; fruit indehiscent, coriaceous, light brown or stramineous, glabrous, samaroid with 1 or, rarely, 2 apical seeds, essentially straight or curved, 6-11 em. long exclusive of stipe 5-15 mm. long, 2-3 em. broad and 1 cm. thick at the seed, the basal portion sterile, alate, compressed, 4-7 cm. long, 1-3 em. broad, 1 mm. thick or less; seed reniform, light brown, resinous, about 15-20 mm. long and 5-8 mm. in diameter. 1a. Myroxylon balsamum (L.) Harms, var. balsamum. Toluifera balsamum L. Sp. Pl. 384. 1753. Type: Specimen not known; Tolü, near Cartagena, Colombia. Myroxylum toluiferum A. Richard, Ann. Sc. Nat. Paris 2: 171, 172. [June] 1824, nom. nov. based on Toluifera balsamum L. Myroxylon toluifera H.B.K. Nov. Gen. 6: 375. [August] 1824, nom. nov. based on Toluifera balsamum. L. Myrospermum toluiferum (A. Rich.) DC. Prodr. 2: 95. 1825. The typical variety is characterized by relatively large fruit, 8-11 em. long, straight or somewhat curved, 2-3 em. broad with the mar- gins approximately parallel. The leaflets are glabrous, acuminate or sometimes acute at the apex. Distribution: In forest, up to about 300 meters eleva- tion, Panamá, Colombia, and Venezuela ; perhaps elsewhere. Local names: Bálsamo, balsam of Tolú, Tolu balsam (general) ; cereipo, estoraque, roble maria, olor, tache (Ven- ezuela) ; useum (Bolivia). The most important difference between the varieties seems to be in the physical and chemical properties of the balsam. The typical variety is mentioned above because it is possible that it occurs in Mexico, particularly in Yucatan. However, because many of the herbarium vouchers are sterile and information concerning the balsam is unavail- able, I am citing all Mexican collections as var. pereirae. 1b. Myroxylon balsamum var. pereirae (Royle) Harms, Notizbl. Bot. Gart. Berlin 5: 95. 1908. Figure 1. 504 Rhodora [Vol. 70 Myrospermum Pereirae Royle, Man. Mat. Med. ed. 2, 414. 1853. TvPE: Jonathan Pereira s.n., El Salvador. Myrospermum Sonsonatense Pereira [as Pareira] ex Oerst. Vid. Medd. Nat. For. Kjbenh. 27. 1855, based on "Myrospermum of Sonsonate" of Pereira, Pharm. Journ. 10: 280. 1851. Myroxylon Pereirae (Royle) Klotzsch, Bonplandia 5: 275. 1857. Toluifera Pereirae (Royle) Baill. Hist, Pl, 2: 383. 1870. Myroxylon sonsonatense [Oerst.] auct. ex Flückiger & Hanbury, Pharmacogr. 179. 1874. Toluifera balsamum L. var. B Pereirae (Royle) Baill. Traité de bot. méd. 676, 1883. The fruit of this variety tends to be smaller than that of the typical variety, 6-8 cm. long, straight or, sometimes, strongly curved, the winged lower portion usually narrower toward the stipe. The leaflets commonly are smaller than average for the species, glabrous or with a trace of pubes- cence along the midvein, the apex acute or, less commonly, breviacuminate. Distribution: In moist or dry forest at elevations up to about 1000 meters, southern Mexico to Panamá (and Colombia?). Introduced elsewhere. MEXICO: Without exact locality, Sessé & Mociño 1039 (F, MA), [Pavon Herb. No. 856] (G). VERACRUZ: “Mataluca near Cordova", Finck s.n. (K, BM). “Km. 10 carretera Cosolapa a El Palmar, a la izquierda de Pueblo Nuevo", Chavelas, Péres, & Sarukhán 24 (us); Chavelas ES-1693 (us). MORELOS: Cuernavaca, Mell 1 (NY). Oaxtepec, Reko s.n. (MEXU, MICH); Miranda 4106 (MEXU); Palacios s.m. (MICH); Rzedowski 19450 (US). Soconusco, Altimirano 276 (MEXU). MICHOACÁN: Coaleoman, Trojes, Hinton 12296 (K, NY, US), 12298 (K, MICH, NY, US). Coalcoman, Villa Victoria, Hinton 12540 (K, US), 13901 (NY, US). Hacienda Coahuayula, Emrick 29 (F). GUERRERO: Montes de Oca, Jalapa, Urbina s.n. (MEXU). OAXACA: Pochutla, Cafetal Calverio, Conzatti, Reko & Makrinius 3079 (US). Cuicatlán, El Faro, Conzatti 2820 (MEXU). Cerro Espino, Reko 3622 (MEXU). CHIAPAS: Without exact locality, Becerra 54 (MEXU). Cacaluta, Escuintla, Matuda 16529 (F, NY, MEXU, MICH), 16554 (F, MEXU, MICH, NY). Piedra Rojada, Enriquez [Miranda No.] 6740 (MEXU). CAMPECHE: Campo Experimental Forestal Tropical “El Tor- mento”, Km. 5, carretera Escárcega a Candelaria, Hernández X. & Chavelas ES-291 (us). “Km. 30 carr. Escárcega-Chetumal”, Chavelas 1968] Leguminosae of Mexico — Rudd 505 & Gómez-Pompa ES-762 (US). “Km, 30, Escárcega-Silvitue, Gómez- Pompa 1327 (MEXU, US). YUCATAN: Without exact locality, Gawmer s.m. (A, POM, US), 716 (us), 776 (BM, F). Merida, Schott 804 (BM, F, US). Local names: Balsamo, bálsamo del Pert, arbol del bálsamo, balsamito, bálsamo de San Salvador, bálsamo negro, cedro chino, chuchupate, chucté, hoitziloxitl [Nahu- atl], naba [Maya], nabal, palo del bálsamo, palo de trapiche, semillas del obispo, yagaguienite (Mexico) ; chirraca (Costa Rica) ; olor, tache (Colombia). The balsam, or resin, is used as incense and in medicinal preparations. The wood is suitable for fine cabinet work. The trees are sometimes planted as shade for coffee planta- tions. 3. MYROSPERMUM Jacq. Myrospermum Jacq. Enum. Pl. Carib. 4, 20. 1760. TYPE: M. frutes- cens Jacq. Colombia. Myrospermum Sect. Calusia Bert. ex DC. Prodr. 2: 94. 1825. Calusia Bert. ex. Steud. Nom. ed. 2, 1: 262. 1840, nomen in synon. Reference: F. Klotzsch, Bonplandia 5: 272-277. 1857. Trees or shrubs, unarmed; leaves alternate, imparipinnate; leaflets alternate, with pellucid dots and dashes; stipules and stipels minute or lacking; inflorescences axillary or terminal, racemose; flowers fairly small; calyx turbinate-campanulate with 5 subequal lobes, valvate in bud; corola with 5 free petals, subequal except for a broader standard; stamens 10, the filaments free, essentially equal, the anthers uniform, ellipsoid, dorsifixed; fruit indehiscent, samaroid, commonly with 1 apical seed; seed reniform, the hilum elliptic, sub- apical. One or two species, from southern Mexico to northern South America. 1. Myrospermum frutescens Jacq. Enum. Pl. Carib. 4, 20. 1760. Type: N. J. Jacquin s.n., Colombia. Figure 1. Bannisteria purpurea Mill. Gard. Dict. ed. 8. 1768, in part, non L. 1753. Type: W. Houstoun smn., “Campechy,” Mexico. Myroxylon frutescens (Jacq.) Willd. Sp. Pl. 2: 546. 1799. Myrospermum emarginatum [Bert. ex Klotzsch] Klotzsch, Bon- plandia 5: 277. 1857. TYPE: C. G. Bertero sn. [Balbis Herb no. 3165]. Colombia. 506 Rhodora [Vol. 70 Calusia (as Caluria) emarginata Bertero ex Klotzsch, Bonplandia 5: 277. 1857. nomen in synon. Tree or shrub, to about 5 m. high; leaves about 10-23-foliolate, the axis 8-17 em. long, puberulent or glabrous; leaflets oblong to elliptic, 1-5 em. long, 0.5-2 em, broad, the apex obtuse to emarginate, the base rounded, the surfaces usually glabrous, sometimes puberulent, the lower surface sometimes glaucous, the secondary venation incon- spicuous; bracts and bracteoles deltoid, 1-2 mm. long, 1 mm. wide, or less; flowers about 12-17 mm. long, fragrant; calyx argenteo- velutinous to subglabrous, 5-6 mm. long, the lobes about 1 mm. long or less; petals white, sometimes with pink, purplish, or yellow mark- ings, pubescent at the apex, otherwise glabrous, spatulate, clawed, the standard about twice as broad as the wing and keel petals; ovary pubescent, at least at the base, 5-7-ovulate (fide Klotzsch), the style glabrous, the stigma terminal, subcapitate, minutely penicillate; fruit 4-5 (-6.5) em. long including stipe about 5 mm. long, 12-15 mm. wide at widest part of wing, the fertile, terminal portion about 15 mm. long and 10 mm. wide; seed about 8 mm. long, 3 mm. wide, and 2 mm. thick. Distribution: Deciduous woods, southern and western México to northern South America. MEXICO: coLima: Near Rio Salado, Miranda 9053 (MEXU) ; Mc- Vaugh & Koelz 1441 (MICH). GUERRERO: Cutzamala, Coyuca, Hinton 7291 (F, GH, K, US). OAXACA: Near Tehuantepec, Alexander 111 (F, GH, NY, US), 168 (NY). CHIAPAS: Rancha La Lanza, near Tiltepec, west of Tonala, Mir- anda 7268 (MEXU, US). Campeche: “Campechy,” Houstoun s.n. (BM). Local names: Cuerillo, guachipilin hediondo (Mexico) ; arco, laurel macho (Costa Rica) ; balsamito, mano de pilón (Colombia). The plants are balsamiferous but less so than those of Myroxylon. 4. ATELEIA (DC.) Benth. 4. Ateleia (DC.) Benth. Comm. Leg. Gen. 37. 1837 (preprint); Ann. Wien. Mus. Naturg. 2: 101. 1838. TYPE: Pterocarpus ateleia DC. Mexico. Pterocarpus section Ateleia DC. Prodr. 2: 419. 1825; Mem. Leg. 393. 1826. 1968] Leguminosae of Mexico — Rudd 507 Ateleia [Sessé & Mocifio ex] DC. Prodr. 2: 419. 1825; Mem. Leg. 394. 1826, nomen in synon. References: R. H. Mohlenbrock, Webbia 17: 153-186. 1962; V. E. Rudd, Contr. U. S. Nat. Herb. 32: 385-411. 1968. Shrubs or trees; leaves alternate, imparipinnate; leaflets alternate or subopposite; stipules apparently lacking or reduced to a tuft of hairs; stipels lacking; inflorescences racemose, sometimes paniculate, axillary or terminal, about 5-20 em. long with about 10 to several hundred flowers; bracts small, deltoid or linear, sometimes persistent, sometimes caducous; bracteoles apparently lacking; flowers 5-14 mm. long; calyx regular, cyathiform, valvate or subimbricate in bud, truncate or subtruncate with 5 lobes or teeth about 0.5 mm. long or less; petal 1, white or yellowish, clawed, cucullate, sometimes expanded, spatulate at anthesis, glabrous or pubescent on the outer face, the margin erose or sinuate; stamens 6-10 (-11), sometimes varying in number in different flowers of the same inflorescence, the filaments alternately subequal in length, shorter than the petal, the anthers uniform, ellipsoid, 1 mm. long or less, dorsifixed; ovary 1-2-ovulate, the stigma essentially sessile, peltate, obliquely terminal; fruit in- dehiscent, samaroid, semiorbicular with a narrow wing along the upper suture, 2-valved, commonly 1-seeded, compressed, stipitate; seeds reddish-brown to dark brown, reniform, the hilum lateral, or- bicular or elliptic. KEY TO SPECIES OF ATELEIA a. Leaflets with lower surface moderately to densely pubescent with crispate or spreading hairs, sometimes glabrescent. b. b. Fruit, including stipe, 3.5-4 cm. long. c. c. Flowers 10-14 mm. long, the calyx 4-6.5 mm. long, the petal pubescent; fruit pubescent, usually glabrescent, the wing along the upper suture very little developed; seed 9-15 mm. long (México; Michoacán). ......1 1. A. arsenii c. Flowers less than 10 mm. long, the calyx 2-3 mm. long, the petal glabrous; fruit glabrous with the wing along the upper suture 1-2.5 mm. wide. d. d. Leaflets mostly rounded at the base, asymmetrical; peti- olules 3-4 mm. long; fruit 1.5-1.8 cm. wide, including wing 2.95 mm. wide (Michoacán). ...... S. A. truncata d. Leaflets mostly cuneate at the base, essentially symmetrical; petiolules 2 mm. long or less; fruit 1-1.5 em. wide including wing 1-2 mm. wide (Campeche). .............. 8. A. gummifera b. Fruit, including stipe, less than 3.5 em. long. e. e. Leaflets mostly rounded at the base, asymmetrical. f. f. Fruit tomentose (Chiapas). ..... 6. A. tomentosa f. Fruit essentially glabrous, the stipe sometimes puberulent. g. 508 Rhodora [Vol. 70 9. Leaflets moderately pubescent below; fruit 2.5-3 cm. long, including stipe, and about 1.5 em. wide (Nayarit; western Jalisco). itinerantes 4. A. standleyana g. Leaflets tomentulose below; fruit 2-2.3 em. long, includ- ing stipe, and 1-1.2 em. wide (Veracruz; Oaxaca; Chia- Dr MM 5. A. pterocarpa e. Leaflets mostly cuneate at base, essentially symmetrical (Campeche) ....cccccccccccsscscssccssssscssesssseccesssceecere 8. A. gummifera a. Leaflets with lower surface glabrous or moderately pubescent with appressed or subappressed hairs. h. h. Fruit 3-4 cm. long, including stipe, and 1.5-2 em. broad includ- ing wing 2-3 mm. wide; leaflets rounded, asymmetrical at base (Nayarit: Isla María Madre). ima 2. A. insularis h. Fruit 2-3.5 em. long, including stipe, and 1-1.5 cm. broad includ- ing wing 1-2 mm. wide; leaflets cuneate to obtuse, essentially symmetrical at base. i. i. Leaflets predominantly ovate, almost as wide as long, the petiolules 2-3 mm. long (Chiapas). ............ 7. A. albolutescens i. Leaflets predominantly oblong, about twice as long as wide, the petiolules 2 mm. long or less (Campeche). s (RR 8. A. gummifera 1. Ateleia arsenii Standl. Contr. U.S. Nat. Herb. 20: 174. 1919. TvPE: Bro. G. Arséne 6655, México. Figure 2. Shrub, to about 1 m. tall; leaves 13-23-foliolate, the axis tomentu- lose, glabrescent, about 10-20 cm. long; leaflets with blades ovate, ovate-oblong, elliptic, or lanceolate-ovate, 1-5.5 cm. long, 0.5-2.5 cm. broad, acute to obtuse, the base obtuse, usually oblique, the upper surface moderately crisp-pubescent, glabrescent, the lower surface moderately crisp-pubescent, the secondary veins moderately conspicu- ous, the petiolules about 1 mm. long or less, crisp-pubescent; bracts deltoid to linear, 1-3 mm. long; flowers 10-14 mm. long; calyx tomen- tulose, 4-6.5 mm. long; petal 10-14 mm. long, 6-8 mm. broad, pubescent on the outer face, usually expanded at anthesis; ovary villous; fruit crisp-pubescent, glabrescent, 3.5-4 cm. long and 1.2- 1.8 cm. broad, the upper margin essentially straight or slightly con- vex, incompletely developed, 1 mm. wide or less, the stipe 1-1.5 cm. long; seed 9-15 mm. long, 6-7 mm. wide, and about 2 mm. thick, the hilum about 0.8 mm. in diameter. Distribution: In oak woods, at elevations of about 1800 to 3000 meters, Known only from México and Michoacán. MEXICO: méxico: Nanchititla, Hinton 3421 (A, F, G, K, MEXU), 6158 (A, F, K, MICH, US), 7621 (A, F, G, K) ; Matuda 37472 (MEXU). Temascaltepec, Hinton 407 (BM), 3526 (F, K), 4240 (A, BM, F, G, K), 5286 (A, BM, F, G, K, NY, US), 6093 (A, BM, G, K, MEXU, NY, US). 1968] Leguminosae of Mexico — Rudd 509 MICHOACAN: Quinceo, vicinity of Morelia, Arséne 2790 (GH, L, NY fragment, Us). “Morelia, flanca du Quinceo, 2500? ou 3200? fruits en 1910,” Arséne 6655 (NY fragment, US TYPE). TELEIA ALBOLUTESCENS ATELEIA GUMMIFERA Fig. 2. Geographic distribution of Ateleia in Mexico. 510 Rhodora [Vol. 70 Local name: Haba de venado (Temascaltepec, México). According to Hinton (no. 3421) this species provides food for deer. 2. Ateleia insularis Standl. Contr. U.S. Nat. Herb. 20: 175. 1919. TYPE: E. W. Nelson 4186, México. Figure 2. Tree or shrub, to about 5 m. tall; leaves 9-15-foliolate, the axis puberulous with appressed or subappressed hairs, glabrescent; leaf- lets with blades ovate, 2-6.5 cm. long, 1-3.5 cm. broad, obtuse, the base obtuse, oblique, the surfaces glabrous except for a few appressed or subappressed hairs, especially along the midvein, the secondary veins moderately conspicuous, the petiolules 2-4 mm. long, puberulent, glabrescent; bracts deltoid, 0.5 mm. long or less; flowers 6-7 mm. long; calyx puberulent, 2.5-3 mm. long; petal 6-7 mm. long and about 4 mm. broad, glabrous, usually expanded at anthesis; ovary essentially glabrous; fruit glabrous, 3-4 em. long, 1.5-2 cm. broad, the upper margin convex with a wing 2-3 mm. wide, the stipe 0.8-1 cm. long; seed about 7 mm. long, 4 mm. wide, and 4 mm. thick, the hilum about 0.8 mm. in diameter. Distribution: Known only from Isla María Madre, at elevations of 700 meters or less. MEXICO: NAYARIT: Isla María Madre, Nelson 4186 (A, F, NY fragment, P, US TYPE); Ferris 5573 (A, DS, US), 5742 (DS, US) ; Howell 10471 (A); Maltby 73 (US); Mason 1843 (XK, US). Essentially the only difference separating this species from A. standleyana and A. truncata is the presence of ap- pressed rather than crispate or patent pubescence. 3. Ateleia truncata Mohlenbrock, Webbia 17: 180, fig. 4, 17. 1962. TYPE: F. Salazar s.n., México. Figure 2. Tree or shrub; 11-13-foliolate, the axis about 9-10 cm, long or more, tomentulose; leaflets with blades elliptic to ovate or obovate, 4-6 cm. long and 1.5-3 cm. broad, obtuse, retuse, the base cuneate to obtuse, oblique, the upper surface crisp-pubescent, glabrescent, the lower surface moderately pubescent, the hairs crispate or somewhat spread- ing, the secondary veins moderately conspicuous, the petiolules 3-4 mm. long, tomentulose; bracts deltoid, 0.5-1 mm. long; flowers 6-7 mm. long; calyx tomentulose, glabrescent, 2.5-3 mm. long; petal about 6-7 mm. long, 4 mm. wide, glabrous, probably expanded at anthesis (mature flowers not seen); ovary glabrous; fruit glabrous, 3.5-4 em. long, 1.5-1.8 cm. broad, the upper margin convex with a wing 2-2.5 mm. wide, the stipe about 1 cm. long; seed (submature) about 7 mm. 1968] Leguminosae of Mexico — Rudd 511 long, 4 mm. broad, and 2 mm. thick, the hilum about 1 mm. in diameter. Distribution: Known only from the type collection. MEXICO: MICHOACAN: Hacienda de la Huerta, near Apatzingán, Salazar s.n. March 6, 1914 (MEXU, US TYPE). Local name: Piojillo. The species is characterized by the relatively large, glab- rous fruit. 4. Ateleia standleyana Mohlenbrock, Webbia 1: 179, fig. 4, 16. 1962. TvPE: J. N. Rose, P. C. Standley, & P. G. Russell 14474, México. Figure 2. Tree, to about 12 m. tall; leaves 9-15-foliolate, the axis about 12- 20 cm. long, velutinous, glabrescent; leaflets with blades ovate to suborbieular, 1-8.5 em. long, 1-5.5 em. broad, obtuse to acute, the base obtuse, oblique, the upper surface minutely crisp-pubescent to subappressed-pubescent along the midvein, otherwise glabrous, the lower surface moderately crisp-pubescent, glabrescent, sometimes glaucous, the secondary veins moderately conspicuous, the petiolules 2-4 mm. long, crisp-pubescent; bracts deltoid, about 1 mm. long; flowers 6-7 mm. long; calyx 2-3 mm. long, tomentulose; petal 6-7 mm. long and about 2 mm. wide, cucullate or expanded at anthesis; ovary pubescent along the margin and at the base, otherwise glabrous; fruit mostly glabrous, 2.5-3 cm. long and about 1.8-1.5 cm. broad, the upper margin concave or convex with a wing 1-1.5 mm. wide, sometimes pubescent at the base of the stigma, the stipe puberulous to subglabrous, about 6 mm. long; seed about 6 mm. long, 4 mm. broad, and 1.5 mm. thick, the hilum about 1 mm. long and 0.9 mm. wide. Distribution: Known from Nayarit and western Jalisco, México, at elevations of 30-450 meters. MEXICO: NAYARIT: “Dry hill, vicinity of Acaponeta, Tepic," Rose, Standley, & Russell 14474 (A, F TYPE, GH, NY fragment, US). 1 km. north of El Cuatante, Valle de Banderas, Rzedowski 17870 (INCB). JALISCO: Soyatán, 30 km. south of Talpa, Rzedowski 15173 (INCB, US). Sta. Lucía, 7 km. north of Llano Grande, McVaugh (Feddema) 21275 (MICH). Local name: Jediondillo (Jalisco). 5. Ateleia pterocarpa Moc. & Sessé ex D. Dietr. Syn. Pl. 4: 1219. 1847. Based on Pterocarpus ateleia DC. Figure 2. 512 Rhodora [Vol. 70 Ateleia pterocarpa [Sessé & Moc. ex] DC. Prodr. 2: 419. 1825; Lég. Mem. 394. 1826, nomen in synon. Pterocarpus ateleia DC. Prodr. 2: 419. 1825; Lég. Mem. 394. 1826. TvPE: Plate 288, a painting from the unpublished *Flora Mexi- cana" of Sessé & Mociño, in the DeCandolle Library (G-nc; F. M. Neg. 30639 ex G-DC). [Lectotype by A. DeCandolle, “Calques des dessins de la Flore du Mexique de Mociño et Sessé qui ont servi des types d'espéces dans le systema ou le prodromus" 2: 288. 1874]. "Habitat in agris Cordovae et in Praedio S. Josephi" (Sessé & Mocino, Fl. Mex. ed. 2, 164. 1894, as Amorpha.) Tree or shrub, to about 7 m. tall; leaves 7-18-foliolate, the axis tomentulose, to about 15 cm. long; leaflets with blades ovate to elliptic, 2-8.7 cm. long, 1-5 cm. broad, obtuse to acute, sometimes retuse, the base rounded or acute, oblique, the upper surface puberu- lent along the midvein, otherwise glabrous, the lower surface mod- erately tomentose, glabrescent, the secondary veins moderately conspicuous, the petiolules 2-3 mm, long, tomentulose; bracts deltoid, 1 mm. long; flowers 4-5 mm. long; calyx tomentulose, about 1.5-2 mm. long; petal 4-5 mm. long and 1.5-2 mm. wide, glabrous; ovary glab- rous except for some pubescence on the stipe; fruit glabrous, 2-2.3 cm. long, 1-1.2 em. broad, the wing 1-1.5 mm. wide, the stipe 4-7 mm. long, sometimes pubescent toward the base; seed about 5 mm. long, 3 mm. wide, and about 2 mm. thick, Distribution: Southern México, in savanna, at eleva- tions of about 200 to 1000 meters. MEXICO: Without exact locality, presumably Veracruz, “in agris Cordovae” or “in Praedio S. Josephi", Sessé & Mociño 2017 (F, MA in part). VERACRUZ: Acatlán, Salazar s.n. (US). “Vallée de Cordova," Bourgeau 1899 (BM, K). Chihuilapan, south of Laguna de Catemaco, Los Tuxtlas, Sousa 2376 (MEXU, US). OAXACA: Chivela, Mell 15 (Ny, US fragment). Temascal, Sousa 1319 (MEXU). Near Presa Alemán, Temascal, González-Quintero 533 (ENCB, US). CHIAPAS: Siltepec, Matuda 1588 (A, K, MEXU, MICH, NY, US). Amatenango del Valle, Matuda 5833 (LL, MEXU, US), 15833 (F). Pishtimbak, north of Tuxtla Gutierrez, Miranda 6029 (MEXU, US). Rancho Lindavista, 24 km. east of Villa Flores, Miranda 5992 (MEXU, US), 6435 (MEXU, US). San Quintin, Sohns 1639 (MICH, US). Habenal, Tenejapa, Breedlove 6487 (DS, US), 7644 (DS, ENCB, US). Local name: Huapinole (Veracruz). Among Mexican Ateleia this species has the smallest 1968] Leguminosae of Mexico — Rudd 513 fruit, with the shortest stipes. The pubescent leaflets are similar to those in A. tomentosa. 6. Ateleia tomentosa Rudd, Contr. U.S. Nat. Herb. 32: 397. 1968. Type: D. E. Breedlove 11395, Mexico. Figure 2. Tree, about 12 m. tall; leaves 9-11-foliolate, the axis tomentose, about 8-12 cm. long; leaflets with blades ovate, 1-6 cm. long and 0.8- 3 em. broad, acute or subacute, the base rounded, usually oblique, the upper surface moderately pubescent, glabrescent, the lower sur- face tomentose, the secondary venation moderately conspicuous, the petiolules 2-3 mm. long, tomentulose; bracts deltoid, about 1 mm. long; flowers about 6 mm. long; calyx tomentulose, about 2.5 mm. long; petal glabrous, about 6 mm. long and 4-5 mm. wide, somewhat expanded at anthesis; ovary pallid-villous; fruit [submature] tomen- tose, 2.5-2.8 em. long and 1.2-1.3 cm. broad, the upper margin straight or concave, the wing about 1 mm. wide, the stipe about 8 mm. long; mature fruit and seed not seen. Distribution: Known only from the type collection. MEXICO: cHIAPAS: “Wooded slope 3 miles southwest of Pinola Las Rosas along road to Soyatitán, Municipio of Venustiano Carranza, elevation 4200 feet," Breedlove 11395 (DS, US TYPE). The distinctive tomentose fruits distinguish this species from the otherwise similar A. pterocarpa. 7. Ateleia albolutescens Mohlenbrock, Webbia 17: 182, fig. 4, 18. 1962. Type: C. A. Purpus 9248, México. Figure 2. Tree or shrub, to about 6 m. high; leaves 5-9-foliolate, the axis puberulent, about 6-10 cm. long; leaflets with blades predominantly ovate, sometimes rhombic to suborbicular, 3-7.5 em. long, 2-5 em. broad, obtuse to acute, sometimes retuse, the base cuneate, the upper surface puberulent along the midvein, glabrescent, sometimes nitid, the lower surface puberulent with whitish, subappressed hairs, glab- rescent, the secondary veins relatively inconspicuous, the petiolules 2.3 mm. long, puberulent; bracts deltoid, about 6 mm. long; flowers about 6 mm. long; calyx tomentulose, 1.5-2 mm. long; petal about 6 mm. long and 2-3 mm. wide; ovary glabrous or ciliate, glabrescent, the stipe usually pubescent; fruit glabrous, 2-2.5 cm. long and 1-1.5 cm. broad, the upper margin convex or almost straight with a wing 1-2 mm. wide, the stipe puberulent, glabrescent, about 5-6 mm. long; seed 4.5-5 mm. long and 3-3.5 mm. wide. Distribution: Known only from the general area of the type collection, in rocky ravines. 514 Rhodora [Vol. 70 MEXICO: CHIAPAS: Hacienda Monserrate, [southeast of Cinta- lapa] Purpus 9248 (F, GH TYPE, NY, UC, US), 10291 (US), 10544 (A, F, NY), 10549 [or 10544?] (GH, UC), 10577 [or 10544?] (A). “Tuxtla Gutierrez-Jalisco” [Arriaga], Purpus 9248 (US). This species appears to be closely related to 4. pterocarpa and A. gummifera. The various sheets of Purpus 9248 apparently do not represent a “pure” collection. Material from two different plants can be found on some sheets. The holotype at GH appears to be a mixed sheet, as is one sheet at US bearing a different locality citation. 8. Ateleia gummifera (DC.) D. Dietr. Syn. Pl. 4: 1219. 1847. Fig- ure 2. Pterocarpus gummifer Bert. ex DC. Prodr. 2: 419. 1825; Mém. 395, pl. 57, f. 1. 1826. Type: C. G. L. Bertero sn., “S. Doming.,” Dominican Republic. Dalbergia gummifera Spreng. ex DC. Mém. Leg. 395. 1826, nomen in synon. Swartzia multijuga A. Rich. Essai Flor. Cuba 457. 1846; in Sagra, Hist. Fis., Pol., y Nat. Cuba 10: 201. 1846; 12: tab. 42. 1846, non Vog. 1837. TYPE: R. de la Sagra?, Cuba. Ateleia cubensis Griseb. Mem. Am. Acad. N.S. 8: 180. 1860, nom. nov. for Swartzia multijuga A. Rich. Ateleia multijuga (A. Rich.) A. S. Hitch. Rep. Mo. Bot. Gard. 4: 80. 1893. Ateleia tumida Mohlenbrock, Webbia 17: 166, fig. 3, 10. 1962. TYPE: M. Fuertes s.n., Dominican Republic, Ateleia gummifera var. cubensis (Griseb.) Mohlenbrock, Webbia 17: 172. 1962. Ateleia parvifolia Mohlenbrock, Webbia 17: 174. 1962. TYPE: E. Ekman 7532, Cuba, Shrub or small tree, to about 7 m. tall; leaves 5-13-foliolate, the axis puberulent, glabrescent, about 5-10 em. long; leaflets with blades elliptic, rhombic, oblong, obovate, or ovate, 1-8 em. long, 0.5-3.5 em. broad, obtuse to subacute, sometimes retuse, the base cuneate to obtuse, the surfaces usually glabrous at maturity, sometimes pubes- cent, especially along the midvein, the hairs appressed or crispate, the secondary veins usually inconspicuous, the petiolules 2 mm. long or less, puberulent, glabrescent; bracts deltoid, commonly 1 mm. long; bracteoles apparently lacking; flowers 5-6 mm. long; calyx tomentu- lose, sometimes glabrescent, 2-3 mm. long; petal 5-6 mm. long, 2-4 mm. wide, cucullate, sometimes expanded at anthesis; ovary puberu- lent along the margin and at the base, otherwise glabrous; fruit glabrous, 2-3.5 em. long and 1-1.5 em. broad, the upper margin with 1968] Leguminosae of Mexico — Rudd 515 a wing 1-2 mm. wide, the stipe 0.5-1.3 cm. long, puberulent or glab- rous; seed 5-7 mm. long, 3-3.5 mm. broad, and about 2 mm. thick, the hilum 1 mm. in diameter. Distribution: México, northern Central America, and the West Indies, at elevations from about sea level to 1000 meters. MEXICO: cAMPECHE: Tuxpeíia, Lundell $49 (A, DS, GH, K, MICH, NY, UC, US), 862 (DS, GH, MICH, NY, UC, US). Local names: Balsamo hediondo, cerezo (Cuba); tush che (British Honduras). This species is relatively wide-ranging and exhibits con- siderable variation, especially in shape and pubescence of leaflets; most specimens have glabrous or sparsely ap- pressed-pubescent leaflets but those from the Yucatan peninsula, eastern Cuba, and Hispaniola tend toward crisp- pubescence. 5. Dussta Krug and Urban ex Taubert Dussia Krug and Urban ex Taubert in Engler and Prantl, Natürl. Pflanzenfam. 3, Abt. 3: 193. 1892. TYPE: D. martinicensis Krug & Urban ex Taubert. Martinique. Vezillifera Ducke. Arch. Jard. Bot. Rio de Janeiro 3: 139. 1922. Type: V. micranthera Ducke. Brazil. Cashalia Standley, Journ. Wash. Acad. Sci. 13: 440. 1923. TYPE: C. cuscatlanica Standl. El Salvador. Reference: V. E. Rudd, Contr. U.S. Nat. Herb. 32: 247-277. 1963. Trees, unarmed; leaves alternate, imparipinnate; leaflets alternate or subopposite; stipules and stipels apparently lacking; inflorescences pseudoterminal, racemose; flowers of medium size, about 15-25 mm. long; calyx campanulate, somewhat oblique with 5 subequal teeth or lobes; corolla papilionaceous with 5 separate petals, pink to purple, the standard pubescent on the outer face; stamens 10, subequal, the filaments essentially separate to the base, or with some adhesion near the base in groups of two to four, the anthers uniform, small, ellipsoid, seeds red, ellipsoid or subeylindrical, the hilum linear, subapical or lateral. About 10 species, in tropical America. 516 Rhodora [Vol. 70 KEY TO SPECIES OF DUSSIA Leaves 5-15-foliolate, the axis about 15-60 cm. long, the leaflets ovate, oblong, or obovate, obtuse or acute, or sometimes acuminate, 3-26 cm. long; bracts lanceolate, entire, the apex acute or acumi- nate, 5-9 mm. long and 0.5-2 mm. wide; bracteoles lanceolate, entire, obtuse or acute, 3-4 mm. long and 0.5-1 mm. wide (Vera- Hu TA NEEN 1, D. mexicana Leaves 11-25-foliolate, the axis about 30-100 cm. long, the leaflets mostly oblong, acuminate, 7-35 cm. long; bracts tridentate to rhombic or lanceolate, erose or entire, 5-10 mm. long and 2-4 mm. wide; bracteoles ovate, entire or dentate, acuminate, 5-7 mm. long and 2-3 mm. wide (Chiapas) teses 2. D. cuscatlanica 1. Dussia mexicana (Standl. Harms, Repert. Sp. Nov. 19: 294. 1924. Figure 1. Ormosia mexicana Standl. Contrib. U.S. Nat. Herb. 23: 436, 1922. TYPE: C. A. Purpus 6326, Mexico. Tree, to about 49 m. high; leaves 5-15-foliolate, the axis about 15-60 cm. long, fulvo-tomentulose, glabrescent, the leaflets alternate or subopposite, the blades 3-26 cm. long, 2-10 cm. broad, ovate to oblong, the terminal leaflet usually obovate, the apex predominantly obtuse, acute or sometimes acuminate, the acumen to about 12 mm. long, the base obtuse to subcordate, often oblique, the upper surface glabrous, the lower surface moderately puberulent with patent or subpatent hairs or somewhat crispate along the major veins, the venation moderately conspicuous, sometimes darkening on drying, the tertiary veins scalariform; inflorescences with axes ferrugino-tomen- tose, the bracts lanceolate, acute to acuminate, 5-9 mm. long, 0.5-2 mm. broad, the bracteoles lanceolate, obtuse to acute, 3-4 mm. long and 0.5-1 mm wide; flowers 18-22 mm. long; calyx ferrugino-tomentulose, 8-10 mm. long, the tube 5-6 mm. long, the teeth 3-4 mm. long; petals pink with white pubescence; fruit minutely fulvo-velutinous, 1- or 2-seeded, 5-6 cm. long and about 2.5 em. broad; seeds not seen but, presumably, like those of D. cuscatlanica, Distribution: Known only from the state of Veracruz, in humid forest, MEXICO: VERACRUZ: Zacuapan, Purpus 6326 (BM, GH, MO, NY, UC, US TYPE of Ormosia mexicana). El Mirador, Purpus 277 (4), 16459 (A, F, K, US); Leibmann 5355 (C, F). Zontecomapan, Andrle E Axtell 5 (us). Tapalapan, Andrle 91 (Us). Camino Misantla- Yecuatla, Gómez-Pompa 952 (MEXU, US), 1162 (MEXU). Cochinitos, Coyame, NE. Laguna de Catemaco, Sousa 2775 (MEXU, US). 1968] Leguminosae of Mexico — Rudd 517 Local names: Frijolillo, jaboncillo, palo de zopilote. It is possible that this and the next species should be synonymous. The differences are weak and, without know- ing the locality of origin of a specimen, it is often difficult to decide which name should be applied. 2. Dussia cuscatlanica (Standl.) Standl. & Steyermark, Field Mus. Pub. Bot. 22: 341. 1940. Figure 1. Cashalia cuscatlanica Standl. Journ. Wash. Acad. Sci. 13: 441. 1923. Type: P. C. Standley 20197. El Salvador. Dussia grandifrons Johnst, Journ. Arn. Arb. 19: 118. 1938. TYPE: A. F. Skutch 2027. Guatemala. Tree, to about 50 m. high; leaves 11-25-foliolate, the axis about 30-100 cm. long, fulvo- or ferrugino-pubescent with crispate or some- what patent hairs, the leaflets alternate or subopposite, the blades 7-35 em. long, 2.5-9 em. broad, commonly oblong, sometimes ovate or obovate, the apex acuminate to acute, the base obtuse, truncate, or subcordate, the upper surface glabrous, the lower surface moderately pubescent with subpatent or crispate hairs, the venation moderately conspicuous, the tertiary venation scalariform; bracts tridentate to rhombic, or lanceolate, erose or entire, acuminate, 5-10 mm. long, 2-4 mm. broad; bracteoles obovate, entire or dentate, acuminate, 5-7 mm. long and 2-3 mm. broad; flowers (15-) 18-25 mm, long; calyx ferrugino-tomentulose, 8-10 mm. long, the tube and teeth about equal in length; petals pink with greenish or purplish markings and white pubescence; fruit fulvo-velutinous, 1- or 2-seeded, 5-10 cm. long, 2.5-3.5 cm. broad and about 1.5-2 em. thick; seeds about 2-3.5 cm. long and 1.3-1.7 mm. in diameter. Distribution: In forest, southern Mexico to Costa Rica, at elevations of 20-2000 meters. MEXICO: cuHiaAPAS: Near Finca Prusia, south of Jaltenango, Miranda 6964 (MEXU, US). Esperanza, Escuintla, Matuda 18585 (MEXU). La Grada, Acacoyagua, Matuda 18640 (MEXU). Local names: Cedillo, matabuey (México); cashal (El Salvador) ; cereza de montana, garvancillo de montana, palo de tigre (Guatemala). In El Salvador the trees of this species are cut for lumber. 518 Rhodora [Vol. 70 6. ORMOSIA Jackson Ormosia Jacks, Trans Linn. Soc. Lond. 10: 360, t. 25-27. 1811, nom. conserv. TYPE: Robinia coccinea Aubl. French Guiana, Toulichiba Adans. Fam. 2: 326. 1763, nom. rejec. Monomial. Ormosiopsis Ducke, Arch. Jard. Rio de Janeiro 4: 61. 1925. TYPE: Clathrotropis (?) flava Ducke. Brazil. Macroule Pierce, Trop. Woods 71: 2. 1942. TYPE: Ormosia cou- tinhoi Ducke. Brazil. Reference: V. E. Rudd, Contrib. U.S. Nat. Herb. 32: 279-384. 1965. Trees, unarmed; leaves alternate, imparipinnate; leaflets opposite; stipules small, deltoid to linear, caducous, apparently lacking in some species; stipels usually lacking, occasionally present in some species; inflorescences terminal or pseudoterminal, racemose; flowers small to medium in size, about 6-25 mm. long; calyx campanulate, hypan- thoid, with 5 subequal teeth or lobes; corolla papilionaceous with 5 separate petals, yellow to blackish purple, the standard glabrous on the outer face; stamens normally 10, alternately subequal, the smaller sometimes sterile or lacking, the filaments separate to the base, the fertile anthers small, ellipsoid, dorsifixed; stigma bilobed, usually lateral; fruit commonly dehiscent, a few species indehiscent, glabrous to velutincus, moderately compressed, 1-6 seeded; seeds [Mexican species] red or bicolored red and black, subglobose, the hilum apical, elliptic. About 100 species, with 50 in tropical America. KEY TO SPECIES OF ORMOSIA, Leaflets glabrous on both surfaces with about 5-9 pairs of secondary veins only approximately parallel, arcuate, irregularly spaced; fruit coriaceous, finely pubescent, glabrescent; seeds red (Vera- cruz; Tabasco) .......... eere 1. O. macrocalyx Leaflets pubescent or subglabrous below with about 10-15 pairs of secondary veins, essentially parallel and regularly spaced; fruit lignous, glabrous or nearly so; seeds red or red and black. Seeds red; fruit glabrous, nitid or subnitid; flowers about 10 mm. long; leaflets with lower surface finely velutinous along the major veins, otherwise finely and sparsely appressed-pubescent (Oaxaca) RM 2. O. isthmensis Seeds red and black; fruit essentially glabrous at maturity but often with a trace of pubescence at the base; flowers 18-22 mm. long; leaflets with lower surface loosely crisp-pubescent (Chiapas) BEE 3. O. schippii 1968] Leguminosae of Mexico — Rudd 519 1. Ormosia macrocalyx Ducke, Arch. Jard. Bot. Rio de Janeiro 3: 137. 1922. Type: A. Ducke 7345. Brazil. Figure 1 Ormosia apulensis Cortés, Flora de Colombia, 61. 1919, nomen nudum. TYPE J. J. Triana 4336. Colombia. Ormosia toledoana Standl. Carnegie Inst. Publ. 461: 64. 1935. TYPE: W. A. Schipp 1052. British Honduras. Ormosia chlorocalyx Ducke, Bol. Téc. Inst. Agron. Belém, 2: 23. 1944. TvPE: A. Ducke 1516. Brazil. Tree, to about 40 m. high; stipules linear, about 5 mm. long, caducous; leaves 7-11-foliolate, the axis 10-45 em. long, sparsely pubescent, glabrate, the leaflets with blades coriaceous to subcori- aceous, ovate to ovate-oblong, 6-19 cm. long, 3-9 em. broad, the apex obtuse to breviacuminate, the base rounded to subcordate, the upper surface glabrous, nitid or subnitid, the lower surface glabrous, the secondary veins incorspicuous; bracts linear, 3-10 mm. long, 1 mm. broad or less; bracteoles subulate, 1-1.5 mm. long; flowers 18-25 mm. long; calyx 8-15 mm. long, pubescent with gray, subappressed hairs, the tube (3-) 8-10 mm. long, the teeth 3-5 mm. long; petals lilac to dark purple; fruit dehiscent, coriaceous, black or brown, glabrescent but often with some fine, fulvous pubescence at maturity, 1-6- (com- monly 2- or 3-) seeded, 3-10 em. long, 2-3.5 em. broad, slightly con- stricted between the seeds, 10-15 mm. thick; seeds red, lustrous, 10-13 mm. long, 10 mm. broad, and 7-8 mm. thick. Distribution: In wet, swampy forest, southern México to the Amazon basin of Brazil, at elevations up to about 100 meters. MEXICO: VERACRUZ: Fortuño, Río Coatzacoalcos, Ll. Williams 8926 (F, G, K, S, U, US). Fortuño, Río Coachapa, Gómez-Pompa 114 (MEXU). TABASCO: Cocoital, Comalealeo, Guzmán s.n. (US). Local names: Colorín, caracolillo (México) ; alcornoque, casique, pernille del monte (Panamá); chocho grande (Colombia) ; tento (Brazil) ; huyruro (Peru). 2. Ormosia isthmensis Standl. Publ. Field Mus. Bot. 17: 264. 1937. Type: LI. Williams 9123. México. Figure 1 Tree, to about 50 m. tall; stipules deltoid, acicular, about 1-2 mm. long, 1-1.5 mm. broad at the base, pubescent; leaves (3-5-) 7-13-folio- late, the axis 9-45 cm. long, pubescent, glabrescent; leaflets with blades coriaceous, ovate, oblong to obovate-oblong, 3-35 em. long, 2-10 520 Rhodora [Vol. 70 em. broad, acute to abruptly acuminate, the acumen to about 10 mm. long, the base obtuse or truncate, the upper surface glabrous, subnitid or nitid, the lower surface finely velutinous along the major veins, otherwise finely and sparsely appressed-pubescent, glabrescent, the secondary veins about 10-12 pairs, moderately raised; bracts and bracteoles linear, 2-8 mm. long; flowers about 10 mm. long; calyx cano- to fulvo-velutinous, 7-8 mm. long, the tube 3-4 mm. long, the teeth about 4 mm. long; petals pink-purplish with white markings; fruit dehiscent, lignous, black or dark brown, glabrous, nitid or sub- nitid, 1-8, commonly 1-seeded, 3-7 cm. long, 2-3 cm. broad; seeds red, 10-13 mm. long, 8-11 mm. wide, and 6-8 mm. thick. Distribution: In rain forest, from southern México to northern Colombia, at elevations of about 5-800 meters. MEXICO: oaxaca: Ubero, Ll. Williams 9423 (BM, F TYPE, G, K, MICH, S, US). Santiago Yaves, Choapan, Reko 9 (F). San Juan La- lana, Choapan, Schultes & Reko 822 (F, GH, NA, UC). Sierra Juarez, Gómez-Pompa, Sharp, & Hernandez 380 (MEXU, MICH, RSA, UC, US). “Arriba de Valle Nacional, rumbo a Oaxaca," Gómez-Pompa 5 (MEXU). TABASCO: Huimarguillo, Chavelas & Gonzalez 134 (MEXU) ; Barlow 30/99 (MICH). Local names: Colorín, mü-sa, palo de Salvador, frijolillo (México) ; acu-té (Guatemala) ; hormiga (British Hondu- ras); alasán (Costa Rica). 3. Ormosia schippii Pierce ex Standl. and Steyerm., emend Rudd, Trop. Woods No. 113: 125. 1960. TvrE: W. A. Schipp 1297. British Honduras. Figure 1 Ormosia schippii Pierce ex Standl. & Steyerm. Fieldiana Bot. 24 (5): 311. 1946, pro parte. Tree, to about 35-40 m. high; stipules not seen; leaves 5-9-foliolate, the axis 10-35 em. long, tomentulose, glabrescent, the blades coriaceous or subcoriaceous, ovate to ovate-oblong or sometimes obovate, 5-27 cm. long, 3-11 cm. wide, the apex acute or breviacuminate, the acumen to about 10 mm. long, the base obtuse to subcordate, the upper sur- face glabrous except for a trace of pubescence along the major veins, the lower suface moderately pubescent, the hairs loosely crispate, the secondary veins raised; bracts lanceolate, acuminate, 8-10 mm. long, 2-2.5 mm. broad; bracteoles linear, 7-8 mm. long, 1 mm. broad; flowers 18-22 mm. long; calyx cinereo- to fulvo-tomentulose, 7-10 mm. long, the tube 4-5 mm. long, the teeth 3-5 mm. long; petals reddish 1968] Leguminosae of Mexico — Rudd 521 purple; fruit dehiscent, sublignous, black or dark brown, essentially glabrous at maturity but often with a trace of pubescence at the base, commonly 1-seeded, 2-3 cm. long, 2.2.5 em. broad, about 1.5 cm. thick; seeds bicolored, red and black, 12-13 mm. long, 10-12 mm. broad, and 7-9 mm. thick. Distribution: In rain forest, generally in swampy places, southern México and British Honduras, at elevations up to about 150 meters. MEXICO: cuiAPAS: Selva Lacandona, Gómez-Pompa 339 (MEXU, us). Between La Arena and Salas, Miranda 8471/1 (MEXU). CAMPECHE: Campo Experimental El Tormento, Escárcega, Marro- quin s.m. (MEXU). Loeal names: Palo macho, carne de caballo (Chiapas) ; bayo, yabo (Campeche); John Crow bead (British Hon- duras). 7. SoPHORA L. Sophora L. Sp. Pl. 373. 1753. LECTOTYPE: S. alopecuroides L. (Brit- ton & Brown, Ill. Fl. N. U. S. ed. 2, 2: 342. 1913). “Oriente.” Broussonetia Gómez-Ortega, Hort. Matr. Dec. 61. 1798, nom. rej., non L’Hér. ex Vent. 1799, nom. cons. TYPE: B. secundiflora Gómez-Ortega. Cultivated, ex México. Patrinia Raf. Journ. Phys. Chim. Hist. Nat. 89: 97. 1819, non Juss. 1807. TvPE: Sophora sericea Nutt. South Dakota. Sophora sect. Eusophora DC. Prodr. 2: 95. 1825. Based on 10 species, including S. tomentosa L. Sophora sect. Pseudosophora DC. Prodr. 2: 96. 1825. TYPE: S. alopecuroides L. “Oriente.” Pseudosophora (DC.) Sweet, Hort. Brit. ed. 2. 122. 1830. Styphnolobium Schott, Wien. Zeitschr. 3: 844. 1830. TYPE: Sophora japonica L. Japan. Veribia Raf. Neogen. 3. 1825; Atlantic Journ. 144. 1832, nom. nov. for Patrinia Raf. Calia Berlandier in Mier-Terán, Mem. Comisión Limites 13. 1832. Type: C. erythrosperma Berlandier. Texas and Mexico. Zanthyrsis Raf. New Fl. 3: 84. 1838. TYPE: Z. paniculata Raf. South Florida and Cuba. Agastianis Raf. New Fl. 3: 85. 1838, nom. nov. for Broussonetia Gómez-Ortega. Dermatophyllum Scheele, Linnaea 21: 458. 1848. TYPE: D. specio- sum Scheele. Texas. Viberia Raf. ex Jacks. Index Kew. 2: 1193. 1895, error for Vezibia Raf. 522 Rhodora [Vol. 70 Trees, shrubs, or perennial herbs; leaves alternate, imparipinnate; leaflets alternate to opposite; stipules commonly deltoid, sometimes lacking; stipels lacking in Mexican Species; inflorescences terminal or sometimes axillary, racemose or paniculate; flowers small to medium in size, about 1-4.5 em. long; calyx campanulate with 5 sub- equal lobes, sometimes subtruncate; corolla papilionaceous, yellow, white, or blue to violet, glabrous, the keel petals commonly joined; stamens 10, free or the filaments united at the base, alternately sub- equal; anthers ellipsoid, dorsifixed, about 1 mm. long; stigma minutely capitate-penicillate; fruit commonly indehiscent, 1-6 (-15) seeded, somewhat torulose; seeds spherical or subellipsoid, red or yellowish to lizht brown, the hilum lateral or subapical, elliptie or orbicular. About 75 species, worldwide in tropical and temperate regions. The generic lectotype, S. alopecuroides, designated by Britton and Brown, has been accepted for the Index Generi- corum. Presumably, that species was chosen from six assigned to Sophora in the first edition of Species Plan- tarum because it alone was cited by Linnaeus in his earlier work, Hortus Cliffortianus, and represented his original concept of the genus. However, Hitchcock and Green (in Proposals by British Botanists 111-119. 1929; Brittonia 6: 117, 1947) and Hutch- inson (The Genera of Flowering Plants 1: 328. 1964) have considered S. tomentosa L. to be the type of Sophora, a view that I favor in the light of DeCandolle's and, later, Sweet's treatments of the genus. As shown in the above list of synonymy, DeCandolle selected S. alopecuroides as the one discordant species to be segregated into his section Pseudosophora, in contrast to the ten species, including S. tomentosa, placed in section Eusophora. A few years later, Sweet raised Pseudosophora to generic status. Thus, we have the incongruous situation that, if the generic concept of Sweet and the typification of Britton and Brown be fol- lowed, Pseudosophora would be reduced to Sophora, the prior synonym, and the true Sophora sensu DeCandolle, would have to be transferred to such later, synonymous genera as Styphnolobium, Calea, and Zanthyrsis, Fortu- nately, except for Styphnolobium, which is sometimes 1968] Leguminosae of Mexico — Rudd 523 recognized as a separate genus, the others are commonly retained as synonyms of Sophora, sensu latior. KEY TO SPECIES OF SOPHORA. Plants herbaceous from a woody root, to about 30 cm. high; flowers 15-20 mm. long; fruit about 5 mm. in diameter (Coahuila; Chi- huahua; San Luis Potosí) ............:. 1. S. nuttalliana Plants woody, shrubs or trees, about 3-12 m. high; flowers 20-30 mm. long; fruit 8-20 mm. in diameter (not known in S. conzattii). Calyx about 6 mm. long, subtruncate; corolla with keel petals separate (Oaxaca) ee 4. S. conzattii Calyx 10-15 mm. long, subtruncate to lobed; corolla with keel petals joined. Fruit (10-) 15-20 mm. in diameter; calyx 10-15 mm. long with lobes 3-5 mm. long (Coahuila; Nuevo León; Tamaulipas; Zaca- tecas; San Luis Potosí) ......... een 5. S. secundiflora Fruit 8-10 mm. in diameter; calyx 10 mm. long with teeth 1 mm. long or less. Flowers about 20 mm. long; calyx with teeth about 1 mm. long (Coahuila; Zacatecas) ....- eH 2. S. purpusii Flowers about 25 mm. long; calyx subtruncate (Clarion Isl., Baja California, Tamaulipas; Yucatan; Quintana Roo, Cozumel Isl) ..............:.. eene 3. S. tomentosa A cultivated species, Sophora japonica L., has not been included in the key. 1. Sophora nuttalliana Turner, Field and Lab. 24: 15. 1956, nom. nov. for S. sericea Nutt. Figure 3 Sophora sericea Nutt. Gen. Pl. 1: 280. 1818, non Andrews 1806. Type: T. Nuttall s.n., South Dakota. Patrinia sericea (Nutt.) Raf. Journ. Phys. Chim. Hist. Nat. 89: 97. 1819. Pseudosophora sericea (Nutt.) Sweet, Hort. Brit. ed. 2. 122. 1830. Vexibia sericea (Nutt.) Raf. Atlan. Journ. 144. 1832. Vibexia sericea (Nutt.) Raf. ex Jacks. Ind, Kew. 2: 1193. 1895, error for Vexibia sericea. Herb, perennial from woody root, commonly 10-30 cm. high; stems pubescent with white, subappressed, antrorse hairs; stipules deltoid, sericeous, 1-5 mm. long, 0.2-2 mm. broad at the base, caducous; leaves 7-25-foliolate, the axis about 3-8 cm. long, white sericeous, the lateral leaflets alternate to subopposite, the blades elliptic or ovate, 3-15 mm. long, 2-10 mm. broad, the apex obtuse, sometimes emarginate, the base obtuse to cuneate, the upper surface essentially glabrous, the lower surface sericeous, the secondary venation not evident; bracts 524 Rhodora [Vol. 70 SOPHORA NUTTALL I ANA SOPHORA PURPUS | | q Oo dee SOPHORA TOMENTOSA SOPHORA CONZATTI I A q ` ) > J ÇA d > 4 ery í PICKERINGIA MONTANA A SOPHORA ~ | —SECUNDIFLORA | Fig. 3. Geographic distribution of Sophora and Pickeringia in Mexico. at base of pedicels linear-deltoid, sericeous, 3-5 mm. long; bracteoles apparently lacking; flowers about 15-20 mm. long; calyx subsericeous, sometimes purplish, somewhat gibbous, 6-9 mm. long, the tube 4-7 mm. long, the teeth deltoid, about 2 mm. long; corolla whitish, the standard reflexei, about one fourth longer than the wings and keel, the keel petals united toward the apex; fruit coriaceous, stramineous, sericeous, glabrescent, terete, torulose, 1-6-seeded, about 2-5 cm. long, 5 mm. in diameter at the seeds, the apex usually attenuate; seeds sublustrous, yellowish to light brown with a darker line along the chalazal side, ellipsoid, about 5 mm. long and 3 mm. in diameter, the hilum lateral, subapical, orbicular, about 0.2 mm. in diameter. 1968] Leguminosae of Mexico — Rudd 525 Distribution: Prairie and rocky hillsides up to about 3000 m. elevation, South Dakota and Wyoming to Arizona and northern Mexico. MEXICO: COAHUILA: Lirios, Palmer 264 in 1880 (GH, NY, US). Rancho Los Angeles, about 40 km. SW. of Saltillo, Gould & Watson 10503 (MICH). COAHUILA ? : Las Vegas, Arséne 18963 (F), 18526 (F). CHIHUAHUA: Madera, Palmer 266 in 1908 (F, GH, NY, US). Rio Grande Valley, Pringle 5366 (MEXU, US), s.n., June 16, 1885 (GH). Paseo del Norte, Pringle 107 (GH), 5233 (MEXU, US). San Ysidro, SW. of Barranca, Mexia 2526 (BM, CAS, GH, MICH, POM, NY, UC, Us). Janos to Guadeloupe Cañon, E. K. Smith s.n. (NY). Below high ridge crests at NW. end of Sierro del Diablo, Stewart 998 (GH). SAN LUIS POTOsÍ: San Luis Potosí, Parry & Palmer 199 (BM, GH, NY, US). Local names: Frijolillo (México); white loco, silky sophora (United States). 2. Sophora purpusii Brandeg. Zoe 5: 235. 1906. Type: C. A. Purpus 1076. México. Figure 3 Shrub; stipules deltoid, sericeous, 1-2 mm. long, caducous; leaves 13-19-foliolate, the axis about 3-5 cm. long, white-velutinous, the lateral leaflets essentially opposite, the blades coriaceous, orbiculate, oblong, or obovate, about 5-8 mm. long, 3-5 mm, broad, the apex obtuse or emarginate, the base rounded, the upper surface moderately ap- pressed-pubescent, glabrescent, nitid, the lower surface sericeous, the secondary veins not evident; bracts at base of pedicels, deltoid, 1-2 mm. long; paired bracteoles subtending the calyx, minute, pubescent; flowers about 20 mm. long; calyx subsericeous, 10 mm. long, the teeth deltoid, about 1 mm. long, the tube about 9 mm. long; corolla white with purplish markings, the standard a little longer than the wings and keel; fruit coriaceous, sericeous, glabrescent, compressed, some- what torulose, 1-6-seeded, about 4-6 cm. long, 8-10 mm. broad at the seeds, the apex acuie; seeds not seen. Distribution: Known only from Coahuila and Zacatecas, México. MEXICO: coAHUILA: Sierra de Parras, Purpus 1076 (F, GH, NY, POM, UC TYPE, US). Dry valley east of Parrás, without collector's name, April 11/47 (GH). 526 Rhodora [Vol. 70 ZACATECAS: North of San Lucas, Lloyd 29 (K, UC, US). Cedros, Lloyd 141 (GH). Local name: Frijolillo. 3. Sophora tomentosa L. Sp. Pl. 373. 1753, non Hort. ex Dippel. 1898. TYPE: P. Hermann s.n. Ceylon. Figure 3 Sophora occidentalis L. Syst. ed. 10, 1015. 1759. TYPE: P. Browne sn. Jamaica. Sophora havanensis Jacq. Enum. Pl. Carib. 20. 1760. TYPE: N.J. Jacquin s.n. Cuba. Sophora littoralis Schrad. in Goett. Gel. Anz. 1: 709. 1821. TYPE: M. Wied-Newwied s.n. Brazil. Sophora arenicola Nees, Flora 4: 297. 1821. TYPE: M. Wied- Newwied s.n. Brazil. Zanthyrsis paniculata Raf. New Fl. 3: 84. 1838. Type: Z. Collins sn. Florida or Cuba. Sophora tomentosa L. B truncata Torr. & Gray, Fl. N. Am. 1: 389. 1840. TYPE: “Dr. Hulse, Dr. Leavenworth." Florida. Sophora tomentosa L. B var. littoralis (Schrad.) Benth, in Mart. Fl. Bras. 15(1): 330, t. 124. 1862, Shrub, to about 3 m. high; stipules linear-deltoid, tomentulose, about 3 mm. long, caducous; leaves 13-21-foliolate, the axis about 12-25 cm. long, puberulent, the lateral leaflets opposite or subopposite, the blades oblong, orbicular, ovate, or obovate, commonly suborbicular, about 2-6 cm. long, 1-4 cm. broad, the apex obtuse, cuneate, or sub- cordate, the upper surface sericeous to glabrous, sometimes nitid, the lower surface sericeous or puberulent, the secondary venation incon- spievous; bracts linear, 3-5 mm. long; bracteoles apparently lacking; flowers about 25 mm. long; calyx pallid-sericeous, about 10 mm. long, subtruncate; petals white to yellow, the standard a little longer than the other petals; fruit subcoriaceous, cinereo-tomentulose, terete, toru- lose or moniliform, 1-15-seeded, 5-15 cm. long and about 8 mm. in diameter at the seeds; seeds sublustrous, light-brownish, spherical or subellipsoid, 5-8 mm. long, 5-6 mm. in diameter. Distribution: Pantropical, usually near the sea, fre- quently on [restricted to ?] limestone. MEXICO: Without exact locality, Sessé & Mociño 1072 (F) ; *Mex. boundary," Schott s.m. (F). BAJA CALIFORNIA: Clarion Isl., Anthony 404 (Ds, GH, US); Barkelew 246 (GH, UC); Barclay s.n. (K); Solis 6 (US), 3602 (MEXU) ; Mason 1562 (CAS, GH, K, US); Townsend s.n. in 1889 (Us); Elmore B4 (DS, 1968] Leguminosae of Mexico — Rudd 527 F, GH, NY, UC, US); Howell 8374 (CAS, DS, GH, K, NY, POM, UC, US); Craig 583 (POM); Carlquist 393 (RSA); Cárdenas s.m. (ENCB). TAMAULIPAS: Washington Beach, Runyon 473 (NY, US). Coastal dunes near Rio Grande, LeSueur 206 (F, GH). Brasos Santiago, Schott 289a (NY). Barra de Tampico, Maury 6489 (MEXU). YUCATAN: Chuburná, Enriquez 683 (MEXU, US). QUINTANA ROO: Tancah, Steere 2512 (F, MICH). Cozumel Isl., San Miguel, Steere 2955 (F); Rudd 2035 (US). Local name: Xoxoak (Mexico); tambalisa (Cuba). This species was reported from Socorro Isl. (Barkelew 246) but the collection probably was made on Clarion Isl. according to I. M. Johnston (Proc, Calif. Acad. Sci. ser. 4, 20: 14, 64. 1931). 4. Sophora conzattii Standl. Contrib. U.S. Nat. Herb. 23: 436. 1922. TYPE: C. Conzatti, B. P. Reko, & E. Makrinius 3171. México. Figure 3 Tree; stipules not seen; leaves 7-9-foliolate, the axis 9.5-13 cm. long, puberulent, zlabrate, the lateral leaflets essentially opposite, the blades coriaceous, ovate or elliptic-ovate, 3.5-5.5 em. long, 2.2-3.5 em. broad, the apex obtuse or subacute, the base obtuse, the surfaces glabrous, the secondary venation reticulate, fairly prominent; in- florescences axillary; bracts deltoid, about 1 mm. long, caducous; bracteoles minute, caducous; flowers 20-25 mm. long; calyx ferrugino- sericeous, subtruncate, about 6 mm. long; petals violet, separate, the standard reflexed, 2 little longer than the wings and keel petals; ovary pallid-sericeous, the style pubescent toward the base, glabrous toward the apex; fruit and seed not known. Distribution: Known only from the type. MEXICO: oaxaca: “Cerro Espina," Pochutla, at 1200 m. eleva- tion, Conzatti, Reko & Makrinius 3171 (US). Local name: Frijolillo. I am not fully convinced that this species is a Sophora put until a fruiting specimen can be examined I have no other suggestion. The leaflets closely resemble those of Sweetia panamensis. The flowers with separate petals 528 Rhodora [Vol. 70 resemble those of Ormosia but lack the characteristic bilobed stigma of that genus. 5. Sophora secundiflora (Gómez-Ortega) Lag. in DC. Cat. Hort. Monsp. 148. 1813. Figure 3 Broussonetia secundiflora Gómez-Ortega, Hort. Matr. Dec. 61, pl. 7. 1798. LECTOTYPE: M. Lagasca s.n. Cultivated Botanical Garden, Madrid, ex Mexico (“è seminibus missis per D. sessé") (Designated by R. MeVaugh in herb. G-pc). Virgilia secundiflora (Gómez-Ortega) Cav. Ic. 5: 1, pl. 401. 1799. Calia erythrosperma Berlandier in Mier-Terán, Mem. Comisión Limites 13. 1832. Type: J. L. Berlandier s.m. (?). Texas and México. Agastianis secundiflora (Gómez-Ortega) Raf. New Fl. 3: 86. 1838. Dermatophyllum speciosum Scheele, Linnaea 21: 459. 1848. TYPE: R. Romer s.n. Texas. Sophora speciosa (Scheele) Benth. in A. Gray, Boston Journ. Nat. Hist. 6: 178. 1850, non Torr. 1856. Sophora sempervirens Engelm. in A. Gray, Boston Journ. Nat. Hist. 6: 178. 1850, nomen in synon. Cladrastis secundiflora (Gómez-Ortega) Raf. ex Jacks. Index Kew. 1: 552. 1893. Sophora secundiflora f. xanthosperma Rehd. Journ. Arn. Arb. 10: 134. 1929. Type: L. W. Nuttall s.n. Texas. Shrub or small tree, to about 12 m. high; stipules minute or lack- ing; leaves 7-11-foliolate, the axis about 6-15 cm. long, sericeous, glabrescent, the lateral leaflets subopposite or alternate, the blades coriaceous, obovate to oblong, about 2-9 cm. long, 0.7-3.5 cm. broad, the apex obtuse, sometimes emarginate, the base obtuse or cuneate, the surfaces sericeous, often glabrate by maturity, the secondary venation reticulate, fairly prominent; bracts and bracteoles linear, the bracts to about 6 mm. long, the bracteoles shorter; flowers 20-30 mm. long, fragrant; calyx pallid-sericeous, 10-15 mm. long, the lobes deltoid, 3-5 mm. lone, the tube hypanthoid, 7-10 mm. long; petals blue to violet-blue, the standard about one-third longer than the other petals; fruit coriaceous, sericeous, sometimes glabrescent, ‘terete, usually torulose, 1-6-seeded, about 8-13 em. long, (1-) 1.5-2 em. in diameter at the seeds, the apex usually acuminate; seeds red (or yellow in f. xanthosperma), sublustrous, subellipsoid, about 10-17 mm. long, 10-15 mm. in diameter, the hilum orbicular to oblong, 2-4 mm. long and 1.5-2 mm. wide. Distribution: Texas and New Mexico southward to Cen- tral Mexico. 1968] Leguminosae of Mexico — Rudd 529 MEXICO: CHIHUAHUA: Mouth of canyon 8 km. NW. of Canon del Rayo, Stewart 985 (GH). COAHUILA: San Lorenzo Canyon, Saltillo, Palmer 539 in 1905 (F, NY, UC, US). Chojo Grande, SE. of Saltillo, Palmer 539 in 1905 (GH), 721 in 1905 (A, F, GH, NY, UC, US). Sa. [de la] Paila, General Cepeda, Hinton 16597 (NY, US). “De la Paila," Rozynski 301 (NY). Sierra Pata Galena, Purpus 1843 (BM, F, GH, UC, US). “S. Anahuac, Sta. Catarina," Hinton 16933 (GH). Sierra Encarnación, Nelson 3899 (NY, US). Sierra Mojada Mts., Jones 169 (POM, US), s.n. (ps). San Lazaro, Wynd & Mueller 136 (A, K, NY, US); Mueller 3100 (GH, UC). “Cafion Espantosa,” Schroeder 49 (GH). Monclova, Sierra de Gloria, Marsh 2003 (F, GH). Muzquiz, Sorpresa Spring, Marsh 348 (F, GH, MEXU). La Esmeralda, Juzepezuk 671 (F). San Antonio de los Alamos, Johnston & Muller 938 (GH). Aguaje del Pajarito, N. of Puerto Colorado, Johnston 8688 (GH). 4 km. W. de El Misterio, Mendiola s.n (ENCB). Sierra de la Madera, “La Cueva,” Charretara Canyon, Johnston 9115 (GH, MEXU). Canyon del Rayo, Sierra del Diablo, Stewart 985 (GH). Cafion de Hidalgo, Sierra Mojada, Stewart 1053 (GH). W. slopes of Sierra del Carmen, Stewart 1689 (GH). Parrás, Palmer 2133 in 1880 (K); Aguirre 366 (MEXU). NUEVO LEON: “New Leon,” Mexican Boundary Survey, Thurber sn. (F). Monterey, Palmer 265 in 1880 (A, GH, K, NY, us); Pringle 3071 [3011?] (A, ENCB, GH, MEXU), 10154 (BM, F, GH, K, MEXU, NY, uc, Us); Johnson & Barkley 16230 M (F, GH); Arséne [Bro. Abbon 245] 6175 (A, US); Mueller & Mueller 504 (MEXU) ; White 1492 (GH) ; Kenoyer 1105 (F). “Mts. 15 mi. W. of Icamole," Safford 1265 (A, us). Between Cienequillas and Pablillo, Mueller & Mueller 888 (A, F), 3602 (MEXU). Cerro Potosi, Galeana, Schneider 1083 (F); Edwards 46 (DS), 76 (F); Mueller & Mueller 518 (A). Near Doctor Arroyo, Shreve & Tinkham 9387 (GH). Lampazos, Edwards 310 (RSA), 414A (F). Chipinque, Sn. Pedro Garza García, Marroquin 696 (ENCB). TAMAULIPAS: Beiween Victoria and Jaumave Valley, Nelson 4443 (F, GH, US); Rozynski 701 (F, NY). La Sardina, Bartlett 10957 (F, US). Cerro de los Armadillos, San José, Bartlett 10388 (A, DS, GH). Sierra de San Lucas, Rozynski 658 (CAS, F, K, NY, US). 9 mi. W. of Pahnilas on road io Tula, M. Johnston 5386 (MEXU). DURANGO: Peñón de Covadonga, cerca de Yerbanís, Peñón Blanco, Parey 3125 (ENCB). ZACATECAS: Conceprión del Oro, Palmer 269 in 1904 (F, GH, K, NY, US); Shreve 9369 (GH). SAN LUIS POTOSÍ: San Luis Potosi, Parry & Palmer 200 in 1878 (BM, F, GH, K, MICH, NY, us). Minas de San Rafael, Purpus 5186 (BM, F, GH, MEXU, NY, UC, US). Chareas, Lundell 5156 (A, F, MICH, uc). Amoles, Graber 222 (US). Cedral, Juzepezuk 413 (F). Mineral de Guadaleazar, Villada 17 (MEXU). "Mi. Mitras," Roybal 18 (UC). 10 km. W. of Guadaleazar, Rzedowski 4924 (ENCB). Cafion E. of 530 Rhodora [Vol. 70 Laguna Seca, Soledad, Rzedowski 4234 (ENCB). Between Charco Blanco and Aguaje de Garcías, Guadalcazar, Medellín 1561 (ENCB). 3 km. E. of Mineral de Catorce, Rzedowski 7263 (ENCB, MICH). 3 km. E. of Corazones, Villa Hidalgo, Rzedowski 8083 (ENCB, MICH). QUERETERO: Pilón, McVaugh 10362 (MEXU, MICH, US). HIDALGO: Cardonal, Ehrenberg s.n. (US); Moore 2166 (GH, UC). 4 km. NW. of Cardonal (El Mayorazgo), Gonzdlez-Quintero 2514 (ENCB). Between Actopán and Ixmiquilpan, Miranda 8996 (MEXU). Sierra de la Mesa, Rose, Painter, & Rose 9094 (US). Km. 145, carre- tera México-Laredo, Takaki s.n. (ENCB). Cardonal, 18 km. NW. of Ixmiquilpan, Gonzdlez-Quintero 2370 (ENCB). 3 km. S. of Santuario, Cardonal, González-Quintero 2095 (ENCB). Puebla: “Cerro de Matzize, near Oaxaca," Purpus s.m. (UC). Local names: Colorín, chilicote, patiol, patol, frijolillo (Mexico) ; Texas mountain laurel, coral bean, mescal bean (United States). The yellow-seeded form cited above is known from Texas and may or may not occur in Mexico. In addition to the five species of Sophora native to Mexico, the Japanese Pagoda Tree, or Chinese Scholar Tree, S. japonica L. (Styphnolobium japonicum ( L.) Schott) has been introduced in the valley of México. There are a num- ber of trees in Chapultepec Park, in Mexico City, and a specimen, Matuda (no. 21189) was collected “en ladera seca, Bosque de los pinos. Ajpsco, D. F." The trees of this species are fairly tall, to about 20 meters high, deciduous, with yellowish-white flowers about 10-15 mm. long, and glabrous, moniliform pods. Three species of Sophora occurring in border states, S. affinis Torrey & Gray in Texas, S. arizonica S. Wats. and S. formosa Kearney & Peebles in Arizona, are as yet un- known in Mexico. Tribe 2. PODALYRIEAE Podalyrieae Benth. Enum. Pl. Hugel 27. 1839; Benth. & Hook. Gen. Pl. 1: 437. 1865. Type: Podalyria Lam. 1968] Leguminosae of Mexico — Rudd 531 Shrubs or perennial herbs; leaves commonly digitately 3-foliolate, sometimes pinnate, sometimes simple, 1-foliolate, or multifoliolate; stipules present or absent; inflorescences terminal or axillary, usually racemose, or the flowers sometimes solitary; corolla papilionaceous; petals free or the keel petals joined; calyx campanulate with 5 sub- equal lobes or teeth, the vexillar lobes sometimes connate, valvate or imbricate in bud; stamens 10, the filaments separate to the base, equal or alternately subequal in length, the anthers usually uniform, ellipsoid, dorsified; ovary 1-many-ovulate, the style glabrous or with a ring or hairs below the stigma; fruit 2-valved, commonly dehiscent; seed reniform or subreniform, the hilum lateral, subapical, some- times carunculate. Only Pickeringia, Thermopsis, and Baptisia are native to America. Only Pickeringia crosses the line into Mexico, in northern Baja California. Thermopsis is found in the United States south to the border states, Texas to Califor- nia, and Baptisia occurs in Texas, but neither genus is known from Mexico. 8. Pickeringia Nuttall ex Torrey & Gray, Fl. N. Am. 1: 388. 1840, nom. cons, non Nuttall 1834. TYPE: P. montana Nutt. in Torrey & Gray, California. Xylothermia Greene, Pittonia 2: 188. 1891, nom. nov. for Picke- ringia Nutt. 1840. Prickothamnus Nutt. ex Taubert in Engler & Prantl, Natürl. Pflanzenfam. 3, Abt. 3: 203. 1892, nomen in synon. The genus is monotypic; hence, the description is that of the species, 1. Pickeringia montana Nutt. in Torrey & Gray, Fl. N. Am. 1: 389. 1840. Type: T. Nuttall s.n., “summits of mountains near Santa Barbara," California. Figure 3 Xylothermia montana (Nutt.) Greene, Pittonia 2: 188. 1891. Xylothermia montana subsp. tomentosa Abrams, Bull. Torrey Bot. Club 34: 263. 1907. TvPE: L. Abrams 3530, “near El Nido, San Diego County,” California. Pickeringia montana var. tomentosa (Abrams) I. Johnst., Con- trib Gray Herb. n.s. 68: 84. 1923. Shrub, to about 2 m. tall, erect or semiprostrate, with spinose branches; stipules not seen, apparently lacking; leaves alternate, digitately 3-foliolate or sometimes unifoliolate, essentially sessile, the 532 Rhodora [Vol. 70 leaflets coriaceous, elliptic to obovate, 5-20 mm. long, 3-10 mm. broad, entire, the apex acute or obtuse, the base cuneate, the upper surface glabrous, nitid or sparsely appressed-pubescent to sericeous or tomen- tulose, the lower surface lighter in color, glabrous to sericeous or tomentose, the secondary venation inconspicuous, the terminal, or middle, leaflet sometimes with a slightly longer petiolule and larger blade than the laterals; inflorescences subterminal [flowers solitary in the axils of the leaves at the extremity of the branches” fide Nuttall], sometimes apparently terminal, racemose, the axes mo- derately pubescent to densely argenteo-tomentose or sericeous; bracts and bracteoles linear, acicular, 2-4 mm, long; flowers 10-20 mm. long; calyx campanulate, valvate in bud, sparsely appressed-pubescent to tomentulose, 5-7 mm. long, the teeth deltoid, 1 mm. long or less; petals free, reddish-purple or, rarely, whitish; stamens alternately subequal in length, about 10-15 mm. long, the anthers ellipsoidal, about 1 mm. long; ovary pallid-tomentose, commonly 6-10-ovulate, the style glabrous, the stigma terminal, minutely capitate-penicillate; fruit dehiscent, coriaceous, light brown, puberulent, linear, compressed, 1-10-seeded, about 2-5 em. long, 4-5 mm. broad, and 2 mm. thick; seeds reniform, sublustrous, brownish-black, 3-4 mm. long, 2 mm. broad, and 1 mm. thick, the hilum minute, subapical. Distribution: In chaparral, north central California to northern Baja California. MEXICO: BAJA CALIFORNIA NORTE: Descanso Valley, Moran 8427 (SD, US). Local names: Chaparral pea, pea chaparral, stingaree bush (California). The vesture of the plants is so variable and with so much intergradation it does not seem desirable to recognize a tomentose variety unless other degrees of pubescence are given equal consideration. The vegetative parts of the one collection thus far known from México are silvery sericeous. SMITHSONIAN INSTITUTION WASHINGTON, D. C. 20560 EVIDENCE FOR THE HYBRID ORIGIN OF PETASITES WARRENII AND P. VITIFOLIUS ALFRED L. BOGLE* The purpose of this paper is to present evidence for the hybrid origin of Petasites warrenii St. John, an endemic to the State of Washington. This interpretation is based on the study of herbarium specimens, on observations in the field of reproductive biology and distribution, and on the progeny from artificial hybridizations involving P. warrenii and its putative parents, P. frigidus var. palmatus ( Ait.) Cronq. and P. sagittatus (Pursh) Gray. A portion of the evidence also suggests that P. vitifolius Greene, which has a much wider distribution across North America, is a prod- uct of the same hybridization complex which has produced P. warren. TAXONOMIC HISTORY The circumboreal genus Petasites Mill. (Compositae- Senecioneae) is represented in North America by four ill- defined native taxa, and two introduced species which have escaped from cultivation (P. japonicus Maxim., and P. hy- bridus (L.) Gaertn., Mey. & Sherb.). The systematics of the indigenous members is based almost entirely on leaf shape, and was last revised by Cronquist (1946, 1953). As a result of his revision the numerous species described from this continent up to that time (cf. Rydberg 1927) were reduced to only two, P. sagittatus (Pursh) Gray (Fig. 1B), and the polymorphie P. frigidus (L.) Fries, with three varieties: var. frigidus (Fig. 1K, L), var. nivalis (Greene) Cronq. (Fig. 1H, J), and var. palmatus (Ait.) Cronq. (Fig. 1A). The following key to the species and varieties is taken from those of Cronquist (1953, varieties of P. frigi- dus; 1955, P. sagittatus.) 1. Leaves evidently lobed, varying to coarsely few-toothed with 5-15 teeth on each side. .............................. (0 5» o» uk LE E on H0E MN P. frigidus *Present address: The Arnold Arboretum, Harvard University, Cambridge, Massachusetts 02138 533 534 Rhodora [Vol. 70 2. Leaves merely coarsely toothed or very shallowly and obscurely lobed: cireumboreal, characteristically arctic, extending south, uncommonly, to central Washington. ..............ss var. frigidus 2. Leaves evidently lobed. 3. Leaves palmately or more often pinnipalmately lobed and veined, the lobes ordinarily not extending more than halfway to the base (or often deeper in western specimens, which are distinctly pinnipalmate) ; leaves seldom more than 2 dm. wide, seldom evidently wider than long; characteristically subarctic or of boreal forests, extending south to the Gaspé Peninsula of Quebec, northern Minnesota, and at high eleva- vations in the mountains to northern Quebec. .... var. nivalis 3. Leaves palmately lobed and veined, the lobes commonly ex- tending well over half way to the base (or in western speci- mens often only about half way to the base); leaves tend- ing to be broader than long, often very large (up to 4 dm. wide); more southern than the other varieties, extending south to California, Minnesota, Michigan, and Massachu- [n RM var. palmatus 1. Leaves varying from merely a little wavy and callous-denticulate to more commonly conspicuously dentate with 20 to 45 teeth on each side. Wet places; Alaska to Labrador, south to northeastern Washington, northern Idaho, Montana to Colorado, northern Minne- sota and Wisconsin. .........cccccccceesssccesscesessceecsescsssccesssseessece P. sagittatus This system was not without precedent. Hooker (1833) first suggested that the North American taxa might best be considered as varieties of a single species. Under the generic name of Nardosmia Cass. he recognized four dis- tinct species (N. frigida Hook., N. corymbosa Hook., N. palmata Hook., and N. sagittata Hook.), but noted that they “are distinguished by very slight characters, and the numerous specimens before me, would, in the varied form of the foliage, lead to the suspicion that they in reality constitute but one species." Herder (1865) reduced Nardosmia palmata and N. corymbosa to varietal status under N. frigida, but his re- vision went unheeded by North American taxonomists, who subsequently described several minor morphological vari- ants as distinct species. Rydberg (1927, pp. 310-315) recognized eleven distinct species, and in the fifteen years following his publication at least two new species and one variety were described. 1968] Petasites — Bogle 535 It was to Herder’s system that Cronquist especially re- ferred in again revising the taxonomy of the group. The interested reader may consult his papers (1946, 1953, 1955) for comments and synonymy. This classification is now found in several major floristic works published within the past dozen years, and its nomenclature is followed here. It has been rejected by Hultén (1950, p. 1583), however, and is not followed by such authors as Anderson (1959), Moss (1960), and Polunin (1959). Cronquist (1946) tentatively placed Petasites warrenii (Fig. 1D) in synonymy under var. frigidus, and later (1955) stated that in leaf form it was “approaching var. nivalis. Perhaps eventually to be recognized as a distinct variety.” He also (1946) reduced P. vitifolius Greene (Fig. 1E, G; incl. P. trigonophyllus Greene) to synonymy under var. nivalis. The peculiarities of form and distribution of Petasites warrenii prompted this investigation of the taxon, and of the possibility of interspecific relationships between P. frigi- dus var. palmatus and P. sagittatus. MATERIALS AND METHODS In the course of this study more than 1000 herbarium specimens of the North American taxa were examined. Those specimens which are considered to have originated in the complex of hybridization described here are listed below under a separate heading. In the field the limits of distribution of Petasites warrenii were ascertained, and observations made on the nature of its inflorescences and reproduction. Live female plants of P. warrenii and P. sagittatus were transferred from the field to cultivation in the greenhouse for later use in arti- ficial pollinations. Artificial pollinations were made both in the field and in the greenhouse in the spring of 1960. In the greenhouse pollen from male plants of Petasites frigidus var. palmatus growing on the road embankment of U. S. 10, about one mile east of Issaquah, Washington, was transferred to flowers of female plants of both P. sagittatus, collected near Bona- 536 Rhodora [Vol. 70 parte Lake, Okanogan Co., Washington, and P. warrenii, collected near Liberty, Kittitas Co., Washington. Pollina- tions were made and plants kept in insect-proof cages. In the field, female flowers of Petasites warrenti, bloom- ing near Liberty, were pollinated by flowers of male inflores- censes of P. frigidus var. palmatus blooming near the sum- mit of Snoqualmie Pass, Washington (a linear distance from Liberty of approximately forty miles). The pollinated inflorescences were protected from insects by covering the scapes with a nylon stocking, the lower, open end of which was tied shut around the scape and supporting wire frame. Pollen was applied to the protruding styles of as many flowers and heads as possible. It could not be determined which styles were in a receptive condition, or how many received pollen. In both locations adjacent flowering in- florescences were left unpollinated as a simple test for natu- ral pollination. Unfortunately, male plants of Petasites sagittatus were not available for making reciprocal crosses, and male plants of P. warrenii are unknown. OBSERVATIONS AND RESULTS Reproductive biology All of the North American Petasites are polygamodioe- cious, rhizomatous, herbaceous perennials, utilizing both sexual and asexual modes of reproduction. Proliferation by means of bracteate rhizomes is predominant. In this way large clones develop from a single vegetative fragment or seedling. Inflorescences of Petasites are variously described as polygamodioecious, subdioecious, or dioecious. All heads ex- amined by the writer of the taxa concerned here have con- tained both bisexual disc florets and unisexual pistillate ray florets in varying numbers, Field observations, however, indicate that the plants are functionally dioecious, and are here termed male and female. Heads of the functionally male scapes contain numerous (up to about 60 in var. palm- atus of Washington) perfect, tubular, disc florets sur- rounded by a single peripheral row of sterile, pistillate, ray florets. The anthers of the disc florets are the sole source 1968] Petasites — Bogle 537 of viable pollen; their ovaries are sterile. In contrast, heads of functionally female inflorescences contain at their cen- ters only a few (1-5) reduced, perfect but sterile, disc florets. These are surrounded by many (up to about 200 in var. palmatus of Washington) fertile, pistillate, ray florets. Seed is formed only in the female inflorescences. Considering the essentially dioecious nature of the plants and the predominance of vegetative reproduction, it is not unusual to find large colonies or clones which are entirely of one sex or the other. It appears significant, therefore, that extensive searching through the colonies of Petasites warrenii over a period of two years has shown that all plants of the taxon are female. This observation also applies to all of the herbarium specimens examined, In addition, innumerable inflorescences were checked at maturity for seed development, but none was found; this suggests the absence of a male pollen parent and of agamospermy. When both types of inflorescence occur in proximity, pol- lination, effected by a variety of insects (including butter- flies, hoverflies, humble-bees, honey-bees, and beetles on var. palmatus in the Puget Sound region), usually takes place and the sexual cycle is completed. Once the pollen has been shed the male inflorescences quickly wilt and die, while the female inflorescences persist, undergoing considerable increase in height. At maturity the involucral bracts of the fruiting heads reflex, allowing the crowded pappus filaments of the many flowers to expand into a conspicuous, cottony mass. The seeds are carried away on the wind, or fall to the ground. In the Puget Sound region seeds of var. palmatus have been observed to germinate while still attached to the re- ceptacle, while in the laboratory fresh seed will ordinarily germinate on moist filter paper within 24 hours. The period of viability appears to be rather short, however, perhaps as little as a few weeks in duration. Somatic chromosome counts by Kruckeberg (Ornduff et al. 1943) on plants from the Pacific Northwest indicate that x = 30 for Petasites warrenii, P. sagittatus, and P. frigidus var. palmatus (P. speciosus). Sorensen and Christiansen n 538 Rhodora [Vol. 70 (1964), however, report counts of 2n — approx. 60 for P. sagittatus (2n — 58, 59, 60), and P. frigidus var. palmatus (2n = 60, 61, 62), in plants from Fort Chimo, Labrador. The latter authors suggest that the basic chromosome num- ber in the genus Petasites is 29, with “A few small chromo- somes, presumably B-chromosomes” accounting for the varying numbers in these and other species of the genus. Herbarium studies Among herbarium specimens the leaf form of Petasites warren (Fig. 1D) can be matched by some specimens among the wide variety of forms once relegated to P. viti- folius (Fig. 1E, G), but now included with P. frigidus var. nivalis (Fig. 1H, J). In some instances the varying speci- mens labeled P. vitifolius have been collected in areas where P. frigidus var. palmatus (Fig. 1A) and P. sagittatus (Fig. 1B) are sympatric and growing in close proximity, as is apparent in the collections of Porsild and Breitung along the Canol Road in Yukon Territory, in the collections of Me- Calla in the vicinity of Banff, Alberta, or in specimens col- lected by the writer and others on the shores of Lake Itasca, Clearwater Co., Minnesota. The ranges of the taxa primarily concerned in this pat- tern of hybridization, as determined from the label data of herbarium specimens and from the writer's observations in the field, are discussed below. Their general distribution in the Pacifie Northwest is illustrated in Fig. 4. Petasites frigidus var. palmatus (Ait.) Cronq. (Fig. 1A) spans the North American continent. It ranges from Newfoundland south to New York, and westward through the border provinces of Canada and the northern portions of Michigan, Wisconsin, and Minnesota, to Vancouver Island, British Columbia. In the continental interior it ex- tends northward to southeastern Yukon, and to the arctic coast of the Northwest Territories, where the segregate spe- cies P. arcticus Porsild (Porsild 1943, p. 74) has been de- scribed on the basis of collections of var. palmatus from the Mackenzie River delta area. On the west coast a robust ecotype, formerly designated P. speciosus (Nutt.) Piper, 1968] Petasites — Bogle 539 Figure 1. A. Petasites frigidus var. palmatus (P. speciosus), B. P. sagittatus, C.-G. various forms of the hybrid P. X vitifolius. C. P. sagittatus X frigidus var. palmatus, D. P. warrenii, E. P. vitifolius, McCalla 9367. Vermillion Lake, Banff, Alberta, (labeled P. palmatus), F. P. warrenii X frigidus var. palmatus, G. P. vitifolius Itasca State Park, Minnesota, H. P. frigidus var. nivalis Mt. Rainier National Park, Washington, J. P. frigidus var. nivalis Mt. Garibaldi, Garibaldi Provincial Park, British Columbia, K. P. frigidus var. frigidus, Coll: Montell Lappland (Finland), L. P. frigidus var. frigi- dus, Spetzman 873. Alaska, Arctic slope. E? 540 Rhodora [Vol. 70 inhabits the cool, moist, western slopes of the Cascade and Coast ranges to altitudes of about 2500 feet, and ecologically favorable places in the valleys of the foothills on the east side of the Cascades in Washington and Oregon (Fig. 4). The variety also has a far-eastern distribution in Sakhalin (Sugawara 1940), where the segregate species P. tate- wakianus Kita. has been designated (Kitamura 1940), and the natural hybrid P. japonicus subsp. giganteus X P. tate- wakianus Kita. described (Kitamura 1942, p. 165). Plants of var. palmatus thrive in the moist soils or wet clays of disturbed sites such as stream banks, road cuts and embankments, or logging roads in the Pacific Northwest, and in the moist soils of the forest floor in Minnesota. The range of Petasites sagittatus (Pursh) Gray (Fig. 1B) extends from Labrador westward through Canada to British Columbia, Yukon, and on into Alaska (see Map no. 322 in Porsild 1957). It occurs in northeastern Washing- ton (Fig. 4), northern Minnesota, and southward sporad- ically in the Rockies from Idaho and eastern Montana to southern Colorado. A large portion of its range overlaps that of var, palmatus, and in some places they may be found growing in close proximity. Plants of the species observed by the writer in Washing- ton, Colorado, and Minnesota have invariably been growing in shallow standing water or marshy situations, with their rhizomes buried to a depth of several inches in the soft mud, The first collection of Petasites warrenii St. John (Fig. 1D) was made by F. A. Warren in 1926 along Swauk Creek, near the town of Liberty, in the Wenatchee Mountains of Kittitas County, Washington (Fig. 2, 4). Field work in the area has revealed its distribution to be restricted to the moist or wet soils of stream banks, or soft muds of marshy alluvial flats along the courses of three streams: the Boulder, Williams, and Swauk creeks (Fig. 2). Boulder Creek drains into Williams Creek, which flows through the old goldminers’ town of Liberty and empties into Swauk Creek. The scattered colonies of P. warrenii found along Swauk Creek and lower Williams Creek have probably 1968] Petasites — Bogle 541 Gy, les LEGEND Paved Highway Improved Gravel Rood Unimproved Dirt Road Trail ~~ Stream B.M. Bench Mork A Forest Service Campground Approximate range of ! Petosites worrenli 174 v? 34 L À 1 f J Approximote scole in miles Figure 2. Distribution of P. warrenii — P. X vitifolius in area of Liberty, Kittitas Co., Washington. developed clonally from fragments of rhizome washed down- stream from the areas of most extensive development along Boulder Creek. The linear distance of distribution along the three streams is roughly six miles. Occasional plants of Petasites frigidus var. palmatus are found in ecologically favorable situations along these streams, and one pistillate clone has been observed growing within a large colony of P. warrenii. These scattered plants probably arise from seeds carried across the mountains on the wind from the main area of distribution on the west side of the Cascades. 542 Rhodora [Vol. 70 Petasites vitifolius Greene (Fig. 1E, G) is known from lower elevations in British Columbia and Alberta, and east- ward to northern Minnesota and the Gaspé Peninsula of Quebec. This distribution covers essentially that portion of the continent in which the ranges of var. palmatus and P. sagittatus coincide. Plants of the taxon utilize the habitats of either putative parent, but do best in the marshy condi- tions typical of P. sagittatus. Results of artificial hybridizations. Seed production in plants artificially pollinated in the field was low. Thirty-seven apparently viable seeds were recovered from inflorescences of Petasites warrenii. Results were considerably better in the greenhouse, where 458 ap- parently good seeds were taken from P. warrenii heads, and 111 from heads of P. sagittatus. Fruits containing viable seeds can be recognized by their larger size, turgid condi- tion, and light- to dark-brown color. Ovaries in which em- bryos fail to develop are small, withered, and lack color. In addition, seed development is accompanied by a consid- erable elongation of the pappus. Visual inspection is suf- ficient for separation. Some of the hybrid seed obtained was immediately germi- nated, and mature plants have since developed and flowered in the greenhouse. These hybrids may be tentatively desig- nated as Petasites sagittatus X frigidus var. palmatus, and P. warrenii X frigidus var. palmatus. They are dis- cussed briefly below, and the forms of their leaves and style branches are compared with those of the parents in figures 1 and 3 respectively. Petasites sagittatus X frigidus var. palmatus. (Fig. 1C). Mature leaves of this F, hybrid are shallowly lobed, with the lobes entire to sinuate-dentate, or again shallowly lobed. In Fig. 1 the general leaf form of this hybrid (Fig. 1C) is compared with those of the other North American taxa, including: Petasites warrenii (Fig. 1D) ; a leaf representa- tive of collections, variously labeled P. vitifolius or P. palm- 1968] Petasites — Bogle 543 atus, made by McCalla in the vicinity of Banff, Alberta (Fig. 1E) ; two specimens of P. frigidus var. frigidus (Fig. 1K, L), to which P. warrenii is attributed; and two leaves of P. frigidus var. nivalis (Fig. 1H, J), to which the vari- ous forms of P. vitifolius are attributed. Ga Gi NA Lee QUAD ann QUON TD MNA IG GË UE RAH Z BPD BIZ, Be = 3 y SA AM Wii SSA RAE P. sagittatus P. warrenii 0.5mm. P. frigidus var. palmatus d (P. speciosus) : Fi Hybrid P. sagittatus X frigidus var. palmatus Figure 3. Comparison of the style branches of parent and hybrid forms. The style branches of the parental taxa, the artificial F, hybrid, and Petasites warrenii are compared in Fig. 3. The styles of P. frigidus var. palmatus (P. speciosus) are clavate, with a very shallow notch about 0.5 mm deep at the tip. The surface of the entire style tip is papillose, with no elongated setae. The styles of P. sagittatus bear more or less oblong branches which are about 1.0-1.5 mm long, with long sweeping hairs clothing the margins and tips. In P. warrenii, and P. sagittatus X frigidus var. palmatus, the style branches are more or less intermediate in length, and terminate in elongate hairs very similar to those in P. sagit- tatus, but their margins are merely papillose. The style 544 Rhodora [Vol. 70 branches of both the artificial and supposed natural hybrids thus compare favorably in form, and are intermediate be- tween those of the parental taxa, Although the styles il- lustrated here were taken from the somewhat reduced, sterile, disc florets of the female heads of the four taxa, comparison with their larger, fertile counterparts in the male heads (with the exception of P. warrenii) shows their form to be representative. The form of the bracts of the inflorescence scape, the margins of the involucral phyllaries, and of the flags of the anther connectives (although not illustrated here) also compare favorably among the artificial and supposed natu- ral hybrids, and are intermediate between those of the parent taxa (cf. illustrations in Bogle 1961). Petasites warrenii X frigidus var. palmatus. (Fig. 1F). If, on the basis of the foregoing morphological evidence, a hybrid origin similar to the above can be accepted for Petasites warrenii, then the combination P. warrenii X frigidus var. palmatus is the equivalent of a backcross of an F, hybrid to the palmate parent. Leaf form among progeny of P. warrenii X frigidus var. palmatus does not exhibit the relative uniformity seen in offspring of P. sagit- tatus X frigidus var. palmatus, and the lobing is deeper than that of the F, hybrids. It was found that the various forms recovered among the backcross progeny compare favorably with many of the variations of P. vitifolius found in nature. The mature leaf form of one of these hy- brids (Fig. 1F) is compared with that of a collection of P. vitifolius from Itasca State Park, Minnesota (Fig. 1G), and with leaves of P. frigidus var. nivalis of alpine habitats in the Cascade mountains of Washington and British Colum- bia (Figs. 1H, J), the variety to which P. vitifolius has been subordinated. Pollen viability Pollen grains from available male plants were stained in lactophenol-aniline blue to obtain a rough measure of pollen viability. Two to four flowers from each inflorescence were 1968] Petasites — Bogle 545 tested, and a minimum of 200 grains were counted for each flower. The number of plants tested is indicated in paren- theses after each binomial. The ranges of stainability per- centages found are as follows: P. frigidus var. palmatus (P. speciosus), (SIE 95-99 % P. sagittatus X frigidus var. palmatus, (10) ..... 29-99 P. warrenii X frigidus var. palmatus, (3)... 29-44 P. vitifolius, (1) 45-63 Although the range of stainability is rather wide in Petasites sagittatus X frigidus var. palmatus, more than half of the counts obtained are in excess of sixty percent. Viability may vary considerably from flower to flower with- in an inflorescence, but in most cases the extremes of range are from different plants rather than from flowers of the same plant. Increased complexity of the genome, as in P. warrenii X frigidus var. palmatus, appears to be accom- panied not only by reduced pollen viability, but by increased variability in the size of pollen grains and in the stainability of their protoplasts; it thus becomes increasingly difficult to distinguish between viable and inviable grains. DISCUSSION The opinions of Hooker, Herder, and Cronquist reflect the impression given by herbarium specimens that the three morphological “peaks” of leaf form embodied in the vari- eties of Petasites frigidus (Fig. 1) are almost completely intergrading. Hooker (1833) maintained P. sagittatus as a species, while suggesting that it, too, intergrades with the other forms. Cronquist considers P. sagittatus sufficiently distinet to be maintained as a separate species. The evidence presented above, however, suggests that P. warrenii, and many of the variants formerly lumped in P. vitifolius but now assigned to P. frigidus var. nivalis, are actually mani- festations of hybridization between P. sagittatus and P. frigidus var. palmatus. Many of the various leaf forms of putative hybrid origin found in nature can be matched by similar forms found among the artificially created hybrid progeny. The mor- phological intermediacy seen in the leaves and style 546 Rhodora [Vol. 70 branches of Petasites warrenii and the artificial F, hybrids is also apparent in their inflorescence bracts, phyllaries, and anther connectives. Such intergradation of form could doubtless be demonstrated in other areas of the continent where hybrids are being formed, but the F, hybrids should be expected to reflect the morphological variation seen in the parent taxa from different geographical areas, The deeper, narrower lobing of midwestern representatives of var. palmatus, for example, may result in deeper lobing in F, hybrids than that apparent in P. warrenii and the arti- ficial F, hybrid illustrated here, both of which reflect the broader, more shallowly lobed west coast phase of var. palmatus (i.e., P. speciosus). Pollen stainability percentages indicate high fertility, or at least some fertility, in plants of both the F, and more complex hybrid generations. Such fertility in nature has probably enhanced the variation among hybrids through WI E [4 D Du ia) , Approximate limits of distribution ~ of four taxa of Petasites in the Pacific Northwest E P. worrenii pM doa A P. frigidus var. palmatus | Lp hes ` d : il P. frigidus var. nivalis WI | Sx P. sagittatus Figure 4. Approximate limits of distribution of four taxa of Petasites in the Pacific Northwest. 1968] Petasites — Bogle 547 the formation of F, and F, combinations, and backcrossing of these to the parent stocks. It is now necessary to remove the taxa of hybrid origin from synonymy under the varieties of Petasites frigidus, and to consider the position of P. sagittatus. Removal of P. warrenii from var. frigidus eliminates the variety from the State of Washington, restricting the southern limit of its range to northern British Columbia. Transfer of P. viti- folius from var. nivalis, however, shrinks the transcon- tinental range of the variety to a much smaller arctic-alpine distribution in northwestern North America, where it then consists only of the former species P. nivalis Greene (Fig. 4) and P. hyperboreus Rydb. Even with P. vitifolius re- moved var. nivalis remains a polymorphic group of unknown origin, possibly arising in hybridization between P. frigidus var. frigidus and var. palmatus, as suggested by Hultén (1950, p. 1583). With regard to the position of Petasites sagittatus it could be argued that there is now a basis for reducing the species to varietal status within the P. frigidus complex if genetic compatibility and interspecific hybridization are to be the criteria for classification. Hultén (1950, p. 1583), however, does not agree with the idea of uniting such diverse forms under a single specific epithet. He would prefer to main- tain the varieties as distinct species and recognize the hy- brids under formula designations. The latter is the course preferred by the writer in the case of P. sagittatus. The link of hybridization between P. sagittatus and P. frigidus var. palmatus is strong. On the basis of four collections from widely scattered points in Alaska and Yukon, Hultén has also proposed hybridization between P. sagittatus and P. frigidus var. frigidus (see map 1177c, Hultén 1950). But in neither case of hybridization is the characteristic form of P. sagittatus “swamped” by the hybrid interme- diates. On the basis of herbarium studies the areas of primary hybridization to the south, at least, appear to be few. The number of probable F, specimens similar to those illustrated here is also few compared with the number representing 548 Rhodora [Vol. 70 apparently more complex hybrid combinations such as Peta- sites warrenii X frigidus var. palmatus. Wide dispersal of wind borne hybrid seed may magnify the extent of actual primary hybridization between the parent species, while persistence of hybrid plants through vegetative reproduc- tion may greatly increase the period of time, as well as the number of individuals, available for further hybridization. P. warrenii is à case in point, Its exact origins are un- known. It may have developed from a chance hybrid seed blown in from some area to the north where the putative parents are hybridizing, or it may have arisen through hy- bridization in the Swauk valley if P. sagittatus occurred there in the past. Occasional female plants of var. palmatus have been found among the colonies of P. warrenii. Should a male plant of var. palmatus become established among the strictly female clones of P. warrenii a new phase of hybridi- zation could begin. In the southern portions of their ranges Petasites sagit- tatus and P. frigidus var. palmatus are distinct morpholog- ically and ecologically, and their hybrid offspring are recog- nizable. The writer feels that extensive field work and detailed study in critical northern areas, to delimit more precisely the varieties and their inter-relationships, should precede expansion of the P. frigidus complex to include P. sagittatus. CONCLUSIONS Petasites warrenii and P. vitifolius (including P. trigono- phyllus) are merely phases of a large and very variable group originating in hybridization between P. sagittatus and P. frigidus var. palmatus. Repeated interbreeding among hybrids, and backcrossing to the parent stocks, has probably oecurred, producing the wide range of variability now seen among the hybrids. The following hybrid designa- tion is proposed to include the intermediate forms: Petasites X vitifolius Greene (pro sp.) = P. sagittatus X P. frigidus var. palmatus. P. warrenii St. John, Res, Stud. State Coll. Wash. 1:109. 1929. 1968] Petasites — Bogle 549 P. vitifolia Greene, Leafl. 1:180. 1906. P. trigonophylla Greene, Leafl. 1:180. 1906. Listed below are the herbarium specimens examined dur- ing this investigation which are considered to be of hybrid origin. The specimens of all taxa examined and annotated are listed in a previous paper (Bogle 1961). Abbreviations used to designate herbaria are those proposed by Lanjouw and Stafleu (1959). Specimens of the artificial hybrids syn- thesized during this study are on deposit in the herbaria of the University of Washington, the University of Minnesota, and the Gray Herbarium of Harvard University. Abbe, Abbe, & Marr 3221 (MIN, US) ; Ahlgren 1032 (MIN) ; Ander- son, J. P. 10414 (sch: Anderson, J. R. (WSU 44648); Becket M-55-23 (MIN); Breitung 1794, 15862 (UAC); Brinkman 4254 (US); Brown 45 (isc); Butters, Huff, & Rosendahl 2755 (MIN); Campbell 150, 569 (CAN); Eastham (UBC 11922a, 11922b) ; Eastwood 476 (CAN, WS), 481 (US, WS); Fernald & Collins 259 (MIN, US); Fernald, Griscom, & Mac- kenzie 26080 (US); Garton 1455 (US), 2641, 8530 (MIN); Gillet 2315 (Us); Harper 3078 (MIN); Hitchcock & Marsh 3283 (wTU, WS); Krivda 1861 (UBC) ; Lakela 1137a (MIN, US), 8081, 8252, 11633, 12882, 18258, 20587 (MIN); Lewis 306 (US); Macoun 6/10/1880 [TYPE of P. vitifolius] (us), (wsu 46231); Marie-Victorin & Rolland-Germain 27567 (US); Marie-Victorin, Rolland-Germain, & Louis-Marie 21232 (US) ; McCalla 2059a, 3704, 7061, 11310 (UAC), 9192, 9235, 9275, 9284, 9367, 9560, 9580 (UBC); Moir 812, 947, 1107, 1435, 2010, 2040 (MIN); Moore & Moore 11028 (MIN); Moss 2006, 2355, 4219, 4264, 8846 (UAC) ; Moyle 948, 1091 (MIN); Nelson 2426 (MIN); Ownbey 3216 (MIN); Porsild & Breitung 9342, 9424 (CAN), 9759 (WTU), 10333 (CAN, ISC), 10334 (sch, 11733 (US); Richards 4763 (MIN); Richardson 14678 (CAN); Rosendahl 4752 (MIN); Rosendahl & Moyle 2218 (MIN); Sandberg 155a [TYPE of P. trigonophyllus] (US, WS), (US 45059), (MIN 212577); Scoggan 5141, 5667 (UAC); St. John, English, Jones, & Mullen 8106 (ws); Taylor & Lewis 392 (UBC); Thompson 5961, 8204 (WTU), 11210 (MONTU, WS, WTU), 14503 (ID, MIN, UAC, UBC, WS) ; Tisdale (UBC 8736); Turner 1725, 4367, 4514 (UBC), 4364, 4367, 4406, 4413 (UAC); Warren 380, 542 [TYPE of P. warrenii], 809 (Ws). ACKNOWLEDGEMENTS The writer wishes to express sincere thanks to Dr. A. R. Kruckeberg, at whose suggestion and under whose guidance this work was begun, I am also indebted to the Society of Sigma Xi for a Grant-in-Aid of Research which has made 550 Rhodora [Vol. 70 possible herbarium studies and field collections both in Min- nesota and abroad. The Department of Botany of the Uni- versity of Minnesota has allowed the use of its greenhouse and herbarium facilities for the continuation of this study. Dr. Kruckeberg, Dr. E. C. Abbe, Dr. G. B. Ownbey, and Dr. T. Morley very kindly criticized earlier drafts of this paper. Finally, I would like to express my gratitude to the curators of the following herbaria for the use of specimens essential to this study: University of Alberta, University of British Columbia, National Museum of Canada, University of Idaho, Iowa State University, University of Minnesota, Montana State University, United States National Museum, University of Washington, and Washington State Univer- sity. DEPARTMENT OF BOTANY UNIVERSITY OF WASHINGTON, SEATTLE LITERATURE CITED ANDERSON, J. P. 1959. Flora of Alaska and Adjacent Parts of Canada. Iowa State Univ. Press. Ames. BoGLE, A. L. 1961. A study of the genus Petasites in the Pacific Northwest, with special reference to Petasites warrenii St. John. M. S. thesis, Univ. Washington. 56 p. CRONQUIST, A. 1946. Notes on the Compositae of the northwestern United States. III. Inuleae and Senecioneae, Rhodora 48: 116- 125. . 1953. Specimens in European herbaria. Leafl. West. Bot. 7(2): 28-31. 1955. Compositae. In C. L. Hitchcock, A. Cronquist, M. Ownbey, and J. W. Thompson. Vascular plants of the Pacific Northwest. Univ. Wash. Publ. Biol. 7(5): 1-343. GREENE, E. L. 1906. Various new species. Leafl. bot. observ. crit. 1:180. HERDER, F. vox. 1865. Plantae radeanae monopetalae, II. Soc. Imp. Natur. Mosc. 38(1). HooKer, W. J. 1833. Flora Boreali-Americana. I. London. HuLTÉN, E. 1950. Flora of Alaska and Yukon. 10. Lunds Univ. Arsskr. N. F. Avd. 2. 46. KITAMURA, S. 1940. Petasites palmatus (and P. tatewakianus Kita. sp. nov.). Acta phytotax. et geobot. 9: 64-66. 1942. Compositae Japonica. III. Mem. Coll. Sci. Kyoto Imp. Univ. Ser. B. 16: 155-292. 1968] Petasites — Bogle 551 LaNJouw, J., and F. A. STAFLEU. 1959. Index Herbariorum. I. The herbaria of the world. Kemink en Zoon N. V., Utrecht. Moss, E. H. 1960. Flora of Alberta. Univ. of Toronto Press, Toronto. ORNDUFF, R., P. H. RAVEN, D. W. KvHos, and A. R. KRUCKEBERG. 1963. Chromosome numbers in Compositae. III. Senecioneae. Amer. J. Bot. 50: 131-139. POLUNIN, N. 1959. Circumpolar Arctic Flora. Oxford Univ. Press, London. PorsiLD, A. E. 1943. Materials for a flora of the continental north- west territories of Canada. Sargentia 4: 1-79. 1957. Illustrated Flora of the Canadian Arctic Archipelago. Natl. Mus. Can. Bull. No. 146. RYDBERG, P. A. 1927. Carduaceae, Senecioneae (pars). North American Flora 34(4). N. Y. Bot. Gard. SPRENSEN, T., and H. CHRISTIANSEN. 1964. Contribution to the Chromosome Cytology of Petasites. Bot. Tidsskr. 59: 311-314. ST. JOHN, H. 1929. New and noteworthy northwestern plants. Part 3. Res. Stud. St. Coll. Wash. 1: 109-112. SUGAWARA, S. 1940. Illustrated Flora of Saghalien. Vol. 4. Tokyo. BIOMETRIC AND TAXONOMIC USES OF CELLULOSE ACETATE PLASTIC EDWARD E. TERRELL Plastics have been used in botany as embedding media and as films, sheets, or bags to hold or cover specimens. The present paper discusses one particular kind of plastic, cellulose acetate, and its versatility and usefulness in taxonomy. The late Prof. N. C. Fassett described (1951, pp. 141-2) his method of mounting small seeds of species of Callitriche on transparent 1" by 114” cellulose acetate slides, These mounts were easy to handle, permanent, and storable in the envelopes attached to herbarium sheets. To attach seeds to slides Fassett used acetone in which pieces of the cellulose acetate plastic were dissolved. This quick-drying solution. also was used to reattach plants to herbarium sheets, Slides were labelled with india ink, Fassett (1949) used slender slivers of the plastic as aids in picking up seeds and other tiny objects by static electricity. The most valuable characteristic of the plastic is its transparency. Inscriptions made on its surface or plant parts mounted on it may be superimposed on plant material, permitting a side-by-side comparison through a dissecting microscope or hand lens. Two other characteristics of the plastic are also important: it may be cut in any size or shape with a paper cutter or shears; its surface may be easily inscribed with lines, numerals, or other markings. Conventional glass slides do not have the latter advantages. Plastic slides prepared in 1961 appear permanent. Clear nail lacquer (“fingernail polish") was used as a glue — it is inexpensive, quick-setting, and has its own applicator brush. Slides (Fig. 1) have been made in two sizes: 3" by 1" for storage in standard microscope slide boxes; and 614” by 1" for storage in double-rowed wooden slide boxes by removing the wooden center partition. The longer slides are more useful for holding long, narrow plant parts, e.g., grass spikes, leaves, or portions of stems. Both sizes 552 1968] Cellulose acetate plastic — Terrell 553 Plate 1399. = - . s Ime Fig. 1. Plastic slides (3" X 1" and 64” X 1") bearing mounted plant material. 554 Rhodora [Vol. 70 of slides are particularly well-suited for taxonomic studies of grasses and other plants with small reproductive parts. Transparent cellulose acetate with polished surfaces may be purchased in 20” by 50” sheets of various thicknesses. Among the most useful thicknesses are .020” (0.5 mm.) to 040” (1 mm.). Commercial “see-thru” rulers are about .020" thick. At present, the retail costs per sheet are about $4.50 for .040" and about $2.70 for .020". Over 300 3" by 1" slides may be cut from one of these sheets. There are at least two suppliers of the plastic in the Washington, D. C. area; it is supposed that any large city would have one or more suppliers. Measurement of Variation Cellulose acetate is useful in making the measurements needed for the study of hybridization by Andersonian methods and in other kinds of biometrical and numerical taxonomic studies. Three general kinds of uses are possible: (1) slides or pieces of plastic marked with lines, angles, etc. as measuring devices, (2) plant material mounted on plastic for visual comparison, (3) plastic cut into the shape of leaves or other parts after tracing an outline of the part. The first two procedures speed up the mechanics of making measurements and improve objectivity. The third procedure has not been tested, but might be useful if "model" leaves representing typical or other states of vari- ation were needed. In an investigation of hybridization between two grass species (Terrell, Hovin, and Hill 1965), a plastie device (Fig. 2) was used to measure glume length. The total range of variation in glume length was first established and then divided into class-groups as follows: 1.5-1.9 mm ; 2.0-2.4 mm; etc. up to 5.5-5.9 mm. The upper limits of class- groups were inscribed on a slide: 1.9 mm; 2.4 mm; 2.9 mm; etc. up to 5.9 mm (such inscriptions rubbed with a red pencil are more easily visible), By overlaying the slide on the samples to be measured the lengths of the glumes were determined as far as the class-groups were concerned. In measuring the samples only “yes-or-no” decisions were 1968] Cellulose acetate plastic — Terrell 555 Plate 1400. 0.9 1.4 1.9 2.4 2.9 3.4 3.9 4.4 4.9 5.4 5.9 _ GLUME LENGTH Fig. 2. Plastic device used in measuring length of glumes (see text). required: if glume lengths in plant 3657-1 were longer than 1.9 mm but shorter than 2.4 mm they fell into the “24 group". (The number, 2.4, was circled on a mimeo- graphed sheet previously set up for recording data.) For statistical purposes all glume lengths in the “2.4 group" were assumed to be at the mid-point of this class-group, namely 2.2 mm. This mid-point convention is believed sta- tistieally valid, especially in view of the small sizes of the class-groups. By using this convention, graphical and sta- tistical treatments were simplified. However, the important point is that the slide as prepared above permitted faster measuring. Parenthetically, it may be pointed out that the slide actually used could have been inscribed like a ruler but only with class-group lines. One of the ways in which plastic slides are most useful is in measuring such difficult-to-measure characteristics as density of hairs. Two conventional ways to measure density would be to: (1) use a plastic ruler and count the number of hairs per linear unit; or (2) cut a hole in a piece of 556 Rhodora [Vol. 70 paper, overlay on the samples, and count the number of hairs per unit area. On the other hand, one may mount the actual plant material on plastic slides. As a preliminary, the total range of density present in the plant material studied is divided into equal-sized class-groups. The plants are surveyed to find plant parts with hair densities repre- senting the upper limits of class-groups. Pieces of stem or other plant parts representing the desired states of density are cut and mounted on a plastic slide. Each piece is given à code number for recording of data. Measuring of the samples involves only “yes-or-no” judgments, made instan- taneously by visual comparison. Preparation of the com- pleted slide takes longer; but, once prepared, the slide permits faster measurements by eliminating the counting part of the operation, only matching is involved. There are other difficult-to-define characteristics, It is not easy to put into words or numerical terms degrees of variation in texture of hairs, appression of hairs, or indu- ration of glumes, for example. If the actual plant parts are mounted on plastie, verbal descriptions such as hirsute are defined in terms of the hairs themselves, Once the initial decision is made as to what will be meant by hirsute, the plant part mounted on the plastic becomes a model or standard for all subsequent comparisons. The standard never changes and does not have to be remembered, only preserved. Samples to be measured are judged by visual matching with the objective standards, not with vague concepts held in the mind. Herbarium and Reference Uses Slides. Slides bearing mounted plant parts are useful also to the monographer studying a plant group or to the person who wishes to have available a representative critical part of a species as an identification-aid. Two of the slides in Fig. 1 were prepared for the latter purpose; they bear typical inflorescences of Hyparrhenia rufa and Dichanthium annulatum, introduced grass species. They were labelled with six letters, the first three letters of the genus and species names (e.g., Dicann— Dichanthium annulatum). A 1968] Cellulose acetate plastic — Terrell 557 Plate 1401. Fig. 3. (above). Plastic herbarium sheet bearing Lolium perenne plant and spikes. Fig. 4. (below). Sheet bearing spikes of six Lolium species. 558 Rhodora [Vol. 70 number followed, in this case U.S.D.A. PI numbers for specimens from which the inflorescences were obtained. Written records of the source data were kept for each slide. Herbarium sheets. A plastic sheet 20” x 50” may be cut into four standard-sized (1114” x 1614”) herbarium sheets (with some plastic left over for slides). When used as herbarium sheets the plastic is more expensive; however, the suggestions below assume only limited numbers of plastic herbarium sheets will be used (for example, one or two sheets per species). In mounting plants on plastic sheets, Archer’s plastic was found to be satisfactory. Botanists sometimes prepare one or two typical “certi- fied" specimens of each species in order to have available identification-aids. These specimens may be kept in a special reference herbarium or in separate folders with conventional herbarium specimens. In either event, if the specimens were mounted on plastic instead of paper sheets, it would be possible to make more effective comparisons with un- known specimens. Even persons with good “form memo- ries” have difficulty comparing details of closely related species without seeing them side-by-side. Fig. 3 shows a more or less typical plant plus selected inflorescences to illustrate much of the variation existing in the species, Lolium perenne, Similar plastic sheets of other species of the genus would obviously be very helpful aids to identifi- cation in a reference herbarium. In Fig. 4 inflorescences of six of the eight species of Lolium occupy one sheet. The information-equivalent of six slides, it compresses much information in a small space. CROPS RESEARCH DIVISION, AGRICULTURAL RESEARCH SERVICE, U. 8. DEPARTMENT OF AGRICULTURE, BELTSVILLE, MARYLAND 20705 LITERATURE CITED Fassett, N. C. 1949. Herbarium technique. Rhodora 51: 59-60. 1951. Callitriche in the New World. Rhodora 53: 137-155. TERRELL, E. E., Hovin, A. W., and Hitt, H. D. 1965. Natural hybrids between Sphenopholis obtusata and Trisetum pensyl- vanicum (Gramineae). Bull. Torr. Bot. Club 92: 169-182. A REVERSION IN POLYMNIA CANADENSIS (COMPOSITAE) AND NOTES ON THE PEDICELLATE DISK FLORETS IN THE GENUS' EDWIN B. SMITH While examining a folder of Polymnia in our herbarium, I came upon an unusual form of P. canadensis L. The speci- men is from Polk County, Arkansas (J. T. Buchholz 245) and a notation on the label reads “infected with galls.” I examined the specimen in detail and found that, while there was evidence in several of the heads of worm damage, the odd appearance of the specimen was due to what has been called a reversion (return to an ancestral condition) of the type exemplified in Hypochoeris radicata (Collins, 1921) and, to a lesser extent, in Taraxacum officinale (Smith, 1965). Figure 1 is a diagrammatic comparison of a normal and a reverted disk floret of Polymnia canadensis. The reverted florets are characterized by a long pedicel tipped by a some- what shortened but normally colored corolla, normal (apparently fertile) anthers, and a sterile secondary pedicel (in place of the normally sterile pistil) which bears a pair of bracts and a terminal fascicle of bracts. In the axils of the bracts in the fascicle, secondary reverted florets were developing, producing compound umbels in the most devel- oped ones. Heads on the specimen did not set fruit. The ray florets (normally fertile) in the reverted heads devel- oped in a manner similar to that of the disk florets, but the corolla in these was greenish and somewhat spathe-like (as in T. officinale; Smith, 1965). The phyllaries and receptacu- lar chaff in the reverted heads were normal. In some of the less developed disk florets, the secondary pedicel ended at the pair of bracts. I noticed, when comparing the normal and the reverted florets, that even the normal disk florets of Polymnia cana- densis have short (ca. 1-3 mm) pedicels, a character to my “The author was the recipient of the NSF Grant GB 6854 during the time of this study. 559 560 Rhodora [Vol. 70 A B Fig. 1. Diagrammatic longitudinal section of a normal and a reverted disk floret of Polymnia canadensis L. Reference line about 5 mm. Anthers are shown in solid black. A. Normal floret: 1 — short pedicel; 2 — sterile pistil with undivided stigma; 3 — whitish corolla. B. Revericd floret: 1 — greatly elongated pedicel; 2 — whitish corolla; 3 — pair cf bracts; 4 = fascicle of bracts. 1968] Polymnia canadensis — Smith 561 knowledge quite rare in the Compositae. Examination of several specimens of P. uvedalia L. disclosed the fact that its disk florets are also normally pedicellate. I am writing this note for two reasons: firstly, to add Polymnia canadensis to the list of species known to exhibit reversion. The close correlation between the morphology of the reverted heads of it and H. radicata (Collins, 1921) is striking. Secondly, to point out that the genus Polymnia (or, at least P. canadensis and P. uvedalia) is characterized in the normal condition by pedicellate disk florets. This second reason has further significance in two respects. It would seem worthwhile to include in the description of the genus Polymnia in our state and regional floras, the fact that the disk florets are pedicellate. This character alone, I believe, would distinguish the genus from all other Compositae in our area. None of the manuals of use in this area mentions the character (Steyermark, 1963; Fernald, 1950; Gleason, 1963; Small, 1933), but it would seem to merit inclusion as one of particular significance. The Compositae are believed by many taxonomists to have originated phylogenetically from umbellate ancestors (Collins, 1921; Bessey System, through the Rubiales, as in Core, 1955) by reduction and finally elimination of the pedicels of the umbels. The pedicellate condition of Polym- nia disk florets supports the theory of the umbellate origin of the Compositae and substantiates the aptness of the term “reversion” for the type of morphological change exhibited in Figure 1. It also marks Polymnia with one of the most primitive characteristics in the Compositae. DEPARTMENT OF BOTANY & BACTERIOLOGY, UNIVERSITY OF ARKANSAS, FAYETTEVILLE LITERATURE CITED COLLINS, J. L. 1921. Reversion in Composites. Jour. Heredity 12: 129-133. Core, E. L. 1955. Plant Taxonomy. Prentice-Hall, Ine., Englewood Cliffs, N.J. FERNALD, M. L. 1950. Gray’s Manual of Botany, 8th ed. American Book Co., N.Y. 562 Rhodora [Vol. 70 GLEASON, H. A. 1963. The New Britton and Brown Illustrated Flora ox the Northeastern United States and Adjacent Canada. 3 vols. Hafner Pub. Co., Inc., N.Y. SMALL, J. K. 1932. Manual of the Southeastern Flora. Science Press Printing Co., Lancaster, Pa. SMITH, E. B. 1965. A reversion in Taraxacum officinale. Trans. Kansas Acad. Sci. 68: 266-268. STEYERMARK, J. A. 1963. Flora of Missouri. Iowa State U. Press, Ames. A “PILLAR”-TYPE BLACK SPRUCE. While riding on Route 101-B in Candia, New Hampshire in Dec. 1967 I happened to see what looked like a black pillar or smoke stack in a swamp. Upon investigation the smoke pipe turned out to be a black spruce (Picea mariana (Mill) BSP.) about 6 in. in diameter and 35 ft. high with a tightly appressed green crown reaching to the ground. The branches were tightly thatched about the stem. There was no evidence of brocming caused by dwarf mistletoe. The crown was hardly wider than 18 in. at any point. About 10 feet southwest from this tree, (which stood alone, with no other trees of similar height closer than 50 feet) grew a black spruce with almost completely different branching habit. It was barely 5 feet high with a pyramidal crown spreading to a diameter of about 10 feet on the ground. It, too, was densely foliaged. No cones were evident on either tree. Black spruces with narrow or irregular crowns are common, but I have never seen one with such a narrow cylindrical crown. Mutants with crowns of this form have frequently been reported in Picea abies L., in fact I have found some myself in northern Sweden. This phenomenon has been discussed by Sylvén (1909), Rubner (1936) and H. Schmidt (1952) among others. Thus Wahlgren (1922) shows in Fig. 184 a “pillar” spruce from Grónsinka National Forest, Sweder and describes it (p. 477) “with short, 1968] Black Spruce — Baldwin 563 Plate 1402. Pillar-form of Black Spruce in Candia, New Hampshire 564 Rhodora [Vol. 70 densely interwoven branches along the stem, which give the tree a pillar-form appearance." Bouvarel (1954) pic- tures a less extreme form which he terms “mutant que- nouille" (distaff or bed-post). Twigs from these black spruces have been sent to Uni- versity of New Hampshire, Yale and the Arnold Arbore- tum for propagating. HENRY I. BALDWIN HILLSBORO, NEW HAMPSHIRE 03244 LITERATURE CITED BovvAREL, P. 1954. Variabilité de L'Épicea (Picea excelsa Link.) dans le Jura Francais. Revue Forestiére Francaise (2): 85-98. (Reprint) RUBNER, K. 1936. Beitrag zur Kenntnis der Fichtenformen und Fichtenrassen I-IV. Tharandter Forstl. Jahrb. 87: 87-91; 101- 188. ScuMiDT, H. 1952. Die Verzweigungstypen der Fichte (Picea abies L.) und ihre Bedeutung für die forstliche Pflanzenzüchtung. Zeit. f. Forstgenetik und Forstpflanzenzüchtung 1 (3): 81-91. SYLVÉN, N. 1909. Studier över granens formrikedom särskilt dess förgreningstyper och deras skogliga värde. Medd. f. Statens Skogsförsöksanstalt 6: 57-117. WAHLGREN, A. 1922. Skogsskötsel. 2nd ed. Nordstedt & Sons Stock- holm 732 pp. NOTES ON FESTUCA ARUNDINACEA AND F. PRATENSIS IN THE UNITED STATES EDWARD E. TERRELL Current regional and state floras of the United States provide varied treatments of the European grasses, Festuca pratensis Huds. (F. elatior L. of American authors; see Terrell 1967) and F. arundinacea Schreb. For example, the eighth edition of Gray’s Manual recognizes simply F. elatior sens. lat. Gleason in Britton and Brown’s Illustrated Flora recognizes var. pratensis (Huds.) Gray and var. arundinacea (Schreb.) Wimm. under F. elatior. Hitch- cock’s Manval of Grasses (1951 ed.; rev. by Chase) recog- 1968] Festuca — Terrell 565 nizes F. elatior and F. arundinacea but treats the latter as of limited distribution in the United States. Other floras employ a great variety of characters of variable worth to separate the two taxa. Generally, there is a scarcity of in- formation in the American floristic literature regarding these European taxa. The purpose of these notes is to point out the following: (1) recent European (or at least British and Scandinavian) floras consider F. pratensis and F. arundinacea as separate species, (2) several characteristics are ideally available to distinguish the taxa, (3) F. arundinacea probably is much more abundant and generally distributed in the United States and southern Canada than current floras indicate. Festuca arundinacea is generally hexaploid (2n=42) ; however, its variants or related taxa may have 2n—28 or 70 (e.g., see Malik 1967). F. pratensis is diploid (2n—14). Both species are indigenous to Europe and Asia, and the former species occurs also in north Africa. Hybrids be- tween the species are usually sterile (Jenkin 1959). Hackel (1882) divided European representatives of this complex into two subspecies, i.e., F. elatior subsp. pratensis and subsp. arundinacea, each with three varieties. How- ever, in the present state of our knowledge it seems best to recognize two species, following such works as Hubbard 1954 and Clapham et al. 1962. Obviously, this complex needs to be studied by modern methods in all parts of its native range. Various literature sources (notably Crowder 1953, 1956, Hubbard 1954, Clapham et al. 1962, and Gillet 1964) were searched for characters to distinguish the taxa. All charac- ters were then tried out on European and American specimens in the U. S, National Herbarium (Smithsonian Institution). The result is the key given below, in which the more significant characters are in italics. Well-collected specimens are necessary; even so, some specimens will be more or less intermediate. Hybrids would be expected to have abnormal anthers or sterile pollen. Illustrations of the lemmas are given by Musil 1963 and Gillet 1964. Huon 566 Rhodora [Vol. 70 (1965) indicated that epidermal characters also are useful in separating the taxa. KEY TO F. PRATENSIS AND F. ARUNDINACEA: Lemmas glabrous, glabrate, or scabrous only at apex, awnless or rarely short-awned; spikelets cylindrical to narrow-oblong; rachilla internodes glabrous or nearly so; shorter branch of each pair of panicle branches bearing only 1-2 spikelets; panicles 10-35 em. long; basal leaf sheaths breaking up and often decay- ing into irregular dark-brown fibers; leaf blades usually 2-6 (-8 )mm. wide; auricles glabrous; plants 30-120 em. high. ........ nm" Festuca pratensis Huds. Lemmas scabrous or short-hispid with minute teeth all over or only on nerves or keels (visible at 10 X or more magnification) or rarely glabrate, short-awned (to 4 mm. long) or awnless; spikelets elliptic to oblong; rachilla internodes scabrous; shorter branch of each pair of panicle branches bearing 3 or more spikelets; panicles 10-50 em. long; basal sheaths tough, whitish to darkish, persistent; leaf blades 3-12 (usually 4-6) mm. wide, usually stiffer and more heavily nerved than blades of F. pratensis; auricles ciliolate at least on lower leaves or sometimes glabrous; plants 45-180 cm. high, usually taller, more robust, and with thicker culms than F. pratensis. |... Festuca arundinacea Schreb. Festuca arundinacea (tall fescue) is currently much used in U. S. agriculture especially in the form of its cultivars ‘Kentucky 31’ and ‘Alta’. Tall fescue is extensively sown to prevent erosion along new highways in the southeastern states, Middle Atlantic states, and parts of the Middle West. To some extent it is employed in the same regions also as a forage crop. In the Pacific Northwest it is primarily a seed crop, secondarily a forage crop. Festuca pratensis (mea- dow fescue) is very little used. The U. S. National Herbarium has about 230 American specimens of F. pratensis (as F. elatior), many of which were collected more than 50 years ago. In contrast, there are only about 30 specimens of F. arundinacea, mostly col- lected since 1940. Both species occur in all sections of the United States and southern Canada. The scarcity of speci- mens of F. arundinacea is due partly to its having been used extensively in agriculture only in more recent years, and to the natural tendency of many collectors to ignore 1968] Festuca — Terrell 567 roadside plants if they appear to have been planted or sown. Field observations in the eastern states suggest that Festuca arundinacea is much more abundant than many current floras and herbarium specimens indicate. In Mary- land and Virginia it is a common feature of the roadside flora and may occur in solid stands for many miles. In eastern North Carolina it was seen also along ditch banks. Smith (1965) found F. pratensis (as F. elatior) in 26 districts in New York State, F. arundinacea in 16 districts. Although the extent to which F. arundinacea establishes itself as a weed from seeds is yet uncertain, it must be con- sidered an increasingly conspicuous part of our roadside flora. CROPS RESEARCH DIVISION AGRICULTURAL RESEARCH SERVICE UNITED STATES DEPARTMENT OF AGRICULTURE BELTSVILLE, MARYLAND LITERATURE CITED CLAPHAM, A. R., T. G. TuTIN, & E. F. WARBURG. 1962. Flora of the British Isles. Ed. 2. Cambridge, England: University Press. XLVIII + 1269 pp. CROWDER, L. V. 1953. A simple method for distinguishing tall and meadow fescue grass. Agron. Jour. 45: 453-454. 1956. Morphological and cytological studies in tall fescue (Festuca arundinacea Schreb.) and meadow fescue (F. elatior L.) Bot Gaz. 117 (3): 214-223. GiLLET, M. 1964. Un nouveau critére de distinction spécifique sur graines entre Festuca pratensis Huds. et Festuca arundinacea Schreb. Ann. Amélior. Plantes 14 (2): 203-206. HacKEL, E. 1882. Monographia Festucarum Europaearum. Berlin: Kassel. 216 pp. HUBBARD, C. E. 1954. Grasses. A Guide to their Structure, Identifica- tion, Uses, and Distribution in the British Isles. Penguin Books, Ltd. XII + 428 pp. Huon, A. 1965. Caractères épidermiques distinctifs des ssp. arundina- cea (Schreb.) Hack. et pratensis (L.) Hack. du Festuca elatior (L.) Hack. Bull. Soc. Bot. France 112 (1-2) : 37-42. JENKIN, T. J. 1959. Fescue species (Festuca L.), in Roemer, T. & Rudorf, W. Handbuch der Pflanzenzuchtung 4 (ed. 2) : 418-434. 568 Rhodora [Vol. 70 MALIK, C. P. 1967. Hybridization of Festuca species. Canad. Jour. Bot. 45: 1025-1029. MusIL, A. F. 1963. Identification of crop and weed seeds. U.S. Dept. Agric., Agric. Hdbk. 219. 171 pp. + XLII pl. SMITH, S. J. 1965. Checklist of the grasses of New York State. N.Y. St. Mus. & Sci. Serv. Bull. 403. 44 pp. TERRELL, E. E. 1967. Meadow fescue: Festuca elatior L. or F. pra- tensis Hudson? Brittonia 19: 129-132. YELLOW-FRUITED CORNUS FLORIDA — CULTIVAR OR FORM? ROBERT D. MACDONALD! During the period from August, 1966 through November, 1967 a number of yellow-fruited* plants of Cornus florida L. were observed in Tennessee. On the basis of these observa- tions, and reports received, it would seem that an adjust- ment in nomenclature (from C. florida cv. Xanthocarpa to C. florida forma xanthocarpa Rehd.) for this taxon may be warranted. During October and November, 1967 five spontaneous plants of Cornus florida bearing yellow fruit were observed. One plant (No. 1, Table I) was located in Monroe County, one (No. 2) in Anderson County, and three (Nos. 4, 5, and 8) in Knox County. One plant (No. 7) which was grown from seed was obtained from Roane County, but the seed source (possibly Warren County) is unknown. Occasional yellow fruits have been reported as being found in large amounts of fruit purchased by commercial nurserymen in Warren and Franklin Counties. This fruit reportedly is collected by local residents from spontaneous trees. These reports are from reliable sources but have not been personally verified. A report was received from a 'Director, John J. Tyler Arboretum, Lima, Pennsylvania 19060. *2.5 Y 8/12, vivid yellow to 10 YR 8/10, moderate orange yellow (color notations are taken from the Nickerson Color Fan with Munsell hue color symbols). 1968] Cornus florida — MacDonald 569 forester of a spontaneous yellow-fruited plant growing in Campbell County. This report was not personally verified. The only record of a yellow-fruited plant of Cornus florida contained in the University of Tennessee Herbarium is that of a plant (origin unknown) in Knox County. This is sheet No. 6997 collected by Dr. C. Rappennecker, November 24, 1948. This plant could not be located. Spontaneous plants with yellow fruit were reported in Swain County, North Carolina, on Hazel Creek near the Tennessee State Line, but these plants have not been per- sonally verified. It is my observation that, on the basis of fruit color alone, it is impossible to distinguish the plants in the trade (cv. Xanthocarpa) from the spontaneous yellow-fruited plants. It has been determined that the difference in color be- tween the red and yellow fruits of C. florida is due to the complete absence of two anthocyanins (peonidin and petuni- din) in the exocarp (Chester and Stone, 1964). The origin of the material analyzed was not given. The yellow-fruited forms of Cornus florida L. would ap- pear to now enjoy only cultivar status. Howard (1961) in his list of cultivar names in Cornus lists (under C. florida), “‘Xanthocarpa’ (Rehder, Jour. Arnold Arb. 2: 179. 1921) * Distinct on the basis of the yellow fruit. Described from material collected in Saluda, North Carolina in 1919." Howard (1967) states, “A yellow-fruited Cornus florida has also been found near Oyster Bay, Long Island, by Hicks & Co. of Westbury." Howard further indicated, “The her- barium supports the idea that both the Wilcox [material from Saluda, North Carolina] and Hicks specimens came from plants under cultivation.” Howard mentioned, in addi- tion, that his notes show no specimens of spontaneous plants with yellow fruit were found in the herbarium of the New England Botanical Club, the Gray Herbarium, or the Arnold Arboretum herbarium. The International Code of Nomenclature for Cultivated Rehder described this taxon as a forma (Cornus florida L. f. xanth- ocarpa, forma nov.). 570 Rhodora [Vol. 70 Plants (1961) states (Article 5), “The term cultivar . denotes an assemblage of cultivated individuals . . .”. The category forma is generally applied to trivial variations occuring among individuals of any population. Fruit color is an example of such variation. Howard (1961) states, in respect to Cornus florida ev. Rubra, “This name should not be used as a cultivar. Al- though Weston described it on the basis of cultivated material, the plant is known in the wild from pre-Linnean references as well as from modern observations and has been treated as Cornus florida forma rubra (Weston) Palm- er and Steyermark.” It is my opinion that yellow-fruited plants of Cornus florida should be treated as forma xanthocarpa Rehd, The name should not be used as a cultivar, since the statement which Howard (1961) applies to C. florida cv. Rubra may (for the most part) be applied here. Table I. Observations regarding yellow-fruited plants of Cornus florida. Herbarium Location Collection No.’ Origin of Plant Height Diameter? (Tenn. ) COFL-XA67-1 Spontaneous 8’ 2” Monroe Co. COFL-XA67-2 Spontaneous 12’ 5” Anderson Co. COFL-X A67-3 Commercial Nursery 5' Twosmall Knox Co. stems COFL-X A67-4 Spontaneous 12’ 4.5" Knox Co. COFL-XA67-5 Spontaneous 18' 4" Knox Co. COFL-XA67-6 Commercial Nursery 6’ Four stems Knox Co. 1” to 2” COFL-XA67-7 Commercial Nursery18" — «15" Roane Co. COFL-XA67-8 Spontaneous 14' 3" Knox Co. LITERATURE CITED CHESTER, W. and C. STONE. 1964. A comparative study of the antho- cyanins in the red-fruited and yellow-fruited flowering dogwood. Bull. Torrey Bot. Club 91: 506. 'Author's personal collection. "Diameter measured 4% feet above ground. 1968] Cornus florida — MacDonald 571 Howarp, R. A. 1961. Registration list of cultivar names in Cornus L. Arnoldia 21: 9-18. HowaRD, R. A. 1967. Personal communication, November 28. International code of nomenclature for cultivated plants. 1961. I.A.P.T. Utrecht, Netherlands. 30 pp. OCCURRENCE OF MURDANNIA SPIRATA IN FLORIDA' The recording of a description of this species in Ameri- can botanical works is deemed important because of its extended distribution from the Asiatic tropics of the East- ern Hemisphere to the subtropics of the Western Hemis- phere. According to Dr. J. P. M. Brenan who determined the specimen, the discovery is the first one out of its native region.” Murdannia spirata (L.) Bruckner. (Aneilema spiratum R. Br.) Plant annual, to 3 dm long; stems slender, soft, ascending, rooting at nodes, diffusely branching in age, glabrous, only the internodal groove viscid-villous; leaves thin, acute, 2-4 cm long, 4-10 mm wide, lanceolate, white- margined; the cordate-rounded clasping base sparsely ciliate or glabrate; inflorescence terminal or axillary, few- flowered, in axils of the uppermost leaves; peduncle elon- gate usually forked into 2, flexuous, divergent, paniculate branches; pedicels 5-9 mm long, distant, subtended by small, acute, clasping, floral bracts; living flowers in full anthesis 4-6 mm long; sepals concave greenish, diffused with pink; petals 4.5-5.5 mm long, pinkish-lavender ; limb suborbicular veined in deeper shades, minutely cremate- denticulate at apex, clearly showing 3 vascular traces fork- ing and reforking into finer traces radiating toward the 1Contribution No. 33 from the Botanical Laboratories, University of ‘South Florida. Aided by grant GB-2742 from the National Science Foundation. “The author is deeply indebted to Dr. J. P. M. Brenan for his determination of the species. 572 Rhodora [Vol. 70 Plate 1403. Murdennta spirata CL.) vdleknet Collier County. Plants filling a hollow in ` sand, left by overturned rec , of Pinus €lllectti, Open La under pines and Sabal palmette, Surrounding vegetation diet by cutting, followed by weed invasion, Imeokale@e City Limite. de Lakela 20612 6 Jan 1955 ollected for the "Flora t Subtropical Florida" Pad ing e Hupi, det. by JPM. Tronan Mec., ie: Murdannia spirata 1968] Murdannia — Lakela 573 margin; stamens opposite the sepals; anthers pastel blue, pollen white; individual grains colorless oval-oblique in face view; filaments tapering to anther connectives from thicker barbate base; hairs beaded; staminodes white, 3- lobulate; style 3 mm long, stigma capitate; capsule ovoid with tardily deciduous style; valves lustrous, chartaceous, apex with persistent style base; seeds angled, averaging 0.8-0.9 in length, typically 4, uniseriate in each locule, contiguous surfaces truncate, the first and the last rounded at free ends; testa pale gray, sparsely verrucose, Murdannia spirata differs from M. nudiflora in softer texture of the plant as a whole; leaf-shape; finer pubes- cence; diffuse, paniculate inflorescence ; blue anthers; color of the petals; more numerous, smaller seeds. The known distribution in Florida is restricted to north- ern Collier County, in the vicinity of Immokalee. A popula- tion of this plant was first noted on the margin of a seasonally wet prairie glade, 6 miles north of Immokalee. The lush green plants dispersed in mucky soil among tufts of Eleocharis and Hydrocotyle presented a different appearance from Murdannia nudiflora (Coll. Lakela 29185, Oct. 5, 1965, in flower and fruit). The plants were encountered in anther site at Immokalee City limits. The colony was centered in a hollow of sand, made by upturned roots of a fallen pine. Over the area on a road off the main street, Pinus Elliottii and Sabal pal- metto, remnants of pine flatwoods, cast an open shade. The plants filled the hollow completely and were largely post- mature but still in excellent fruit. Photo. (Coll. Lakela 29412, Jan. 6, 1966.) These sites were conveniently located for follow-up study and collecting. The period of anthesis is approximately from September-December. Flowers remain open for from two to three hours. Duplicates of dried specimens with one living plant were mailed to Kew, England. OLGA LAKELA UNIVERSITY OF SOUTH FLORIDA TAMPA, FLORIDA 33620 574 Rhodora [Vol. 70 REFERENCES CONSULTED BRENAN, J. P. M. 1952. Notes on African Commelinaceae. Kew Bulletin, 179-190. GAMBLE, J. S. and FISCHER, C. E. C. 1931. Flora of the Presidency of Madras, Part IX, 1540-1544. NAIRNE, ALEXANDER Kan 1894. The Flowering Plants of Western India, 353-354. EMENDATIONS IN UNITED STATES PHYSALIS U. T. WATERFALL An investigation of Physalis in Mexico, Central America and the West Indies, undertaken during the past four years, including a trip in August, 1965 to study types and other historically significant specimens in European herbaria, has resulted in the necessity of reinterpreting some of the taxa of this genus in the U.S. In addition, further observation indicates the desirability of a change in status for another taxon. Physalis hederaefolia Gray, var. puberula Gray was in- cluded in var. hederaefolia in the earlier treatment (Water- fall, 1958). In studying Mexican material it seemed that collections with vesiture of short, antrorse hairs only, or with varying amounts of glandular-capitate trichomes, might well be referred to var. puberula, those with vestiture including long, jointed hairs to var. hederaefolia (Waterfall, 1967). The two may be intermixed in much of their range, but var. puberula is more western. Examples of var. puberula include: CALIFORNIA: SAN DIEGO CO.: “Cleveland collection" 6993, between Walkers and Jacumba, June 23, 1885 (sp); Daniel Cleveland 6931, July 3, 1884 (sD); ARIZONA: COCHISE CO.: Wyatt W. Jones sin num, Don Luis, Aug. 18, 1956 (sp); YAVAPAI CO.: Charles W. Harbison 41.220, Kirkland Junction, May 29, 1941 (sp). Physalis philadelphica Lamarck, var. immaculata Water- fall Rhodora 69: 215. 1967. This small-flowered variety 1968] Physalis — Waterfall 575 with essentially immaculate corollas was described from Chihuahua, Jalisco, Guanajuato and Sonora. Specimens at hand from the United States are: ARIZONA: COCHISE co.: Booth & Jones sin num, Don Luis, Aug. 26, 1955 (sp); Wyatt Jones sin num, McNeal, Sept. 6, 1962 (SD). Physalis cordata Miller includes the material with large, glabrous, acuminate-toothed fruiting calyces originally re- ferred by the author (1958) to P. pubescens var. glabra which proves to be only an extreme of P. pubescens var. pubescens (Waterfall, 1967). Representative collections at hand are: ARKANSAS: HOT SPRINGS co.: Demaree 30279, foothills, Magnet Cove, Butterfield, Oct. 28, 1950 (OKLA); CALIFORNIA: SAN DIEGO CO.: Gardner 9820, Lakeside, Sept. 26, 1945 (sp); FLORIDA: POLK CO.: J. Cecil King sin num, Lake Wales, Sept. 3, 1958 (OKLA) ; LOUISIANA: ASCENSION PARISH: Anna Ashby 36, very frequent in cane field, 4 miles from Donaldsonville, Nov. 23, 1956 (OKLA); LA FOURCHE PARISH: Joseph Ewan 18877 pro parte, S. P. RR. Station, Raceland Junction, Oct. 20, 1956 (OKLA). Physalis turbinata Medicus has longer, narrower calyx lobes, and larger fruiting calyces than P. pubescens sens. str. to which it is closely related (Waterfall, 1958). Much of the material labelled P. turbinata in American herbaria is P. cordata Miller, a glabrous, or near-glabrous species with long fruiting pedicels, quite different from the hairy species specified by Medicus. Representative collections at hand are: FLORIDA: POLK CO.: Me- Farlin 5924, Bartow, June 23, 1931 (cas); TEXAS: ANGELINA CO.: Hamby 1582, 3 mi w of Diboll, Aug. 9, 1940 (US); BOWIE CO.: Correll 31272, hardwood hills, among grasses in open woods, about 2 miles north of Texarkana, July 12, 1965 (LL); CAMERON C0.: Correll 14856, La Palmas Plantation, about 4 miles southwest of Brownsville, in palm grove, Oct. 4, 1952 (LL); HARRIS CO.: Correll 31889, in rich woods of meadow along stream near Addicks, Oct. 6, 1965 (LL); HIDALGO CO.: Fleetwood 3112, Santa Ana National Wildlife Refuge south of Alamo, Sept., 1959 (TEX); WILLACY CO.: Johnston 542221, sparse mesquite brush over sacahuiste, 9 miles west of Redfish Bay, Nov. 28, 1954 (LL). Physalis pubescens L., var. missouriensis (Mackenzie & Bush) Waterfall, comb. et stat. nov., P. missouriensis Mac- kenzie & Bush, Trans. Acad. Sci. St. Louis 12: 84-85. 1902. 576 Rhodora [Vol. 70 This taxon seems easily differentiated from P. pubescens var. pubescens by its immaculate corolla and smaller an- thers. Nevertheless, the relationship seems close enough that it might be better to consider it a variety of that species. DEPARTMENT OF BOTANY AND THE RESEARCH FOUNDATION OKLAHOMA STATE UNIVERSITY STILLWATER, OKLAHOMA LITERATURE CITED WATERFALL, U. T. 1958. A Taxonomic Study of the Genus Physalis in North America north of Mexico. Rhodora 60: 107-114; 128- 142; 152-173. WATERFALL, U. T. 1967. Physalis in Mexico, Central America and the West Indies. Rhodora 69: 82-120; 203-239; 319-329. ALISMA GRAMINEUM IN VERMONT WM. D. COUNTRYMAN Although common in the western United States north of the 40th parallel and in the adjacent Prairie Provinces of Canada, Alisma gramineum K. C. Gmel. has been known in eastern North America only from the St. Lawrence River valley and the eastern end of Lake Ontario. Asa Gray was apparently the first person to collect Alisma gramineum east of the Mississippi River. His specimen, collected at Ogdensburg, St. Lawrence County, New York, is deposited in the Gray Herbarium; unfortunately the label bears no date and, therefore, it can be stated with certainty only that it was collected before January 1888, the date of Gray's death. In Canada, Brother Marie-Victorin collected the species near Longueuil and at Montreal, Quebec in 1918. Fernald and Wiegand (1923) reported the plant from a cove of Lake Ontario near Chaumont, Jefferson County, New York. Muenscher and Maguire (1931) and Muenscher (1932) found it growing at a number of stations along the St. Law- rence River in St. Lawrence and Franklin Counties, New York. More recently Raymond and Kucyniak (1948) have reported Alisma gramineum as “a pesky weed around Mon- treal: aa. now progressively invading the Richelieu River Valley." Such a rapidly spreading species could soon be expected to make its appearance in the Champlain Valley. It is not surprising, therefore, that I found it in Vermont growing along the shore of Lake Champlain at Shelburne Bay in 1961. Subsequent collections have shown it to be present along the shores of the Lake as far south as the Town of Addison, Addison County, Vermont. It is noticeably more common in the northern portion of the lake. In 1966, after diligent searching, I found only a few plants in Addison, Vermont and none at all south of that town. I have not collected on the New York side in the northern portion of Lake Champlain and do not know if Alisma gramineum 577 578 Rhodora [Vol. 70 grows there; it would seem most probable that it does. Recently I have been privileged to examine the specimens of Alisma gramineum in the herbarium of the Plant Re- search Institute, Department of Agriculture, Ottawa. These collections show that the species is well established on the Canadian side of the St. Lawrence River from Lake Ontario as far downstream as St. Angéle de Laval, Quebec, a distance of approximately 250 miles. Although the plant has now successfully invaded the Richelieu River and Lake Cham- plain it is interesting to note that it has not yet been found in any other tributaries of the St. Lawrence. Both submersed plants with ribbon-like leaves and emersed plants with lanceolate to elliptic blades were found at most of the Vermont stations. When fruiting, Alisma gramineum is easily recognized by its tricarinate achenes in contrast to the bicarinate achenes of A. triviale Pursh and A. subcordatum Raf. All of the specimens cited below were collected by me in western Vermont along the shores of Lake Champlain. I have deposited duplicates of all specimens cited in the her- barium of the New England Botanical Club. ALISMA GRAMINEUM K. C. Gmelin. VERMONT. GRAND ISLE COUNTY: Alburg, occasional, submersed in bay north of Rouse’s Point toll bridge, 11 August 1964 952; Isle La Motte, Jordan Bay, 10 August 1964 948; North Hero, rare, near North Hero Station, 27 August 1965 1359; Grand Isle, Simms Point, The Gut, terres- trial specimens blooming earlier than those growing submersed, 14 August 1964 958; CHITTENDEN COUNTY: Colchester, west of R.R. tracks, Colchester Point, 5 October 1964 1161; Shelburne, Shelburne Bay, 11 August 1961 98; Charlotte, sandy shore at Town Farm Bay, 30 July 1965 1323; ADDISON COUNTY: Ferris- burg, occasional at Basin Harbor, 18 August 1966 1547; Panton, Arnold Bay, 13 September 1964 1086; Addison, scarce, only about 12 plants seen in one-half mile of beach, Potash Bay, 21 August 1966 1555. DEPARTMENT OF BIOLOGY NORWICH UNIVERSITY NORTHFIELD, VERMONT 05663 1968] Alisma gramineum — Countryman 579 LITERATURE CITED FERNALD, M. L. and K. M. WIEGAND, 1923. Notes on Some Plants of the Ontario and St. Lawrence Basins, New York. Rhodora 25: 206. MUENSCHER, W. C. 1932. Aquatic Vegetation of the Oswegatchie and Black River Watersheds. In: A Biological Survey of the Os- wegatchie and Black River Systems. New York State Conserva- tion Department, Biological Survey No. VI, Section VIII: 215. MUENSCHER, W. C. and BASSETT MAGUIRE. 1931. Notes on Some New York Plants. Rhodora 33: 165. RAYMOND, MARCEL and JAMES KUCYNIAK. 1948. Six additions to the Adventitious Flora of Quebec. Rhodora 50: 179. Volume 70, No. 784, including pages 457-600, was issued December 31, 1968. 580 Rhodora [Vol. 70 VOLUME 70 1968 Errata for RHODORA . 92 line 12, 15 should be 16. . 32 line 12 mucronota, should be curtipendula. 33 line 20 for Flagelliformis, read flagelliformis. 43 line 31 for Astralagus, read Astragalus. 52 line 8 delete the quotation marks before the word Nuttall. 53 line 41 read Ownbey for Ownby. 39 line 5 for Catesboei, read Catesbaei. 61 caption figs. 9-16 line 1 Lasioneuron, should be lasioneuron. . T2 line 7 for liguils, read ligules. 75 caption fig. 22 line 2 Illinoensis should be illinoensis. 75 line 1 Smilax illinoensis should be in italics. 101 line 22 Virginica should be virginica. 135 line 7 for yonger, read younger. No. 782 165 line 2 gracillima should be in italics. 170 line 35 for acquatic, read aquatic. 176 line 6 for aerolata, read areolata. 202 line 29 for Brauni, read braunii. 207 line 12 warnstorifianum should be warnstorfianum. 212 lines 1, 11, 12, 26 for Sasketchewan, read Saskatchewan. 225 line 37 for dichotomous read dichotomus. 232 line 7 for squammatus, read squamatus. 233 line 29 for Cichoreae read Cichorieae. 236 lines 31, 34 for vellutinum, read velutinum. 286 line 24 for Banhinia, read Bauhinia. 908 line 9 for woud, read would. No. 783 394 line 26 for enedmic, read endemic. 401 line 37 for Talinumm engesii, read Talinum mengesii. 405 line 32 for allopolploidy, read allopolyploidy. 405 line 37 for teretrifolium, read teretifolium. 414 line 38 for rachus, read rachis. 430 line 23 for apears, read appears. 431 line 19 for ludoviviana, read ludoviciana. . 443 line 29 for halopila, read halophila. . 443 line 31 for Seripus, read Scirpus. VOUCUCUCU UCUCUCU CU UCUGS PUPP PP PPP EY UCUCUCUOU UPD INDEX TO VOLUME 70 New scientific names and combinations are printed in bold face type Abies balsamea 83, 90, Variation in Cone Morphology of Balsam Fir 83, var. balsamea 90, var. phanerolepis 83, 87, 90, 91; Fraseri 90; lasiocarpa 83 Acacia 129, 130, 287, 331, 492 Acanthospermum australe 232, 238; hispidum 232 Acer rubrum 224 Achillea borealis 180-182, 184, 186, 187; millefolium 210, ssp. lanulosa var. megacephala 204 Acosmium 499; lentiscifolium 499 Actinomeris 99 Adenostemma brasilianum 229, 231, 233, 235 Adenostylinae 231 Adiantum 2, 3 Agastache anethiodora 38; nepe- toides 38; scrophulariifolium 38 Agastianis 521; secundiflora 528 Ageratinae 231 Ageratum conyzoides 235 Agoseris glauca 50 X Agroelymus palmerensis 471 X Agrohordeum 472 Agropyron 467, 471, 472; lati- glume 470; pungens 442; seri- ceum 470, 471; smithii 451; trachycaulum 210, 470, 471; violaceum 470, 471 Agrostis scabra 205, 208 Alabama, Allium ampeloprasum in 303 Alaska, Hybridization within the Triticeae of 467 Alisma gramineum 577, 578; sub- cordatum 578; triviale 578 Alisma plantago 171; plantago- aquatica 171 Allionia hirsuta 33; linearis 33 Allium ampeloprasum 303, in Alabama 303; reticulatum 28; stellatum 28; textile 28; tricoc- cum 285 Alloplectus 413; ambiguus 414 Alnus oblongifolia 465 Alocasia sp. 414 Alomia fastigiata 231, 235; micro- carpa 231 Alternaria 385 Alyssum 300; alyssoides 298, 299; calycinum 298, 299; campestre 299; desertorum 298; Intro- duced in North America 298; maritimum 299; minimum 298; minus var. micranthum 299, 300; petraeum 28, 299; saxatile 299; sect. Alyssum 298-300; sect. Psilonema 298; strigosum 299, 300, ssp. strigosum 299; szowitsianum 299 Amazon, The Vegetal Ingredients of the Myristicaceous Snuffs of the Northwest 113 Ambrosia polystachyus 232, 238 Amellus spinulosus 50; villosus 49 Ammophila arenaria 223; brevili- gulata 225 Amorpha 512; canescens 29; mi- crophylla 29; nana 29 Ampelamus albidus 431 Amphianthus 396; pusillus 394, 396 Anadenanthera 120, 128, 331; per- egrina 114, 115, 117-119, 125, 127, 128, 131, 132, 144, 331-333 Anaphalis margaritacea var. in- tercedens 179 Anderson, C. E. and J. H. Bea- man, Status of the Genus Cym- ophora (Compositae) 241-246 Aneilema spiratum 571 Anthemideae 482 583 584 Anthodiscus 335; obovatus 335 Anthurium dominicense 414 Aplopappus spinulosus 370 Apocynaceae 326, 327, 417, 427 Apodanthes 426 Arabis arenicola 206, 209, 211, var. arenicola 211, var. pubes- cens 211; drummondii 180, 187, 452; hirsuta var. pycnocarpa 180-182; lyrata 205; missouri- ensis 452 Araceae 412-414, 417 Aralia racemosa 182 Arctic-Alpine and Boreal Dis- junct Species, Habitat Studies of 176 Arctic-Alpine Plants in Montana, New Records of Disjunct 103 Arctostaphylos alpina ssp. rubra 212 Arenaria caroliniana 290 Arethusa bulbosa 196, 198, 206- 208 Argus, George W., Contributions to the Flora of Boreal Sas- katchewan 200-211 Aristida 30; oligantha 30; pallens 30 Aristolochia 324 Armeria maritima var. 205, 206, 209, 211 Arnica chionopappa 184 Artemisia arbuscula ssp. thermo- pola 286; campestris ssp. bo- realis 206, 209, 211; cernua 30; dracunculoides 30; dracunculus 30; nutans 30; sect. Tridentatae 286; stelleriana 223; tripartita var. hawkinsii 286 Asclepiadaceae 415, 427 Ascyrum hypericoides 225 Aspidium goldianum 457, 458 Aspilia 241 Asplenium 2, 346; platyneuron 452; trichomanes 188, 346; vir- ide 182 Aster 221, 448, 444; squamatus interior Rhodora [Vol. 70 232, 237 Astereae 237 Asteridae 425 Astragalus 492; carnosus 29; caryocarpus 29; crassicarpus 29; gracilis 30; melanocarpus 50; missouriensis 30; physodes 29; tenellus 30, 43 Ateleia 499, 506, 507; albolutes- cens 508, 509, 513; arsenii 507- 509; cubensis 514; gummifera 507-509, 514, var. cubensis 514; insularis 508-510; multijuga 514; parvifolia 514; pterocarpa 508, 509, 511-514; standleyana 508-511; tomentosa 507, 509, 513; truncata 507, 509, 510; tumida 514 Atriplex patula var, hastata 443, 448, 449, 451 : Aulacomnium palustre 207; sp. 212 Aurinia petraea 299; 299 saxatilis Baccharidastrum 232, 233; noto- bellidiastrum 231, 232, 237 Baccharis 232; genistelloides 237; helichrysoides 231, 237; oxyo- donta 231, 237 Baldwin, Henry L, A “Pillar’- type Black Spruce 562-564 Bamberg, S. A. and R. H. Pemble, New Records of Disjunct Arc- tic-Alpine Plants in Montana 103-112 Banisteria 318, 326; Caapi 318, 324, 326; purpurea 505 (as Ban- nisteria) Banisteriopsis 318, 323, 324, 326- 328, 330; Caapi 318, 319, 323- 325, 328, 330; inebrians 323, 328; quitense 323; Rusbyana 325, 328 Baptisia 531; leucophaea 45 Bartonia 31; decapetala 31; nuda 32; polypetala 32 1968] Bassia hirsuta 443, 444 Batschia canescens 460; decum- bens 31; fimbriata 31; longi- flora 31, 460 Bauhinia coulteri 286, A White Flowered Form of, from Ta- maulipas, Mexico 286, f. albi- flora 286; sect. Bauhinia 286 Beaman, J. H. and C. E. Ander- son, Status of the Genus Cym- ophora (Compositae) 241-246 Beetle, A. A., Another Note on Sagebrush Taxonomy 286 Begonia 413; decandra 414 Begoniaceae 414 Berchemia scandens 41 Berry, Wendell H. Jr., Range Ex- tension of Muhlenbergia soboli- fera in New Hampshire 452 Bertholletia excelsa 135 Betula glandulifera 210, 212 Bidens segetum 238 Blechnum 2 Bogle, Alfred L., Evidence for the Hybrid Origin of Petasites warrenii and P. vitifolius 533- 551 Botrychium dissectum var. ob- liquum 485 Bouteloua curtipendula 32 Bozeman, John R. and John F. Logue, A Range Extension for Hudsonia ericoides in the Southeastern United States 289-292 Britton, Donald M., The Spores of Four Species of Spinulose Wood Ferns (Dryopteris) in Eastern North America 340- 347 Bromeliaceae 412, 415, 417 Bromus pumpellianus 204; tecto- rum 442 Broussonetia 521, 528 Bumelia confertiflora 31; lanugi- nosa var. oblongifolia 31; ob- 521; secundiflora Index to Volume 70 585 longifolia 31 Bunchosia 322 Burk, C. John, A Floristic Com- parison of Lower Cape Cod, Massachusetts and the North Carolina Outer Banks 215-227 Bursera 246 Butomus umbellatus 171 Byrsonima 321 Cactaceae 410 Cactus ferox 33; flagelliformis 33; fragilis 34, 431; mamillaris 33; viviparus 33, 431 Caesalpiniaceae 493 Caesalpinioideae 493, 496 Calamagrostis canadensis 210; inexpansa 204; neglecta 204, 205, 208 Calea 233; multipega 238; pin- natifida 238 Calia 521, 522; erythrosperma 521, 528 Calliandra 492 Calliergon stramineum 207 Callitriche 552 Calochortus nuttallii 37 Calopogon pulchellus 197, 198 Calusia 505; emarginata 506 Calymenia angustifolia 33; nycta- ginea 32; pilosa 33 Camassia esculenta 44, 45; qua- mash 45; scilloides 45 Campanula rotundifolia 179, 181, 210 Campsis chinensis 67; radicans 67 Carduus undulatus 431 Carex 208, 212, 221; “aborigi- num" 420; ablata 420; aquatilis 205, 207, 210, 212; atrata 421; atratiformis 182, 184; aurea 210; capillaris var. major 180; chalciolepis 421; concinna 210; epapillosa 421; foenea var. tu- berculata 419; heteroneura 421, var. chalciolepis 421, var. epa- 586 pillosa 421, var. brevisquama 421; intumescens 179; limosa 206; luzulina 420, var. ablata 420; maritima 206, 209, 211; michauxiana 206-208; microp- tera var. crassinervia 420; olig- osperma 206; parryana 420, 421, var. brevisquama 420, 421; pauciflora 206; paupercula 207; potosina 419; rostrata 206; rus- byi 419; stenochlaena 106; tri- sperma 207; vallicola 419, var. rusbyi 419 Carices, Notes on Rocky Moun- tain 419 Carphephorus 474-477, 480-482; bellidifolius 477, 478, 483; car- nosus 477-479, 483; corymbosus 431, 477, 478, 483; odoratis- simus 483; paniculatus 483; pseudoliatris 474, 477-479, 483; tomentosus 477, 478, 481, 483 Carya ovata 452 Caryocar 335; gracile 335 Caryophyllidae 425, 427 Caryota sp. 414 Cashalia cuscatlantica 515, 517 Cassiope mertensiana 110; tetra- gona 103, 105, 108-110 Cathartolinum bahamense 440; bracei 440; corallicola 440; lig- nosum 440 Ceanothus americanus 452 Ceratophyllum demersum 171 Chamaecyparis thyoides 217 Chamaedaphne calyculata 193 Chaptalia integerrima 233, 239; mandoni 240; nutans 233, 240 Chase, Sherret S., A Case of AI- binism and of Presumptive Self-Compatibility in Ulmus 294-296 Cheiranthus asper 34; asperus 34 Chloranthaceae 416 Chloris mucronata 32 Chrysocoma odoratissima 483; Rhodora [Vol. 70 paniculata 483 Chrysopsis villosa 301, var. vis- cida 301, 302; viscida 302 Cichorieae 233 Cicuta virosa 171 Cirsium undulatum 431 Cissus 415; trifoliata 415 Cladonia spp. 290 Cladrastis secundiflora 528 Clathrotropis (?) flava 518 Clethra alnifolia 290 Clibadium 232; armani 231, 232, 238; surinamense 232 Clusia griesbachiana 415 Coleman, James R., A Cytotax- onomic Study in Verbesina (Compositae) 95-102; Chromo- some Numbers in Some Bra- zilian Compositae 228-240 Comandra livida 210 Commelinidae 426 Compositae 222, 228, 241, 561, Chromosome Numbers in Some Brazilian 228 Connarus 335 Connecticut, Two Species New to 455 Conyza chilensis 231, 232, 238 Coprosmanthus 55, 57, 247; ecir- rhatus 265; herbaceus 249, var. ecirratus 58, 265; lasioneuron 253; peduncularis 249; tamni- folius 260 Corallorhiza odontorhiza 485 Cornus canadensis 278-281, De- velopmental Variability of, in Northern New England 278, f. alpestris 279, f. elongata 278, 279, 281, f. purpurascens 278, var. alpestris 279, 281, var. in- termedia 278, 279; florida 568, 569, cv. Rubra 570, cv. Xantho- carpa 568, 569, f. rubra 570, f. xanthocarpa 568; suecica 278, 279; X unalaschkensis 278, 279 Corydalis bulbosa 296-298, A Note on 296; intermedia 297; Eupatorium 1968] solida 297, ssp. solida 298 Coryphantha missouriensis 33; vivipara 33 Countryman, gramineum in 579; Wolffia in shire 491 Cristaria coccinea 41 Cronquist, Arthur, The Evolution and Classification of Flowering Plants (Book Review) 422-428 Croton punctatus 225 Cryptantha celosioides 32 Cryptogramma stelleri 180, 187 Cyathea 2 Cymophora 241; pringlei Status of the Genus 241 Cynoglossum glomeratum 32 Cyperaceae 222 Cyperus aristatus 399, 403-405; granitophilus 394, 3977, 399, 403- 405; inflexus 403; polystachyos 225; sp. 32 Cyrilla racemiflora 290 Cystopteris bulbifera 180, 187; fragilis var. laurentiana 181, 184 Cytisus rhombifolius 34 Wm. D. Alisma Vermont 577- New Hamp- 241; Dactyloctenium aegyptium 32 Dalbergia gummifera 514; laevi- gata 500 Dalbergieae 497 Dalea aurea 35; candida 47; en- neandra 35, 431, 432; laxiflora 35, 432; parviflora 30; purpurea 47 Dermatophyllum 521; speciosum 521, 528 Deschampsia caespitosa ssp. ber- ingensis 204 DeWolf, Gordon P., Note on Cory- dalis bulbosa 296-298 Diamorpha cymosa 399, 401 Dichantheum annulatum 556 Diervilla lonicera 182 Digitalis purpurea 46 Index to Volume 70 587 Dilleniidae 425 Diodia virginica 34 Diospyros virginiana 224, 225 Dipteryx 337 Distichlis spicata 442, 444-446; stricta 451 Distylum 425 Donia squarrosa 51 Draba arabisans 180-182; vegica 184, 185 Dracocephalum 34; cordatum 431; cuspidatum 34, 35; parviflorum 34 Dracopsis amplexicaulis 353 Drepanocladus 210; sp. 212; un- ciatus 207 Drosera anglica 206-208; rotundi- folia 206, 208 Dryas integrifolia 103, 105, 108; ssp. integrifolia 105 Dryopteris 2, 221, 340, 346; as- similis 340, 346, 347; Borreri 2; campyloptera 340-346; car- thusiana 340; dilatata 340-347; fragrans 346, var. remotiuscula 188; goldiana 457; intermedia 340-346; marginalis 182; spi- nulosa 340, 342-346; The Spores of Four Species of 340 Dudley, Theodore R., Alyssum (Cruciferae) Introduced in North America 298-300 Duroia 335; petiolaris 336 Dussia 499, 515, 516; cuscatlan- tica 501, 516, 517; grandifrons 517; martinicensis 515; mexi- cana 501, 516 nor- Eaton, Richard J., The Diagnostic Value of Fresh Mature Drupes of Viburnum recognitum and V. dentatum 288-289 Echinacea 358; purpurea 49 Echites puberula 36 Eclipta alba 238, 239 Elaeagnus angustifolius 35; ar- 588 gentea 35, 36; argenteus 35; commutata 35 Eleocharis 221, 285, 573; halo- phila 442, 443; nitida 207 Elephantopus 229; mollis 229, 235 Eleutheranthera 241, 246; ruder- alis 242 Elizabetha 130, 188, 144, 154; Duckei 154; princeps 135, 139, 153, 154, 333 Elodea canadensis 171, 203 x Elyhordeum arcuatum 468-470, 472 Elymus 467, 471, 472; arenarius 467; glaucus 470, 471; hirsutus 470, 471; innovatus 470, 471; mollis 468, 470, 471, ssp. mollis 204; sibiricus 467-471 Emilia coccinea 233, 239 Empetrum nigrum 205 Enslenia albida 431 Epidendrum ramosum 415 Epilobium glandulosum 182 Epipactis Helleborine 196, 198 Equisetum fluviatile 205; scir- poides 210 Erechtites hieracifolium 233, 239 Erianthus brevibarbis 49 Erigeron hyssopifolius 182, 184 Eriogonum flavum 26, 27, 36; sericeum 36 Eriophorum gracile 206; vagina- tum ssp. spissum 207 Ervum multiflorum 30 Erysimum asperum 34; tanum 36 Erythroxylon Coca 119 Escherichia coli 422 Euonymus obovatus 431 Eupatorieae 229, 235, 474, 475 Eupatorium 231; altissimum 455; ballotaefolium 231, 236; calli- lepis 236; congestum 236; hys- sopifolium 455; intermedium 231, 236; laetevirens 236; lani- gerum 236; macrophyllum 236; orbiculatum 236; pallescens mon- Rhodora [Vol. 70 236; serotinum 455; Two Spe- cies New to Connecticut 455; velutinum 231, 236 Euphorbiaceae 410 Euryale ferox 170, 171 Euthamia 50 Evolvulus argenteus 36; pilosus 36; sessiliflorus 36 Ewan, Joseph, John Goldie and the Republication of his Diary 457-461 Faba 496 Fabaceae 492, 493, 495, 496 Faboideae 493, 495-497 Fagus grandifolia 224 Ferula foeniculacea 36; pubescens 36 Festuca arundinacea 564-567, Notes on, in the United States 564; baffinensis 103, 106, 109, 110; brachyphylla 103, 105, 106; elatior 564-567, ssp. arun- dinacea 565, ssp. pratensis 565, var. arundinacea 564, var. pratensis 564; rubra 203-205, ssp. richardsonii 204, 205, var. glabrata 204, 205, 208, var. arenaria 204, 205, var. prolifera 204; ovina 105; pratensis 564- 567, Notes on, in the United States 564; prolifera 204; vivi- para 103, 105, 106 Florida, Occurrence of Murdan- nia spirata in 571 Forsteronia portoricensis 415 Fragaria virginiana 210 Fraser’s 1813 Catalogue, On the Names in 25 Freeze-dry Technique for the Plant Collector, A 410 Fritillaria alba 37; americana 37; atropurpurea 37 Fumaria bulbosa 297, * cava 297, B intermedia 297, Y solida 297 Gaillardia aristata 51; grandiflora 51; pulchella 223 1968] Galapagos Islands, A New Spe- cies of Physalis from the 408 Galegeae 498 Galinsoga 245 Galinsoginae 241 Galium septentrionale 210 Gaura angustifolia 225; coccinea 37 Gaylussacia dumosa 291 Genisteae 497 Gerardia flava 460; maritima 443, 444; quercifolia 460 Gesneriaceae 414 Glaux maritima 443, 448 Glyceria melicaria 442 Glycyrrhiza 40; lepidota 39 Goldie, John and the Republica- tion of his Diary 457 Gonolobus hirsutus 37 Gonyza notobellidiastrum 232 Goodyera pubescens 197, 198 Gramineae 222 Granite Outerop Communities of the Southeastern Piedmont, Plant Speciation Associated with 394 Graustein, Jeannette E., Thomas Nuttall Naturalist Explorations in America 1808-1841 (Book Review) 429 Grindelia 50; squarrosa 51 Grinnellia americana 292, 294 Gutierrezia 50 Guttiferae 415 Guzmannia berteroniana 415 Habenaria lacera 195, 198; psy- codes 195, 198; virescens 195; viridis var. bracteata 195, 198 Haemadictyon 325, 326; amazon- icum 326 Haematoxylon 492 Halenia deflexa 210 Hamamelidaceae 425, 426 Hamamelidae 425 Haplopappus spinulosus 50 Hargraves, Paul E., Two Natural- Index to Volume 70 589 ly Occurring Abnormalities of the Diatom Podocystis adriatica 292-294 Harms, Vernon L., Application of the Name Heterotheca viscida 301-303 Hebert, Henry J.-C., Generic Con- siderations concerning Car- phephorus and Trilisa (Com- positae) 474-485 Hedeoma glabra 431 Hedyosmum 416; arborescens 416 Hedyotis procumbens 306-310, var. hirsuta 308, the Status of 306, var. procumbens 308 Hedysareae 497 Heliantheae 232, 238, 353 Helianthemum corymbosum 225 Helenieae 481 Helenium nudiflorum 454, A Yel- low-centered Form of 454, f. homochromum 454 Heliconia 413, 416; bihai 416 Hepaticae 316 Hermann, F. J., Notes on Rocky Mountain Carices 419-421 Heteranthera dubia 171 Heterotheca 301; fulcrata 301; sect. Heterotheca 301; subaxil- laris 220; viscida 301, 302, Ap- plication of the Name 301; villosa 301 Hevea 119 Hieracium canadense 181; com- mersonii 231, 240; pilosella 179, 182 Hillia parasitica 416 Hodgson, H. J. and Wm. W. Mitchell, Hybridization within the Triticeae of Alaska; A New x Elyhordeum and Comments 467-473 Holmstedt, Bo and Richard Evans Schultes, De Plantis Toxicariis e Mundo Novo Tropicale Com- mentationes. II. The Vegetal Ingredients of the Myristi- 590 caceous Snuffs of the North- west Amazon 113-160 Hordeum 467; brachyantherum 468, 469, 471; jubatum 451, 467- 472 Hounsell, R. W. and E. C. Smith, Contributions to the Flora of Nova Scotia. IX. Habitat Stud- ies of Arctic-Alpine and Boreal Disjunct Species 176-191 Houstonia procumbens 306; ro- tundifolia 306 Howard, Richard A., A Freeze- dry Technique for the Plant Collector 410-418 Howe, M. Dorisse, A Yellow-cen- tered Form of Helenium nudi- florum 454 Hudsonia 290; ericoides 289-291, A Range Extension in the Southeastern United States for 289; montana 291; tomentosa 225, 443 Hydrocharis morsus-ranae 171 Hydrocotyle 573; bonariensis 225 Hymenopappus 481, 482 Hyparrhenia rufa 556 Hypericum canadense 291; densi- florum 291; surculosum 38 Hypochaeris brasiliensis 233, 240; gardneri 233, 240; radicata 233, 240, 559, 561 Hyssopus anethiodorus 38; ani- satus 38 Hystrix patula var, Bigeloviana 452 Ilex coriacea 290; glabra 290; opaca 224, 225; vomitoria 225 Illinois, Solidago ` gymnosper- moides in 287 Inga 331 Inuleae 232, 238 Inulopsis 232, 233; scaposa 231, 238 Ipomoea sagittata 225 Isoëtes macrospora 202, 211; Rhodora [Vol. 70 melanospora 394; muricata var. braunii 202, var. hesperia 203 Iva annua 451 Jaegeria hirta 231-233, 239 Johnson, Kermit L., Spiranthes ovalis, new for Kansas 485 Jones, Samuel B. Jr., Allium am- peloprasum in Alabama 303; An Example of a Vernonia Hybrid in a Disturbed Habitat 486-491 Judd, William W., Studies of the Byron Bog in Southwestern Ontario. XXX. Distribution of Orchids in the Bog 193-199 Juglans nigra 432 Juncus 221; albescens 103, 104; alpinus ssp. nodulosus 208; ar- ticulatus 225; balticus 208, var. littoralis 205; biflorus 225; biglumis 108, 104; dichotomus 225; georgianus 394; gerardi 442-444, 446, 448; stygius ssp. americanus 208, var. ameri- canus 206, 207; triglumis 104 Juniperus virginiana 399 Jussieua angulata 38; leptocarpa 38 Justicia 156, 333; Adhatoda 156; pectoralis 156, var. stenophylla 135-139, 143, 144, 155, 156, 333 Kalmia angustifolia var. liniana 291 Kanimia 231, 231, 236 Kansas, Spiranthes ovalis, new for 485 Knobloch, Irving W., A Case of Mistaken Identity, or the Mys- terious C. E. Lloyd 462-466 Kobresia bellardii 104; myosu- roides 103, 104 Lactuca 40; integrifolia 40; ludo- vicianus 431; oblongifolia 40; pulchella 40 caro- 233; strobilifera 1968] Lakela, Olga, Occurrence of Mur- dannia spirata in Florida 571- 574 Larix laricina 206, 207, 212 Lathyrus myrtifolius 40 Leguminosae 492-496; of Mexico — Faboideae. I. Sophoreae and Podalyrieae 492 Leiophyllum 290; buxifolium 290, 291, var. Hugeri 291, var. pros- tratum 290 Lemna minor 491 Lemna minor 171; trisulca 171 Lepachys 350, 354; columnaris 350, 369, var. pulcherrima 350, 376, var. tagetes 383; columni- fera 369, var. pulcherrima 376; mexicana 350, 355; peduncu- laris 350, 364, var. picta 350, 366; pinnata 350, 357; pinnati- fida 357; tagetes 350, 383 Leptolobium 499; dasycarpa 499 Lester, Donald T., Variation in Cone Morphology of Balsam Fir, Abies balsamea 83-94 Leucampyx 482 Leucopsis 232 Leymus 467, 470 Liatris 474, 478, 479, 482; belli- difolia 474; corymbosa 431, 474; graminifolia 40; odoratissima 474; paniculata 40, 474; tomen- tosa 474 Liliidae 426 Lilium andinum 39; Catesbaei 39; philadelphicum 39, var. andi- num 39; umbellatum 39 Limonium nashii 443 Linum bahamense 439, 440, Com- plex, The 439, f. bahamense 440, f. corallicola 440, var. bahamense 439, 440, var. bracei 439, 440, var. corallicola 439- 441; bracei 440; corallicola 441; lewisii 38, var. saxosum 39; lignosum 440; perenne 38, ssp. lewisii 38, var. saxosum 39, Index to Volume 70 591 var. americanum 38 Liquiritia 39, 40; lepidota 39 Lithospermum canescens 460; de- cumbens 31; incisum 31, 460; longiflorum 31 Litrisa 474, 475, 481 Lloyd, C. E., A Case of Mistaken Identity or the Mysterious 462 Lobularia maritima 299 Logue, John F., and John R. Bozeman, A Range Extension for Hudsonia ericoides in the Southeastern United States 289-292 Lolium perenne 557, 558; spp. 557, 558 Lomatium foeniculaceum 36 Lonicera villosa 210 Loteae 497, 498 Lotoideae 495, 497 Ludwigia leptocarpa 38; maritima 225 Lycopodium annotinum 216; caro- linianum 225; lucidulum 216; selago 186, 188 MacDonald, Robert D., Yellow- fruited Cornus florida — Culti- var or Form? 568-571 Macroule 518 Magnolia virginiana 290 Magnoliaceae 425 Magnoliatae 425 Magnoliidae 425 Magnoliophyta 422 Malouetia 336; Tamaquarina 327 Malpighiaceae 318, 321 Malus “Prince Georges" 296 Malva coccinea 41 Mamillaria vivipara 431 Mangaly, Jose K., A Cytotaxono- mic Study of the Herbaceous Species of Smilax: Section Coprosmanthus 55-102, 247- 278 Massachusetts, A Floristic Com- parison of Lower Cape Cod, 592 and the North Carolina Outer Banks 215; Seasonal Occur- rence and Ecology of Salt Marsh Phanerogams at Ipswich 442 Matelia carolinensis 37 Meehania cordata 431 Megapterium oklahomense 43 Melilotus officinale 451 Mentha arvensis var. villosa 210 Mentzelia decapetala 31; nuda 32; oligosperma 41 Marilãinen, Jouko, Najas minor AlL, in North America 161-175 Mexico, A White flowered Form of Bauhinia coulteri from Ta- maulipas 286; Leguminosae- Faboideae I. Sophoreae and Podalyrieae of 492 Microseris cuspidata 50 Mikania 231; campanulata 236; capricorni 236; conferta 237; pachylepis 237 Mimosa 126, 128, 287, 492; aca- cioides 125-128; peregrina 331 Mimosaceae 493 Mimosoideae 493, 495, 496 Mirabilis hirsuta 33; linearis 33; nyctaginea 32 Mitchell, Wm. W. and H. J. Hodg- son. Hybridization within the Triticeae of Alaska: A New X Elyhordeum and Comments 467-4738 Mitella nuda 210 Monotropa uniflora 485 Montana, New Records of Dis- junct Arctic-Alpine Plants in 108 Moraceae 333 Moronobaea spp. 122 Morton, Gary H., Chaff on the Receptacle of Solidago juncea 304-306 Muhlenbergia ` sobolifera Range Extension in Hampshire of 452 452, New Rhodora [Vol. 70 Murdannia nudiflora 573; spirata 511-573, Occurrence in Florida of 511 Murdy, W. H., Plant Speciation Associated with Granite Out- crop Communities of the South- eastern Piedmont 394-407 Musa 413; sapientum 416 Musaceae 412, 416 Musineon divaricatum var. hook- eri 49 Mutisieae 233, 239 Myrica pensylvanica 443, A New Form of 453, f. parvifolia 453, f. pensylvanica 453 Myriophyllum exalbescens 203 Myristica sebifera 152; theiodora 150, 152 Myristicaceae 129, 147 Myristicaceous Snuffs of the Northwest Amazon 113, 115, 116, 118-123, 129, 130, 132, 143, The Vegetal Ingredients of 113- 160 Myrospermum 498, 505; emar- ginatum 505; frutescens 501, 505; Pereirae 504; sect. Calusia 505; sect. Myroxylon 502; son- sonatense 504; toluiferum 503 Myroxylon 492, 493, 498, 502, 506; balsamum 502, var. balsamum 502, var. pereirae 501, 503; Pereirae 504; peruiferum 502; sonsonatense 504; toluifera 503, toluiferum 503 171, Najas flexilis 167, 170, 171, 203; gracilima 161-166; graminea 161; marina 167, 168, 170, 171; minor 161-171, in North Amer- ica 161; tenuissima 171, 172 Nardosmia corymbosa 534; frig- ida 534, var. corymbosa 534, var. palmatus 534; palmata 534; sagittata 534 1968] Neale, James J., Two Eupatorium Species New to Connecticut 455 Nebraska, Sporobolus texanus Vasey in Lincoln, 450 Nemexia 53, 56, 58, 247; bilt- moreana 58, 269; cerulea 249; ecirrhata 265; herbacea 249, var. melica 253; hugeri 58, 267; lasioneuron 253; nigra 249; tamnifolia 260; tenuis 253 New England, Developmental Variation of Cornus canadensis in Northern 278; Salt Marsh Vaucheriae 274 New Hampshire, Range Exten- sion of Muhlenbergia sobolifera in 452 New Hampshire, Wolffia in 491 North America, A Definitive New Book on the Flora of North- western 580 North America, Alyssum (Cruci- ferae) Introduced in 298; Najas minor in 161; The Spores of Four Species of Spinulose Wood Fern (Dryopteris) in Eastern 340 North Carolina, A Floristic Com- parison of Lower Cape Cod, Massachusetts and the Outer Banks of 215 Nothocrax urumutum 336 Notholaena lumholtzii 465 Nova Scotia, Contributions to the Flora of, IX. Habitat Studies of Arctic-Alpine and Boreal Disjunct Species 176-191 Nymphaea tetragona ssp. lei- bergii 210 Nyssa 217; sylvatica 224 Obeliscaria 350, 354; columnaris 369, var. pulcherrima 376; ped- uncularis 364; pinnata 357; pulcherrima 376; tagetes 383 Obelisteca 349, 354 Index to Volume 70 593 Oenanthe aquatica 171 Oenoplea 41; volubilis 41 Oenothera 66; acaulis 42; alata 43; albicaulis 42; cespitosa 42; macrocarpa 42, 43, var. incana 43, var. oklahomensis 43; mis- souriensis 42, var. incana 43; scapigera 42; serrulata 42; nut- tallii 42 Olmedia 121 Olmedioperebia sclerophylla 333 Ontario, Studies of the Byron Bog in Southwestern, XXX. Distri- bution of Orchids in the Bog 193-199 Opuntia fragilis 34, 431; mis- souriensis 33; polyacantha 33 Orchidaceae 412, 415 Ormosia 499, 518, 528; apulensis 519; chlorocalyx 519; coutinhoi 518; isthmensis 501, 518, 519; macrocalyx 501, 518, 519; mex- icana 516; schippii 501, 518, 520; toledoana 519 Orobus diffusus 43; dispar 43 Orthopappus 229; angustifolius 229, 231, 235 Ormosiopsis 518 Oxycoccus quadripetalus 206-208 Oxyria digyna 185 Oxytropis acuminata 44; bertii 44 lam- Palmae 414 Panicum 221; virgatum 442 Papilionaceae 496, 497 Papilionatae 497 Papilionoideae 495, 497 Parthenium hysterophorus 239 Passiflora lutea 47 Patrinia 521; sericea 523 Paullinia 322 Pellaea 1, 3, 21, 22; andromedi- folia 3-6, 8-15, 19-22; atropur- purea 3, 7, 8; Comparison of 232, 594 Sexual and Apogamous Races in the Fern Genus 1; inter- media 3, 5, 7, 11-16, 18, 20, 22; myrtillifolia 11, 14; ovata 8, 5, 7, 11-17, 22; rufa 11, 14; sagit- tata 3, 8, var. cordata 8, var. sagittata 8; sect. Pellaea 3; ternifolia 3, 11-13, 15-16, var. ternifolia 3, 22 Pemble, R. H. and S. A. Bamberg, New Records of Disjunct Arc- tic-Alpine Plants in Montana 103-112 Penstemon albidum 47; angusti- folium 46; bradburyi 46; caeru- leum 46; cristatum 46, 47; erianthera 46, 47; glaber 47; grandiflorum 46; teretiflorum 47 Perezia 233, 234; cubataensis 233, 240 Petalostemon candidum 47; viola- ceum 47 Petasites 533, 536, 538; arcticus 538; frigidus 533, 545-547, var. frigidus 533-535, 539, 548, 547, var. palmatus 533-548, var. nivalis 531-535, 538, 539, 543- 547; hybridus 533; hyperboreus 547; japonicus 533, ssp. gigan- teus 540; nivalis 547; palmatus 539, 542; sagittatus 533-540, 542-548; speciosus 537-539, 543, 545, 546; tatewakianus 540; trigonophyllus 535, 548; viti- folia 549; vitifolius 533, 535, 538, 539, 542-545, 547, 548, Evidence for the Hybrid Ori- gin of 533; X vitifolius 541, 548; warrenii 533, 535-548, Evi- dence for the Hybrid Origin of 533 Phacelia 398; dubia 397-399, 405, var. georgiana 397-399, 405; maculata 394, 897-399, 405; subg. Cosmanthus 398 Phalangium esculentum 44, 45; Rhodora [Vol. 70 quamash 44, 45 Phaseoleae 498 Phaseolus 492 Philodendron aff. bipinnatifidum 417; krebsii 417 Phlox drummondii 223 Phragmites communis 442, 443 Physalis 574, A New Species from the Galapagos Islands of 408, Emendations in United States 574; clarionensis 410; cordata 410, 575; galapagoénsis 409, 410; hederaefolia var. hederaefolia 574, var. puberula 574; missouriensis 575; phila- delphica var. immaculata 574; porrecta 410; pubescens 575, var. glabra 575, var. missouri- ensis 575, var. pubescens 575, 576; turbinata 575 Physostegia 35 Phytochemical Research, Some Impacts of Spruce's Explora- tions on Modem 313 Picea abies 562; mariana 206, 207, 210, 212, 562 Pickeringia 531; montana 514, 531, var. tomentosa 531 Pilostyles 426 Pinguicula vulgaris 184, 212 Pinus 224; Elliottii 573; palustris 290; rigida 220, 224; serotina 290; taeda 224 Piperaceae 126 Piptadenia 118, 125, 330, 331, 333; Niopo 331; peregrina 114, 117, 118, 125, 126, 128, 331, 333 Pitcairnia angustifolia 417 Pithecellobium 492 Plantago maritima 448, 444, 448 Pluchea suaveolens 231, 232, 238 Plumeria 417; alba 417 Poa arida 451; glaucantha 181 Podalyria mollis 45 Podalyrieae 497, 530 Podocoma 233; hirsuta 231, 232, 238 1968] Podocystis adriatica 292, 293, spathulata f. anastomosans 293 Two Naturally Occurring Ab- normalities of the Diatom 296; Pogonia ophioglossoides 196-198 Polygonaceae 410 Polygonella polygama 290 Polygonum 221; ‘aviculare var. littorale 443; viviparum 210 Polymnia 561, Notes on the Pedi- cillate Disk Florets in 559; can- adensis 559-561, A Reversion in 559; uvedalia 561 Polystichum acrostichoides 220; braunii var. purshii 182 Populus balsamifera 210 Portulaca pilosa 397; smallii 394, 397, 405 Potamogeton acutifolius 170; crispus 170, 171; friesii 203; gramineus 203; natans 170, 203; pectinatus 171; pusillus 171; trichoides 171; vaginatus 203; zosterifolius 203 Potentilla concinna 45; fruticosa 184, 185, 210; humifusa 45; mollis 45 Poulsenia 121 Prestonia 326-328; 324, 326-328 Prickothamnus 531 Primula misstassinica 186 Prunus caroliniana 225; susque- hanae 452 Pseudosophora 521, 522; sericea 523 Psoralea argophylla 48; cuspidata 47; esculenta 47; incana 48; macrorhiza 47 Pteris 2, 3, 8; biaurita 2; cretica 2; quadriaurita 2 Pterocarpus ateleia 506, 511, 512; amazonica gummifer 514; sect. Ateleia 506 Pterocaulon alopecuroides 231, 232, 238; undulatum 232 Purdy, Maud H. 1874-1965, 283 Index to Volume 70 595 Pycnanthemum dichotomum 45; pilosum 45 Pyrola secunda 210 Qualea 337 Quercus 221, 224; alba 224; fal- cata 224, 225; georgiana 394; ilicifolia 220, 224; incana 290; laevis 290; laurifolia 224; mar- garetta 290; nigra 224; phellos 224; stellata 225; velutina 224; virginiana 224, 225 Rafflesia 426 Ranunculus reptans 205 Ratibida 348-350, 353, 370, 371; A Monograph of the Genus 348; columnaris 369, f. denudata 369, 371, f. pulcherrima 376; columnifera 48, 348, 367, 369, 370, 376, 386-388, f. columnifera 349, 350, 355, 368, 369, 389, f. pulcherrima 349-351, 355, 376- 378, 389, var. appendiculata 369, 370, var. breviradiata 369, 370, var. pulcherrima 376, var. tubularis 369, 370; latipalearis 349-352, 355, 380, 381, 383, 386- 388; mexicana 348-352, 354-357, 386-389; peduncularis 348, 351, 362, var. peduncularis 349, 350, 352, 355, 363, 364, 367, 386-389, var. picta 348-350, 355, 365, 366, 386-389; picta 366; pinnata 348- 352, 354, 355, 357-359, 386-389; sulcata 369; tagetes 348-352, 355, 383-389, var. cinera 383 Retinophyllum 336; concolor 336; speciosum 336; truncatum 336 Reveal, James L., On the Names in Fraser’s 1813 Catalogue 25- 54 Rhamnus volubilis 41 Rheo 66 Rhexia virginica 176 Rhus radicans var, rydbergii 182 Rhynchospora alba 206-208; glo- 596 bularis var. typica 397; saxicola 394, 397, 405 Ribes aureum 48; hirtellum 182; longiflorum 48; triste 210 Richards, Edward Leon, A Mono- graph of the Genus Ratibida 348-393 Robinia coccinea 518 Rocky Mountain Carices, Notes on 419 Roegneria 467 Rogers, C. M., The Linum ba- hamense Complex 439-441 Rosa rugosa 223 Rosidae 425 Rourea 335 Rubiaceae 416 Rubus 221; acaulis 210 Rudbeckia 353, 359; amplexicaulis 353; columnaris 48, 369, 370; columnifera 48, 369, 370; glo- bosa 383; pinnata 350, 357; pur- purea var. serotina 49; tagetes 350 Rudd, Velva E., Leguminosae of Mexico — Faboideae. I. Soph- oreae and Podalyrieae 492-532 Rumex crispus 451 Ruppia maritima 442, 443, 446 Sabal minor 220, 225; palmetto 573 Saccharum brevibarbe 49 Sagittaria sp. 171 Salicornia 444, 447, 448; bigelovii 443, 444, 447, 448; europaea 448, 444, 447; rubra 451; stricta 447 Salix 210; bebbiana 210; brachy- carpa 204, 205, 208; maccalliana 210; myrtillifolia 210; pellita 210; planifolia 204, 205, 208, 210; silicicola 208, 209; turnorii 205, 208 Salt Marsh Phanerogams at Ip- swich, Massachusetts, Seasonal Occurrence and Ecology of 442 Rhodora [Vol. 70 Salvia 287; 465 Sambucus pubens 182 Sanguinaria canadensis 216 Sapindaceae 322 Sarracenia purpurea 206-208 Saskatchewan, Contribution to the Flora of Boreal 200 Satureja arkansana 431; vulgaris 179 Saxifraga aizoides 184, 185; aizoon var. neogaea 180-182; tolmiei 107 Scheuchzeria palustris var. amer- icana 206 Schizachne purpurascens 184 Schizaea pusilla 216 Schultes, Richard Evans, Some Impacts of Spruce’s Amazon Explorations on Modern Phy- tochemical Research 313-339; and Bo Holmstead, De Plantis Toxicariis e Mundo Tropicale Commentationes II. The Veg- etal Ingredients of the Myris- ticaceous Snuffs of the North- west Amazon 113-160 Scilla esculenta 44, 45 Scirpus 221; americanus 442-444; cespitosus var. delicatulus 180, 184, 186; hudsonianus 207; lacustris 170 Scrophulariaceae 396 Sedum 399, 401, 403; pusillum, 394, 399-401, 403, 405; rosea 182, 184, 185; smallii 399-401, 405; stellatum 401 Selaginella arenicola ssp. acan- thonota 290; selaginoides 210, 212 Senecio 233; adamantinus 231, 239; brasiliensis 239; erisithali- folius 239 Senecioneae 233, 239 Seseli divaricatum 49; 49 Sicyos 50; angulatus 50 elegans sonoriensis lucidum 1968] Sideranthus 49, 370; integrifolius 49; pinnatifidus 50 Silene acaulis 206, 209, 211, ssp. acaulis var. exscapa 209, f. athabascensis 209; antirrhina 452; menziesii 210 Smilacina trifolia 206 Smilax 55-57, 62, 72, 73, 77, 78, 247; acirrha 265; auriculata 225; balbisiana 271; bilt- moreana 58, 59, 61, 65, 66, '70, 73, 74, 76, 77, 79, 84, 248, 254, 266, 268-270; blancoi 271; bona- nox 225; china 57; cordifolia 271; diversifolia 58, 253; ecir- rhata 58, 59, 65, 68-72, 74-81, 247, 248, 263-266, 268, 270, var. biltmoreana 270, var. hugeri 268; herbacea 56, 57, 59-63, 67- 74, 77-79, 248-250, 254, 257, 258, 261, 265, 271, ssp. crispifolia 249, var. crispifolia 58, var. ecirrhata 58, 265, var. inodora 253, var. lasioneuron 253, var. peduneularis 249, var. pulveru- lenta 257, var. Simsii 249; his- pida 77, 271; hugeri 58, 59, 61, 64-66, 70, 71, 74, 75, 77-79, 81, 248, 254, 266-268, 270; illinoen- sis 61, 62, 64, 68, 70-72, 75, 77, 79-81, 248, 263; inermis 260; lanceolata 260; lasioneuron 57- 59, 61. 63, 68, 70, 71, 73, 75, 77, 79, 248, 250, 253, 254, 257, 258, 263; leptanthera 58, 260; macrophylla 72; moranensis 271; peduncularis 56, 249; pseudo-china 57, 59, 61, 62, 64, 67, 70, 71, 73, 74, 77, 79, 248, 250, 260-262, 271; pulverulenta 56, 57, 59, 61, 62, 64, 70, 71, 74, 75, 17-79, 248, 250, 257, 258; riparia 78; salicifolia 271; sect. China 57, 247; sect. Coprosman- thus 55-58, 60, 63, 72, 73, 77, 78, 247, A Cytotaxonomic Study of the Herbaceous Species of Index to Volume 70 597 55, 247; sect. Eusmilax 57; sect. Macranthae 57, 247; sect. Nemexia 57, 247; sect. Pleio- smilax 56, 57, 247; sect. Smilax 57, 247; subg. Nemexia 247; tamnifolia 57, 260, 261; tel- feriana 271; tenuis 58, 253; zeylanica 271 Smith, E. C., and R. W. Hounsell, Contributions to the Flora of Nova Scotia. IX. Habitat Stud- ies of Arctic- Alpine and Boreal Disjunct Species 176-191 Smith, Edwin B., A Reversion in Polymnia canadensis (Com- positae) and Notes on the Pedi- cellate Disk Florets in the Genus 559-562 Solidago 221, 304; chilensis 231, 232, 238; flexicaulis 179; gym- nospermoides 287, in Illinois 287; juncea 305, Chaff on the Receptacle of 305; media 287; multiradiata 184; paucifloscu- losa 290; sempervirens 443, 444 Sonchus 40; ludovicianus 431; pulchellus 40 Sophoreae 497, 498 Sophora 498, 499, 521-523, 527, 530; affinis 530; alopecuroides 521, 522; arenicola 526; ari- zonica 530; conzattii 523, 524, 527; formosa 530; havanensis 526; japonica 521, 530; littoralis 526; nuttalliana 29, 528, 524; occidentalis 526; purpusii 523- 525; sericea 521, 523; sect. Eusophora 521, 522; sect. Pseu- dosophora 521, 522; secundi- flora 523, 524, f. xanthosperma 528; sempervirens 528; speciosa 528; tomentosa 521-524, 526, B truncata 526, 8 var. littoralis 526 i, Spartina alterniflora var. glabra 442, 444, 445; cynosuroides 225; 598 patens 275, 276, 442-445, 448; pectinata 442 Specularia perfoliata 452 Spergularia marina 443, 444 Sphaeralcea coccinea 41 Sphagnum 193, 195-198, 206-208; centrale 207; recurvatum 206; spp. 206, 207; teres 207; warn- storfianum 207 Sphenopholis intermedia 187 Spiranthes cernua 197, 198; ovalis 485, New for Kansas 485; ro- manzoffiana 206, 212 Spirodela polyrhiza 171 Sporobolus texanus 450, 451, in Lincoln, Nebraska 450 Stachys foeniculum 38 Stellaria arenicola 204, 205; caly- cantha 210; longifolia 210 Stevia 231; conmixta 237; decus- sata 237; veronicae 231, 237 Stipulicida setacea 290 Stratiotes aloides 170 Strychnos 335 Stuckey, Ronald L., Biography of Thomas Nuttall; A Review with Bibliography 429-438 Stylophorum diphyllum 432 Stylosanthes racemosa 49 Styphnolobium 521, 522; japoni- cum 530 Suaeda 448; depressa 451; lin- earis 443, 444, 448; maritima 443, 448; novaezelandiae 448; richii 225, 443, 444, 448 Subularia aquatica 202, ssp. americana 211, ssp. aquatica 211 Surirella 292 Svenson, Henry K. Maud H. Purdy 1874-1965, 283-286 Sycios 50; angulata 50 Sycopsis 425 Symphonia spp. 122 Symphyopappus 231, 233; cunea- tus 231, 237; polystachyus 231, 237 Rhodora [Vol. 70 Swartzia multijuga 514 Sweetia 499; fruticosa 499; pan- amensis 500, 501, 527 Tagetes patula 48 Talinum 402; mengesii 399, 401- 403, 406; teretifolium 399, 401- 403, 405 Tanacetum huronense var. floc- cosum 204 Taraxacum officinale 179, 559 Taxodium 217 Teeri, James A., Developmental Variability of Cornus canaden- sis in Northern New England 278-282 Terrell, Edward E., Biometric and Taxonomic Uses of Cellulose Acetate Plastic 552-558; Notes on Festuca arundinacea and F. pratensis in the United States 564-568 Tetracentron 426 Tetrapteris 330; methystica 329, 330 Thalictrum polygamum 179, 184 Thalesia 499 Thelypteris 2 Theobroma subincanum 122 Thermia rhombifolia 34 Thermopsis 531; mollis 45; rhom- bifolia 34 Thuraria 51; herbacea 50 Tillandsia usneoides 225 Toluifera 502; balsamum 502, 503, var. 8 Pereirae 504; Pereirae 504 Torreyochloa pallida var. fer- naldii 205 Toulichiba 518 Trachelospermum difforme 36 Tradescantia rosea var. graminea 290 Trapa natans 171 Trichilia sp. 130 Trichogonia 233; 237 gardneri 231, 1968] Tridax 241, 245; accedens 241, 242, 245, 246; dubia 242, 245, 246; oligantha 241, 242, 245, 246; pringlei 241 Trifolieae 497, 498 Trifolium arvense 443 Triglochin maritima 442 Trilisa 474-476, 480-482; carnosa 475; odoratissima 474-478, 480, 481; paniculata 474-482 Triticeae 467, 470, 471; of Alaska, Hybridization within the 467 Trixis 233, 234; divaricata 233, 240; pinnatifida 231, 233, 240 Trochodendrales 425, 426 Trollius laxa 285 Troximum ciliatum 50; cuspi- datum 50; glaucum 50; margin- atum 50 Tryon, Alice F., Comparison of Sexual and Apogamous Races in the Fern Genus Pellaea 1-24 Typha angustifolia 171; latifolia 171 Ulmus 295; A Case of Albinism and of Presumptive Self Com- patibility in 294; alata 295; americana 295; carpinifolia 295; glabra 295; X hollandica 295; var. Klemmer 295; ja- ponica 295; laevis 295; thomasii 295; wilsonii 295 Ungar, Irwin A., Sporobolus tex- anus Vasey in Lincoln, Nebras- ka 450-451 Uniola latifolia 51 United States, A Range Exten- sion for Hudsonia ericoides in the Southeastern 289, Emenda- tions in Physalis in the 574, Notes on Festuca arundinacea and F. pratensis in the 564 Utricularia intermedia 208; sp. 208 291; Vaccinium corymbosum Index to Volume 70 599 crassifolium 290 Vallisneria americana 171 Vaucheria 274; arcassonensis 274- 276; compacta 274, 276, var. koksoakensis 275, 276; coronata 274; intermedia 274-276; litorea 274; minuta 274; piloboloides 274, var. compacta 276; thuretii 274; vipera 274 Vaucheriae, New Marsh 274 Verbesina 95, 96, 99; A Cytotax- onomie Study in 95; alternifolia 96, 98-102; aristata 96-102; en- celioides 100, 101; glabrata 239; hybrids 97-99, 101; lindenii 96, 100; longifolia 96, 98-102; oc- cidentalis 96-98, 100-102; oligo- cephala 96, 100, 101; rothrockii 96-102; sect. Lipactinia 95; sect. Ochractinia 100; sect. Pterophyton 96-100; sect. Sau- benetia 100; sect. Sonoricola 100; sect. Verbesinaria 97-100; sect. Ximenesia 100; serrata 96, 100; virginica 96, 100, 101; warei 96-98, 100, 102 Verbesininae 241 Vermont, Alisma gramineum in 577 Vernonia 229, 486; altissima 486- 490; angustifolia 486-490; cog- nata 235; diffusa 235; fascicu- lata 486; hybrid 486-490, An Example in a Disturbed Habi- tat of a 486; X illinoensis 486; missurica 486; polyanthes 231, 235; scirpoides 235 Vernonieae 229, 235 Vexibia 521; sericea 523 Vexillifera 515; micranthera 515 Vibexia 521; sericea 523 Viburnum dentatum 288, 289; recognitum 288, 289, The Diag- nostie Value of Fresh Mature Drupes of 288, var. recognitum 288 England Salt rm, 600 Vicia 496, americana 43 Vicieae 497 Viguiera porteri 394 Vincetoxicum acanthocarpus 37 Virgilia bicolor 51; fimbriata 51; grandiflora 5l; secundiflora 528 Virola 118-122, 127-129, 136-139, 144, 145, 147, 338, 335; calo- phylla 119, 130, 147, 148, 150, 333; calophylloides 119, 130, 132, 149, 150, 333; elongata 152; theiodora 130, 132-135, 140, 141, 145, 146, 150-152, 332 Vitaceae 410, 415 Vitis campestris 51, riparia 52 Vriesia sintenisii 417 Walker, James W., The Evolution and Classification of Flowering Plants by Arthur Cronquist (Book Review) 422-428 Waterfall, U. T., Emendations in United States Physalis 574- 576; A New Species of Physalis from the Galapagos Islands 408-409 Webber, E. E., New England Salt Marsh Vaucheriae 274-277; Seasonal Occurrence and Ecol- ogy of Salt Marsh Phanero- gams at Ipswich, Massachu- setts 422-450 41 Rhodora [Vol. 70 Wedelia 233; pilosa 231, 239; subvelutina 239 Wells, Theodore W., A New Form of Myrica pensylvanica 453 Wilbur, Robert L., The Status of Hedyotis procumbens var. hir- suta (Rubiaceae) 306-311 Winteraceae 427 Wolffia in New Hampshire 491; columbiana 491; punctata 491 Wolffia punctata 171 Woodsia 346; alpina 184; glabella 182, 184; ilvensis 180-182, 187, 188 Woodwardia areolata 176 Wunderlin, R. P., A White Flow- ered Form of Bauhinia coulteri from Tamaulipas, Mexico 286- 287; Solidago gymnosper- moides in Illinois 287-288 Xylothermia 531; montana 531, Ssp. tomentosa 531 Yucca aloifolia 225; angustifolia 52; glauca 52; gloriosa 52, 225 Zanthoxylum clava-herculis 225 Zanthyrsis 521, 522; paniculata 521, 526 Zingiberaceae 412, 413 Zygadenus elegans 210