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 ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. Associate Editors VOLUME 69 1967 The New England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. RY ^ LIBRA 7 APR 14 1967 3 Hodova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Clnb, by EARLOW REFERENCE ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL , STUART KIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER gon om Editors ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. Vol. 69 January-March, 1967 No. 777 CONTENTS: Diploid Dryopteris dilatata from Quebec CORALE M ESO ay LY OBERE RM nice ee 1 Taxonomic Fern Notes. V. New Combinations in Peruvian Species of Thelypteris Rolla Tryon ................ m 5 A Reconsideration of Sidopsis Rydberg and Notes on Mal- vastrum A. Gray (Malvaceae) David M. Bates ................ 9 Notes on the Flora of Coós County, New Hampshire, II BAUM TE Harre a E iiiv RR RED RID 29 Studies in the Eupatorieae, (Compositae) I-III BONO saw UM EL idi HEEL ME 35 (Continued on Inside Cover) The Nem England Botanical Club, Ine. IE a An c Botanical Museum, Oxford 5t.. Cambridge 38, Mass. CONTENTS: — continued The Identity of Psoralea multijuga Ell. (Leguminosae) Robert L. Wilbur |... ener rnnt nnne nennen enne 48 Southeastern Limit of Fagus grandifolia Daniel B. Ward 51 Flagellates and Desmids New to Massachusetts Edgar E. Webber ...... eene 55 Stipulicida in Gray’s Manual Range A. M. Harvill, Jr. .... 59 Additions to the Flora of the Bimini Island Group, Bahama . Islands William R. Stimson eene 60 Allium ampeloprasum in Mississippi Thomas M. Pullen .... 61 An Annotated Bibliography of Mexican Ferns (Review) Ramon Rida eene eene nnne nenne eerte nennen 62 News Announcement Stanwyn G. Shetler ................. 63 New Taxa in the Genus Alsophila Ramon Riba .................. 65 A Correction Frank C. Sey Mow ....... eene 69 Indications of Possible Mid-Cenozoic Hybridization in the Aspens of the Columbia Plateau Burton V. Barnes ........ 70 Physalis in Mexico, Central America and the West Indies U. T. Waterfall... eee nnne 82 RHODORA.—A quarterly journal of botany, devoted primarily to the flora of North America and floristically related areas. Price, $6.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) $1.80. Back vol- umes 1-58, with a few incomplete, can be supplied at $5.00 per volume. Volunie 59— available at $6.00. Somewhat reduced rates for complete sets can be obtained upon application. Scientific papers and notes, relating directly or indirectly to the plants of North America, will be considered by the editorial com- mittee for publication. Articles concerned with systematic botany and cytotaxonomy in their broader implications are equally accept- able. All manuscripts should be double-spaced throughout, Please conform to the style of recent issues of the journal. Illustrations can be used only if the cost of engraver's blocks is met through the author or his institution. Forms may be closed five weeks in advance of publication. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Albion R. Hodgdon, Dept. of Botany, Nesmith Hall, University of New Hampshire, Durham, New Hampshire. Subscriptions and orders for back issues (making all remittances payable to RHODORA) should be sent to RHODORA, Botanical Museum, Oxford Street, Cambridge, Mass. 02138. Second Class Postage Paid at Boston, Mass. Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 69 January-March No. 777 DIPLOID DRYOPTERIS DILATATA FROM QUEBEC DONALD M. BRITTON In eastern North America, ferns allied to Dryopteris dilatata (Hoffm.) A. Gray include D. spinulosa (O. F. Muell.) Watt (4X), D. intermedia (Muhl.) A. Gray (2X), D. dilatata (Hoffm.) A. Gray (2X) from the Lake Superior basin, and D. campyloptera Clarkson (4X) from the Ap- palachians,. The last two are referrable to D. spinulosa var. americana (Fisch.) Fernald of Gray's Manual (Fernald, 1950). The distribution of D. campyloptera has been con- sidered to be from high elevations in the mountains of North Carolina and Virginia in the south, north to Maine, Nova Scotia, the Gaspé Peninsula, Newfoundland, Labra- dor, Ungava (Fort Chimo) (Dutilly, Lepage and Duman, 1953), south along the east coast of Hudson's Bay and James Bay (Dutilly, Lepage and Duman, 1958; Dutilly and Lepage, 1963) to Mt. Sherrick (Potter, 1934), and Lake Kelvin (Dutilly and Lepage, 1963). Also, there have been a few collections in interior Quebec (Dutilly and Lepage, 1964). Cytological determinations of chromosome number have been made by Wagner (1963) from material collected at Mountain Lake, Virginia; Walker (personal communica- tion) from material collected in Rutland Co., Vermont and Quebec Co., Quebec; Britton (1962) from Mt. Washington, N. H. and Britton (unpublished) from Yarmouth, Nova Scotia. These have all been tetraploid. Diploid specimens are now well known from the Lake 1 2 Rhodora [Vol. 69 Superior basin. I have studied cytologically nearly 100 plants from the Pigeon River on the Minnesota border of Ontario to Batchawana River which is approximately 50 miles north of Sault Ste. Marie, Ontario. The diploid taxon in Ontario was mapped by Soper (Britton and Soper, 1966). Wagner and Hagenah (1962) refer to this taxon as "Lake Superior ‘D. dilatata’”’ and suggest that it may be a dis- tinct species, whereas I consider it to be conspecific with D. dilatata of British Columbia (Britton, 1962), which in turn is probably conspecific with the diploid D. dilatata of Great Britain, called D. assimilis S. Walker (Walker, 1961 ; Walker and Jermy, 1964). The discovery in the Clay Belt of Quebec, near Amos, of a member of this alliance was reported by Morton in Baldwin et al. (1962). These were named by Morton, D. dilatata ssp. campyloptera (Baldwin, 1962; Morton, 1965). Through the kindness of Father André Asselin of Amos College, I was sent in 1965 two plants from Lac Beaudoin near Amos. These were both diploids, not tetraploid as one might expect from the published ranges given above. To study more plants of this taxon and to search for the tetra- ploid taxon, Dr. Hugh Dale and the author visited Amos in June. Under the able guidance of Father Asselin, we col- lected specimens from Lac Beaudoin, Lac Beauchamp and Lac Legendre. Twelve specimens of D. dilatata were studied cytologically and were diploid (»—41). No tetraploids were found, and no hybrids of D. dilatata were discovered, al- though D. spinulosa, D. intermedia and D. X triploidea were also abundant in the area. Wagner and Hagenah (1962) have published a key to separate the diploid taxon from the tetraploid, but it is quite unsatisfactory for my material. According to them, the diploid has an average exposure length of 33-37 microns, whereas the range for averages of different tetraploids is 37-40 microns. Ten collections of the diploid from Lake Superior studied by this author had average values from 36-41 (average of lengths of 20 exospores in Permount) whereas the tetraploids from New Hampshire and Nova Scotia varied from 37-41. There was a cluster of values 1967] Dryopteris — Britton 3 for different diploids around 37 and 38 microns, and one specimen was as high as 41.4. One tetraploid was as low as 37.4. Wagner (1963) reported a tetraploid which was as low as 35.9. Accordingly, it is impossible to separate the diploid from the tetraploid by average spore lengths alone. The other characteristics in the key have also failed to separate the diploids from the tetraploids, Out of the first 30 collections of the diploid laid out at random on a lab- oratory bench, 17 were not “ovate and very broad", but matched the shape of some of the tetraploids. Nine did not have a petiole to midrib ratio approaching one to one ("average 95%, range 75-115°%” Wagner and Hagenah, 1962). Sixteen did not have approximate to overlapping segments, and the petiole scales were variable in most. I know of no simple way to distinguish between the two taxa other than by chromosome number. It will require further cytological work to determine the ploidy of the Quebec, Labrador, Gaspé and Newfoundland plants which have been referred to as D. spinulosa var. americana (Fisch.) Fernald. However, it is clear that the diploid which is allied to D. dilatata and which is so abun- dant along the north shore of Lake Superior, extends nearly 300 miles to the east. This is less than 100 miles south of Lake Kelvin, and is approximately 300 miles W. N. W. of Quebec Co., Quebec where the nearest tetraploid was col- lected. It may well be, that it is the diploid which extends up the coast of James Bay to the tree line at Fort Chimo in Ungava. If this is the case, the tetraploid may be the less common of the two taxa in Quebec. A further possibil- ity is the occurrence of a region in Quebec where the diploid and tetraploid occur together. The specimens collected by Morton (Baldwin et. al., 1962; Morton, 1965) in the Clay Belt and ascribed to sub-species campyloptera (Clarkson) Morton should be identified as Dryopteris dilatata. DEPARTMENT OF BOTANY, UNIVERSITY OF GUELPH, GUELPH, ONTARIO, CANADA, 4 Rhodora [Vol. 69 LITERATURE CITED BALDWIN, W. K. W., et al. 1962. Report on Botanical Excursion to the Boreal Forest Region in Northern Quebec and Ontario. Natl. Mus. of Canada, Dept. of Northern Affairs and National Re- sources, Canada. BRITTON, D. M. 1962. Dryopteris dilatata (Hoffm.) A. Gray in North America. Rhodora 64: 207-212. and J. H. Soper. 1966. The cytology and distribu- tion of Dryopteris species in Ontario. Canad. Jour. Bot. 44: 63-18. DuTILLY, A., and E. LEPAGE. 1963. Contribution à la flore du versant sud de la Baie James, Québec-Ontario. Contr. of the Arctic Inst., the Catholie Univ. of America, Washington, D.C. No. 12F, 1-199. 1964. Randonnée botanique à travers la péninsule Québec-Labrador. Le Naturaliste Canadien 91: 197-240. , and M. DuMAN. 1953. Contribution à la Flore du Bas- sin de la Baie d'Ungava. Contr. of the Arctic Inst., the Catholic Univ. of America. No. 4F, 1-104. 1958. Contribution à la flore des îles (T.N.O.)et du versant oriental (Qué.) de la Baie James. Contr. of the Arctic Inst., the Catholic Univ. of America, No. 9F, 1-199. FERNALD, M. L. 1950. Gray's Manual of Botany, American Book Co., New York. MonTON, C. V. 1965. Recent Fern Literature. Am. Fern Jour. 55: 85-89. PorrER, D. 1934. Plants collected in the southern region of James Bay. Rhodora 36: 274-284. WAGNER, W. H. Jr. 1963. Pteridophytes of the Mountain Lake Area, Giles County, Virginia, including notes from Whitetop Mountain. Castanea 28: 113-150. and D. J. HAGENAH. 1962. Dryopteris in the Huron Mountain Club area of Michigan. Brittonia 14: 90-100. WALKER, S. 1961. Cytogenetie studies in the Dryopteris spinulosa complex. II. Am. Jour. Bot. 48: 607-614. and A. C. JERMY. 1964. Dryopteris assimilis S. Walker in Britain. Brit, Fern Gaz. 9(5): 137-140. TAXONOMIC FERN NOTES. V. NEW COMBINATIONS IN PERUVIAN SPECIES OF THELYPTERIS. ROLLA TRYON The following new combinations are made in order to have suitable names for the species of Thelypteris in Peru. Many of the Andean species of the genus that are widely distributed in tropical America have been transferred to Thelypteris by, among others, Ching, Morton, and Proctor. However, few of the strictly Andean species that properly belong in Thelypteris have been placed there, In addition to the basionym, the name in Dryopteris and sometimes other synonyms are given for purposes of reference or ex- planation. Thelypteris aspidioides (Willd.), comb. nov. Ceterach aspidioides Willd. Sp. Pl. 5: 137. 1810. Dryopteris aspidioides (Willd.) C.Chr. Ind. Fil. 253. 1905. Thelypteris assurgens (Maxon), comb. nov. Dryopteris assurgens Maxon, Jour. Wash. Acad. Sci. 34: 24. 1944. Thelypteris Bangii (C.Chr.), comb. nov. Dryopteris Bangii C.Chr. Danske Vid. Selsk. Skr. VII, 4: 333. 1907. Thelypteris biformata (Rosenst.), comb. nov. Dryopteris biformata Rosenst. Fedde Repert. 7: 300. 1909. Thelypteris boqueronensis (Hieron.), comb. nov. Dryopteris boqueronensis Hieron. Hedwigia 46: 329. 1907. Thelypteris coarctata (Kze.), comb. nov. Aspidium coarctatum Kze. Bot. Zeit. 1845: 287. Dryopteris coarctata (Kze.) C.Chr. Ind. Fil. 258. 1905. Thelypteris consobrina (Maxon & Morton), comb. nov. Dryopteris consobrina Maxon & Morton, Bull. Torrey Club 65: 356. 1938. Thelypteris deflexa (Presl), comb. nov. Nephrodium deflexa Presl, Rel. Haenk. 1: 36. 1825. Dryopteris Lindigii C.Chr. Ind. Fil. 275. 1905, nom. mov. for N. deflexum Presl; not Dryopteris deflexa (Kaulf.) C.Chr. 1905. Thelypteris Lindigii (C.Chr.) Alston, Jour. Wash. Acad. Sci. 48: 283. 1958. Thelypteris densa (Maxon), comb. nov. Dryopteris densa Maxon, Jour. Wash. Acad. Sci. 34: 25. 1944. Thelypteris dumetorum (Maxon), comb. nov. Dryopteris dumetorum Maxon, Jour. Wash. Acad, Sci. 34: 26. 1944. 5 6 Rhodora [Vol. 69 Thelypteris ensiformis (C.Chr.), comb. nov. Dryopteris ensiformis C.Chr. Danske Vid. Selsk. Skr. VII, 10: 269. 1913. Thelypteris falcata (Liebm.), comb. nov. Meniscium falcatum Liebm. Danske Vid. Selsk. Skr. V, 1: 183. 1849, Dryopteris faleata (Liebm.) C.Chr. Danske Vid. Selsk. Skr. VII, 10: 270. 1913; not Dryopteris falcata (L.f.) O.Ktze. 1891. Dryopteris Jurgensenii (Fée) Maxon & Morton, Bull. Torrey Club. 65: 360. 1938. Thelypteris furva (Maxon), comb. nov. Dryopteris furva Maxon, Jour. Wash. Acad. Sci. 34: 24. 1944. 'Thelypteris gigantea (Mett.), comb. nov. Meniscium giganteum Mett. Fil. Lechler. 1: 19. 1856. Dryopteris gigantea (Mett.) C.Chr. Ind. Fil. 267. 1905; not Dryop- teris gigantea. (Bl.) O.Ktze. 1891. Dryopteris simplicifrons C.Chr. Ind. Fil. 486. 1906, nom. nov. for M. giganteum Mett. Thelypteris glandulosolanosa (C.Chr.), comb. nov. Dryopteris glandulosa-lanosa C.Chr. Dansk. Bot. Ark. 9 (3): 6l. 1937. Thelypteris Herzogii ( Rosenst.), comb. nov. Dryopteris Herzogii Rosenst. Meded. Rijks Herb. Leiden 19: 15. 1913. Thelypteris Jamesonii (Hook.), comb. nov. Nephrodium Jamesonii Hook. Sp. Fil. 4: 66. 1862. Dryopteris Jamesonii (Hook.) C.Chr. Danske Vid. Selsk. Skr. VII, 10: 227. 1913. Thelypteris laevigata (Kuhn), comb. nov. Phegopteris laevigata Mett. ex Kuhn, Linnaea 36: 112. 1869. Dryopteris laevigata (Kuhn) C.Chr. Ind. Fil. 273. 1905. Thelypteris Leprieurii (Hook.), comb. nov. Nephrodium Leprieurii Hook. Sp. Fil. 4: 106. 1862. Dryopteris Leprieurii (Hook.) O.Ktze. Rev. Gen. 2: 813. 1891. Thelypteris leucothrix (C.Chr.), comb. nov. Dryopteris leucothrix C.Chr. Smiths. Misc. Coll. 52: 377. 1909. Thelypteris Linkiana (Presl), comb. nov. Grammitis Linkiana Presl, Tent. Pterid. 209. 1836, based on the description of Gymnogramma polypodioides by Link, Hort. Bot. Berol. 2: 50. 1833; not Gymnogramma polypodioides (Raddi) Spreng. 1827. Gymnogramma diplazioides Desv. Mém. Soc. Linn. Paris 6: 214. 1827, as diplazoides. Dryopteris diplazioides (Desv.) Urban, Symb. Antill. 4: 21. 1903; not Dryopteris diplazioides (Mett.) O.Ktze. 1891. Dryopteris Linkiana (Presl) Maxon, Jour. Wash. Acad. Sci. 14: 199. 1924. 1967] Fern Notes — Tryon T Thelypteris diplazioides (Desv.) Proctor, Bull. Inst. Jamaica Sci. Ser. 5: 59. 1953; not Thelypteris diplazioides (Moritz ex Mett.) Ching, 1936. Thelypteris longifolia (Desv.), comb. nov. Meniscium longifolium Desv. Mém. Soc. Linn. Paris 6: 223. 1827. Dryopteris longifolia (Fée) Hieron. Hedwigia 46: 351. 1907, illegit. Dryopteris Desvauxii Maxon & Merton, Bull. Torrey Club 65: 369. 1938, nom. nov. for M. longifolium Desv. Thelypteris lugubriformis (Rosenst.), comb. nov. Dryopteris lugubriformis Rosenst. Fedde Repert. 7: 299. 1909. Thelypteris Macbridei (Maxon), comb. nov. Dryopteris Macbridei Maxon, Jour. Wash. Acad. Sci. 34: 25. 1944. Thelypteris macrotis (Hook.), comb, nov. Nephrodium macrotis Hook. Sp. Fil. 4: 86. 1862. Dryopteris macrotis (Hook.) O.Ktze. Rev. Gen. 2: 813. 1891. Thelypteris membranacea (Mett.), comb, nov. Phegopteris membranacea Mett. Fil. Lechler. 2: 22. 1859. Dryopteris membranacea (Mett.) C.Chr. Ind. Fil. Suppl. 35. 1913. Thelypteris mollis (Mett.), comb. nov. Phegopteris mollis Mett. Ann. Sci. Nat. V, 2: 242. 1864. Dryopteris permollis Maxon & Morton, Bull. Torrey Club 65: 372. 1938, nom. nov. for P. mollis Mett.; not Dryopteris mollis (Jacq.) Hieron. 1907. Thelypteris nemoralis (Sod.), comb. nov. Nephrodium nemorale Sod. Crypt. Vasc. Quit. 267. 1893. Dryopteris nemoralis (Sod.) C.Chr. Ind. Fil. 279. 1905. Thelypteris nervosa (Kl.), comb. nov. Polypodium nervosum Kl. Linnaea 20: 386. 1847; not Polypodium nervosum Boj. Hort. Maurit. 419. 1837, nomen nudum. Dryopteris nervosa (Kl.) C.Chr. Ind. Fil. 279. 1905. Thelypteris nitens (Desv.), comb. nov. Polypodium nitens Desv. Mém. Soc. Linn. Paris 6: 240. 1827. Dryopteris nitens (Desv.) C.Chr. Danske Vid. Selsk. Skr. VII, 10: 142. 1918; Thelypteris Pavoniana (Kl.), comb. nov. Polypodium Pavonianum Kl. Linnaea 20: 386. 1847. Dryopteris Pavoniana (Kl.) C.Chr. Ind. Fil. 283. 1905. Thelypteris peruviana (Rosenst.), comb. nov. Dryopteris peruviana Rosenst. Fedde Repert. 7: 298. 1909. Thelypteris phacelothrix (C.Chr. & Rosenst.), comb. nov. Dryopteris phacelothrix C.Chr. & Rosenst. Fedde Repert. 11: 56. 1912. Thelypteris pilosula (Mett.), comb. nov. Aspidium pilosulum Mett. Fil. Hort. Bot. Lips. 130. 1856, based on the description of Aspidium lasiesthes by Mett. op. cit. 90; not \Aspi- dium. lasiesthes Kze. 1850; not Aspidium pilosulum Wall. Cat. no. 337. 1829, nomen nudum. 8 Rhodora [Vol. 69 Dryopteris pilosula Hieron. Hedwigia 46: 332. 1907. Dryopteris pilosula (Mett.) C.Chr. Danske Vid. Selsk. Skr. VII, 4: 277. 1907. The earlier publication of the name Aspidium pilosulum by Kunze, Linnaea 23: 229. 1850, was without description. Accordingly, the “new combination" by Hieronymus in Dryopteris, with description, is actually a new name for a new species, Christensen's acceptable com- bination, then, becomes a later homonym. Thelypteris pteroidea (Kl.), comb. nov. Polypodium pteroideum Kl. Linnaea 20: 389. 1847. Dryopteris pteroidea (Kl.) C.Chr. Ind. Fil. 287. 1905. Thelypteris Rolandii (C.Chr.), comb. nov. Dryopteris Rolandii C.Chr. Danske Vid. Selsk. Skr. VII, 10: 258. 1913. Thelypteris Rosei (Maxon), comb. nov. Dryopteris Rosei Maxon, Smiths. Misc. Coll. 65 (8): 10. 1915. Thelypteris Rosenstockii (C.Chr.), comb. nov. Dryopteris Rosenstockii C.Chr. Danske Vid. Selks. Skr. VII, 4: 304. 1907. Thelypteris Rusbyi (C.Chr.), comb. nov. Dryopteris Rusbyi C.Chr. Smiths. Misc. Coll. 52: 390. 1909. Thelypteris Standleyi (Maxon & Morton), comb. nov. Dryopteris Standleyi Maxon & Morton, Bull. Torrey Club. 65: 368. 1938. Thelypteris subandina (C.Chr. & Rosenst.), comb. nov. Dryopteris subandina C.Chr. & Rosenst. Fedde Repert. 12: 472. 1913. Thelypteris tristis (Kze.), comb. nov. Polypodium triste Kze. Linnaea 9: 47. 1834. Dryopteris tristis (Kze.) O.Ktze. Rev. Gen. 2: 814. 1891. Thelypteris Warmingii (C.Chr.), comb. nov. Dryopteris Warmingii C.Chr. Danske Vid. Selsk. Skr. VII, 10: 227. 1913. GRAY HERBARIUM, HARVARD UNIVERSITY. A RECONSIDERATION OF SIDOPSIS RYDBERG AND NOTES ON MALVASTRUM A. GRAY (MALVACEAE) DAVID M. BATES The monotypic genus Sidopsis was based by Rydberg (1932) on Sida hispida Pursh (1814). Although Rydberg erroneously cited Elliott (1822) as the author of S. hispida, Elliott, himself, clearly attributed the species to Pursh, leav- ing no doubt as to the correct authority of the binomial. Similarly, there is no doubt that Rydberg intended his genus to apply to a small, yellow-flowered annual of the midwest- ern and eastern United States, generally recognized either as Malvastrum angustum Gray (1849) or as Sphaeralcea angusta (Gray) Fern. (1939), for he cited Gray's name in synonymy, and his description certainly applies to that spe- cies. Though Rydberg's intention is clear, the fact that he chose to base his genus on S. hispida creates a problem, for Pursh's species has not been typified nor has there been agreement on its identity, and consequently the status of Sidopsis is questionable. It is doubtful whether this prob- lem will ever be solved to everyone's satisfaction, since type material apparently has not been preserved; however, tak- ing all facets of the problem into consideration, it is possible to arrive at a reasonable determination of S. hispida. Ir- respective of the nomenclatural disposition of S. hispida and in turn that of Sidopsis, there is the question of whether or not the species to which Rydberg applied his name de- serves generic recognition. Pursh described S. hispida in the following manner: *S. hispido-pilosa; foliis lanceolatis serratis, pedunculis soli- tariis axillaribus longitudine petiolorum, calyce exteriore filiformi. In sandy plains of Georgia. Lyon v.s. in Herb. Lyon. Flowers yellow." Consideration of Pursh’s entry in- volves several factors, the first being its applicability to Malvaceae of the eastern United States, for though both the collector and the site of the type collection have been 9 10 Rhodora [Vol. 69 questioned, there seems little doubt that the collection must have been taken from the eastern or at most midwestern regions of the country. The comparison of Pursh's descrip- tion with that which I have drawn from collections of M. angustum. reveals no serious discrepancies. Malvastrum angustum is copiously but not conspicuously pubescent with rather coarse, stiff, mostly appressed hairs. If the term “pilose” is taken in the sense of denoting open-hairiness, as it may be, then M. angustum may be characterized as his- pid-pilose. The leaves of M. angustum are often lanceolate although they tend, in general, to be linear to narrowly ob- long-lanceolate. Likewise they are serrate, remotely so, with only four to six serrations on either margin. The flowers are solitary in the axils, but, as in many Malvaceae, they often appear glomerate on greatly reduced axillary branches. The presence or absence of these axillary branch- lets may only be a reflection of the age and/or vigor of the plants and not an indication of genetic differences. In any event, collections of M, angustum have often been made that have only solitary, primary axillary flowers. The pedicels of this species, however, are usually shorter than, rather than equal to, the subtending petiole — at least at anthesis — but in fruit they often exceed the petiole slightly. The involucral bracts are filiform and the corollas are yellow. Finally, Pursh described S. hispida as a perennial, whereas, M. angustum is certainly an annual, but the distinction be- tween annual and perennial duration is often not easily made, and the rather ligneous character of the lower stems may have been misleading. Besides the description itself, a further characterization of S. hispida is implied by Pursh’s inclusion of it in the genus Sida L. rather than in Malva L. During Pursh’s time these genera were distinguished primarily by the presence or absence of involucral bracts, these being present in Malva and absent in Sida. Even today this is a valid difference with the few exceptions which for other reasons are not pertinent to this discussion, Pursh must have been aware of this difference; he cited, as the basis of his genera, Schreber’s edition of Genera Plantarum (1791) where this 1967] Sidopsis — Bates 11 distinction was clearly made. It can only be assumed that in aspect Pursh's plant was so Sida-like that he chose to emphasize this feature rather than the presence of involu- cral bracts in assigning it to a genus. This is further brought out by Pursh's placement of S. hispida between S. spinosa L. and S. rhombifolia L., species which may be con- sidered representative of the mean expression of the genus. Among the North American species of Malvaceae, only those of the genus Malvastrum Gray (1849), as I shall de- fine it below, exhibit both this Sida-like aspect and fit Pursh's description in other respects. It is from among these that S. hispida must be sought. Besides M. angustum, five other species of Malvastrum occur in the United States. Two of these, M. aurantiacum (Scheele) Walp. and M. bicuspidatum (S. Wats.) Rose, are found only west of the Mississippi and probably do not need to be considered. Nevertheless, they may also be eliminated on morphological grounds: both have broadly ovate leaves, and the former has stalked, deltoid-ovate involucral bracts, while the latter has dense, canescent pubescence. The other three species are confined to southern regions of Florida and Texas but may occasionally be found in coastal regions of the Gulf Coast. They occur south of the probable collection site of S. hispida, but, because of their weedy nature, it is not in- conceivable that sometimes they might be carried further north. The leaves of all three species are predominantly ovate to broadly ovate and only rarely, in the inflorescence, might they be termed lanceolate. One, M. americanum (L.) Torr., has in addition dense, elongate, spicate inflorescences which are at variance with Pursh's characterization. The other two species, M. coromandelianum (L.) Garcke and M. corchorifolium (Desr.) Britton ex Small, although amply distinct from M. angustum, can be distinguished from it with respect to Pursh's description only on the basis of the leaf conformation. It seems unlikely, however, that Pursh would have redescribed M. coromandelianum, for that spe- cies was already well known (as Malva tricuspidata Aiton). In any event, a consideration of the problem on a purely morphological basis demonstrates that S. hispida corre- 12 Rhodora [Vol. 69 sponds more closely to M. angustum than to any other species of Malvastrum. Other questions pertinent to the problem of identity of S. hispida concern the geographical area of the type collec- tion and the identity of the collector. Both have been ques- tioned (Kearney, 1950; Ewan and Ewan, 1963), largely because no species answering the description of S. hispida has been collected in Georgia. I have not seen any speci- mens nor am I aware of any reports of M. angustum, M. coromandelianum, or M. corchorifolium from either the Carolinas or Georgia. Further, Professor Wilbur H. Dun- can of the University of Georgia has informed me that he has no record of any of these species from Georgia. It would appear, whether or not S. hispida and M. angustum are actually synonymous, that S. hispida was not collected in Georgia, at least as it is presently bounded. If they are synonymous, then it is also unlikely that the collection was made in “sandy plains,” for M. angustum. is primarily adapted to dry, limestone barrens. Ewan and Ewan (1963), accepting S. hispida and M. angustum as the same, sug- gested that Aloysius Enslen and not John Lyon was respon- sible for the type collection. Enslen's collections, taken through Georgia and lower Louisiana, were thought more likely to include S. hispida than those of Lyon. But the Enslen herbarium in the Naturhistorisches Museum, Vienna, does not include any specimens of S. hispida, ac- cording to the information kindly supplied to me by Mr. Kurt Fitz of that institution, nor are any species of Sida included in a handwritten catalogue of 1822 enumerating the Enslen collections. Perusal of John Lyon's Journal (Ewan and Ewan, 1963) shows that he had the opportunity to collect M. angustum, and there is therefore, no compelling reason to credit the collection to another. In the summer of 1803 Lyon crossed the Chattahootchee River near what is now Columbus, Georgia, and journeyed westward to Tuckaubatchee on the Tallapoosa River. This locality is now in Alabama but in 1803 was still part of Georgia. These travels took Lyon just south of Tallapoosa County where Mohr (1901) reported a 1967] Sidopsis — Bates 13 collection of M. angustum taken by E. A. Smith in August 1873 from the central pine belt. Lyon, himself, reported of his travels on August 20th, the day before reaching Tucka- batchee, that he “proceeded along a high, dry barren ridge in the forepart of the day.” Certainly both the area and the habitat were favorable for a collection of M. angustum (and in “Georgia,” no less). Perhaps a more likely site for the type collection, however, was in the vicinity of Nash- ville, Tennessee, an area where M. angustum has been col- lected rather commonly. Lyon visited this region in 1807 and again in 1809 and in both instances had ample opportu- nity to collect this species. With respect to the only other possibility, M. coromandelianum, it is evident that Lyon's travels were entirely to the north of its natural range. Ex- cept for the remote probability of Lyon's happening upon a chance introduction, it is impossible for him to have col- lected it. The consideration of evidence now available permits one of two choices. Either S. hispida may be accepted as the same as M. angustum, or it may be treated as a dubious name. Although Gray (1849), probably following Hooker (1834), originally felt that these two names might apply to the same species, he later (1897) changed his mind and excluded S. hispida from M. angustum. Fernald (1939) likewise rejected the two names as being synonymous, stating, *In view of the complete doubt about the identity of Sida hispida Pursh, which antedated Elliott by seven years and which was presumed by the latter author to be his plant [It could not have been, for Elliott clearly had a species of Sida.], it is quite unwise to force upon the plant of dry barrens and hills of the Mississippi basin the name of an unidentified plant of Georgia and possibly South Caro- lina. I am, therefore, retaining for the plant of the Missis- sippi basin the first name which unquestionably belongs to it. If and when Pursh's type is found and positively identi- fied with Sphaeralcea angusta [Fernald made the transfer from Malvastrum to Sphaeralcea for nomenclatural, not taxonomic, reasons as will be brought out below] Pursh's name will be justified; at present its use would be question- 14 Rhodora [Vol. 69 able." It is unlikely, however, that Sida hispida will ever be typified, for Lyon’s herbarium apparently has not been preserved and the few extant collections which can be at- tributed to him do not include this species; thus, there will always be grounds for rejecting S. hispida if one is disposed to do so. I feel, however, that the evidence available is strong enough so that one may state with reasonable cer- tainty that S. hispida and M. angustum have been applied to the same species. If one accepts the premise that the type collection was not made in Georgia proper, and this seems logical in view of the fact that no species answering the description of S. hispida is known from that state, then there is no obstacle in accepting these names as synonymous. On a morphological basis they correspond more closely than any of the other possibilities discussed, and the known travels of John Lyon put him in a favorable position to make the type collection. The question then turns to the generic disposition of the species. For reasons which will be summarized below, I prefer to recognize it in Malvas- (rum, as M. hispidum (Pursh) Hochreutiner, rather than in Sidopsis or some other genus. In 1849 Gray proposed a number of generic changes in the Malvaceae, among them the establishment of Malvas- ‘rum, erected largely to provide a clearer distinction be- tween Sida and Malva. Although Gray was successful in giving both Sida and Malva more precise definition, his de- scription of Malvastrum was vague and the genus rapidly became a repository for diverse elements with little in com- mon except carpels with solitary, ascending ovules, style branches equal in number to the carpels with capitate rather than decurrent stigmas, and, in general, an involucre of three bracts. Even at its inception Malvastrum included representatives of at least three genera. Of the eight spe- cies enumerated by Gray, the first three, M. coccineum (Pursh) Gray, M. munroanum (Dougl.) Gray, and M. gros- sulariaefolium (Hook. and Arn.) Gray, belong to Sphaer- alcea St. Hil.; the fourth, M. fremontii Torr. ex Gray, to Malacothamnus Greene; and the last, M. angustum, if con- sidered distinct, to Sidopsis. The remaining three belong 1967] Sidopsis — Bates 15 to Malvastrum sensu stricto. In subsequent years many spe- cies, principally from the Americas, were added, and oc- easionally from this large complex small genera were segregated, e.g., Tarasa Phil, Urocarpidium Ulbr., Mala- cothmanus Greene, and Eremalche Greene. In general, these gained little recognition until Kearney (1935, 1951, 1955) and more particularly Krapovickas (1951, 1954, 1957a, 1957b) were able to show, largely through a correlation of chromosome numbers and morphology, with emphasis on carpel characters, the validity of recognizing additional gen- era in what had been considered Malvastrum. Their redefi- nition of Malvastrum restricted it to about twelve species which, except for the now ubiquitous M. coromandelianum and M. americanum, are found only in the Americas. I would restrict Malvastrum even further and remove from it M. lacteum (Ait.) Garcke and M. subtriflorum (Lag.) Hemsley. This concept of Malvastrum has left many species formerly included in the genus in taxonomic limbo, and this rather considerable residue presents many interesting prob- lems. I am currently preparing a monograph of the South African species previously referred to Malvastrum, as well as a short paper giving generic recognition to M. lacteum and M. subtriflorum. There remain, however, the acaules- cent species from the Andes and a miscellaneous assemblage of species, principally from South America, that require in- dividual evaluation. Even in a restricted sense, however, Malvastrum is not free from problems, for it has not been typified properly to preserve its usage. Malvastrum was conserved over the South African Malveopsis Presl, and the lectotype chosen by Miss Green (1935) was M. coccineum. But as I have pointed out above, M. coccineum is a Sphaeralcea, and therefore, Malvastrum is a nomenclatural synonym of Sphaeralcea when so typified. It was on this basis and not for taxonomic reasons, that Fernald (1939) transferred M. angustum to Sphaeralcea, for there is no doubt that Mal- vastrum and Sphaeralcea are generically distinct. Never- theless, if the lectotype is not changed, a new name must be found for those species still referred to Malvastrum. 16 Rhodora [Vol. 69 Kearney (1947, 1955) recognized this and suggested that M. coromandelianum should be selected as the lectotype. In this conclusion he has been supported by Krapovickas (1957b) and Borssum Waalkes (1960). To my knowledge such a proposal has never been formally submitted to the General Committee, but, for reasons I shall elucidate in a forthcoming note, I doubt that M. coromandelianum can be accepted as the lectotype. Fortunately, there is an alter- native solution which I believe is acceptable and which will preserve the usage of Malvastrum; and I am therefore con- tinuing to use the name. Interestingly, this whole problem would not have arisen if more thought and effort had been given to working out the identity of Malveopsis. Having reviewed the pertinent references and available herbarium material, I am convinced that Malveopsis, based on Malva anomala Link and Otto, may be regarded as a synonym of Anisodontea, Presl, a name I am reviving for those South African species previously referred to Sphaeralcea. Perhaps the most serious handicap in attempting to de- limit and to evaluate genera and their relationships within the tribe Malveae is the absence of adequate descriptions at both a specific and generic level. Considering the redefini- tion of Malvastrum, it seems appropriate to include here an adequate working description of the genus before dis- cussing briefly the relationships of the species included in it. I have incorporated in this description data from my description of M. hispidum, but, as will be evident from the discussion of this species’ relationship to other species of the genus, this does not greatly extend the perimeters of Malvastrum. Malvastrum Gray, Mem. Am. Acar., ser. 2, 4: 21. 1849. Sidopsis Rydberg, Fl. Pr. Pl. Centr. N. Am. 541. 1932. Annual or perennial herbs or subshrubs, generally less than 1 (-2.5) m. tall, the branches spreading or erect; variously pubescent, the hairs simple or stellate, often pustular-based and generally with few radiate or appressed arms. Leaf blades mostly 1.5-9 cm. long, linear-lanceolate to oblong, lance- or deltoid-ovate, unlobed or ob- scurely 3-lobed, acute or sometimes obtuse, crenate to serrate, sub- cordate, truncate, or cuneate, ventrally plane or with nerves 1967] Sidopsis — Bates 17 impressed, glabrate to copiously pubescent, dorsally with nerves gen- eral raised, more densely pubescent; petioles mostly shorter than the blades; stipules 3-9 (-14) mm. long, filiform or subulate (or asymmetrically oblong-ovate), often caducous or drying early. Flowers axillary and/or terminal, solitary, in cymose clusters, or in dense spikes, sometimes conspicuously bracteated, erect at anthesis and in fruit, opening briefly in the late morning or early afternoon. In- volucral bracts 3, spreading or appressed, mostly free or only basally adnate to the calyx tube, filiform to narrow-lanceolate or oblanceolate, 3-8 mm. long, but sometimes elliptic-ovate, then 7 X 3.5 to 10 X 5 mm. long and broad, or stalked and with an abruptly expanded, del- toid-ovate blade, 4-7 mm. long, up to 4.5 mm. broad. Calyx generally angled or winged in bud, often hispid, 4-7 (-11) mm. long at anthesis, to 10 (-14) mm. long in fruit, often drying brown, scarious, the lobes 2.5 X 2 to 7 X 5 mm. long and broad, deltoid to cordate-ovate, acute or acuminate, principal nerves 3; nectaries 5, free, usually obscure. Corollas yellow to orange-yellow, mostly scarcely exceeding the calyx; petals 3 X 2 to 9 X7 (to 15 X 11) mm. long and broad, obovate, at the apex slightly oblique, shallowly emarginate, narrowed evenly to a non-auriculate base, glabrous except for few spreading hairs along the basal margin or at the union with the staminal col- umn. Staminal column yellowish, included, up to 3 (-6) mm. long, glabrous or puberulent at the base between the petals; filaments terminal or arising over the upper 1 mm. or so, to 1 (-2) mm. long; anthers 5 to ca. 36 (-51), yellowish then reddish after dehiscence. Style branches 5-16, exserted and recurved within the anthers before or at anthesis; stigmas capitate, expanded, papillate. Carpels as many as the style branches, free from one another in a flat whorl about a central columella, each with a single, ascending ovule. Fruit exposed or clasped by the calyx, 3-10 mm. in diameter, reddish-brown or tan, sometimes glabrous but generally with few to many, erect, simple or bifurcate, hispid hairs and often with stellulate pubescence as well; mature carpels varying from ca. 1 X 1 to 4.5 X 4.5 mm. long and broad, angular-orbicular to transverse-elliptic in outline, laterally compressed, armed or unarmed, the lateral edge rounded or sharp, sometimes raised in a slight wing or prolonged dorso-apical- ly into ascending, horizontal or deflexed, laterally flattened knobs or spines, the lateral walls often fused ventro-apically and sometimes forming an indehiscent or partially dehiscent erect knob or spine, smooth below or with a raised, transversely radiate pattern of thick- enings, the dorsal wall often canaliculate; carpels indehiscent, par- tially dehiscent along the midvein basally, or completely dehiscent into two free valves, glabrous within, with or without a small ventro- apical endoglossum. Seeds 0.9 to 2.8 mm. long, asymmetrically to nearly symmetrically reniform, glabrous, black or reddish-brown or -black. Chromosome number, n= 6, 12, 18. 18 Rhodora [Vol. 69 It is difficult to estimate how many species Malvastrum actually comprises, for the species or species-complexes in- cluded in it have not been carefully studied. The above de- Scription has been drawn principally from notes made during my tenure as a National Science Foundation Post- doctoral Fellow at the British Museum (Natural History) and the Royal Botanie Gardens, Kew. My primary purpose at that time was to attempt to estimate generic variation within the Malvaceae tribe Malveae, and although such a survey naturally involved observations of a great number of species, its orientation was not that of monographic work at a specific level, and therefore was not particularly occu- pied with delimiting species. On the other hand, such a survey could not help but lead to conclusions concerning the constitution of genera. In the case of Malvastrum, it seems that the species, on the basis of morphological comparisons, conveniently fall into four groupings or series. The largest of these series in terms of apparent number of species, and for this reason perhaps to be considered the modal expression of the genus, includes the following spe- cies; M. coromandelianum (L.) Garcke [M. tricuspidatum (Ait.) Gray], 2n — 24 (Skovsted, 1935, 1941, Krapovickas 1949, 1951, 1954, Roy & Sinha 1961) ; M. corchorifolium (Desr.) Britton ex Small [M. rugelii S. Wat.]; M. amer- icanum (L.) Torr. [M. spicatum (L.) Gray], 2n — 24 (Skovsted, 1935, 1941, Krapovickas, 1951, 1954) ; M. inter- ruptum K. Schumann, 2n — 36 (Krapovickas, 1954) ; and M. spiciforum (Hassler) Krapov., 2n = 12 (Krapovickas, 1954). Morphological differences within this series are as- sociated primarily with the character of the inflorescence, pubescence, and size and ornamentation of the carpels, but within this grouping there is a continuity of expression largely expressed in the modal values given in the generic description. M. corchorifolium probably should not be recognized at a specific level, for it differs from M. coro- mandelianum only in having carpels that are spineless or essentially so. Throughout the Caribbean islands, numerous forms occur with varying spine development so that any line drawn between the two species is quite arbitrary. A 1967] Sidopsis — Bates 19 difference of similar magnitude separates M. americanum and the Argentinian M. interruptum. In the former the spikes are naked and the flowers are subtended by bifid bracts, whereas in M. interruptum the inflorescences are leafy and the bracts are in pairs. Cytologically, however, these two species are distinct, M. americanum being tetra- ploid and M. interruptum hexaploid. Within this series the only diploid species of the genus has been reported: M. spiciflorum endemic to Argentina. Except for the fact that chromosome numbers of all species of the genus are apparently based on the number 6, nothing is actually known of the cytological relationships among them, though it may be predicted that some genomic dupli- cation must exist. On morphological grounds there seem to be no stronger correlations between species at a given ploidy level than between species at different levels, and certainly it would be unwarranted at this time to speculate on phylo- genetic relationships on the basis of the few counts avail- able. Nevertheless, it seems improbable that the diploid M. spiciflorum, notable for its extremely small, unarmed car- pels, reduced number of stamens, and small floral parts, is ancestral to any of the known species, The other series include species which vary from this modal grouping in a few conspicuous characters, but in no case do they become so extreme as to suggest that further generic segregation might be called for. One of these series comprises the North American (Texas) M. aurantiacum (Scheele) Walp. [M. wrightii Gray] and the South Amer- ican (Bolivia, Paraguay, Argentina) M. amblyphyllum R. E. Fries, 2n — 24 (Krapovickas, 1954). These species are distinctive primarily in their large floral parts and fruits. In M. amblyphyllum the involucral bracts may be linear- lanceolate but sometimes are elliptic-ovate, whereas in M. aurantiacum the bracts are stalked, with an abruptly ex- panded blade. In both species the bracts nearly equal the calyces, which even in flower are often 10 mm. long. Sim- ilarly, the petals are larger and the anthers more numerous. The carpels vary from 3.5-4.5 mm, long and broad, each with two dorso-apical projections and a spinescent and 20 Rhodora [Vol. 69 fused ventral region. The flowers are mostly solitary in the axils, but in M. amblyphyllum may reach three in number on reduced axillary branchlets. Despite a rather high degree of similarity in conspicuous characters, it is probable that this series does not represent a natural alliance. In general aspect, particularly in the actual conformation of the car- pels, M. amblyphyllum seems to tie in with the M. coro- mandelianum series, whereas it is more difficult to relate M. aurantiacum to any of the other species in particular. The third series may be designated the M. scoparium series, for this was the first species named in a complex of species or forms running from western North America (Arizona) to Bolivia and Peru and the Galápagos Islands. Besides M. scoparium (L'Hér.) Gray, names most commonly encountered which belong to this complex are M. bicuspi- datum (S. Wats.) Rose, M. guatemalense Standl. and Steyerm., M. scabrum (Cav.) Gray, M. depressum (Benth.) Svenson, M. dimorphum J. T. Howell, M. scoparioides Ulbr., and M. guaraniticum Hassler. It is difficult to suggest where species lines ought to be drawn in this series, but it may not be amiss to suggest that all should be included un- der M. scoparium with the major variants, if they can be shown to have geographical continuity, recognized at a sub- specific level. This series of forms has as its unifying char- acteristic a carpel conformation in which there are two dorso-lateral spines projecting outward in more or less hor- izontal fashion with a deep acute sinus between them. Ven- tro-apically there may be no projection or a small to quite prominent awn. Variation in the prominence of the awns or spines and considerable differences in pubescence (for the most part an unreliable taxonomic criterion in the Mal- veae) as well as some miscellaneous differences, such as the length of the petals or number of flowers in the inflores- cence, have formed the basis for recognizing most of the Species. Skovsted (1935) reported for M. scoparium a chromo- some number of Zn — 24, but later (1941), for M. scabrum, a species of the same complex, reported 2n = 42. I have examined voucher material of M. scoparium, kindly lent by 1967] Sidopsis — Bates 21 Dr. Skovsted of the Botanical Museum and Herbarium, Copenhagen, and it is correctly determined. Unfortunately there is no material of M. scabrum, but there is little doubt that the identification was erroneous and was probably only that under which the species was received as seed from the botanical garden in Bucharest. Not only is the count at variance with all others reported in the genus, but the illus- tration of the mitotic complement presented by Skovsted (1941) is at variance with those which he and, later, Krapo- vickas (1954) have given for other species of Malvastrum. Skovsted (1941) also pointed out that species of Malvas- trum with 24 somatic chromosomes (the only true species of Malvastrum which he had) are distinctive cytologically. Not only are the chromosomes distinct in their size and shape, but also they are arranged in irregular but charac- teristic fashion during mitosis. It is more than likely that in this case “M. scabrum” was actually a species of Malva. The number 2n — 42 is common in that genus, and the compar- ison of illustrated karyotypes shows them to be remarkably similar. The last series to be considered includes only M. hispidum. Rydberg?s description of Sidopsis in his floristic account of the plains and prairies of central North America was brief and included no discussion of the reason for providing this species with generic recognition. He did note in his descrip- tion, however, the completely dehiscent nature of the car- pels. Kearney (1951) stressed this character, together with the Sida-like aspect of the species, to maintain it as gener- ically distinct. Sidopsis, however, is not the only genus of Malveae characterized by ascending ovules in uniovulate, completely dehiscent carpels. Others are Malacothamnus Greene, Tarasa PhiL, and Nototriche Turez. The first of these belongs to a phylogenetically distinct alliance (Bates, 1963) including Phymosia Desv. and Iliamna Greene, with a base chromosome number of x — 17; Tarasa and Noto- triche belong to a series of genera including among others Sphaeralcea St. Hil., Urocarpidium Ulbr., and Eremalche Greene, with a base number of x — 5. In these generic groupings it is not simply the mode of carpel dehiscence 22 Rhodora [Vol. 69 which delimits genera. Admittedly, the nature of carpel dehiscence is a useful diagnostic feature, but, in fact, these are natural genera differing from one another in many char- acters of somewhat more subtle nature. It is probable that carpel characters, in general, have been accorded greater importance than they deserve in classification of the Mal- veae. This certainly has been the case in Sidopsis. In car- pel characteristics M. hispidum differs from other species of the genus only in having a smaller number — 5-6 instead of (7-) 8 (-16) — and complete dehiscence, Variation in carpel numbers is often greater in other related genera, and in Sida there is a similar pattern with some species 5-car- pellate but most 7- or more-carpellate. Similarly, the mode of carpel dehiscence loses much of its importance in view of the nearly complete dehiscence in M, americanum. The car- pels of this species dehisce along the midvein from the ventro-basal junction with the columella around the dorsal wall and nearly to the ventro-apical junction. Here, how- ever, dehiscence is stopped by a slight fusion of the valves. In what forms a genus of remarkable uniformity, there seems no justification for emphasizing these slight carpel- lary differences, (better reflective of specific limits), to maintain Sidopsis as distinct from Malvastrum. With the exception of the fruiting characters only the narrow leaves of M. hispidum actually broaden the limits of Malvastrum, but these correspond to the basic unlobed pat- tern common to all species. In addition, M. hispidum shares with other species particular characteristics, such as the unusual 4-armed hairs of M. coromandelianum, or with all species a similar pattern of floral structure and probably floral behavior. Although I have had the opportunity of observing only M. coromandelianum, M. americanum, and M. hispidum in a living state, there is every reason to be- lieve, considering the relationship of style branches and anthers at or just before anthesis, that autogamy is the commonest form of reproduction in the genus, It cer tainly is in the above three species. In fact, M. hispidum is often cleistogamous and the other species may be functionally so, for even before the flower opens, in the late morning or 1967] Sidopsis — Bates 23 early afternoon, the pollen has been shed and has coated the stigmas. Besides a high degree of morphological similarity to other species of Malvastrum, M. hispidum shares the same base chromosome number. Counts have been made from micro- sporocytes of bud collections made in Kentucky (Bates 2704), Missouri (Bates 2688, 2690, 2691), and Kansas (Bates 2702). Localities for these collections are listed in the citation of specimens and voucher material is deposited in the L. H. Bailey Hortorium. In all cases, 2n — 36. At metaphase pairing was complete, and segregation in ana- phase 1 showed no abnormalities. This is the second hexa- ploid reported in Malvastrum; however, M. hispidum and M. interruptum are distinct morphologically and quite likely have been derived independently. The description of M. hispidum following has been drawn primarily from collections made during the summer of 1965. These collections, in themselves, provide a reasonable sampling of the species through most of its range, but have been supplemented by examination of dried specimens from the following herbaria: the L. H. Bailey Hortorium and the Wiegand Herbarium, Cornell University; the Gray Her- barium, Harvard University ; the Herbarium, University of Illinois; the Royal Botanic Gardens, Kew; the British Mu- seum (Natural History) ; and the Botanischer Garten und Institut für Systematische Botanik der Universitat Zürich. To those who permitted me to examine material in their care, I wish to express my appreciation. Malvastrum hispidum (Pursh) Hochr. Ann. Conserv. Jard. Bot. Genève 20: 129. 1917. Sida hispida Pursh, Fl. Amer. Sept. 452. 1814. Malvastrum angustum Gray, Mem. Am. Acad., ser. 2, 4: 22. 1849. Malva perpusilla Nutt. ex Gray, loc. cit., pro syn. Malveopsis hispida (Pursh) Kuntze, Rev. Gen. Pl. 1: 72. 1891. Sidopsis hispida (Pursh) Rydb., Fl. Pr. Pl. Centr. N. Am. 541. 1932. Sphaeralcea angusta (Gray) Fern., Rhodora. 41: 435. 1939. Erect, annual herbs 1.5-4.5 (-6) dm. tall, simple or branched, the branches ascending, arising from near but not at the base, progres- 24 Rhodora [Vol. 69 sively shorter distally, rarely exceeding the leader; stems and branches slender, terete, greenish distally, generally reddish-brown proximally, copiously pubescent with appressed, mostly 4- (rarely 6-) armed hairs, the arms to 0.75 (-1.2) mm. long, generally in sub- parallel pairs oriented parallel to the long-axis of stems and branches. Leaf blades yellowish-green, up to 5.5 cm. long X 1.3 cm. broad, linear to narrow-oblong or lanceolate, the apex acute or obscurely mucronate, the base cuneate to narrow-obtuse, the margins usually with 4-6 pairs of evenly spaced, bristle-tipped serrations, rarely sub- entire, the nerves 1-3 from the base, the surfaces plane, mostly in- conspicuously appressed-pubescent above and below, the hairs 4-armed above, 4-7-armed below, simple to 4-armed marginally; petioles gen- erally less than 1/3 the length of the blades; stipules spreading, 2-7 mm. long, narrow-subulate to filiform, simple-ciliate, drying and turn- ing brownish early. Flowers erect at anthesis and in fruit, solitary in the axils but sometimes glomerate, especially distally, by the re- duction in internode length; pedicels erect, mostly less than 10 mm. long at anthesis, to 17 mm. long and finally patent in fruit. Involu- cral bracts 3, free, (2.6-) 4-6.8 mm. long, filiform, linear, or very narrow-subulate, mostly thickened with a slight groove ventrally, dorsally hispid with simple to 2- or 4-armed hairs, in fruit becoming brownish, reflexed. Calyx at anthesis (2.8-) 4.3-7.2 mm. long, in fruit 9-12.5 mm. long, copiously hispid with sessile, appressed, 4-6-armed hairs, the lobes in bud through fruit plieate-winged at the margins, overhanging the whitish tube, (1.2-) 2.8-5.8 mm. long X (1-) 2-5.8 mm. broad at anthesis, 8-10 mm. long X 8-12 mm. broad in fruit, (narrowly to) broadly cordate-ovate, abruptly acuminate to subcus- pidate, marginally simple- or bifurcate-ciliate, puberulent within, the tube glabrous within with 5 obscure nectaries at the base opposite the lobes; drying brown and scarious. Corolla included within or slightly exceeding the calyx, the petals yellow, 2.8 X 1.5 to 4.5 X 3.2 mm. long and broad, obovate, obliquely shallow-emarginate apically, gradually cuneately narrowed to a non-auriculate base, but sometimes with a few simple, hispid hairs along the basal margins. Staminal column shorter than the petals, ca. 2-3 mm. long, yellow, glabrous or glabrate, the filaments to 0.5 mm. long, terminal; anthers 8-13 (-16), bright yellow. Ovary composed of 5 (-6) free carpels in a flat whorl, each with a single ascending ovule. Style to 4 mm. long, glabrous, the branches as many as the carpels (occasionally a branch may be bifurcate at the apex), recurved within the anthers before or at an- thesis; stigmas capitate, papillate, yellow. Fruit 5.6-7.6 mm. in diam. in a star-like whorl with broad, deep sinuses between adjacent car- pels. all free surfaces copiously pubescent with sessile, erect, simple or bifurcate hairs, often with a reddish base, these at the apex up to 1 mm. long; carpels yellowish- to reddish-brown, 2.6 X 2.6 to 3 x 3.5 mm. long and broad, more or less orbicular but the base often trun- cate, laterally compressed with rounded lateral edges, plane without 1967] Sidopsis — Bates 25 lateral venation; midvein present; smooth within, lacking an endoglos- sum; loculicidally dehiscent, from the ventro-apical and -basal edge to the dorsal edge, into two free valves. Seeds 2.1 X 2 to 2.5 X 2.5 mm. long and broad, rounded and nearly symmetrically reniform, glabrous, reddish-black or sometimes grayish. Chromosome number, n = 18. TYPE. In the absence of authentic Lyon material, Pursh's descrip- tion (1814) may serve as type. DISTRIBUTION. East of the Mississippi M. hispidum occurs in Kentucky, Tennessee, and Illinois and has been reported from Tallipoosa County, Alabama (Mohr, 1901). In Ken- tucky it has been collected just southwest of Lexington, and in Tennessee seems to be confined to the area about Nash- ville. It has not been collected since before the turn of the century in Illinois (Jones & Fuller, 1955) but at that time was found in northern regions of the state about Ottawa and Rock Island. To the west of the Mississippi the species is found in its greatest concentration in Missouri and east- ern and central Kansas, To the north it reaches Iowa, but I have seen records only for the immediate vicinity of the Mississippi, and although it is reported from Nebraska (Steyermark, 1963, among others) I have not seen any col- lections actually made there. To the south of Missouri it may be found in the Ozarks of northern Arkansas, and it has been collected south of Oklahoma City, Oklahoma. This latter area is just to the north of the Red River where Nut- tall is reported by Gray (1849) to have collected the species. Throughout its range M. hispidwm is exceeding local. Al- though populations may consist of dense stands, most of those which I have seen could be included in a few square meters. The largest population observed, (Bates 2672) which in fact consisted of a number of sub-populations, could be bounded by a square with sides about twenty meters. With the exception of my collection (2700) in southeastern Kansas where I found this species growing in clay-loam, all were taken in areas about limestone outcrops, either in crevices of the outcrops themselves or in soils (often cherty) overlying them. It seems that this type of habitat is most common to the species, although Steyermark (1963) reported it may occasionally be found in “open al- 26 Rhodora [Vol. 69 luvial ground in valleys, and along gravel bars." It is prob- able that M. hispidum is only adventive in such habitats. REPRESENTATIVE SPECIMENS. KENTUCKY. Jessamine Co.: High Bridge, dry limestone soil along roadsides and waste fields, 17 Aug. 1942, McFarland 13 (Bn, CU, GH); Common along roadside, Kentucky highway 33, 0.3 mile e. of United States highway 68, 26 Aug. 1965, Bates 2705 (BH). Woodford Co.: Road's edge, Kentucky highway 33, 0.8 mile west of Woodford-Jessamine county line, 26 Aug. 1965, Bates 2704 (BH). TENNESSEE: Davidson Co.: Vicinity of Nashville, in August, Gattinger (BM, CU, GH, K); Cedar glades near La Vergne, 27 July 1938, Svenson 9362 (GH). Rutherford Co.: Railroad track, La Vergne, July 1894, Ruth (ILL); Dry cedar glade, La Vergne, 20 Aug. 1922, Svenson 202a. (GH). Wilson Co.: Limestone barrens, Leb- anon, 30 Aug. 1947, Sargent 93 (GH). ILLINOIS: La Salle Co.: Bar- rens, Ottawa, 29 Sept. 1882, Seymour (ILL), Sept. 1881, Boltwood (GH). Rock Island Co.: Growing abundantly among the debris of an old limestone quarry near Rock Island Arsenal, in 1866, Parry (GH), in 1862, (BH). IOWA: Muscatine Co.: Banks of the Mississippi River near Davenport, Barnes (ILL). MISSOURI: Callaway Co.: Common in loamy soil of limestone outcrop, county highway AA, ca. 1.0 mile e. of junction with United States highway 54, 6 miles n. of Jefferson City, 22 Aug. 1965, Bates 2688 (BH). Greene Co.: Dorchester near Springfield, 7 Sept. 1887, Blankenship (cu). Jasper Co.: Locally abundant in limestone barrens near Webb City, 25 Sept. 1920, Palmer 19155 (GH); Rocky barrens, Webb City, 6 Aug. 1920, Bush 9068 (GH). Jackson Co.: Common in barrens, Independence, 6 July 1900, Bush 768 (K); Rocky places, Lee’s Summit, 14 Aug. 1898, Mackensie 323 (GH), 13 Sept. 1927, Bush 11550A (K, Z); Greenwood barrens, 30 Aug. 1905, Bush 3285 (GH); Rare in crevices of limestone, road cut, Missouri highway 7, 0.8 mile n. of junction with United States highway 50, 22 August 1965, Bates 2690 (BH); Locally common in loamy soil of limestone ridge, road cut, Missouri highway 150, ca. 1.3 miles e. of Greenwood, 22 Aug. 1965, Bates 2691 (BH). Lincoln Co.: Cuivre River State Park, common in crevices of limestone outcrop, Missouri highway 147, 0.3 mile n. of junction with Missouri highway 47, 21 Aug. 1965, Bates 2672 (BH); Cuivre River State Park, banks and flats above dry limestone creek bed, below Missouri highway 147, 0.3 mile n. of junction with Missouri highway 47, 21 Aug. 1965, Bates 2686 (BH). St. Louis Co.: Stony hills, Eggert (ILL); St. Louis, in 1832, Drummond (lectotype Malvastrum angustum Gray, GH; iso- types BM, K); St. Louis, Englemann (GH, K). ARKANSAS: Boone Co.: Harrison, rocky hillsides, 24 Oct. 1914, Palmer 6910 (CU). KANSAS: Bourbon Co.: 7 miles s. of Uniontown, 20 Sept. 1937, Horr & Frank- lin E180 (GH, ILL); Rare along roadside in clay-loam, Kansas, high- way 3, just s. of Petersburg, 24 Aug. 1965, Bates 2700 (BH); Locally 1967] Sidopsis — Bates 27 common in limestone soils, Kansas highways 7, 39, 5 miles w. of Hiat- ville, 24 Aug. 1965, Bates 2702 (BH). Douglas Co.: Without locality, 16 Aug. 1895, Hitchcock (GH). Reno Co.: Without locality, July 1891, Carlton (ILL). Riley Co.: Without locality, stony hills, 27 Aug. 1895, Norton (GH). OKLAHOMA: Johnston Co.: Tishomingo, granite barrens, 10 Sept. 1914, Palmer 6491 (cv). BAILEY HORTORIUM CORNELL UNIVERSITY, ITHACA, NEW YORK LITERATURE CITED BarEs, D. M. 1963. The genus Malacothamnus. Ph.D. thesis, Uni- versity of California, Los Angeles. BorRSSUM WAALKES, J. VAN. 1960. The typification of the genus Mal- vastrum. Taxon 9: 212-213. ELLIOTT, S. 1822. Sida. In A sketch of the Botany of South Carolina and Georgia 2: 159-162. J. R. Schenck. Charleston. EWAN, J. and N. Ewan. 1963. John Lyon, nurseryman and plant hunter, and his journal, 1799-1814. Trans. Am. Philos. Soc., n. ser. 53(2): 1-69. FERNALD, M. L. 1939. New species, varieties and transfers. Rhodora 41: 423-461. GRav, A. 1849. Malvaceae. In Plantae Fendlerianae Novi-Mexicanae. Mem. Am. Acad., ser. 2, 4: 15-25. — 4 1897. Malvastrum.: In Synoptical Flora of North America. 1(1): 308-313, GREEN, M. L. 1935. In International Rules of Botanical Nomencla- ture, ed. 3. p. 145. Gustav Fischer, Jena. Hooker, W. J. 1834. Notice concerning Mr. Drummond's collections, made chiefly in the southern and western parts of the United States. Journ. Bot. Hooker 1: 183-202. Jones, G. N. and G. D. FULLER. 1955. Malvastrum. In Vascular Plants of Illinois. p. 322. University of Illinois Press, Urbana. KEARNEY, T. H. 1935. The North American species of Sphaeralcea subgenus Eusphaeralcea, Univ. Calif. Pub. Bot. 19: 1-128. «1947. Type of the genus Malvastrum. Leafl. West. Bot. 5: 23-24. -——.—.-———— 1950. Notes on Malvaceae. op. cit. 6: 51-52. — 4 1951. The American genera of Malvaceae. Am. Midl. Nat. 46: 93-131. ————.————-4 1955. Malvastrum A. Gray — a re-definition of the genus. Leafl. West. Bot. 7: 238-241. KRAPOVICKAS, A. 1949. Las especies de “Sphaeralcea” de Argentina y Uruguay. Lilloa 17: 179-222. 28 Rhodora [Vol. 69 . 1951. Monteiroa, nuevo género de Malvaceae. Bol. Soc. Argent. Bot. 3: 235-244. - 1954. Estudio de las especies de “Anurum”, nueva sección del género “Urocarpidium” Ulbr. (Malvaceae). Dar- winiana 10: 606-636. 1957a. Nümeros cromosómicos de Malvaceas Americanas de la tribu Malveae. Rev. Agron. Noroeste Agr. 2: 245-260. 1957b. Las especies de Malvastrum sect. Malvas- trum de la flora Argentina. Lilloa 18: 181-195. MoHn. C. 1901. Malvastrum. In Plant life of Alabama. Contr. U.S. Nat. Herb. 6: 616. PunsH, F. 1814. Sida and Malva. In Flora Americae Septentrionalis. 2: 452-454. Roy, R. P. and R. P. SINHA. 1961. Meiotic studies in some malva- ceous species. Curr. Sci. Bangalore. 30: 26-27. RYDBERG, P. A. 1932. Sidopsis. In Flora of the Prairies and Plains of Central North America, p. 541. New York Botanical Garden, New York. SCHREBER, J. C. D. voN. 1791. Sida and Malva. In Genera Planta- rum. 2: 463, 466. Francofurti ad Moenum. SKOVSTED, A. 1935. Chromosome numbers in the Malvaceae, I. Jour. Genetics 31: 263-296. ~ 1941. Chromosome numbers in the Malvaceae. II. Compt. Rend. Lab. Carlsb. Physiol. 23: 195-242. STEYERMARK, J. A. 1963. Sphaeralcea. In Flora of Missouri. p. 1050. Iowa State University Press, Ames. NOTES ON THE FLORA OF CO6S COUNTY, NEW HAMPSHIRE, II. During the summer of 1965 I made more than a thousand collections in Coós County and after these were processed this winter I finally had a chance to evaluate the catalogue of Pease's Flora of Northern New Hampshire. The county is divided into 43 organized and unorganized townships which vary greatly in the number of taxa cited from them. At the top of the list are Shelburne and Randolph, each being rep- resented by over 800 taxa. At the other end of the scale are thirteen unorganized townships which are represented by less than 100 taxa each, Dix Grant is cited five times, Cole- brook Academy Grant three and Pease had seen no speci- mens from Erving Location. For this reason Erving Location became my first objec- tive last summer. Through the kindness of the International Paper Company and Mr. W. A. Ruck I was able to travel their Phillips Brook Road and make collections in the loca- tion at three periods during the summer. The township is now better represented than several townships located on main highways. Colebrook Academy Grant seemed to be the logical target for 1966. However townships are not always the stable entities we picture them to be. This grant was a small area bounded on three sides by Pittsburg and on the fourth side by the Province of Quebec. It no longer appears on the U. S. G. S. map of the quadrangle, having apparently been absorbed by Pittsburg. The loss of this township to the county is compensated by the appearance on the more recent maps of the Mount Washington region of Chandler Pur- chase. This is an unlikely-looking strip five miles long and only a half mile wide extending from Millen Hill near Jefferson Notch southward to Mount Pleasant in the Presi- dential Range. I have seen no specimens collected in Chandler Purchase. Town lines are not sacrosanct. On editions of the U. S. G. S. maps published since about 1925 many of the town boundaries have been drastically changed from those shown 29 30 Rhodora [Vol. 69 on earlier editions. On the 1896 sheets for Mount Wash- ington and Crawford Notch the western boundary of Cutts Grant extends up the western side of the Dry River valley, crosses the ridge of the Presidential Range between Mount Pleasant and Mount Franklin and continues to a point northwest of the summit of Mount Monroe. The northern boundary follows a zig-zag course to Boott Spur and the eastern boundary extends down the eastern side of Dry River. Thus Cutts Grant includes a portion of the Crawford Path, the summits of Mount Franklin and Mount Monroe and the headwall of Oakes Gulf, some of the prime botanical areas of the Mount Washington region. On the Crawford Notch sheet of 1946 the western boundary of Cutts Grant is at least a half mile east of the ridge of the Presidential Range. Its northernmost point is at an elevation of about 4100 feet on the southern slopes of Mount Franklin. The northern boundary extends across to Gulf Peak, about a mile south of Boott Spur. Little if any of the present limits of the grant is above timberline. Prof. Pease in his citations retains the old boundaries of the township; thus of over one hundred taxa cited from the grant the number collected from within its present limits can be counted on the fingers of one hand and none of them are strictly mountain species. The alpines were all collected in Sargent Purchase. The southern boundary of Hadley Grant has also been shifted on the more recent maps. While the number of taxa involved is much less, a number cited from Hart Ledge in this township were actually collected in Hart Location which is in Carroll County rather than Coós, This is un- fortunate since some of the taxa are southern elements unknown elsewhere in Coós County. Some interesting finds in my 1965 collections and a few changes which should be made in the Pease Flora follow. Equisetum arvense L., forma campestre (C. F. Schultz) Klinge. Millsfield, moist ditch beside road to Big Millsfield Pond, Harris 27271b (65). This form wi*h a strobilus terminating a green vegetative stem is new to the County. X Esuisetum litorale Kuhlewein. Erving Location, moist ground, Harris 27520 (65). The third township for the County. Equisetum variegatum Schleich., var. variegatum. Erving Location, 1967] Coós County — Harris 31 wet ground, Harris 27815 (65). This taxon appears in Pease's list of Species Exclusae. My collections of this and the previous species were verified by Dr. Rolla Tryon. Pease 39615 (59) from Stratford cited in the Flora as var. Jesupi is also var. variegatum and var. Jesupi should be retired to the list of Species Exclusae. Bromus Dudleyi Fern. Martin Location, damp woods Dolly Copp, Harris 1822 (33). The third collection from the County; better late than never! Eragrostis poaeoides Beauv. Berlin, roadbed of Grand Trunk Rail- way, Harris 27677 (65). Pease 29851 (42) from Milan cited as E. pectinacea should be changed to this species. These are the second and third collections from the County. Lolium multiflorum Lam., var. diminutum Mutel. Millsfield, road- side, Harris 28262 (65). First collection in the County. Triticum aestivum L. Milan, roadbed of Grand Trunk Railway, Harris 27618 (65); Berlin, roadbed of Grand Trunk Railway, Harris 27685 (65). One previous collection from the County. Secale cereale L. Columbia, roadside Cilley Hill, Harris 28306 (65); Cambridge, roadside rte. #16, Harris 27254 (65). New to the County. Digitaria sanguinalis (L.) Scop. Columbia, roadside U. S. rte. i3, Harris 28351 (65). Third and most northern collection from the County; D. Ischaemum is the common crab-grass for the area. Echinochloa crusgalli (L.) Beauv., forma longiseta (Trin.) Farw. Millsfield, roadside rte. #26, Harris 28279 (65). New to the County. Scirpus atrovirens Willd., var. georgianus (Harper) Fern., forma viviparus Vict. Erving Location, moist ground, Harris 28179 (65). Three previous collections from the County. Arisaema Stewardsonii Britt. Colebrook, damp ground along Mo- hawk River, Harris 27241 (65). A new native species for the County. Cypripedium acaule Ait. forma acaule. Millsfield, moist woods on road to Big Millsfield Pond, Harris 27345 (65); Dummer, moist woods on woodroad to Dummer Ponds, Harris 27229 (65). While Pease indi- cates that forma albiflorum is general in the County, he cites no collections of the pink-flowered form from north of Randolph. There is no question that forma albiflorum is the commoner of the two in the northern portion of the County but a little searching will usually disclose that the typical form is present as well. Salix pentandra I. Pease 33769 (48) from Colebrook cited as S. fragilis is S. pentandra, a new introduced willow for the County. Ulmus americana L., forma pendula (Ait.) Fern. Dummer, woods Dummer Hill, Harris 27736 (65). New to the County. Polycnemum majus (Schimp.) R. Br. Berlin, roadbed of Grand Trunk Railway, Harris 27670 (65). While my collection of this non- descript species is somewhat immature I am sure that there is no question of its identification. The genus is not included in Gray's Manual but does appear in the New Britton & Brown on the basis 32 Rhodora [Vol. 69 of collections made in Ontario. The species is not represented in the herbarium of the New England Botanical Club and there are no specimens from North America in the Gray Herbarium. My collection is a good match of European material in the Gray Herbarium. Mirabilis nyctaginea (Michx.) MacM. Berlin, near Grand Trunk Railway, Harris 27629 (65). This collection adds the Nyctaginaceae to the flora of Coós County. The station is well established on the outskirts of Berlin and must have been growing there for several years. Lychnis Flos-cuculi L. Carroll, roadside rte. #302 one quarter mile east of Bethlehem line, William. Countryman 840 (65). First collec- tion from the County. Silene pumilio Wulf. The station of this species on the Ammonoosuc Ravine Trail, reported by me last year, came through the winter well. This summer I visited the Mountain Valley Nursery in Jefferson Meadows and found the plant for sale there. Thus a source of supply for this obviously planted species was not far from Mount Washington. Silene Cserei Baumg. Berlin, roadbed of Grand Trunk Railway, Harris 27649 (65). One previous collection from the County. This species grows with, and closely resembles, S. Cucubalus. It is prob- ably more common in New England than present collections indicate. Ranunculus repens L. Erving Location, moist ground, Harris 27473 (65). The third collection for the County. Chelidonium majus L. Berlin, roadside Prospect Street, Harris 27628 (65). A flourishing colony; it seems strange that there has been only one other collection of this species made in the County. Arabis laevigata (Muhl.) Poir. This species should be placed in the list of Species Exclusae; the only collection, Lancaster, Pease 39124 (58), is A. glabra. Ribes hirtellum Michx., var. saxosum (Hook.) Fern. Northumber- land, Hodgdon, Lincoln & Steele 11239 (59). New to the County. Pyrus Aucuparia (L.) Gaertn. Whitefield, Bray Hill Road, Pease 34381 (49). New to the County. Potentilla Robbinsiana Oakes. The entire Mount Washington popu- lation of this rare species is located in Sargent Purchase, not Cutts Grant as cited in Pease. In June of 1964 I moved six plants from the acre or two on which it occurs to another area which resembles its home site. In June 1965 five of the plants had survived the winter and I found the remains of the sixth; only one of the sur- vivors bore flowers. Filipendula rubra (Hil) Robins. Dummer, swale Dummer Hill, Harris 27742 (65). One previous collection from the County. This is an escape from cultivation. The farms on Dummer Hill were all abandoned fifty years ago and the swale in which I found the plant is a considerable distance from any of these old farms, The colony consisted entirely of basal leaves. When I planted some hardy peren- 1967] Coós County — Harris 33 nials at my camp on the hill about fifteen years ago I accidentally introduced a few plants of this species and my plants too produce only basal leaves. Rubus ravus Bailey. R. tavus, the last taxon on page 190, should read È. ravus. Lotus corniculatus L. Milan, pasture Spruceville, ‘Harris 27867 (65). This makes a fourth township for this species which was added to the County flora last year. Oxalis europea Jord., forma willicaulis Wieg. Milan, roadbed of Grand Trunk Railway, Harris 27614 (65). The second collection from the County. Viola papilionacea Pursh, var. albiflora Grover. Dummer, Dummer Hill, Harris 27188 (65). This violet has escaped from the garden and become thoroughly established in the lawn at our camp; new to the County. Epilobium angustifolium L., forma albiflorum (Dumort.) Haussk. Erving Location, roadside bank, Harris 27803 (65). A fourth town- ship for the County. I normally take a dim view of reporting every white-flowered form seen of common species but this produced the most spectacular floral display I found in Erving Location. The bank was covered entirely by tall plants of this color form for a distance of about 150 feet. Panax trifolius L. Dummer, along road up Dummer Hill, Harris 27180 (65). The third township in the County for this species. Convolvulus sepium L., forma coloratus Lange. Berlin, roadside rte. #16, Harris 27625 (65). Prof. Pease did not differentiate be- tween this form and forma sepium. In his citations for the species the following are definitely forma coloratus: Colebrook 10887 — M 3927 (07); Northumberland 12212 (09); Lancaster 12122 (09); Ber- lin 16127 (14); M Location 12186 — M 4074 (07); Whitefield Deane (96). Mentha arvensis L., forma lanata (Piper) S. R. Stewart, Stark, ledges Crystal Falls, Harris 28354 (65). The second collection for the County. Veronica peregrina L., var. peregrina. Jefferson, weed in beds of Mountain Valley Nursery, Harris 27703 (65). New to the County. Rhinanthus Crista-galli L. Randolph, Bog Dam road between Camp 18 and Camp 19, Hurris 27713 (65). Species previously known only from Pittsburg. Rudbeckia serotina Nutt., forma homochroma (Steyerm.) Fern & Schub. Dummer, our field Dummer Hill, Harris 27851 (65). New to the County. Rudbeckia serotina, forma pleniflora (Moldenke) Fern. & Schub. Dummer, our field Dummer Hill, Harris 27852 (65). Second collection from the County. 34 Rhodora [Vol. 69 Leontodon autumnalis L., var. pratensis (Link) W. D. J. Koch. G Grant, Pease 40005 (62). New to County. Hieracium paniculatum L., forma glandulosum R. Hoffm. Dummer, Dummer Hill, Harris 28231 (65). Second collection for the County. STUART K. HARRIS, BOSTON UNIVERSITY STUDIES IN THE EUPATORIEAE, (COMPOSITAE) E. — HI R. M. KING Since the death of Prof. B. L. Robinson over thirty years ago, the tribe Eupatorieae has not been critically studied. During this period great numbers of collections, much re- search, and many modern techniques have become available to botanists. The present study initiates an attempt to in- corporate these new materials and methods in formulating a modern conspectus of the genera and species of the tribe. Grateful acknowledgement is extended to the directors of the following herbaria who generously loaned me this ma- terial: A, CAS, DS, G, GH, K, MICH, NY, P, PH, POM, RENO, RM, RSA, SMU, UC, US, USFS, and W. I. PLEUROCORONIS R. M. King and H. Robinson, Fhytologia 12: 468-469. 1966. (Type species, Hofmeisteria pluriseta A. Gray.) Woody sub-shrub. Leaves simple to compound, petiolate, serrate, opposite in lower portions of plant becoming alternate above. Pedun- cles not strongly differentiated, bearing alternate leaves. Inflores- cence polycephalic. Heads turbinate or cylindrical, 25-30-flowered Phyllaries in multiple series, outer one short, ovate, inner lanceolate. Corollas slender, regular, 5-lobed, white, tubular, glandular on the outer surface. Anthers elongate with large erect appendages, exothe- cial cells isodiametric or slightly wider than long. Pollen spherical, tricolpate, distinctly spinose. Style branches elongate, slightly en- larged at the tip, cells at the surface slightly bulging; basal node of style glabrous. Pappus of 3-16 long scabrate setae with intervening short erosely dentate squamae, sometimes bearing numerous addi- tional setae in a distinct inner series. Achenes narrowly ellipsoid to obpyramidal, 4-5-ribbed, lateral surfaces densely pubescent, sub- cuticular cells without minute punctations except on the ribs; car- popodium somewhat rounded, usually asymmetric, cells vertically elongate, basal foramen shallow, indistinct. Basic chromosome num- ber, as determined from one species, X — 9 pairs. The species of Pleurocoronis can be distinguished by the following key Leaves. dissected |... seen errant roro aot ERR sledge P. gentryi Leavos Simple ee qeapumaesene seruo ro on nano soon) SHE e Rea LR Ch UP Pha 36 Rhodora [Vol. 69 Petioles more than twice as long as leaf blades, leaves small, narrowly ovate with usually five or less teeth ............... eee P. pluriseta Petioles always less than twice as long as leaf blades, leaf blades deltoid to reniform, usually with 15 or more marginal teeth ............ MM P. laphamioides [)1:$2$9 Figs. 1-9, Leaves of Plewrocoronis. Figures 1-4, P. pluriseta. 1. Orcutt 1889 (US). 2. Woodcock 1889 (US). 3. Coville & Funston 207 (us). 4. Brandegee 1889 (US). Figures 5-9. P. laphamioides. 5. Mason 1975 (us). 6. Wiggins 17080 (US). 7. Wiggins 17080 (US). 8. Johnston 3630 (US). 9. Palmer 208 (US). 1. Pleurocoronis gentryi (Wiggins) R. M. King & H. Robinson Phyto- logia 12: 470. 1966. Hofmeisteria gentryi Wiggins Contr. Dudley Herb 4: 25. 1950. MEXICO: BAJA CALIFORNIA: Sierra Giganta above Puerto Escondido, April 21, 1988. H. S. Gentry 3742 (Holotype ps! Isotype GH!). Plants 4-6 dm tall. Stems few-branched, striate, gray-barked with dense glandular hairs on the younger portions; evanescent in older regions. Leaves thin. Petioles glandular, slender 12-15 mm long. Blade broadly ovate in outline 12-22 mm wide, 15-25 mm long, bipin- nately or tripinnately divided; densely hispid on both surfaces. Heads turbinate, 10-12 mm tall (including style branches) ca 25-flowered. Phyllaries ca 25, mostly 3-nerved. 2-7 mm long, 0.5-1 mm broad. Florets 5-6 mm long. Stamens 5. Anthers (including appendages) ca 1.5 mm long ca 240 u wide, appendage ca 385 u long ca 220 u wide. Filaments 2.5-3 mm long ca 50 u wide. Style branches purplish, great- ly exserted at maturity ca 2.5-3 mm long ca 225 u wide. Setae 4-6, 5-6 mm long. Squamae 4-6, 1-1.5 mm long. Achenes narrowly pris- matic, dark brown ca 3 mm long. Pollen ca 18 uw in diameter. Chromo- some number not determined. SPECIMENS EXAMINED: MEXICO: BAJA CALIFORNIA: 10 mi s of Mission Dolores, Wiggins, Carter, & Ernst 300 (DS, MICH). Arroyo de San Matias near Rancho San Matias, Carter & Sharsmith 4207 (uc). Arroyo Gabilan, Carter & Ferris 4080 (UC). vicinity of Porte- zuelo, e of La Victoria, Carter & Ferris 3910 (UC). Pilon de las Parras, w of Loreto, Carter & Sharsmith 4236 (UC). Arroyo Tabar, w of Puerto Escondido, Carter & Sharsmith 4247 (UC). Mesa de Alta Gracia, sw of La Cumbre de Alta Gracia, Carter 4890 (UC). 2. Pleurocoronis pluriseta (A. Gray) R. M. King & H. Robinson. 1967] Eupatorieae — King 37 TAi BAJA CALIFORNIA, SUR, MEXIC Plate 1338. Pleurocoronis gentryi Holoytpe (Ds) Gentry 3742. Phytologia 12: 469. 1966. Hofmeisteria pluriseta A. Gray, Pacif. Rail. Rep. 4: 96, 1857. UNITED STATES: Arizona: Bill Williams Fork, J. M. Bigelow in 1853-54. (Holotype GH! isotypes; NY!; P!; PH!). Hofmeisteria viscosa Nelson Bot. Gaz. 37: 263. 1904. UNITED STATES: NEVADA: “The Pockets”, April 30, 1902, Leslie N. 38 Rhodora [Vol. 69 Fig. 10. Plewurocoronis pluriseta (From A. Gray, Pacific Rail. Rep. 4: 96, 1857). Goodding 671 (Holotype RM!). Plants 2.5-5 dm tall. Stems few-branched, erect, striate, with eva- nescent puberulence. Leaves thin. Petioles with glandular hairs; nar- row, 2-5 cm long. Blades glabrous or rarely with glandular hairs, 1967] Eupatorieae — King 39 ovate to lanceolate, 1.5-5 mm wide, 3-10 mm long, entire or more frequently with 5 or less teeth. Petioles always twice as long as the blades. Inflorescence a loose corymb. Heads 6-11 mm tall (including style branches), ca 30-flowered. Phyllaries ca 25, outer surface glandular; striate, the apex acute, 0.5-1 mm broad, 2.5-6 mm long. Florets 4-5 mm long. Stamens 5. Anthers (including appendages) ca 1.5 mm long, ca 220 u wide, appendage ca 220 u wide, ca 330 u long. Filaments ca 2 mm long ca 65 u wide. Style branches, ca 1.5 mm long. Squamae 10-12, 1-2 mm long. Setae 10-12(16), 2.5-5 mm long. Achenes narrowly prismatic; dark brown to black, 3-4 mm long. Pollen ca 18 » in diameter. Chromosome number N = 9 II (Raven unpublished). REPRESENTATIVE SPECIMENS EXAMINED: UNITED STATES: ARIZONA: COCONINO CO.: Havasupai Canyon, Clover 6085 (MICH). Separation Canyon, Clover 4219 (MICH). MOHAVE Co.: Grand Canyon National Monument, W. P. Cottam. May 5, 1952 (CAS). YUMA co.: Tule Tank, Shreve 5936 (ps); Benson & Darrow 10804 (POM). Palm Canyon, Benson & Darrow 10853 (POM). Sierra de la Gila, Gila Desert, Monnet 1088 (CAS, P, US). CALIFORNIA: IMPERIAL CO.: Painted Gorge, Carisso Mts., Ferris & Rossbach 9633 (DS, F, GH, RSA, UC). Between Coyote Wells and Jacumba, MacFadden 14461 (CAS). 4.9 m n e of Mountain Springs, Raven 11550 (SMU). Indian Picture Rocks, F. M. Reed 5358 (P). INYO CO.: Pleasant Canon, Hall & Chandler 6917 (POM, UC). Surprise Canyon, Ferris 7994 (A, DS, POM, UC). Titus Canyon, Eastwood & Howell 7773 (cas). Echo Canyon, Coville & Gilman 146 (US). Death Valley, Coville & Funston 207 (us). Inyo Mts., Austin 461 (UC). East of Ophis Mine, Slate Mts., Alexander & Kellogg 1133 (ps, UC). Darwin Springs, C. L. Hitchcock 6208 (Ds, NY, RSA, UC) Saline Canyon, Train 671 (US). RIVERSIDE CO.: Andreas Canon Stark 4983 (RSA, TEX, UC). Painted Canyon, Peirson 7173 (RSA); Munz & Keck 9346 (GH, POM); McGregor 735 (ps); Ferris & Rossbach 9572 (DS, F, GH, MICH, RSA, UC). Tahquitz Canyon, Peirson 1213 (RSA). 5 m s w of Palm Desert, Breedlove 1887 (Ds). Cotton- wood Springs, Stark 3751 (DS, RSA, UC, US). Palm Springs, Parish 4122 (DS, GH, NY, PH, UC, US). Palm Canyon, F. M. Reed 3866 (uc) 3870 (UC). Marshall Cañon, 11 km w of Coachella, Hall 5795 (DS, F, G, GH, NY, PH, POM, UC). Near Indio, Hall 5992 (Ds, G, UC, US). Joshua Tree National Monument, Hitchcock & Muhlick 23261 (uc). Eagle Canon, 4 m east of Palm Springs, E. C. Jaeger March 18, 1922 (ps). Devil's Canyon, Johansen & Ewan 7107 (DS, UC). SAN BERNARDINO co.: 3.8 mi e of Hayden, Wolf 10753 (DS, G, NY, RSA, TEX, UC). Carlyle Mine, Pinto Mts., east of 29 Palms, Munz 15648 (DS, GH, POM, UC, US). Calico, Parish 9813 (A, DS, G, GH, UC). Sheep Springs Canyon, Hitch- cock & Muhlick 23293 (Ds, G, UC). Camp Cady, Hall 6130 (ps, uc). 39 m from Needles, Ferris 7230 (DS, GH, NY, POM). 1 m s of 29 Palms, Alexander & Kellogg 952 (DS, vc, US). Whipple Mts., Alexander & Kellogg 1226 (ps, UC). 10 m NW of Riggs, Munz & Hitchcock 10957 40 Rhodora [Vol. 69 (POM). Painted Canon, Howell 3299 (CAS). 10 m e of Daggett, Munz & Keck 7864 (NY, POM). 17 m from Vidal on old road to Needles, Wolf 3190 (DS, RSA, UC, US). SAN DIEGO CO.: 15 mi e of Banner, Gould 2270 (RSA, SMU, TEX). Palm Canon, Gander 5320 (SMU, UC) ; Epling & Robison April 8, 1932 (Ds, RSA, UC, US). San Felipe Wash, Fast- wood 2279 (CAS, GH, POM, UC, US). The Narrows, Gander 9212 (RSA). NEVADA: CLARK CO.: Valley of Fire, Maguire, Maguire & Maguire 5011 (GH, UC) ; 5012 (GH, UC) ; Clokey 5954 (DS, G, MICH, NY, PH, POM, RENO, SMU, TEX, UC, US, USFS, W); Train 1885 (DS, F, RENO, UC); Eastwood & Howell 9010 (CAS, NY, UC). Black Canyon, 5 m s of Hoover Dam, Ferris 13384 (DS, NY). North of Cottonwood Peak, Gullion 537 (UC). Petrified forest, Canyon w of Logan, Heller 10460A (A, DS, F, G, GH, NY, US). Hiko Wash, Train 1370 (RENO, SMU, UC). Hemingway Wash., Lake Mead, Maguire 17724 (POM, UC). Lake Mead, Clover 8234 (MICH, SMU). LINCOLN CO.: Muddy Valley, Kennedy & Gooding 54 (DS, NY, PH, UC, US). UTAH: “southern part of state" Dr. Edward Palmer in 1877 (Ny, US). MEXICO: BAJA CALIFORNIA: Cucopa Mts., MacDougal 127 (Ny). San Luis Gonzaga Bay, I. M. Johnston 3326 (CAS, GH, NY, UC, US). El Canon de los Osos, 4 m n of Gaskills Tanks, Epling & Darsie February 21, 1933 (uc). Santa Maria, T. S. Brandegee May 15, 1889 (F, GH, PH, UC, US). % m s of Puertocito, Wiggins & Wiggins 15867 (DS, MICH, TEX). SONORA: Cholla Bay, Punta, Penasco, Raven 11665 (NY). 3a. Fleurocoronis laphamioides (Rose) R. M. King & H. Robinson var. laphamioides Phytologia 12: 470. 1966. Hofmeisteria laphamioides Rose, Contr. U. S. Nat. Herb. 1:79, 1890. MEXICO: BAJA CALIFORNIA: San Pedro Martir Island, February 13, 1890, Edward Palmer 148 (holotype US!; isotypes CAS!; F!; GH!; NY!). Hofmeisteria pluriseta A. Gray var. laphamioides I. M. Johnston Proc. Calif Acad 4. 12: 1186. 1924. Plants 0.5-1.5 m tall. Stems often branched, erect, striate, glabrous or rarely with glandular hairs. Leaves thin. Petioles glandular, narrow 1-2.5 cm long. Blades glandular, deltoid to reniform or orbic- ular, 4-28 mm wide and long, margin indented usually with 15 or more teeth. Petioles always less than twice as long as blades. In- florescence in corymbs or cymes, Heads 8-12 mm tall (including style branches), ca 30-flowered. Phyllaries ca 25, striate, 0.5-1 mm broad, 2.5-6 mm long. Florets 4-5 mm long. Stamens 5. Anthers (including appendages) ca 1.5 mm long, ca 220 u wide, appendage ca 220 „u wide, ca 330 u long. Filaments ca 2-2.5 mm long ca 65 u wide. Style branches ca 1.5 mm long, ca 250 u wide. Squamae 10-12, 1-2 mm long. Setae 10-12, 2.5-5 mm long. Achenes narrowly ellipsoid; dark brown to black, 3-4 mm long. Pollen ca 18 u in diameter. Chromosome number not determined. 1967] Eupatorieae — King 41 i SOM m t * $ is HERB. ONT SPARES DEFARTMESI GBOGGEHTETERE Plate 1339. Pleurocoronis laphamioides Holotype (us) Palmer 148. SPECIMENS EXAMINED: MEXICO: BAJA CALIFORNIA: Salsi- puedes Island, /. M. Johnston 3521 (CAS). San Marcos Island, I. M. 42 Rhodora [Vol. 69 Johnston 3630 (CAS, GH, NY, UC, US); Moran 3955 (DS, UC). Angel de La Guarda Island, I. M. Johnston 3376 (cas). San Pedro Martir Island, J. M. Johnston 3157 (CAS, F, GH, K, UC, US); I. M. Johnston 3162 (CAS, DS, F, GH, NY, UC, US); Dr. Edward Palmer 406 in 1887 (GH, US). San Bartolome Bay, Rose 16213 (Ny, US). Las Animas Bay, I. M. Johnston. 3516 (CAS, GH, NY, UC, US). San Francisquito Bay, Rose 16755 (NY, P, US). Santa Rosalia, Dr. Edward Palmer 208 in 1889 (GH, US). 10 m e of La Purisima, Wiggins 11459 (GH). Bahia de los Angeles, Wiggins & Wiggins 14889A (pns). Bahia de los Angeles, Wiggins & Wiggins 14820 (CAS, DS, MICH, TEX). Cerro Tordillo and vicinity, Gentry 7452 (DS, MICH); Gentry 7453 (DS, MICH). vicinity of Aguaje de San Andreas, Gentry 7463 (DS, MICH). Rinconado de San Hipolito, Gentry 7805 (ps). Las Tinajas, Gentry 7545 (DS, MICH) 4 m s of Barrill, Harbison 41602 (RSA). SONORA: Kino Point Long 80a (GH, US). Libertad, Graham 3831 (ps). Punta Sargento, Wiggins 17080 (US). San Pedro Nolasco Island, Moran 4043 (DS). 3b. Pleurocoronis laphamioides (Rose) R. M. King & H. Robinson var. pauciseta (I. M. Johnston) R. M. King, comb. nov. Hofmeisteria pluriseta var. pauciseta I. M. Johnston, Pro. Calif. Acad. 4. 12: 1187. 1924. MEXICO: SoNoRA: San Pedro Nolasco Island, April 17, 1921. J. M. Johnston 3134 (Holytype CAS! ; Isotypes GH! ; NY!; UC!; us!). Hofmeisteria laphamioides Rose var. pauciseta (I, M. Johnston) Blake, Contr. U. S. Nat. Herb. 23: 1430. 1926. The variety pauciseta differs from the variety laphami- oides in having only 5 or 6 setae on the pappus instead of 7-16. SPECIMENS EXAMINED: MEXICO: BAJA CALIFORNIA: Isla Carmen, Wiggins 17512 (us). Espiritu Santo, Rose 16868 (US). Purisima, Brandegee in 1889 (NY). 10 m e of La Purisima, Wiggins 11459 (ns). SONORA: San Pedro Nolasco Island, Moran 4043 (UC). The following collections may be regarded as intermediate between P. pluriseta and P. laphamioides and I have annotated the following specimens as P. pluriseta: UNITED STATES: ARIZONA: PIMA co.; Dripping Wells, Puerto Blanco Mt., Gould, Darrow & Haskell 3019 (CAS, GH). MEXICO: BAJA CALIFORNIA: Isla Angel de La Guarda, Moran 7254 (DS). Turtle Bay, Revillagigedo Islands, Mason 1975 (F, K) and I have annotated the following specimens as P. laphamioides. UNITED STATES: ARIZONA: PIMA C0.: Dripping Springs, Darrow, Haskell, & Gould 2435 (CAS). McClintock 52-80 (cas). Dripping Wells, Puerto Blanco Mts., Gould, Darrow & Haskell 3019 (vc). MEXICO: BAJA CALIFORNIA: Turtle Bay, Revillagigedo Islands, Mason 1975 (CAS, Us). SoNoRA: 7 m n w of Quitovac, Wiggins 8328 (DS, MICH, US). 1967] Eupatorieae — King 43 M. tenuis C. anomalochaeta Fig. 11. Distribution of Pleurocoronis, Malparia, and Cartero- thamnus. MALPARIA S. Watson, Proc. Amer. Acad. 24: 54, 1889 (Type species, Malparia tenuis S. Watson). Slender annual to slightly woody sub-shrub. Leaves simple, narrow, mainly sessile, sub-entire, opposite in lowermost nodes becoming alter- nate above. Peduncles not strongly differentiated. Inflorescence poly- cephalic. Heads nearly cylindrical few, rather small, ca 30-flowered, in loose cymose panicles. Phyllaries in several series, lanceolote, the tips acute to acuminate. Receptacle flat, naked. Corolla slender, regular, tubular, 5-lobed, glandular on the outer surface, Anthers elongate with large erect appendages. Pollen spherical, distinctly spinose. Style branches filiform, thickened toward apex and exserted at maturity. Pappus in a single series, squamae overlapping setae. Achenes slender, prismatic, 5-ribbed. Carpopodium with a very large foramen, often born laterally. Cells of carpopodium quadrate. Chromosome number determined as X — 10 pairs. (Raven unpub- lished) 44 Rhodora [Vol. 69 1. Malparia tenuis S. Watson Proc. Am. Acad. 24: 54. 1889 MEXICO: BAJA CALIFORNIA: Near Los Angeles Bay, Dec. 1887 Dr. Edward Palmer 567 (Holotype, GH!; Isotypes NY!; US!). Hofmeisteria tenuis (Wats.) I. M. Johnston Proc. Cal. Acad. Sci. 4, 12: 1188. 1924. Plants 1-4 dm tall. Stems erect, striate, often reddish, sparingly branched to very branched in certain large individuals, finely pu- berulous to nearly glabrous. Leaves puberulous to nearly glabrous, up to 5.0 em long, 0.3-3.0 mm wide. Heads 5-6 mm in diameter, 7.0 i 1 Plate 1340. Malparia tenuis Holotype (GH) Palmer 567. 1967] Eupatorieae — King 45 mm to 1.4 cm tall (including the projecting florets). Phyllaries 20-25, puberulous, 12-15 mm wide, 0.4- 1.2 cm long. Florets 20-30, ochro- leucous or whitish; corolla narrow, tinged with pink, 5-6 mm long; stamens 5, anthers 2 mm long (including appendage) ca 0.25 long; wide, pappus setae 3, reddish brown, 5 mm long, pappus squamae 3, reddish brown, 0.5 mm long, pollen spherical, ca 18 ,& diameter, achenes dark brown or black, 3 mm long. Chromosome number deter- mined as n — 10 pairs. (Raven unpublished) REPRESENTATIVE SPECIMENS EXAMINED: UNITED STATES: CALIFORNIA: IMPERIAL CO.: upper end of Painted Gorge, Carisso Mts., Ferris & Rossbach 9614 (DS). Fish Mt. E. C. Jaeger April 10, 1926 (PoM). Split Mt. Canyon, Fish Creek Mts., Ferris 9696 (DS). Split Mt., Colorado Desert, T. S. Brandegee April 1905 (Ds, UC, US). Mt. Signal, Abrams 3189 (DS, NY, POM); E. M. Wilson April 3, 1949 (DS). SAN DIEGO CO.: Yagin Pass, Borrego Valley, Ripley & Barneby 10077 (CAS). MEXICO: BAJA CALIFORNIA: Bahia de San Luis Gonzaga, Wig- gins & Wiggins 18166 (CAS, DS, MICH). 18 m s of San Filipe, Wiggins &Wiggins 15815 (DS, MICH, TEX). 35 m n of San Felipe, Wiggins & Wiggins 18161 (CAS, DS). 15 m n of San Felipe, Wiggins & Wiggins 15785 (Ds). b m w of Barril, Wiggins 7830 (DS, F, GH, MICH, NY, UC, US). near Barrie, Shreve 6991 (MICH, PH, UC). Puerto Refu- gio, Angel de la Guarda Island, Copp 135 (ps). Angel de la Guarda Island, J. M. Johnston 4205 (CAS). 2.5 m w of Agua Dulce, Moran 10215 (DS). Las Tinajas, Gentry 7609 (DS, MICH, UC). 65.5 m s of San Felipe on the road to Laguna Chapala, Wiggins & Ernst 697 (DS, MICH) 3 m n of San Felipe, Olmsted 1167 (RSA). 6 m s e of Tres Enriques, Thorne & Henrickson 32579 (MICH, RSA). 10 m n of Santa Rosalia, Reed 6261 (POM). 2.7 m n of San Felipe, Raven 14764 (DS, RSA). Cucopa Mts., MacDougal 175 (NY). Bahia de Los Angeles, Cowan 2304 (Us). San Francisquito Bay, I. M. Johnston 3563 (CAS, GH, K, NY, UC). SoNoRa: Libertad Bay Inn, Long 73 (US). CARTEROTHAM NUS R. M. King gen. nov. (Compositarum — Eupatorieae — Eupato- riinae) Suffrutex; Capitula homogama discoidea; flores omnes her- maphroditi, fertiles, regulares; antherae ad apicem appendiculatae; receptaculum paleaceum; achaenia 5 costata. Species typica: Carter- othamnus anomalochaeta R. M. King Carterothamnus anomalochaeta R. M. King sp. nov. Suffrutex; pauci-ramosus, erectus; folia alterna, lamina tenuis, orbicularis vel reniformis, profunde fissa, 1-5-2 cm lata, 1-1.25 cm longa, petiolus usque ad 1.5 em longus; inflorescentia simplex; receptaculum paleis caducis inter flores instructum; involucri bracteae 30-40, multi- seriatae, aequales, usque ad 2.5 mm longae, paleis ca 25, ca 4 mm longis; corolla glabra; pollen asperatum; achaenia 5 angulata, pris- 46 Rhodora fVol. 69 matica, 1-1.2 mm longa; pappus polysetaceus, seta unica elongata ceteris brevibus. Scandent subshrub; stems few-branched, erect, terete, striate, glandular pubescent in younger portions; leaves alternate, blades thin, orbieular or reniform, 3-7 cleft two thirds or more in to base, lobes oblong, glandular pubescent on both surfaces, 1.5-2 cm wide, 1-1.25 em long, petioles up to 1.5 em long; inflorescence monocephalic; peduncles terete, striate, minutely glandular, up to 3 cm. long, rarely with setaceous bracts; heads discoid, campanulate, ca 7 mm high (including style branches), ca 7 mm wide, ca 50 flowered; receptacle convex with a slight depression in the center, paleaceous; phyllaries 30-40, in 3-4 series, thin, unequal, glandular on the outer surface, ovate-lanceolate, apex acute, margins erosely dentate, ca 1 mm wide, up to 2.5 mm long; pales ca 25, narrowly lanceolate, ca 4 mm long; florets perfect, glabrous fertile, regular, 5 lobed, white?, sta- mens 5; anthers appendaged, obtuse, ca 100 u wide, ca .75 mm long; style branches 2, greatly exserted at maturity, club-shaped, 1.5-2 mm long; style tube ca 1.5 mm long; pollen tricolpate, spherical, distinctly spinose, ca 13 u in diameter; achenes prismatic, 5-ribbed, ribs sectors 1-1.2 mm long, dark brown; carpopodium symmetrical, basal foramen prominent with a clean margin, cells quadrate, rather thick-walled; pappus in one series, consisting of about 10-12 short setae 0.5-0.75 mm long and one long barbed seta ca 2.5 mm long; ehromosome number not determined. Carterothamnus possesses two characteristics that would indicate relationship with Hofme'steria, its monocephalic inflorescence and glabrous corolla. Carterothamnus differs from Hofmeisteria in the shape of its carpopodium, surface of its pollen, number of flowers in each head, shape of the phyllaries, and especially in the possession of a paleaceous receptacle. Paleaceous receptacles are rare in the tribe Eu- patorieae, being found in only 3-4? other genera. The pap- pus of C. anomalochaeta seems unique in the Eupatorieae. It is approached only by certain species of Stevia which have one long seta and several short squamae. The genus is named in honor of Annetta Carter, distin- guished botanist of the University of California at Berkeley. Holotype: Annetta Carter 4812. (US mo. 2, 466, 373) Arroyo del Potrero, between Portezuelo de la Cuesta de los Dolores and Rancho Agua Escondido (ca 90 km easterly from Villa Constitucfon). Sierra de la Giganta, Baja California, Sur, Mexico, ca. Lat. 25° 06' N., Long. 110? 57' — 110° W. Altitude ca 450 meters, A scandent clump on north facing cliff. In the “narrows” of the canyon. 19 October 1967] Eupatorieae — King 4T "3 i Sk SE BAJA CALIFORNIA, SUR, MEXICO Sierra de la Giganta à acandast elump on north-facing elitr, . del Poorern, betwgen Porterae! : SE Plate 1341. Carterothamnus anomalochaeta Holotype (Us) Carter 4812. 1964. Isotype at UC. Paratype: same locality Carter 5147 July 17-20 1966. CATONSVILLE COMMUNITY COLLEGE CATONSVILLE, MARYLAND THE IDENTITY OF PSORALEA MULTIJUGA ELL. (LEGUMINOSAE). ROBERT L. WILBUR' It was with a sense of guilt that I noted in Barneby's most impressive treatment of the North American species of Astragalus (Mem. N.Y. Bot. Gard. 13: 544-545. 1964.) that the identity of Psoralea multijuga Elliott was still in question nine years after I had examined the type. Vail (Bull Torrey Club 21: 118. 1894) reported that Elliott's type had been examined by both Drs. Britton and Small and had “proved to be Astragalus glaber Michx.” Astragalus glaber Michx. is a later homonym of an Old World species named by Lamarck in 1783; the southeastern species is now called Astragalus michauxii (Kuntze) F. J. Hermann. Small (Man. Se. Fl. 710. 1933.) listed P. multi- juga Ell in the synonymy of Tium michauxii (Kuntze) Rydb. Weatherby (Rhodora 44: 249. 1942.) in a paper con- cerned with the types in Elliott’s Herbarium felt that in spite of its resemblance to the species now known as A. michauxii, its inflorescence appeared quite different. Wea- therby concluded that he **did not recognize it; [and that] : it should have further investigation." Barneby (op. cit. p. 545.) felt that ‘‘Elliott’s description of its ‘trailing stem’ and especially its monospermous germ are impossible to reconcile with the erect growth habit and 30-40 ovules" of A. michauxii. He pointed out that since there was an Astragalus multijugus DC. dating from 1825, there was little likelihood that Elliott's name could affect the nomenclature of this or any other Astragalus. In pass- ing it might be noted that Elliott did not describe P. multi- juga as possessing a "trailing" stem but one "apparently 1-2 feet high" and, lacking fruit, he merely surmised from the appearance of the young "germ" [--ovary] that it ‘Grateful acknowledgement is made for NSF Grant 18799 which was held while this paper was prepared. 48 1967] Psoralea — Wilbur 49 would be 1-seeded. Elliott apparently had no first hand familiarity with any of the included species of Astragalus with the possible exception of A. villosus. Even A. cana- densis was described as prostrate. I had the privilege of examining the type specimen of P. multijuga in the Charleston Museum (CHARL) in Au- gust 1957. The specimen there bears a manuscript name, as reported by Weatherby, with the genus questioned. It also bears the name of the Mr. Gourdine who had collected it "some years ago" in the vicinity of Abbeville, South Carolina. Both the collector and the locality were indicated by Elliott in his Sketch. Elliott's P. multijuga is not Astragalus michauxii (=A. glaber Michx.) but A. canadensis L. The specimen is the top of a young plant about 2 dm long with two flowering racemes and two more with flowers in bud. The calyx lobes of the type are about 3.5-3.7 mm long and about twice as long as the moderately appressed villous calyx tube (c. 1.5-1.7 (2.2) mm long). The leaflets are sparsely strigillose beneath and the stipules are about 6.5 mm long. Elliott reported that "under the microscope" the leaves beneath were "apparently covered with minute black glands." The leaves do appear minutely mottled but the spots are defi- nitely not psoralioid glands but appear to be cellular occlu- sions which caused some puckering upon drying. The specimen hence seems to fall within the range of variation allowed in A. canadensis L. or A. canadensis var. caro- linianus (L.) M. E. Jones. However, as previously reported (N.C. Agri. Expt. Sta. Tech. Bull. 151. p. 225. 1963.), the proposed varieties of this species in eastern North America lack discrete geographical ranges and the variation of morphological characters upon which the varieties were based seems almost continuous. Barneby (op. cit. p. 604.) stated that there “is evidently some correlation between certain types of variation and dispersal patterns, but no reliable method of separating named geographic races . . bad Elliott (Sk. Bot. S.C. & Ga. 2: 225-226. 1822.) included five species of Astragalus (A. caroliniamus, A. canadensis, A. glaber, A. obcordatus and A. villosa as a species of 50 Rhodora [Vol. 69 Phaca). 'There are specimens of at least the first four species in Elliott's herbarium but it is not surprising that he failed to identify the imperfect type specimen of P. multijuga with the two scrappy specimens from Muhlen- berg now in his herbarium representing A. carolinianus and A. canadensis. DUKE UNIVERSITY DURHAM, NORTH CAROLINA SOUTHEASTERN LIMIT OF FAGUS GRANDIFOLIA DANIEL B. WARD' The American beech, Fagus grandifolia Ehrh., occurs throughout eastern North America from eastern Wiscon- sin and Upper Michigan to Cape Breton Island, Nova Scotia, south to eastern Texas and western Florida (Little, 1953; Fowells, 1965). It reaches its southeastern limit in a series of isolated populations in Columbia County and northern Alachua County, Florida, as noted by Kurz & Godfrey (1962). These isolated stands have now been found to be somewhat more extensive than previously be- lieved, and without exception to be in a state of rapid de- cline. The source of a recent erroneous extension of range into central peninsular Florida has been traced. American beech apparently is entirely absent from the lower Coastal Plain of southeastern Georgia (Duncan, 1950). It is common in southwestern Georgia and in the panhandle of Florida west of Leon County, but is again lacking in the low, pine flatwoods fringing the Florida Gulf Coast. East of Leon County beech seemingly does not occur for approximately 70 miles, to within a short distance of Lake City, Columbia County. The absence of an otherwise widespread species in the area between Leon and Columbia counties is a frequent feature of plant and animal distribution in Florida. This area, bordered to the east by the Suwannee River and containing toward the west the much smaller Aucilla River, was occupied during Pleistocene interglacials by an embayment of the Gulf of Mexico (MacNeil, 1950). At times of maximum inundation this embayment extended across southeastern Georgia, isolating peninsular Florida from the mainland by the “Suwannee Strait.” The disjunct populations of American beech in Columbia and Alachua counties are all south and east of this long- '] am indebted to W. D. Rice of Lake City for guidance to several of the beech stands noted here and to Marshall Crosby of Gaines- ville for assistance with the measurement of the Santa Fe grove. 51 52 Rhodora [Vol. 69 vanished barrier. They are not widely scattered, but occur rather in two clusters, near Lake City, Columbia County, and between Bland and Sante Fe, Alachua County’. All Columbia County stands are small and show little or no evidence of natural reproduction, A stand in Lake City, well known around the turn of the century and the probable source of several early collections (the earliest — A. W. Bitting, 22 June 1892 [FLAS]), is now reduced to a few trees near Alligator Lake. Scattered trees are said to occur on the Doug Epperson farm about 2 miles west of Lake City. Two trees, the larger 18 inches dbh., have been seen along the bank of a dry ox-bow channel of Falling Creek, 5 miles northwest of Lake City. Further south, in Alachua County, two stations are known for American beech. One, on the Carl Doke farm, one-half mile east of Bland, consists of a solitary large tree; although fruits are produced abundantly, no seedlings have been observed. The second Alachua County stand forms what is believed to be the southeasternmost stand for American beech. It covers several acres along the slopes and on the precipitous banks of a small, forking stream, a tributary of the Santa Fe River, north and south of Florida Highway 236, 2 miles west of the town of Santa Fe. Over most of the area the trees are widely scattered, but along a portion of the stream course and on the terrace between two small forks it is the dominant species, to the near exclusion of other woody plants. Within the Santa Fe stand 115 trees have been measured "The range of Fagus grandifolia appears in the past to have ex- tended at least 40 miles south and 35 miles east of the present southernmost station. Dr. W. A. Watts, Trinity College, Ireland (pers. comm.), in analyzing a twelve meter core from the bed of Mud Lake, east of Ocala, Marion County, has found indisputable Fagus pollen in quantities too large to attribute to long-distance wind drift. Between the 11 and 12 meter levels Fagus made up one to two percent of the tree pollen, and trace amounts were found up- ward in the core to a level of approximately 5.8 meters. Watts esti- mates that these horizons date far back into the Pleistocene, and are very probably older than the last interglacial period. 1967] Fagus grandifolia — Ward 53 and mapped. The tally is not exhaustive for specimens in the smaller diameter classes, but is probably close to com- plete for the larger individuals. Excluding those under 4 inches dbh., the average diameter for the remaining 97 was 14.8 inches. Three of these were greater than 25 inches in diameter, with the largest attaining 27.4 inches. The distribution of individual trees in diameter classes ap- proximated a normal curve, with far fewer small individ- uals than would be anticipated in a vigorously reproducing population; only 7 trees were tallied with diameters be- tween 2 and 6 inches. The distance and time by which these populations have been separated from the extensive stands of the Pleistocene mainland have not resulted in consequential morphological divergence. W. H. Camp, during his never-published study of variation within Fagus grandifolia, annotated Lake City material with a subspecific designation corresponding to var. caroliniana (Loud.) Fern. & Rehder, the Coastal Plain variant sometimes known as white beech. He simi- larly annotated specimens from throughout West Florida, including isotypic material of forma mollis Fern. & Rehder from Tallahassee, Leon County. A recent mapping of the distribution of American beech (Fowells, 1965) has indicated a disjunct station in west- central peninsular Florida. Through the courtesy of E. L. Little, this record was learned to be based upon a specimen on deposit at the University of North Carolina (UNC). This specimen has since been examined; it is labeled “Crystal Springs, Pasco County, Florida, April 18, 1915, T. G. Har- bison 11829," and it is indisputably American beech. A Crystal Springs does exist in Pasco County, but neither beech nor the association in which it is likely to occur is to be found there at the present time. With the assistance of H. E. Ahles, Harbison's 1915 field books were perused. In that year Harbison did travel through Florida, as well as other southern states, and his notes do not always spec- ify his whereabouts on a given date. However, his rail- road schedules, livery rentals, certain non-Florida or non-Peninsular species collected simultaneously with the 54 Rhodora [Vol. 69 Fagus (Quercus borealis, Carya pallida), and other indi- cations of his route, clearly demonstrate that Harbison's Crystal Springs was the town by that name in Copiah County, Mississippi, where American beech occurs abun- dantly. DEPARTMENT OF BOTANY UNIVERSITY OF FLORIDA, GAINESVILLE LITERATURE CITED DUNCAN, W. H. 1950. Preliminary reports on the flora of Georgia. 2. Distribution of 87 trees. Am. Midl. Nat. 43: 742-761. FowEkE.Ls, H. A. 1965. Silvies cf forest trees of the United States. U.S. Dept. Agr. Handb, 271: 172-180. Kurz, H., & R. K. GODFREY. 1962. Trees of Northern Florida. Univ. of Florida Press, Gainesville, 311 pp. LITTLE, E.L. 1953. Check list of the native and naturalized trees of the United States. U. S. Dept. Agr. Handb. 41: 183-184. MACNEIL, F. S. 1950. Pleistocene shore lines in Florida and Georgia. U. S. Geol. Surv. Prof. Paper 221-F. FLAGELLATES AND DESMIDS NEW TO MASSACHUSETTS EDGAR E. WEBBER The many lakes, ponds, and marshy areas in Massachu- setts provide numerous and varied habitats for phycological studies. The freshwater algal vegetation of the eastern and southeastern regions of the state has been investigated rather thoroughly (Webber, 1963). Published accounts of the algae of western Massachusetts, however, appear to be limited to those by West (1889) and Webber (1964). A variety of interesting flagellates and desmids has been collected from two locations in the Amherst, Massachusetts area, These locations are a roadside seepage area adjacent to State Route 9, one mile southeast of Amherst, and Haw- ley Bog, Hawley, Mass. Because many of the algae collected from these two lo- cations have not been recorded from Massachusetts, a re- port of their presence in the state is warranted to attain accurate distributional data for such species. Flagellates were determined after Prescott (1962), while Irénée-Marie (1938) was consulted for the majority of des- mid identifications. EUGLENOPHYTA Euglena acus var. rigida Hueb. Plate I. fig. 1. Cells 624 X 1004, elongated, tapering posteriorly, rigid while in motion, with two rod-shaped paramylon bodies. Common. Amherst. Phacus helikoides Poch. Plate I, figs. 2, 3. Cells 49-554 X 100-1114, elongate-pyriform, twisted throughout but tapering posteriorly to a straight caudus slightly less than % the length of the cell, periplast longitudinally striated, one large and centrally placed circular paramylon body. Common. Amherst. P. longicauda (Ehrenb.) Dujardin Plate I, fig. 4. Cells 31-624 X 112-1364, ovoid, broadly rounded anteriorly, taper- ing posteriorly to a straight caudus as long as the cell body, periplast longitudinally striated, one central circular paramylon body. Com- mon. Amherst. P. pyrum (Ehrenb.) Stein Cells 18.6 u X 37g, ovoid-pyriform, tapering posteriorly to a 55 56 Rhodora [Vol. 69 Figs. 1-6, all X 450. Fig. 1: Euglena acus var. rigida; figs. 2, 3: Phacus helikoides; fig. 4: P. longicauda; fig. 5: Hemidinium nasutum; fig. 6: Micrasterias pinnatifida forma rhomboidea. straight caudus, anteriorly with a conspicuous papilla on either side at the point of flagellar attachment, periplast ribbed and spiralling posteriorly to the right, paramylon not observed. Common. Hawley. Leptocinclis acuta Prescott 1967] Flagellates — Webber 57 Cells 12u X 314, pyriform, tapering posteriorly, periplast spirally striated, paramylon as two lateral plates. Rare. Amherst. First record for Massachusetts. Trachelomonas armata var. longispina (Playf.) Defl. Test 31u X 44-554, broadly rounded, beset with short spines an- teriorly and a mixture of short and long spines posteriorly. Common. Amherst. T. horrida Palmer Test 24u X 44u, ovoid, completely covered with numerous spines of varying length. Common. Amherst. T. volvocina Ehrenb. Test 24u dia. spherical and smooth. Common. Amherst. PYRRHOPHYTA Hemidinium nasutum Stein Plate I, fig. 5. Cells 18.64 X 31g, elliptical and flattened, transverse furrow in- complete. Common. Amherst. First record for Massachusetts. CHLOROPHYTA Because many of the desmids collected are ubiquitous and, there- fore, well known, they are presented in list form. Those particular species new to Massachusetts are so indicated. Desmids were collected at the Amherst location, unless otherwise noted. Closterium abruptum West C. Kuetzingzi Breb. C. libellula Focke var. intermedium Roy & Bis. C. setaceum Ehrenb. C. ulna Focke Staurastrum avicula Breb. var. subarcuatum (Wolle) West S. biennianum var. ellipticum Wille S. controversum Breb. (Irénée-Marie, 1988, pl. 55, fig. 12; pl. 56, fig. 11.) First record for Massachusetts. S. erytoserum Breb. S. dejectum Breb. S. Dickei Ralfs. Hawley. S. furcatum Ehrenb. var. pisciforme (Breb.) Turner (Irénée-Marie, 1938, pl. 55, figs. 9, 10.) First record for Massachusetts. S. gracile Ralfs S. Johnsonii West & West var. depauperatum G. M. Smith (Irénée- Marie 1938, pl. 53, figs. 2, 3.) First record for Massachusetts. S. paradoxum Meyen S. pentacerum (Wolle) G. M. Smith. Hawley. S. polymorphum Breb. S. setigerum Cleve. Micrasterias pinnatifida (Kuetz.) Ralfs 58 Rhodora [Vol. 69 M. pinnatifida forma rhomboidea Brunel Plate I, fig. 6. Rare, Hawley. First record for Massachusetts. M. rotata (Grev.) Ralfs Pleurotaenium nodosum (Bailey) Lundell P. nodulosum Breb. Xanthidium antilopaeum (Breb.) Kuetz. var. hebridarum West & West DEPT. OF BOTANY, PENNSYLVANIA STATE UNIVERSITY THE BEHREND CAMPUS, ERIE, PA. LITERATURE CITED IRENEE-MARIE, FR. 1938. Flore desmidiale de la region de Montreal. Laprairie, Canada. Prescott, G. W. 1962. Algae of the western Great Lakes area. Re- vised edition. W. C. Brown Co., Dubuque, Ia. WEBBER, E. E. 1963. The ecology of some attached algae in Wor- cester County, Massachusetts. Amer. Midl. Natur. 70:175-186. 1964. A rare bluegreen alga from Massachusetts. Rhodora. 66:163-164. WEsT, W. 1889. List of desmids from Massachusetts, U. S. A. Jour. Roy. Microscop. Soc. pp. 16-21, STIPULICIDA IN GRAY'S MANUAL RANGE — Pine barrens on white sands along the Blackwater River in Isle of Wight, Nansemond, and Southampton counties of Vir- ginia yielded many species which were then new to the Manual Range when Professor M. L. Fernald and his col- leagues collected in this region more than twenty years ago. Dr. Henry K. Svenson and the writer were working on the distribution of some of these species which extend into Virginia from the south when we discovered an area in the southwestern corner of Isle of Wight County where Fernald apparently had not collected. Here is an oak-pine “barren” with turkey oak and loblolly pine the dominant trees and understories of ericaceous shrubs, creeping blue- berry, Vaccinium crassifolium, and Pyzidanthera bar- bulata. Dr. Svenson discovered a low, annual member of the Caryophyllaceae to be common in open areas of white sand and the plant was previously unknown to him and the writer, It is Stipulicida setacea Michaux and is now known from Virginia, the Carolinas, Georgia, Florida, and Mississippi. The genus is endemic to the southern coastal plain and has a second species ranging from Florida to Mississippi. Some associates of the Stipulicida in Virginia are Gay- lussacia baccata, Vaccinium elliottii, V. tenellum, and V. vacillans. Where turkey oak stands are dense, understories include Kalmia angustifolia, Pyxidanthera, and Vaccinium crassifolium, but Stipulicida is absent. Specimens of the Stipulicida, Svenson and Harvill 13932, will be deposited in the Gray Herbarium and the herbaria of Longwood College, University of North Carolina, and Virginia Polytechnic Institute. The number belongs to the junior collector’s series. A. M. HARVILL, JR. LONGWOOD COLLEGE, FARMVILLE, VIRGINIA 59 ADDITIONS TO THE FLORA OF THE BIMINI ISLAND GROUP, BAHAMA ISLANDS. — Howard (Ecol. Monogr. 20:317-349, 1950) listed the flora of the Bimini Island Group in his excellent study of the vegetation of those islands. Recent field work on the Biminis has turned up the following additions: 1. N Paspalum millegrana Schrad. South Bimini Island. Along dirt road between airport and Federal Aviation Agency Station. Common. Also occurring in scattered clumps all along road from ferry landing to airport. 13 July 1964. Stimson 722. Eleocharis geniculata (L.) R. & S. South Bimini Island. Growing in shallow water at edges of and along shore of small pond on northern end of the island north of road to airport. Common in this place. 12 July 1965. Stimson 708. Tillandsia aloifolia Hook. South Bimini Island. *Black- land” forest between Cavelle Pond and north coast of island. Only one plant seen here. 30 March 1965. Stim- son 1110. South Bimini Island. “Blackland” forest north of road to airport about 1/5 mile west of airport. Common. 31 March 1965. Stimson 1119. Tillandsia balbisiana Schultes. South Bimini Island. Edge of halophytic Avicennia swamp near north edge of eastern lobe of Cavelle Pond. Epiphytic on fallen branch. 30 March 1965. Stimson 1112. Specimens of these collections are deposited in the her- barium of Duke University; duplicates have been distrib- uted to other herbaria. WILLIAM R. STIMSON DEPARTMENT OF BOTANY, DUKE UNIVERSITY, DURHAM, NORTH CAROLINA 60 ALLIUM AMPELOPRASUM IN MISSISSIPPI — The discovery of Allium ampeloprasum L. as a wild plant means that Mississippi is the seventh state in the eastern United States in which this species is known to occur. Fernald (1950), Gleason (1952), and Gleason and Cronquist (1963) listed the species as occurring in Virginia. James (1956) added records for Tennessee. North Carolina and South Carolina were included by Ahles, Bell, and Radford (1958) and Browne (1961, 1963) reported the species from Kentucky and Illinois. During the summers of 1964 and 1965, field work con- nected with my study of the vascular flora of Mississippi revealed A. ampeloprasum at three sites. At all locations the plants were flourishing. Apparently this native of Europe is well able to compete with our native vegetation and establish itself as a weed in this part of the country. Specimens cited: Mississippi. Tishomingo County. Road- side ditch between Iuka and J. P. Coleman State Park, June 18, 1964. Thomas M. Pullen 64469. Lafayette County. Dry roadside bank in Holly Springs National Forest about two miles south of Denmark Fire Tower. May 31, 1965. Thomas M. Pullen 6589. State Highway 30 right-of-way four miles northeast of Oxford. June 22, 1965. John M. Herr, Jr. Voucher specimens are deposited in the herbarium of the University of Mississippi. THOMAS M. PULLEN' DEPARTMENT OF BIOLOGY, THE UNIVERSITY OF MISSISSIPPI, UNIVERSITY. "This work was supported by the Committee on Faculty Research of the University of Mississippi. I wish to thank Dr. E. T. Browne, Department of Botany, University of Kentucky for verifying my identification. 61 62 Rhodora [Vol. 69 LITERATURE CITED AHLES, H. E., C. R. BELL, and A. E. RADFORD. 1958. Species new to the flora of North or South Carolina. Rhodora 60: 10-32. BROWNE, E. T. 1961. Some new or otherwise interesting reports of Liliaceae from the southeastern states. Rhodora 63: 304-311. 1963. Allium ampeloprasum L. in Illinois. Castanea 28: 170-172. FERNALD, M. L. 1950. Gray's Manual of Botany. ed. 8. American Book Co.: New York. GLEASON, H. A. 1952. The New Britton and Brown Illustrated Flora. New York Botanical Garden: New York. , and A. Cronquist. 1963. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. Van Nostrand: Princeton, N. J. JAMES, R. L. 1956. Introduced plants in northeast Tennessee. Cas- tanea 21: 44-52. AN ANNOTATED BIBLIOGRAPHY OF MEXICAN FERNS George Neville Jones. University of Illinois Press, Urbana and London, 1966. pp. XXXIII-297. $5.00 Students of American ferns have been given a firm foundation by this publication of Dr. Jones. The bibliog- raphy proper (authors and titles) occupies 237 pages with more than 1200 entries with approximately 3000 cross references. The feature that immediately commends itself is the brief paragraph after almost each item, indicating its relation with the ferns of Mexico (in this book the term "ferns" is used in a wide sense, comprising all the crypto- gams belonging to the Tracheophyta), and obviously en- hances the value of the bibliography. Titles of periodicals are abbreviated following the prin- ciples and rules laid down in Appendix V, Guide to the Citation of Botanical Literature, in the 1956 edition of the International Code of Botanical Nomenclature, and some additional principles derived from L. Schwarten and H. W. Rickett, Abbreviations of Periodicals cited in the Index to American Botanical Literature (Bull. Torrey Club 74: 348-356. 1947). In addition to the selection of authors-titles of taxonomic and phytogeographic significance many articles on mor- phology, ecology, exploration, economie botany, bibliog- 1967] Book Review — Riba 63 raphy, biography, etc. have been included in the General Index. A Geographical Index has entries referring to each state of Mexico. A Biographical Index lists names and dates of 180 botanists associated with collections or papers about Mexican ferns. Finally, a Systematic Index cites the genera of ferns with pertinent references. There can be no doubt that the author’s statement “Bot- anists and others not particularly interested in the ferns of Mexico may find this biblography useful" will be amply fulfilled. This publication will be a real aid to those bota- nists interested in the study of American ferns in general and Mexican ferns in particular. RAMON RIBA, GRAY HERBARIUM, HARVARD UNIVERSITY. INSTITUTO DE BIOLOGIA, UNAM, MEXICO CITY NEWS ANNOUNCEMENT — FLORA NORTH AMER- ICA LAUNCHED — FLORA NORTH AMERICA, as the project will be called, was officially launched on 30 January 1967 when the newly formed Editorial Committee held its first meet- ing at the Smithsonian Institution in Washington, D. C. This three-day meeting, convened by William L. Stern (Smithsonian), Chairman pro tem. of the Steering Com- mittee, was attended by all members of the Editorial Com- mittee: Peter H. Raven, Chairman, Stanford University Stanwyn G. Shetler, Secretary, Smithsonian Institu- tion John H. Beaman, Michigan State University Kenton L. Chambers, Oregon State University Robert Kral, Vanderbilt University Walter H. Lewis, Missouri Botanical Garden John T. Mickel, Iowa State University Roy L. Taylor, Canada Department of Agriculture, Ottawa John H. Thomas, Stanford University Also attending were Robert F. Thorne (Rancho Santa Ana 64 Rhodora [Vol. 69 Botanic Garden), Chairman of the Advisory Council, and Vernon H. Heywood (University of Liverpool), Secretary of FLORA EUROPAEA, who served as a consultant in the dis- cussions and gave a concluding public lecture, “FLORA EU- ROPAEA, Its Conception and History,” on 1 February. The purpose of the project is to prepare a concise diagnostic manual to the vascular plants of the continental United States, Canada, and Greenland, and the Editorial Commit- tee dealt at least in a preliminary way with a large range of questions concerning the roles of the respective commit- tees, the functioning of the Editorial Committee and its secretariat, the solicitation of authors and advisors, and the format, arrangement, timetable, and funding for FLORA NORTH AMERICA. It is expected that the first 12 to 18 months will be occupied getting the project fully organized and the working procedures implemented. This will be followed by the second phase of intensive writing and editing for the first volume. Tentatively, four volumes, followed by a fifth comprising a theoretical symposium on the North American flora, are anticipated. The whole effort is expected to last 12-15 years. A full progress report on FLORA NORTH AMER- ICA will be published at an early date. The Editorial Com- mittee wil convene its next meeting at College Station, Texas, in August 1967, when the American Institute of Biological Sciences holds its annual meetings at the Agri- cultural and Mechanical College of Texas. The first meet- ing was financed by the Smithsonian Office of Systematics (Richard S. Cowan, Director). STANWYN G. SHETLER SMITHSONIAN INSTITUTION, WASHINGTON, D. C. 20560 NEW TAXA IN THE GENUS ALSOPHILA RAMON RIBA During a taxonomic revision of the Alsophila swartziana group two new species and a variety were found, which are described as follows. Alsophila tryonorum Riba, spec. nov. Figs. 1-4 Species A. conjugatae et A. trichiatae affinis. Petiolus crassus aculeatus squamatus valde pubescens, squamis lanceolatis basaliter truncatis bicoloribus denticulis fuscis tenuibus, trichomatibus bi- formibus rigidis et catenulatis. Lamina bipinnato-pinnatifida 3.0-3.5 m. longa 1.4-2.0 m. lata. Pinnae sessiles lineari-ovatae 70-100 cm. longae 35-40 cm. latae, rhachides secundariae pubescentes trichomati- bus biformibus rigidis et catenulatis, pinnulae sessiles pinnatifidae lanceolatae longe acuminatae 13-20 cm. longae 1.5-2.7 cm. latae, costae subter squamatae et pubescentes, squamis tenuibus parvis orbiculatis apiculatis, trichomatibus biformibus aliquot rigidis plerumque flexo- osis multis catenulatis. Segmenta supra pubescentia in costulis venis et inter venis, subter valde pubescentia in costulis et venis, tricho- matibus rigidis ad margines vix ciliatis, venae 11- 13-jugae obliquae venulis 1- 2-jugis, sori 7- 10-jugi ad furcam venarum vel basaliter plerumque in venulis basalibus inserti. Holotypus: Cuesta de Fusagasugá, Dept. Cundinamarca, COLOMBIA, 2540 m. Feb. 18, 1940, Cuatrecasas 8036 (US). Paratypi: VENEZUELA, Aragua: Colonia Tovar and vicinity, Fend- ler 49 (pars) (F, GH, MO, US), Pittier 9345 (GH, NY, VEN, US); be- tween El Portachuelo and Ocumare, Pittier 11394 (VEN, US) ; between Maracay and Choroni, Pittier 13902 (GH, VEN, US). COLOMBIA. An- tioquía: Páramo de Sonsón, Bro. Daniel 3437 (US); La Ceja, Bro. Daniel 682 (US); Cundinamarca: Cuesta de Fusagasugá, Cuatrecasas 8071 (F, GH, US); Tolima: Caldas Termales, Santa Rosa de Cabal, Dryander 2742 (Us); Honda-Ruíz, prov. de Soledad, Mayar 75 (US); Cauca: Popayán, Yepez 175 (F, US) ; Valle del Cauca: Hoya del Río Digua, Cuatrecasas 14982 (F, GH); Putumayo, between San Francisco and Sachamate, Ewan 16673 (GH, US). ECUADOR. Napo: Río Chingual, Acosta 13247 (F); Sodiro (NY); Nanegal, Sodiro (US). This species shows an evident relation with A. trichiata Maxon because of the similarity of the distribution of the pubescence. However, in A. trichiata the rigid trichomes are longer and the catenate trichomes of the petiole are not so abundant as in A. tryonorum; the petiole scales in A. 65 66 Rhodora [Vol. 69 Figs. 1-4. Alsophila tryonorum Riba. 1. — Pinnule, X 0.5; 2.— Ultimate segments, X 1.5; 3. — Border of crozier scale, X 20; 4.— Berder of petiole scale, X 20. All from the holotype. Figs. 5-7. Also- 1967] Alsophila — Riba 67 trichiata have a margin with contiguous dark denticula- tions, while in A. tryonorum the margin has very few dark denticulations or none, except in crozier scales where the dark denticulations are always present. Another relation is with A. conjugata Spruce in the similarity of the petiole scales together with the abundant winged trichomes and bullate scales in the lower surface of the segments. How- ever, the petiole in A. conjugata is glabrous, while in A. tryonorum it is strongly pubescent with rigid and catenate hairs. This species is named after Dr. Rolla M. Tryon and Dr. Alice F. Tryon for their contributions to the taxonomy of the ferns. Alsophila estelae Riba, spec. nov. Figs. 5-7 Petiolus gracilis aculeatus squamatus, pubescens solum adaxiale latere, squamis lanceolatis, basaliter brunneis denticulis fuscis, trich- omatibus rigidis. Lamina bipinnato-pinnatifida 1.2-1.5 m. longa 55-60 em. lata, rhachis pubescens. Pinnae sessiles, lineari-ovatae acuminatis 25-30 em. longae 10-11 em. latae, rhachides secundariae pubescentes trichomatibus rigidis, pinnulae sessiles pinnatifidae lineari-lanceo- latae acuminatae 5-6 cm. longae 1.0-1.2 cm. latae costae subter pubescentes et vix squamatae, trichomatibus biformibus rigidis et catenulatis, squamis tenuibus parvis orbiculatis. Segmenta supra pubescentia in costulis, subter pubescentia in costulis et venis tricho- matibus rigidis et adpressis, squamata in costulis, squamis bullatis, ad margines vix ciliatis, venae 7-8-jugae obliquae plerumque simplices, sori 2- 4-jugi in medio venarum. Holotypus: Mount Horeb, ca. 3/4 mile South (air) of Green Hill, on trail to Woodcutters Gap, Portland Parish, JAMAICA, 3700 ft. De- cember 31, 1966, Riba 214 (MEXU). Isotypus: GH. Paratypi: Mount Hybla, Cinchona, St. Thomas Parish, J. P. (Jamaican Plants) 49 (GH, IJ); same locality as holotype, December 28, 1966, Riba 196 (GH, MEXU). Alsophila estelae is one of the most graceful tree ferns in Jamaica. At first sight it can be easily confused with A. swartziana, a species that grows in the same place. How- ever, the slender trunk (3-4 m. tall and ca. 4-6 cm. thick) and the crown of small leaves make that species clearly dis- phila estelae Riba. 5. — Pinnule, X2; 6.— Ultimate segments, X 5; 1. — Border of petiole scale, X 60. All from the holotype. Fig. 8. Alsophila scabriuscula Maxon var. guatemalensis Riba, ultimate seg- ments, X 3. From the holotype. 68 Rhodora [Vol. 69 tinctive in the field. Alsophila estelae is easily differentiated from A. swartziana (beside the characters already men- tioned) by the usually simple rather than forked veins and by the position of the sorus when, rarely, the veins are forked. Then the sorus is borne below the fork in A. es- telae rather than at the fork as in A. swartziana. In addi- tion, A. estelae lacks the flat scales or winged trichomes near the base of the sori which are characteristic of A. swartziana. This species is named for Mrs. Maria Estela Riba. Alsophila scabriuscula Maxon var. guatemalensis Riba, var. nov. Fig. 8 Differt a forma typica sinu inter segmenta fertilia acuto vel anguste quadrangulari segmentis 6-7 mm. longis. Segmenta crenato- dentata vix duplicato-dentata, costulae et venae subter pubescentes trichomatibus longis flexuosis, venae 7- 8-jugae, sori 5- 6-jugi, re- ceptaculum parvum sessile paraphysibus longis flexuosis. Holotypus: Between Ixcan and Finca San Rafael, Sierra de los Cuchumatanes, Dept. of Huehuetenango, Guatemala, July 24, 1942, Steyermark 49417 (F). Paratypi: Alta Verapaz, between Cubilquitz and Hacienda Yaxcab- nal, Steyermark 44829 (GH, US); Izabal, along Rio Frío, Steyermark 39942 (F, US); along Río Tameja, Steyermark 41802 (GH, US); Sierra del Mico, between Los Amates and Izabal, Kellerman 7163 (F). The differences between this variety and var. scabriu- scula are in the sinuses, which in the variety guatemalen- sis are acute to narrowly quadrangular, and in the smaller size of the leaves and their parts. In the variety scabriu- scula the sinuses are widely quadrangular and the leaves are larger with the ultimate segments up to 17 mm. long. Although the two varieties occur in the same general region, they are distinct as demonstrated by the characters given above. Additional collections in the future may show intergradation between the two variants, but with the ma- terial that I have, they can stand as varieties, This paper was completed during the tenure of a John Simon Guggenheim Fellowship, which is gratefully ackowl- edged. GRAY HERBARIUM, HARVARD UNIVERSITY AND INSTITUTO DE BIOLOGIA, MEXICO 1967] Correction — Seymour 69 A CORRECTION Since the publication of the article entitled Botanical Notes from the Pringle Herbarium I (Rhodora 68:514ff. 1966), my attention has been called to an error which ought to be corrected. Eleusine indica (L.) Gaertner was reported from Hartford, Vermont. “Hartford Ctr., Vermont" is the way the hand-written or hand-printed label reads. It is not in the hand of the collector himself. However, the same collector, E. J. Dole, collected three other specimens of the same species, of the same number, on the same date (or exactly one year later!) in Hartford, Connecticut. It is probable that the specimens all came from the same locality, namely, Hartford, Conn. The error must have occurred in copying the label. The abbreviations, Vt. and Ct., look much alike. Also the handwriting of the collector, which is on one of the labels, is very difficult to read. Consequently Eleusine indica remains unknown in Vermont. FRANK C. SEYMOUR PRINGLE HERBARIUM. UNIVERSITY OF VERMONT, BURLINGTON INDICATIONS OF POSSIBLE MID-CENOZOIC HYBRIDIZATION IN THE ASPENS OF THE COLUMBIA PLATEAU BURTON V. BARNES Trembling aspen, Populus tremuloides Michaux, is the most widely distributed tree species in North America (Strothmann and Zasada 1957). Although several species and varieties of western aspens have been recognized, Popu- lus tremuloides Michx. for many years has been the accepted name embracing this genetically variable taxon (Little 1953). One would expect major physiological differences and some related morphological differences throughout its vast range; an indication of the former is given by Pauley, Johnson, and Santamour (1963). They reported the results of a seed source experiment of aspens initiated in 1950-1952 in eastern Massachusetts, in which aspens from the North- eastern States, Lake States, Northern, Central, and South- ern Rocky Mountains, Washington, and the Yukon territory were compared. Although the initial survival per- centage was acceptable to high (41-100%), all western sources exhibited slow growth and were clearly ill-adapted to the environment of Weston, Massachusetts. Only 2 of the 549 individuals were alive in 1962. This experiment suggested that surprising differences may exist between western and eastern aspens. The purpose of this paper is to introduce other evidence indicating that certain individuals as well as a population of aspen from the Columbia Plateau' may not be as closely related to eastern quaking aspen as we have tacitly assumed, Two clones of quaking aspen from northern Idaho resem- bling hybrids between P. grandidentata and P. tremuloides 'For purposes of this paper, the “Columbia Plateau," because of its relevance in mid-Cenozoic geology and flora, is used to designate the general area from which the author's collections and Populus fossils collected by numerous investigators were made. As used here it em- braces what are now parts of California, Nevada, Oregon, Washing- ton, Idaho, and Montana (Fenneman 1931). 70 1967] Populus — Barnes 71 in Lower Michigan are hereinn described. Morphological differences between a group of aspen clones from northern Idaho, Washington, and western Montana, considered here as part of the eastern edge of the Columia Plateau, are com- pared and contrasted with clones from Michigan. METHODS AND MATERIALS During the summers of 1959 and 1960 collections of leaves and twigs of aspen clones for a personal study her- barium were made in northern Idaho, western Montana, and Washington as time and circumstances permitted. Observa- tions of growth habit, phenology, and sex were also made. Opportunities for detailed study of the leaf and twig sam- ples were delayed until March of 1966 when I had returned to the University of Michigan and resumed the investiga- tions of natural hybridization in the aspens. Examination of the clones of the western specimens led to an unexpected dis- covery, namely that two of the western clones were evident- ly hybrids between P. grandidentata and P. tremuloides in leaf and bud characteristics. This discovery prompted a systematic study of 30 western clones sampled. The early leaves from the central portion of shoots 1-3 in long were selected and the blade width and length, petiole length, and number of teeth per leaf side were determined. Up to 20 leaves were taken per clone for measurement; the average number was 18. The terminal buds of at least four shoots of each clone were examined and the presence and distribution of pubescence observed and recorded, Com- parisons were made between these western clones and the bigtooth and quaking aspen clones sampled in northern Michigan (Barnes 1959). The latter group consisted of 21 bigtooth clones and 31 trembling aspen clones growing on two sites. Means of blade width and length, petiole length, and number of teeth for the two groups are not entirely comparable due to minor differences in sampling. However, the site differences are so great that only very general com- parisons of leaf size can be made in any event. Number of teeth and pubescence on terminal buds have proved to be more reliable characteristics, and are thus apparently better phenotypic indicators of hybridity in Michigan aspens than 72 Rhodora [Vol. 69 LOWER MICHIGAN POPULUS TREMULOIDES POPULUS GRANDIDENTATA Ponce PO 45 P. GRANDIDENTATA X P. TREMULOIDES PR 9% OF tese COLUMBIA PLATEAU P. TREMULOIDES See OO oo oo 00 eo ec 06 59110 59113 59115 59123 59128 scale 0 5 cm Fig. 1. Comparison of leaf size, shape and dentation of P. grandi- dentata, P. tremuloides and hybrids between them from Lower Michigan and “P. tremuloides” from the Columbia Plateau. 1967] Populus — Barnes 73 the leaf dimensions. Pubescence of buds collected Aug. 10- 13, 1966 was studied on 50 clones of quaking aspen in Iron Co. and 20 clones in Washtenaw Co., Michigan. RESULTS AND DISCUSSION Silhouettes of leaves from the central portion of shoots from clones 59122 and 59127 from Kootenai County, Idaho are illustrated in Fig. 1 (lower left). Representative leaves of Lower Michigan P. tremuloides, P. grandidentata, and natural hybrids? between them and leaves from six clones of quaking aspen from Idaho and Washington are also shown. It will be noted that Idaho clones 59122 and 59127 are like various natural hybrids in Michigan (Fig. 1). Al- though the Michigan hybrids exhibit great morphological variability, the most consistent diagnostic morphological characters are those of tooth number and bud pubescence which in presumed F, hybrids are approximately inter- mediate between those of the parents. There is, however, considerable variation even in these traits, and it is not yet known what amount of variation in these traits is attributable to maternal effects, introgression, and additive and non-additive gene action, For the aspens studied in Lower Michigan the average number of teeth of P. grandi- dentata clones was approximately 10, natural hybrids 18, and P. tremuloides 30. Means and ranges of tooth number and the means of leaf dimensions of the respective popula- tions are presented in Table 1. Populus grandidentata exhibited dense pubescence on all bud seales. Terminal buds in winter condition of P. tremu- loides of Lower Michigan sites were glabrous except for rare clones having minute pubescence on the basal scales. Nearly all the terminal buds from the 70 clones collected in mid-August 1967 exhibited sparse pubescence on the upper portion of the basal bud scales. Many P. tremuloides clones showed bud scales which have ciliate margins. Nat- ural hybrids have pubescence on almost all bud scales but it is usually much less dense than in P. grandidentata; the ?The term “natural hybrid" is used here to include F, hybrids, later generation segregates and introgressants; we are as yet unable to determine precisely the parentage of each hybrid clone. 74 Rhodora [Vol. 69 pubescence tends to become more dense on the basal scales. Using leaf dimensions and morphology, and bud char- acteristics, one would be clearly unable to separate the two Idaho clones (59122 and 59127, Fig. 1) from natural aspen hybrids of Lower Michigan. The average number of teeth per leaf side of both western clones was 19.5. The average tooth number of 25 natural Michigan hybrids reported by Barnes (1961) was 17.6. The Idaho clones had moderate to dense pubescence on the lower terminal bud scales, and in 59122 the pubescence extended downward completely coating the shoot produced the year of collection. No pu- bescence was observed on the upper bud scales except for a few fine hairs along their margins. This condition is somewhat different from most Lower Michigan hybrids. The latter typically have fine pubescence on both lower and upper bud scales. In pubescence the two Idaho clones close- ly resembled the putative introgressant Michigan clone 310, but both western clones had markedly fewer teeth (Fig. 1). The most striking difference between all of the Michigan and the specimens of western trembling aspens was the presence of moderate to dense pubescence on the basal scales of terminal buds of all clones of the latter. Some western clones exhibited pubescence not only on lower scales but the upper as well. Furthermore, clone 59108 ex- hibited pubescence on the shoots produced the year the col- lection was made. This feature has never been reported or observed for immature or mature shoots of any eastern P. tremuloides. All of the western collections were made in July, August, or September, so that this apparently was not merely juvenile pubescence which would disappear later. The axillary buds, where present, were also pubescent on the lower scales, but to a lesser degree than on terminal buds. It is unknown to what extent aspen clones of the Colum- bia Plateau or other geographic regions of the West ex- hibit pubescent buds. At the request of the author, Mr. R. Dennis Harr sent a sample of buds from eight clones of quaking aspen from Larimer County, Colorado. Terminal buds of these clones were much more glabrous than those 1967] Populus — Barnes 15 of the Columbia Plateau collections, but two did have a mod- erate amount of pubescence on the lower scales. Minute pubescence was also found on lower terminal bud scales on most of the Colorado specimens. Terminal buds of Lower Michigan quaking aspens were even less pubescent. The Colorado and Michigan collections seem more alike than are the Colorado and Columbia Plateau clones. A comparison of the leaf dimensions and tooth number per leaf side of 30 western clones and the Lower Michigan clones in Table 1 indicates no major or systematic difference between them. Leaf dimensions are apparently highly in- fluenced by the local environment, and hence clones from sites in both regions should be grown under controlled con- ditions to assess the differences. Tooth number is probably somewhat less susceptible to environmental modification. Tooth number, however, is positively correlated with leaf size. The correlation coefficient of the relationship between leaf area and tooth number of P. tremuloides on the Pell- ston plain site (indicated in Table 1) was significant, r — .65 (P<.05). Also, the size of leaves and number of teeth of clones on the Moraine site are larger than the re- spective attributes of clones on the Pellston plain. Since tooth numebr is positively correlated with leaf size one might expect the western clones, the leaf sizes of which are larger, to have as many or more teeth than the clones from the Michigan sites. However, the number of teeth of the western clones was somewhat less (25.3) than that of clones from both of the Michigan sites (30.5). The difference was not statistically significant (P>.40) because of the great variation within both groups. These data do indicate, how- ever, that some clones have markedly larger and fewer teeth than would typically be expected of Michigan P. tremu- loides; other clones exhibited no difference whatsoever. Obviously, systematic studies throughout the range of P. tremuloides coupled with progeny tests under controlled conditions are needed to evaluate the differences and deter- mine their causes. Although the western clones were originally sampled and cursorily identified as typical of eastern P. tremuloides, 76 Rhodora [Vol. 69 systematic differences are now clearly evident, Descrip- tions of western aspens in floras and dendrology texts pro- vide no evidence of differences in pubescence of the buds (Jepson, 1909; Kirkwood, 1930; Longyear, 1927; Preston, 1940; Sudworth, 1908). Illustrations of western aspen leaves, however, do indicate the lower tooth numbers dis- cussed. Sudworth (1908), for example, illustrated both an eastern P. tremuloides and a western P. tremuloides from Colorado. Number of teeth per leaf side of the eastern aspen was approximately 33 in contrast to about 22 for the Colorado aspen. This, however, might be due to chance sampling of clones not representative of the populations. How can we account for the observed differences between the western and Michigan populations of quaking aspen? One explanation would be clinal evolutionary differentiation along gradients running from west to east. Another hypoth- esis is hybridization and flow of genes from a western taxon or taxa having pubescent buds and few large teeth into the mid-Cenozoic quaking aspen. It is not the pur- pose of this paper to attempt an answer to this question through lengthy speculative arguments. However, a brief review of the taxonomy of western aspens is made below and some evidence indicating that hybridization of aspens could have taken place in the Columbia Plateau is presented. The western aspen, P. aurea, was described by Tidestrom (1911) as a species which “formed forests throughout Colorado, Utah, and adjoining territory. . . ." This taxon differed from P. tremuloides Michaux in floral characteris- ties and in the fall coloration of leaves. The leaves were described as serrate and the buds (“gemmis”) were termed glabrous, gummy, and conical. Other authors doubted the validity of this taxon (Daniels, 1911; Preston, 1940; Clokey, 1951) and it was subsequently considered synonymous with P. tremuloides Michx. by Little (1953). The plants orig- inally described by Tidestrom were from the central and southern Rockies and may be more closely related to east- ern P. tremuloides than the populations in the Columbia Plateau. The collections by Harr having buds somewhat 1967] Populus — Barnes rii similar to Michigan populations of P. tremuloides tend to support this. Tidestrom (Piper and Beattie 1915) also recognized an- other species, P. vancouveriana Trelease. He distinguished this species from eastern P. tremuloides and western P. aurea by the “peculiar dentition of the leaves. The teeth are much larger than in any of its immediate allies and besides being crenulate are depressed so that each tooth viewed from the edge forms a double curve.” The number of teeth was not reported. The buds were described as “smooth, conical, gummy, dark purple." Without examina- tion of the type specimen it is difficult to draw any firm conclusions about the extent to which this taxon resembles clones of the Columbia Plateau. One explanation for the existence of the larger-toothed, pubescent Idaho clones is that they are recent hyrids or introgressants between western P. tremuloides and the eastern P. grandidentata. The westernmost occurrence of P. grandidentata is in southeastern Manitoba (Slabaugh 1958). Thus the nearest localities are about 1,000 miles northeast of the Idaho “P. tremuloides” clones. Since P. grandidentata sheds pollen later than P. tremuloides where the two are sympatric and under the prevailing westerlies, this hypothesis seems most unlikely. Another possibility is that there were situations in the mid-Cenozoic period parallel to the ones now taking place commonly in the southern part of Lower Michigan — condi- tions promoting natural hybridization and introgression be- tween P. grandidentata and P. tremuloides. This hypothesis would necessitate the existence of a western taxon similar to the present-day eastern P. tremuloides in having many teeth and essentially glabrous buds, and the contemporane- ous presence of a large-toothed aspen with pubescent buds like those of P. grandidentata, During the Miocene period, a mixed deciduous forest, the *Arcto-Tertiary Geoflora," characterized the Columbia Plateau (Chaney 1959). Several species of Populus, at least two of which were aspens, i.e, members of Section Leuce, flourished in this region together with beech, maples, oaks, 78 Rhodora [Vol. 69 ashes, and willows. An aspen leaf termed Populus lind- greni Knowlton by LaMotte (1936)? and a leaf described by Chaney and Axelrod (1950)* as Populus washoensis Brown both closely resemble the leaves of modern P. grandi- dentata. Chaney and Axelrod (1959) considered the Popu- lus lindgreni Knowlton as illustrated by LaMotte (1936) to be synonymous with P. washoensis Brown. The leaf named P. voyana by Chaney and Axelrod (1959)* is similar to leaves of modern P. tremuloides. Thus, if the two aspens were sympatric, hybridization might have taken place and some genes from the large-toothed aspen became in- corporated in the populations of trembling aspen of the Columbia Plateau. After some hybridization between these species, extinction of the typical large-toothed aspen mem- ber took place along with most of the other mesophytic members of the community due to the rise of the Cascade Mountains. Further gene exchange was thus prevented so that unless first generation hybrids have been perpetuated asexually, the present quaking aspen clones could be intro- gressants possessing some genes of the fossil species. Before accepting the above hypothesis, other facts should be investigated. Besides such leaves as resemble the modern P. grandidentata, the fossil record also exhibits a variety of leaves of the genus Populus which have been termed Populus eotremuloides Knowlton and P. lindgreni Knowlton. These Chaney and Axelrod (1959) considered to be close- ly related to P. balsamifera and P. heterophylla, respec- tively. Most of the specimens illustrated by Chaney (1959), LaMotte (1936), and Smith (1941) do not resemble typical modern P. tremuloides of Michigan. Their illustrations re- semble P. tremula and hybrids similar to those between Michigan P. grandidentata and P. tremuloides. The array of forms, including early leaves, late leaves, and leaves from ?Plate 4, Fig. 1. Populus lindgreni Knowlton. Pleisotype. U.C. Mus. Palaeobot., No. 956. ‘Plate 18, Fig. 7. Populus washoensis Brown. Hypotypes U.C.M.P., Paleobot. Coll. No. 2927. Blue Mountains. "Plate 18, Fig. 4. Populus voyana Chaney and Axelrod. Hypotype, U.C.M.P., Paleobot. Coll. No. 562. Trout Creek. 79 Populus — Barnes 1967] 'peurquioo says ure[d uojs[[oq pue eureiopy Ulo4g sau0[D, "ue3rqory ‘AjunoD jeuruj ‘MPA ‘NLEL '9g uonoeg ‘yN PUL gg uonoeg 'agc "ues 'Ajunor) ue34oqeq) ‘MEN ‘NISL '9 WONS “HAS, | | $$-6I 6'€6 68 | 8c | o's | 08 uoydurqgse A, ‘BUR UOW | | | U134SƏM 'ogep[ ULay}ION sopio]nuoa4] "qd 68766 GOS [een | OP | Cry | TS ,UBSIYIT, Tamo Sapiojnwas] "d 68-22 huoc NI dar | EPILET TZ s | | ued uojsppoq ‘UEZIN sapio]nuo4] "d 98-86 | Lze ge | gr TS | OL ƏS ƏUILIOW ‘UBSIYOI səpomuaət} ‘d er ror g'e | ES | TL SI YS 9euredopy 'uegrqorW — »203uopipuna4D ^q | | asuey | Ua l suq (Sue | | PIM | seuo[) EDU PCIE B apis je9[ əd 9poneq | ƏPLrig | Pra _| jo UOI}BIO'T uoxerL Ww9ep jo 419qunw uo ur uorsuour(q Jeo ioquinN 'uoydurtuse A, pue vuejuopy foqep pue uezrqorpy wolf seuo[p uadse Jo soAeo[ uo 47397 JO iequinu pue suorsuoudp zesg TE qe Table 1. Leaf dimensions and number of teeth on leaves of aspen clones from Michigan and Idaho, Montana and Washington. 80 Rhodora [Vol. 69 various positions along the shoot, indicate a need for thorough study of Populus species in the Arcto-Tertiary- Geoflora of the Columbia Plateau and their relationships to aspens of other regions of the West, as well as those of eastern North America. The Asian race of P. tremula, known as P. tremula var. davidiana Schneider and a taxon resembling the Asian aspen with pubescent buds, P. sieboldii Miq., may also have been in the Arcto-Tertiary- Geoflora. Graham (1965) reported that leaves of P. washo- ensis Brown resemble both P. grandidentata and the P. tremula of eastern Asia. Thus, the aspens of the Columbia Plateau may be the progeny of a complex of several related species whose influence may yet be tracable through bio- systematic studies, Another important problem to be resolved is the time and circumstances of hybridization if, in fact, such did occur. Did it occur before the Arcto-Tertiary Geoflora reached the Columbia Plateau, or did it occur initially in Miocene times in the Columbia Plateau when conditions were such that gene exchange was possible? Finally, the possibility of actual existence of a large-toothed, pubescent aspen in modern-day stands having relicts of this Miocene flora should be explored. SCHOOL OF NATURAL RESOURCES UNIVERSITY OF MICHIGAN, ANN ARBOR LITERATURE CITED BARNES, BURTON V. 1959. Natural variation and clonal devolop- ment of Populus tremuloides and P. grandidentata in northern Lower Michigan. Ph. D. Thesis, University of Michigan, 334 p. 1961. Hybrid aspens in the lower peninsula of Michigan. Rhodora 63(755): 311-324. CHANEY, RALPH W. 1959. Miocene floras of the Columbia Plateau. Part I. Composition and interpretation. P. 1-134, Jn Contribu- tions to paleontclogy. Carnegie Institution of Washington Publ. No. 617. and DANIEL I. AXELROD. 1959. Miocene floras of the Columbia Plateau. Part II. Systematic considerations. P. 135-224, In Contributions to paleontology. Carnegie Institu- tion of Washington Publ. No. 617. 1967] Populus — Barnes 81 CLOKEY, IRA. W. 1951. Flora of the Charleston Moutains, Clark County, Nevada. Univ. Calif. Publ. in Botany 23. 260 p. DANIELS, FRANCIS POTTER. 1911. The flora of Boulder, Colorado, and vicinity. Univ. Mo. Studies 2 (2): 1-311. FENNEMAN, NEVIN M. 1931. Physiography of western United States. McGraw-Hill Book Co., Inc. New York, 534 p. 103 p. + 22 plates. GRAHAM, ALAN. 1965. The Sucker Creek and Trout Creek Miocene floras of Southeastern Oregon. Kent State Univ. Bull. 53(12): 103 p. + 22 plates. JEPSON, WILLIS LINN. 1909. The trees of California. Cunningham, Curtis and Welch, San Francisco, 228 p. Kirkwoop, J. E. 1930. Northern Rocky Mountain trees and shrubs. Stanford University Press, California, 340 p. LAMOTTE, ROBERT SMITH. 1936. The Upper Cedarville flora of northwestern Nevada and adjacent California. 5: 57-142 + 14 plates, In Contributions to paleontology, Middle Cenozoic floras of western North America. Carnegie Institution of Washington Publ. No. 455. LITTLE, ELBERT L., JR. 1953. Check list of native and naturalized trees of the United States (including Alaska). U.S.Dep.Agr. Ag. Handbk. 41. 472 p. LONGYEAR, BURTON O. 1927. Trees and shrubs of the Rocky Moun- tain region. G. P. Putnam’s Sons, New York. 244 p. PAULEY, S. S., A. G. JoHNSON, and F. S. SANTAMOUR, JR. 1963. Results of aspen screening tests: I. Seed sources of quaking aspen (P. tremuloides Michaux). Minn. For. Notes No. 136. 2 p. PIPER, CHARLES V., and KENT R. BEATTIE. 1915. Flora of the northwest coast. The New Era Printing Co., Lancaster, Pa. 418 p. PRESTON, RICHARD J. 1940. Rocky Mountain trees. Iowa State College Press, Ames, Iowa. 285 p. SLABAUGH, PAUL E. 1958. Silvical characteristics of bigtooth aspen (Populus grandidentata). U.S.F.S. Lake States For. Expt. Sta. Station Paper 63. 16 p. SMITH, HELEN V. 1941. A Miocene flora from Thorn Creek, Idaho. Am. Mid. Nat. 25: 473-522. SUDWORTH, GEORGE B. 1908. Forest trees of the Pacific coast. U.S. Govt. Printing Office, Washington. 441 p. TipESTROM, IVAR. 1911. Notes on Populus, Plinius. Am. Mid. Nat. 2: 29-35. PHYSALIS IN MEXICO, CENTRAL AMERICA AND THE WEST INDIES: * U. T. WATERFALL Linnaeus established Physalis in 1753 with a total of nine species. The genus was known to authors before Species Plantarum under Solanum as, for example, “Solanum vesi- carium indicum minimum" (Hermanno, 1679), and under Alkekengi, as illustrated by "Alkekengi indicum minimum" (Tournefort, 1716). Miller (1768) described several species from Mexico, al- though few of his names have been identified with Mexican species in any account of the flora written since that time. Reinstated are the neglected Miller names P. cordata and P. maxima for relatively common species, with further dis- cussion presented later. The complexity of the genus, and the wide range of varia- bility frequently encountered, the ambiguity of many of the earlier descriptions, as well as the unavailability of some of the types to those not so fortunate as to be able to make a trip to Europe to study them, have resulted in some duplica- tion of names, as will become evident on perusal of the treatment of the species. An example of a completely unrecognized name in the area under consideration is P. glutinosa Schlecht. Type material (HAL) shows this to be the very distinctive taxon ‘National Science Foundation Grant GB-3642 provided funds for the author's travel to study types, and other historically significant specimens, in major European herbaria in August, 1965. Special thanks are due the curators, and other officials, of the herbaria of the British Museum (Natural History) and the Royal Botanic Gardens, Kew in London, and those in Stockholm, Berlin, Brussels, and Paris for their aid, kindnesses, and other consideration. 'In specimen citations, the symbols used to indicate herbaria are the standard ones of Lanjouw and Stafleu (1964) with the addition of the symbol MEXP for the Herbarium of the Instituto Polytécnico Nacionál, Laboratorio de Botánica Fanerogámica, Mexico, D. F. To the Curators of all these herbaria the author is most grateful for the loan of their valuable specimens. 82 1967] Physalis — Waterfall 83 known in American herbaria as Cacabus mexicanus S. Wats., or as Physalis eximia Standley. Some specific names might better be forgotten, if that were possible, or relegated to nomina confusa, For example, P. Coztomatl Dunal (1852) might very well be a form of P. viscosa L., but, observing the long, attenuate lobes of the flowering calyx, as shown on a photograph of the Sessé and Mocifio Plate, one might with equal logic conclude it to be P. orizabae Dunal. Physalis dentata Dunal (l. c.) is probably P. acutifolia (Miers) Sandwith ; however, the Field Museum photograph, allegedly of this species, from “Types of the Berlin Her- barium”, is not the large-flowered, long-pedicelled species from the Pavon Herbarium in the Boissier Herbarium at Geneva as shown by the Field Museum photograph, Nega- tive No. 31435. The Berlin specimen, which I did not see personally when I was there, appears to be P. chenopodif olia Lam., a completely different species. No study of the genus can be made without considering the limitations of related genera. Somewhat varying dispo- sitions have been made of them in the past. One of the problems concerns the placement of those species having flowers in axillary fascicles, but fruits in enlarging, inflated calyces. These vary from species having axillary fascicles of 3-7 flowers, sometimes reduced to one, as in P. arborescens L. (P. mayana Standley) to those with (1-) 2-5 flowers per fascicle as in P. chimalteca Standley and Steyermark. In this group of species the corolla is more deeply lobed than is usual in the genus. In Wettstein (1891) these species would key to Withania, a genus said by him to occur in the tropics and subtropics of the Old World. In Standley (1924) they would key to Athenaea (although Standley and Steyer- mark have described several of them under Physalis) ; in Sleumer (1950) they would go to Withania emend. They are here referred to Physalis. The equation of Cacabus mexicanus S. Wats. with Physalis glutinosa Schlechtendal is not intended to imply that Cacabus is synonymous with Physalis. Physalis glutin- osa has enlarging fruiting calyces, much inflated about the 84 Rhodora [Vol. 69 fruit in the manner characteristic of Physalis. Its corolla is campanulate, similar to those found in many species of Physalis, but larger than usual. In Cacabus, sens, str., the calyx rather closely invests the fruit. Not only does the well-developed fruiting calyx of Caca- bus hondurensis (Donn.) Smith remove it from that genus, but the peculiar seeds mentioned by Standley (1931) seem to justify his disposition of it as the monotypic genus Eutheta. Margaranthus in fruit cannot be distinguished from small-fruiting species of Physalis, as the fruiting calyx en- larges similarly in both, but the pentangular-urceolate corol- la of Margaranthus is quite distinctive. Dunal (1852) placed Margaranthus with genera of the Solanineae having longitudinally dehiscent anthers. Wettstein (1891) placed it in the Lyciinae. Sleumer (1950) returned it to “Solaninae Wettst. em.” To the present author it seems so closely related to Phy- salis that it should not be placed in a different subtribe. However, it is, for the time at least, not treated with Phy- salis. Measurements given here are to be considered as gen- eralized indication of size, but not necessarily as absolute quantities. No doubt in natural populations some deviation from them will be found. Width-measurements given for corollas mean the greatest width, i. e., the width across the limb of fully expanded corollas. The length of calyces in- cludes the calyx lobes, although these are usually given separately, too. PHYSALIS L., Species Plantarum 182. 1753; Alkekengi Tourn. ex Hall, Enum. Stirp. Helv. 2:508. 1742; Herschellia Bowdich, Excurs. Mader. 159. 1825; Quincula Raf., Atl. Journ. 145. 1832; Alicabon Raf., Sylva Tellur. 56. 1838; Pentaphiltrum Reichb., Das Herbarienbuch 121. 1841; Boberella Krause, in Sturm, Fl. Deutschl. ed 2 (10):54. 1903. Plants annual or perennial with herbaceous stems, rarely arbores- cent, from a few centimeters tall to 3 meters tall, equally wide and densely branched with stems as much as 3 em in diameter; rhizoma- tous structures present in some species; leaves petioled, alternate, but sometimes two or even three apparently together due to internode 1967] Physalis — Waterfall 85 reduction; leaf blades usually ovate, but varying to linear; leaf margins from toothed, often coarsely and irregularly so, to entire; vestiture including long multicellular (*jointed") hairs, short hairs, stipitate or sessile glands, or branched or stellate trichomes, often more than one kind intermixed on the same plant; flowers pedicelled, usually single in the leaf-axils, but sometimes several and fascicled, rarely some of them in false racemes due to leaf and internode re- duction; flowering calyces varyingly 5-lobed, on pedicels shorter than the calyx to several times its length; corollas plicate, from rarely tubular-expanded and apically constricted to often more or less cam- panulate to campanulate-rotate, or the limb even more or less reflexed ; corollas usually yellowish, often with 5 darker, contrasting macula- tions occupying varying amounts of the area above the lower tubular part, these maculations usually discrete, sometimes more or less coa- lescing, solid in color, or each consisting of several to many separate spots; sometimes the maculations not strongly contrasting, sometimes absent; varying amounts of hair usually present in the tube near the point of filament-divergence, sometimes as 5 hairy pads, sometimes very few hairs present, or none, sometimes these areas more or less glandular; stamens 5, anthers dehiscing longitudinally, oblong or linear-oblong to ovate, yellow to blue, violet or greenish blue, or some- times so tinged or lined; filaments slender and filiform to nearly as wide as the anthers; style filiform, the stigma capitate to nearly trun- cate; fruit a 2-carpellate, few-seeded to many-seeded berry, which, however, often has a thin dryish pericarp; fruit usually sessile in the calyx which is usually inflated around it, but the berry rarely grow- ing large enough to be tightly invested by the calyx; sometimes a short gynophore is present; fruiting calyx often invaginated basally; fruiting pedicels usually elongating; seeds nearly orbicular to reni- form, usually more or less punctate to irregularly reticulate. KEY 1. Corolla 5-lobed; flowers several to 1 or 2 in leaf axils. 2. Corolla 8-12 mm long and 10-15 (-20) mm wide. 3. Flowers usually several in the leaf axils; fruits may be single; stem hairs more or less spreading. 4. Stem hairs dendritic. ............. 1. P. arborescens. 4. Stem hairs simple. ............ A 7. P. melanocystis. 3. Flowers usually 1-2; fruits usually single; stem hairs an- trorsely appressed (also see P. melanocystis var. cernua), some simple, some more or less dendritic... tnr pue dicesiseunseseepesstoencosee to RR 2. P. porphyrophysa. 5. Calyx lobes more or less elongate. 6. Calyx prominently hairy at anthesi8. .........» 3. P. amica. 6. Calyx essentially glabrous. ..... en 4. P. amphitrica. 5. Calyx subtruncate to more or less short-lobed. 7. Calyx prominently hairy at anthesis. .......... 5. P. calidaria. 86 Rhodora [Vol. 69 7. Calyx essentially glabrous. ...... 6. P. chimalteca. l. Corolla apically 5-angled with shallow sinuses, or subtruncate; flowers solitary (excepting sometimes, by stem and leaf reduction, in P. angustior and P. aggregata). 8. Corolla campanulate with a long limb, to tubular-campanulate, to subtubular and apically constricted, 22-35 mm long. 9. Corolla limb not expanded. 10. Plant nearly glabrous. ...... sees 8. P. constricta. 10. Plant prominently vestite, the hairs at least partly capi- tate-glandular. ....... sess 9. P. campula. 9. Corolla limb more or less expanded. ll. Herbage greenish. ........ sss 10. P. glutinosa. ll. Herbage cinereous-vestite, ......sssssn 11. P. cinerea. 8. Corolla open-campanulate to rotate or reflexed-rotate at full anthesis, smaller (except P. amphitrica). 12. Corolla bluish; indument scale-like. ................ 52. P. lobata. 12. Corolla yellowish to whitish to bluish-tinged, rarely bluish, but then indument not scale-like. 13. Fruiting calyx 10-ribbed, subequally 10-angled, or teretish. 14. Indument of stellate hairs, rarely partly dendritic. een eee ehe ee eseeeee t enne ene na ahh enses ener herr e sss s el e reru 12. P. viscosa. 14. Indument various but not stellate, sometimes dendritic or partly so, sometimes none. 15. Flowering pedicels shorter than the calyx to 2-3 times its length, usually 1-10 mm long, but in rare cases some of them becoming 15-30 mm long. 16. Anthers large, (2-) 2.2-4 mm long. 17. Vestiture of stems, petioles and pedicels in part of spreading, jointed hairs 2-3 min long. 18. Leaf blades ovate, usually broadly so, some- times narrowly so. 19. Blades of principal cauline leaves (8-) 12-20 em long, smaller ones one-third to one-fourth as large. 20. Some of the hairs tipped with reddish glands; throat of ccrolla densely hairy; anthers blue. 21. Corollas 25-35 mm wide. saec. 1——— HH 14. P. stapelioides. 21. Corollas 15-22 mm wide. ......... sss ERE 25. P. Sancti-josephi. 20. Hairs not tinped with reddish glands; throat of corollas nearly glabrous ; anthers yellow. saser 22. P. luteoanthera. 19. Blades of principal cauline leaves usually 2-6 cm. long. 22. Only a few long hairs on stems and peti- 1967] Physalis — Waterfall 87 oles; principal leaf blades usually 2-6 cm long. 23. Flowering calyces prominently hairy; anthers with only a tinge of blue. ....... DT PNE MEE tent A 21. P. gracilis. 23. Flowering calyces with only a few long hairs; anthers blue. ......... d bain der rm 43. P. philadelphica. 22, Stems, leaves and petioles with abundant long hairs; leaf blades usually 1.5-2 cm long. 24. Anthers bluish, or so tinged or lined; leaf blades entire or with 1-3 coarse teeth on each side. 25. Leaf blades oval to ovate, entire, prin- cipal ones 5-6 em long. ....... DUC OE MIDI E HUE 29. P. latecorollata. 25. Leaf blades ovate to hastate-ovate or hastate-deltoid with 1-3 coarse teeth on each side. ....... 30. P. philippensis. 24. Anthers yellow; leaf blades irregularly several-toothed. .... 33. P. tehuacanensis. 18. Leaf blades lanceolate-ovate to linear-lanceolate. Fi cdedies dads qaareacenadwadideauecdtds celttedcceeanadea 31. P. caudella. 17. Vestiture various, or absent, but not including jointed hairs 2-3 mm long. 26. Anthers yellow, not bluish tinged or lined. 27. Corolla prominently dark maculate. 28. Leaf blades ovate, sometimes apically at- tenuate. 29, Flowering calyx with short multicellular hairs 0.2-0.3 mm long. .... 16. P. ingrata. 29. Flowering calyx with multicellular hairs ca. 0.5-1.0 mm long (or glabrous in some extremes). 30. Hairs (when present) ca. 1 mm long, more or less spreading; leaves often longer than the internodes. s.e... NE eir I 21. P. gracilis 20. Hairs of flowering calyx ca. 0.5-0.7 mm long, antrorsely curled; principal cau- line leaves often one-half to one-third as long as the internodes. .......- REPRE Accra EE 41. P. longicaulis. 98. Leaf blades ovate-lanceolate to linear- lanceolate. .............. eee 32. P. virginica. 88 Rhodora [Vol. 69 27. Corolla immaculate, or with spots only sliehtly differentiated in color. 31. Fruiting calyces usually 2-3 em long. 32. Fruiting calyx nearly filled by berry, open apically. ................ 26. P. Muelleri, 32. Fruiting calyx inflated around the berry, apically closed or nearly closed, ............. 3333223239 4E PE PYETS VR AREE PURSE TATE 36. P. hederaefolia. 31. Fruiting calyces ca. 1 em long. 33. Leaf blades deltoid to ovate-deltoid, ex- panding abruptly from the petiole. ........ RR 94. P. microphysa. 33. Leaf blades lanceolate to lanceolate- ovate, gradually expanding from the petiole. cece 42. P. Rydbergii. 26. Anthers bluish, purple, violet or greenish-blue, or so tinged or lined (sometimes fading in drying). 34. Some of the flowers in few-flowered, termi- nal, apparently false racemes, formed by shortening of internodes and reduction of leaves. 35. Corollas 13-15 mm wide. a, REA 17. P. aggregata. 35. Corollas 20-29 mm wide. ........ es Spssesestanesuuusseecssuniesssuseeuseeyyerss 24. P. angustior. 34. Flowers solitary. 36. Corollas prominently marked with 5 dark patches, these sometimes ccnfluent basally. 37. Flowering calyx with divergent hairs. 38. Fruiting calyx evenly covered with hairs, or hairs more abundant on prin- cipal veins, but also scattered between them. 39. Fruiting calyx usually 30-40 mm long and 20-25 mm wide; anthers ca. 4 mm long. 40. Leaf blades more or less cordate, principal ones usually 6-9 cm long, apices often somewhat lunate- attenuate. ........ 23. P. peruviana. 40. Blades rarely semicordate, usually 4-5 em long, apices sometimes slightly one-sided, but not lunate- attenuate. .................. 18. P. lassa. 39. Fruiting calyx (15-) 20-25 (-30) mm long and (11-) 13-18 (-20) mm 1967] Physalis — Waterfall 89 wide; anthers 2.5-3 (-4) mm long. 41. Corollas 10-30 mm wide when fully open; maculations 3-7 mm long. 42. Corollas 2-3 cm wide; flowering calyces 8-9 mm wide at base of lobes. | * e 24. P. angustior. 42. Corollas 10-18 mm wide; flower- ing calyces 4-7 (-8) mm wide at base of lobes. 43. Principal leaf blades 6-10 cm long .... 25. P. Saneti-josephi. 43. Principal leaf blades 2-7 cm long. 44. Vestiture of flowering calyx including jointed hairs 0.5-1 mm long. .... 27. P. sordida. 44. Flowering calyx densely vestite with jointed hairs 1-L5*mm long een MIU 28. P. Pennellii. 41. Corollas 8-9 mm wide; maculations ca.idcmmslongl Pedes bratia Ai 46. P. minimaculata. 38. Fruiting calyx with a few hairs, often on ribs, or none. 45. Fruiting pedicels 1-2 cm long. 46. Anthers dark blue to purplish. .... MERERI USE 31. P. caudella. 46. Anthers yellowish, tinged or lined with light blue or violet. ................ "oer 32. P. virginiana. 45. Fruiting pedicels 2-5 mm long. ........ Pisin pesuanaveaedansenctites 43. P. philadelphica. 37. Flowering calyx with appressed hairs or none. 47. Blades of principal leaves ca. ovate. 48. Hairs of stems, petioles and pedicels antrorse; plants perennial. 49. Fruiting pedicels 4.5-5 cm long. ERE 3 13. P. Mevaughi. 49. Fruiting pedicels 1-2 cm long. 50. Leaf surfaces with a few scat- tered appressed hairs, or gla- brala D es 15. P. orizabae. 50. Surfaces or young leaves, or all leaves, greyish with abundant, 90 Rhodora [Vol. 69 more or less antrorsely ap- pressed hairs. ........sssse TM 19. P. chenopodifolia. 48. Shorter hairs of stems, petioles and pedicels retrorse, sometimes none. .... ——"— 43. P. philadelphica. 47. Leaf blades lanceolate to linear-lanceo- late. 5l. Many of the leaf blades more or less hastate, or with 1-3 large teeth, basal or as far up as the middle of the blade. .............. 20. P. hastatula, 5l. Leaves neither hastate, nor with 1-3 large teeth on basal half of blade. "A 32. P. virginiana. 36. Corollas immaculate, or with spots light colored and not strongly contrasting. 52. Stems, petioles and pedicels with spread- ing hairs, some gland-tipped. .................... MMMMMMMNNNNMMMMMMMMMMMMM 27. P. sordida. 52. Stems, petioles and pedicels with few, short, appressed hairs, or none. 53. Fruiting pedicels 6-7 mm long; hairs retrorse. .......... 44. P. michoacanensis. 53. Fruiting pedicels (7) 10-25 mm long; hairs retrorse or none. 54. Vestiture more or less abundant and greyish. ........ 19. P. chenopodifolia. 54. Plant with few short hairs, or gla- brous. re 45. P. angulata. 16. Anthers small, 0.5-1.8 (-2) mm long. 55. Corollas (6-) 5-15 (-20) mm wide and 5-9 mm long. 56. Stems spreading-hairy. 57. Corollas 15-20 mm wide; anthers dark blue. M 35. P. parvianthera. 57. Corollas 8-10 mm wide; anthers yellow. ........ RR 42. P. Rydbergii. 506. Stems glabrous, or with few, short, upwardly appressed hairs. 58. Principal leaves 1.2-4 mm long, broadly ovate to reniform. sesser 39. P. Mimulus. 58. Principal leaves 5-11 em long, ovate, lanceo- late-ovate, or narrower. ...... 45. P. angulata. 55. Corollas small, 5-7 mm wide and 5-7.5 mm long. 59. Fruiting calyx 20-35 mm lone. 1967] Physalis — Waterfall 91 60. Fruiting calyx with smooth ribs. .................... T reer E a E 43. P. philadelphica. 60. Ribs of fruiting calyx with more or less tooth-like lamellations developing at the bases of some of the hairs, or replacing the hairs. EEEE E tc E EA, a a 59. Fruiting calyx 5-20 mm long. 61. Fruiting calyx 13-20 mm long. ........................ OPERE MOL A E RO SONNEN 48. P. Lagascae. 61. Fruiting calyx 5-11 mm long. 62. Corolla 5-10 mm long; fruiting pedicels (5-) 10-15 mm long, usually longer than the fruiting calyx. ......... 51. P. sulphurea. 62. Corolla ca. 3.5 mm long; fruiting pedicels ca. 5 mm long, shorter than the fruiting CRY Wier nine due 49. P. microcarpa. 15. Flowering pedicels usually (10-) 15-40 mm long, several times the length of the flowering calyx. 63. Anthers yellow. 64. Corolla yellowish, sometimes with slightly con- trasting spots, sometimes drying bluish. 65. Leaf blades ovate to narrowly lanceolate and sometimes more or less hastate, margins entire or with 1-2 usually obscure teeth on each side. STET REN 37. P. glabra. 65. Leaf blades usually ca. ovate, if lanceolate then usually with several teeth on each margin; stems more or less vestite. .... 38. P. crassifolia. 64. Corollas purplish, violet or bluish, the throat AEI sh SAHNE ae E CAS CINQUE NEN 40. P. purpurea. 63. Anthers bluish or bluish-green, at least on their margins. 66. Corollas campanulate-funnelform; anthers usually 1-2.5 mm long, not twisted. ........ 45. P. angulata. 66. Corollas rotate to reflexed-rotate; anthers usually 2.5-3.5 mm long, sometimes twisted at maturity. Ew MI MEME EU dm 50, P. acutifolia. 13. Fruiting calyx 5-angled. 67. Flowering calyx large (5-) 7-10 (-14) mm wide at base of lobes. 68. Hairs more or less dendritic or branching. .................. 68. Hairs not dendritic. 69. Calyx lobes narrow, long-attenuate. 70: Leaves cordate. 5... 2. 08 9$. 54. P. Greenmanii. 70. Leaves not cordate, varyingly narrowed basally. 55. P. Pringlei. Besestsssssasssssssossssossonsesssseooseecssssomoocestootn 92 Rhodora [Vol. 69 69. Calyx lobes deltoid or ovate to lanceolate, acuminate. 71. Leaf blades ca. ovate. 72. Stem hairs long, spreading, jointed; stems her- baceous. 73. Fruiting calyx only a little longer than wide. RPEERRRRRRRRRERRR 55. P. Pringlei. 73. Fruiting calyx 1.5-2 times longer than wide. mM 57. P. subrepens. 72. Stems glabrescent, the few hairs short, ap- pressed; stem lignescent. ...... 62. P. jaliscensis. 71. Leaf blades deltoid-ovate, sometimes subhastate, principal ones 15-25 mm long and 12-20 mm wide. m" 30. P. philippensis. 67. Flowering calyx usually 2-5 (-6) mm wide at base of lobes. 74. Fruiting calyx narrow, about twice as long as wide. 75. Flowering pedicels 8-45 mm long. 76. Corollas prominently maculate; anthers bluish. 77. Pedicels 15-40 mm long; herbaceous vine. PERRA 58. P. volubilis. 77. Pedicels 8-15 mm long; stems lignescent. MER 61. P. lignescens. 76. Corollas immaculate, or spots not contrast- ing; anthers yellow or greenish-yellow. .... "MORE 59. P. viridoflava. 75. Flowering pedicels 3-5 mm long. mE 56. P. angustiphysa. 74. Fruiting calyx as long as wide to 1.5 times longer than wide. 78. Vestiture dense, of more or less branching, jointed hairs. ........... sese 77. P. vestita. 78. Hairs unbranched, or none. 79. Fruiting calyx densely greyish short-hairy (as usually all herbage); gynophore and inside of fruiting calyx capitate-glandular. nctenosenanssesetecsessoscoesoaseaoosssevesoesossess 68. P. ignota. 79. Fruiting calyx glabrous or variously ves- tite, but not densely greyish short-hairy; inside of fruiting calyx not capitate- glandular excepting sometimes in P. tur- binata. 80. Calyx lobes caudate-subulate to subulate - attenuate. 81. Flowering pedicels 15-75 mm long; stems, petioles and pedicels usually 1967] Physalis — Waterfall 93 more or less pilose, longer hairs 1.5-4 mMM lone s 63. P. maxima. 81. Flowering pedicels usually 1-15 mm long; stems, petioles and pedicels glabrous to variously vestite, but not long-pilose. 82. Corollas immaculate, or with obscure markings with edges sometimes dif- fuse. 83. Corollas small, 4-8 mm long; well- developed fruiting calyces usually 40:65 MINTO LOND, eeel atiina eis a O eae, 64. P. nicandroides. 83. Well-developed corollas (8-) 10-15 mm long and (10-) 15-25 mm wide; fruiting calyces usually 20-35 mm long. .... 69. P. foetens. 82. Corollas with small contrasting ma- culations ea. 2 mm long and 1 mm wide, or smaller in P. latiphysa. 84. Corollas 4-6 mm long; anthers 1-1.8 mm long. 85. Fruiting calyx 15-23 mm wide. MERE D 70. P. subulata. 85. Fruiting calyx (25-) 30-40 mm wide. ............ 67. P. latiphysa. 84. Corollas (8-) 10-15 mm long; anthers (1.5-) 2-3 mm long. ........ HOPED CH TAA P. foetens. 80. Calyx lobes ovatish to narrowly tri- angular - attenuate, but not subulate. 86. Fruiting calyx hairy. 87. Well-developed fruiting pedicels 20- 45 cm long. 05 5 cB ul ii 60. P. longipedicellata. 87. Fruiting pedicels 3-12 mm long. 88. Corolla maculate, spots strongly contrasting. 89. Hairs of stem short, antrorsely appressed or curled; flowering calyx two-thirds or three- fourths divided into narrowly lanceolate-triangular lobes 4- 5 mm long; fruiting calyx ob- scurely U-RHEIeQ. vene M E 19. P. chenopodifolia. 89. Stem-hairs more or less spread- 94 Rhodora [Vol. 69 ing; flowering calyx usually di- vided no more than half way. 90. Leaf blades more or less translucent when dry; some of the hairs with brownish glands. .... 75. P. leptophylla. 90. Leaves thicker, opaque; hairs usually without brownish glands. 91. Fruiting calyx 15-30 (-32) mm long. 92. Anthers bluish, 93. Lobes of fruiting calyx ovate-deltoid to lanceo- late. .. 74. P. pubescens. 93. Lobes of fruiting calyx narrowly lanceolate, tips approaching subu- late. 71. P. angustiloba. 92. Anthers yellow, 1.2-1.6 mm long. ................ eene eee 72. P. turbinatoides. 91. Fruiting calyx 30-40 mm long. ........ 73. P. turbinata. 88. Corolla immaculate, or with spots not contrasting. 94. Anthers blue. .... 69. P. foetens. 94. Anthers yellow, sometimes tinged with green or blue. ........ eeseeeee eet ee eren nennen 76. P. hylophila. 86. Fruiting calyx glabrous or essentially so. 95. Fruiting calyx 25-35 mm long. 96. Lobes of fruiting calyx 7-13 mm long. 97. Corolla maculations prominent; fruiting calyx usually gradually terminated apically. .................... NEMMMMRRERRRN 65. P. cordata. 97. Corolla maculations not promi- nent, their edges often diffuse; fruiting calyx more or less por- reet. ....eseeessss 66. P. porrecta. 96. Lobes of fruiting calyx 4-6 mm long. .............. 78. P. clarionensis. 95. Fruiting calyx 10-18 mm long. 98. Anthers 1.3-2 mm long; fruiting 1967] Physalis — Waterfall 95 calyx 15-18 mm long. Eoen T er oe AE i: VAM 79. P. minuta. 98. Anthers ca 1 mm long; fruiting calyx 10-13 mm long. 99. Young parts sparingly short- hairy; leaf blades ovate to oblong-ovate or ovate-lanceclate. Dip eae EEN 80. P. carnosa. 99. Plant glabrous; leaves reniform to-ovate. 5. 39. P. Mimulus. THE SPECIES 1. Fhysalis arborescens L., Species Plantarum, ed 2:261. 1762; P. mayana Standley, Field Mus. Publ. Bot. 8:42. 1920. Shrubby, 0.5-2 m high, more or less tomentcse with somewhat dendritic hairs, leaf blades ovate to rhombic ovate, somewhat acuminate, entire to irregularly few-toothed, sometimes undulate, rarely salient-toothed; blade gradually narrowed to a petiole 1-4 em leng; principal blades (6-) 9-12 cm long and (4-) 6-8 em wide; lower leaf surfaces usually densely dendritic-hairy or stellate- hairy, upper surfaces less so; flowers in axillary fascicles of 3-7, sometimes reduced to 1, maturing at different times, the inflorescence probably representing a reduced branch, often maturing only 1 fruit; flowering calyx mcre or less herbaceous, oblong-hemispheric or short- oblong to nearly square, 5-7 mm long, divided above into ca. ovate lobes ca. 3 mm long, on pedicels 6-12 mm long; ccrolla 8-12 mm long, prominently dark-maculate, divided into lobes 3-5 mm long; anthers yellow, ca. 4 mm long and 1.5 mm wide, their filiform filaments ca. 2 mm long; fruiting calyx ovoid, 2-3.5 em long and 1.8-2.5 cm wide, slightly to densely covered with simple or 2-branched to stellate trichomes; fruiting pedicels 10-15 mm long; berry 7-11 mm in diam- eter, sessile on the somewhat invaginated calyx-base. SELECTED SPECIMENS. MEXICO: cAMPECHE: Houstoun HS 295 76, left side (BM); VERA CRUZ: 2 km south of Tampico, Feb. 1910, Palmer 379 (BM, F, GH, NY, US); YUCATAN: Chichankanab, Gaumer 24381, type collection of P. mayana Standley, (Type: F, Isotype: US); Mérida-Uxmal read, Lundell 8083 (MICH, US). The Houstoun collection in the Sloane Herbarium of the British Museum can be taken as the LECTOTYPE of the spe- cies. Linnaeus (l.c. supra) referred to Miller's Dictionary, t. 206, f. 2, and stated “Habitat in Campechia". The eighth edition of Miller's Dictionary refers to the same figure and says, “The eighth sort was discovered by the late Dr. Hous- toun growing naturally in Campeachy from whence he sent seeds to England". 96 Rhodora [Vol. 69 This collection is obviously the same species later de- scribed from Yucatan as P. mayana. 2. Physalis porphyrophysa Donnell-Smith, Bot. Gaz. 61: 377. 1916. Shrub, 1-3 m tall, *subscandent" according to some collectors; young stems densely vestite with short, antrorsely curled hairs, glabrescent; leaf blades ca. ovate, slightly rhombic-ovate, or some- what narrower, margins entire to slightly and irregularly undulate, larger ones 8-15 cm long and 5-7 cm wide, gradually narrowed to a short petiole ca. 1-3 em long; hairs of the lower leaf surfaces abun- dant, simple, variously branched and stellate, the few on the upper surfaces usually simple; flowers in axillary fascicles of (1-) 2-4 (only a few flowering specimens seen), often only one maturing a fruit; flowering calyx 4-5 mm long, herbaceous, divided above into more or less triangular lobes ca. 2 mm long, pedicels 5-10 mm long, filiform, only slightly thickened near the base of the calyx; no ex- panded corollas seen, but some collectors state “flowers yellow”; a dissected pre-anthesis corolla appears slightly maculate, has oblong anthers ca. 3 mm long on short filiform filaments; fruiting calyx 10- ribbed, usually 25-35 mm long and 18-25 mm wide, sometimes pur- plish at maturity, on pedicels 6-12 mm long; berry ca. 12-15 mm in diameter, subsessile or on a short stipe up to 1.5 mm long on the invaginated calyx-base. i SPECIMENS SEEN. GUATEMALA: Uaxactun, arroyo, Mar. 20, 1931, Bartlett 12163 (MICH, US); Apr. 7, 1931, Bartlett 12519a (MICH, US); on Maya ruins, Apr. 26, 1931, Bartlett 12759 (MICH, US); south bank of Motagua River, June 19, 1909, Deam 6357 (GH, MICH, Us, VT); vicinity of Zacapa, Dec. 1906, Pittier 1754 (Type: US); Zacapa, small subscandent shrub in hedge, Oct. 7-16, Standley 74616 (F, NY, US). 3. Physalis amica Standley & Steyermark, Field Mus. Publ. Bot. 23: 231. 1947. Herbaceous, ca. 1-7 m tall, branched, stems villose with jointed hairs, tending to be glabrescent; leaf blades ca. ovate somewhat acumi- nate, larger ones 6-7 em long and 4-5 cm wide, their petioles ca. 1 cm long; leaf surfaces with scattered, jointed, more or less appressed flattened hairs, these sometimes reduced to bases only; flowers in axillary fascicles of 2-5 (probably reduced branches), sometimes only one; flowering calyx somewhat chartaceous, ca. 7 mm long, divided above into lanceolate or lanceolate-ovate lobes 1.5-3 mm long; pedi- cels ca. 1 cm long, prominently thickened upward to a width of 2-4 mm at base of calyx; pedicels and calyx villosulose; corolla nearly rotate, ca. 15 mm long and 30 mm wide, its lobes 5-8 mm long, broadly triangular to ovate-triangular, without obvious dark spots, matted-hairy in the tube; anthers bluish, oblong, ca. 3 mm long and half as wide; filaments filiform, ca. 3 mm long; fruit unknown, 1967] Physalis — Waterfall 97 GUATEMALA: vicinity of San Rafael, upper narrow reaches of stream, Feb. 20, 1940, Steyermark 36175 (Type: F). 4. Physalis amphitrica (Bitter) Standley & Steyermark, Field Mus. Publ. Bot. 23: 229. 1947; Saracha amphitrica Bitter, Repert. Sp. Nov. 20: 362. 1924. Herbaceous, 1-3 m tall, branched, stem sometimes weak and trail- ing, or said to be shrubby by a few collectors, glabrous, or with a few hairs when young; leaf blades ovatish, acuminate, the larger ones usually 8-15 cm long and 4-8 em wide, their petioles 1-2.5 cm long; leaf surfaces with a few short, jointed, flattened hairs, these some- times reduced to mere bases; flowers in axillary fascicles of 2-5, sometimes only 1 evident, but abscission scars of others present; calyx at anthesis somewhat thin-chartaceous, glabrous outside, or with lobes slightly short-hairy, on pedicels 8-20 mm long, filiform, enlarged apically beneath the calyx; corolla yellowish or greenish- yellow, 20-25 mm long, ca. twice the length of the calyx, divided ca. half way into ovate to lanceolate, sometimes slightly acuminate lobes, with greenish-purplish markings near their bases, matted-hairy inside near the tube; anthers bluish to purplish, ca. 3-4 mm long, on filaments ca. 5 mm long; fruiting calyx seldom collected, 3.5 cm (immature) to 5.5 cm long and 3.5 cm wide, reticulate-veined, much inflated around the small berry. SELECTED COLLECTIONS. MEXICO: CHIAPAS: Siltepec, Jan., 1937, Matuda 254 (F, MICH, US) : Boquerón, Sept., 1913, Purpus 6671 (BM, GH, F, NY, UC, US). GUATEMALA: Chiquihuite, branched herb or shrub 5-10 ft., damp forest, Mar. 8, 1939, Standley 68133 (F, US) ; above Mujulia, Feb. 1, 1941 (F); moist rocky slopes around waterfall, Río Negro, Volcán Tajumulco, Steyermark 37660 (F, NY). 5. Physalis calidaria Standley & Steyermark, Field Mus. Publ. Bot. 23: 231-232. 1947. Herb, 1-2 m tall, young stem with few to many short brownish antrorsely curled hairs, glabrescent; leaf-blades ovate to elliptic, gradually to abruptly acuminate, their margins entire to slightly acuminate, larger ones 8-15 cm long and 5-7 cm wide on petioles 15-35 mm long; blades sparsely appressed short-villosulose above and more or less curled-villosulose to appressed-villosulose on the abaxial surface, more abundantly so on midribs and veins, petioles similarly vestite; flowers in axillary fascicles of usually 4-5, some of them sometimes abscissed at petiole bases; calyx short-hairy on outside, subtruncate, usually 4-10 mm long, its apex sinuate-margined with sepal-projections ca. 1.5-2 mm long, broadly expanding to the tube; corolla light yellow, with 5 slightly contrasting darker spots, rotate- campanulate, 15-20 mm long and 20-35 mm wide when fully expanded, very short-hairy externally and with 5 more or less confluent, densely hairy pads surrounding and above the point of insertion of the filaments; corolla divided about half-way into ovate to deltoid-ovate 98 Rhodora [Vol. 69 lobes; anthers 2.5-3 mm long, on slender filaments 4-5 mm long, glabrous or with a few basal hairs; fruiting calyx glabrous, 10- angled, the 5 principal ones more prominent and basally auricled, a prominent reticulum between the 10 ribs; berry 8-10 mm in diameter, the surrounding calyx 25-30 mm long and 22-25 mm wide. COLLECTIONS SEEN. GUATEMALA: Dept. of Quetzaltenango, 8000 ft, Feb. 4, 1941, Hunnewell 17233 (GH); slopes of Volcán de Zunil, Feb. 3, 1941, Standley 85932 (F); Fuentes Georginas, western slope of Volcán de Zunil, 2850 m, Mar. 4, 1939, Standley 67488 (Type: F); between Todos Santos and Finca el Porvenir, lower to middle slopes of Volcán Tajumulco, Mar. 1, 1940, Steyermark 36988 (F). 6. Physalis chimalteca Standley & Steyermark, Field Mus. Publ. Bot. 23: 232. 1947. Coarse herb, possibly woody below, herbarium specimens 0.5 m long, being tops of plants; stems glabrous, or slightly vestite above with appressed and more or less antrorsely curled hairs; leaf blades ovate, acuminate, larger ones 8-12 cm long and 4-8 cm wide, margins entire or with 1 to 3 large teeth or very shallow lobes on each side, petioles of larger leaves 1-2 cm long; leaf surfaces with a few flat- tened, jointed, usually appressed hairs, these sometimes reduced to bases only; flowers in axillary fascicles of (1-) 2-4, only one matur- ing a fruit in the few seen; calyx at ‘anthesis thin-chartaceous, glabrous externally, the 5 shallow lobes marginally ciliate, the pedi- cel filiform, erect 8-10 mm long; corolla rotate to reflexed-rotate, densely covered externally with short, antrorsely appressed hairs; corolla 15-20 mm long, the upper 8-10 mm divided into lanceolate or ovate-lanceolate lobes, each basally with an area of a few dis- crete dark spots, below which matted hairs extend into the tube; anthers 2.5-3 mm long on filiform filaments 3-5 mm long; fruiting calyx 2.5-5 cm long and 2-3 em wide, glabrous, reticulate-veined, berry oily, ca. 1 cm in diameter, sessile on the invaginated calyx. SPECIMENS SEEN. MEXICO: cuHrapas: Siltepac, Jan. 1937, Matuda 250, (F, MICH, US). GUATEMALA: Volcano Actenango, Feb. 7, 1907, Kellerman 6610 (F); San Miguel Uspantan, April 1892, 'Heyde et Lux 3435 (F); above Las Calderas, common on slopes of Volcán de Acatenango, Jan. 3, 1930, Standley 61803 (Type: F); ibid, Standley 61965 (F); slopes of Volcán de Zunil, wet forested quebrada, Feb. 3, 1941, (F). 7. Physalis melanocystis (Robinson) Bitter, Repertorium Spe- cierum Novarum 20: 369-370. 1924; Withania melanocystis Robinson, Proc. Amer. Acad. 26: 171. 1891. Woody, branched, height not determinable from collections; vesti- ture of short, unbranched, spreading or antrorse-appressed hairs; leaf blades lanceolate to ovate-lanceolate or rhombic-lanceolate, prin- 1967] Physalis — Waterfall 99 cipal ones 4-10 em long and 1.5-5 cm wide on petioles 5-20 cm long; flowers in axillary fascicles of (1-) 3-6; corolla 8-10 mm long and 12-18 mm wide, its limb rotate or rotate-reflexed when fully expanded, divided above into ovate to deltoid-ovate lobes 3-6 mm long; macula- tions not strongly contrasting; anthers yellowish, 3-4 mm long, on filaments 1-2 mm long; flowering pedicels 4-10 mm long; fruiting calyx 17-25 mm long and 15-18 mm wide on pedicels 10-15 mm long; berry spheric, 10-13 mm in diameter. Ta. var. melanocystis Vestiture of short, unbranched, dense, more or less spreading hairs, principal leaf blades 4-7 cm long and 1.5-3 em wide on petioles 5-10 mm. long. SPECIMENS SEEN. MEXICO: SAN LUIS POTOSI: Tamasopo Canyon, 1890, Pringle 3285 (GH, US); rocky limestone woods, Espi- nazo del Diablo, Tamosopo Canyon, Aug. 7, 1934, Pennell 17959 (us); Bitter (l.c. supra) cites Schiede 1189 (B) from Papantla, VERACRUZ. "b. var. cernua (Donn.-Smith) Waterfall, comb. et stat. nov., Athenaea cernua Donn. Sm., Bot. Gaz. 48: 297. 1909. A var. melanocystis differt caulibus trichomatibus curtis antrorso- adpressis; foliis majoribus 7-10 cm longis et 3-5 cm latis, petiolis 1-2 em longis. SPECIMENS SEEN. GUATEMALA: shaded bank of Motagua River, Gulan, June 15, 1909, Deam 6279 (G, US). MEXICO: GUER- RERO: Acapulco, Dec. 1955, Paray 1813 (MEXP). Although the two varieties are similar, the appressed vestiture and larger leaves of var. cernua seem to adequate- ly distinguish it from the more northern var. melanocystis. Future collecting should find intermediate localities for the species. 8. Physalis constricta Waterfall, sp. nov. Herba subglabra; foliis ovatis, majoribus 5-6 cm longis et 30-45 mm latis, margine inaequaliter magnodentato vel integerrimo, petio- lis 15-25 mm longis; calycibus anthesi 11-12 mm longis et 6-7 mm latis, lobis triangularibus vel ovatis, 3-4 mm longis; pedicellis 4 cm longis; corollis 20-26 mm longis, mediis ca. 10 mm latis, sursum gradatim attenuatis, apertis constrictis, 5 mm latis; antheris 4 mm longis, filamentis filiformibus; calycibus fructiferis 28 mm longis et 20 mm latis, inflatis. MEXICO: oaxAca: in 1842, Ghiesbreght s.n. (Type: P). This distinctive species is characterized principally by its long, semitubular corollas, 20-26 mm long, expanded to a width of ca. 10 mm above the calyx and gradually nar- rowed above, then constricted into an aperture ca. 5 mm wide, essentially unlobed, and without a limb, and by its 100 Rhodora [Vol. 69 long, erect pedicels, ca. 4 cm long; although the plant is nearly glabrous, the stem has a few short jointed hairs present, and the upper surfaces of some of the leaves have a few capitate-glandular ones present. If the corolla both expanded and constricted more abrupt- ly, and if it were pentangular above, it would be reminiscent of a large species of Margaranthus. The corolla is not fun- nelform and Ipomoea-like as in Cacabus, sens. str.; the fruiting calyces are similar to those found in Physalis. One has the alternatives of creating for it a new genus in the Solaneae-Lyciinae of Wettstein in Die Natürlichen Pflanzen- familien, of placing it in Margaranthus as an anomalous species, or of treating it as an anomalous species of Phy- salis. Believing that the latter disposition better indicates its possible relationships, the author is so treating it. 9. Physalis campanula Standley & Steyermark, Field Mus. Publ. Bot. 23: 18-19. 1943. Branching herb, only ends of branches collected; herbage covered with spreading jointed hairs of varying lengths, some capitate- glandular; stems glabrescent; leaf blades ca. ovate, the larger ones 5-7 em long, sometimes with 1-3 irregularly shaped teeth or lobes on each margin, attenuate; most branch leaves 3-5 cm long with entire margins; petioles 1-4 cm long; flowers solitary; flowering calyx nearly oblong, 8-10 mm long and 7-8 mm wide, the upper 3-5 mm differentiated into deltoid acuminate lobes 3-5 mm long with ciliate margins; corolla immaculate, tubular, somewhat swollen above the calyx and slightly constricted at the minutely 5-dentate apex, 17-18 mm long and 8-9 mm wide; anthers ca. 3 mm long, bluish, on fila- ments 5-7 mm long, equally inserted near the base of the corolla with tomentose areas extending 1-2 mm above the point of insertion; fruit- ing calyx (only 1 approaching maturity) 25 mm long and 17 mm wide, the berry ca. 1 cm in diameter. GUATEMALA: Volcan Tacana, along Quebrada Canjula between Sibinal and Canjula, Feb. 18, 1940, Steyermark 36067 (Type: F). 10. Physalis glutinosa Schlechtendal, Linnaea 19: 310-311. 1846; Cacabus mexicanus S. Watson, Proc. Amer. Acad. Arts & Sci. 18: 127. 1883; Physalis eximia Standley, Field Mus. Publ. Bot. 17: 273- 274. 1937. Herbaceous, or perhaps somewhat woody below, branched, as much as 1.3 m tall and broad, or “in crevices of cliffs, half-woody, hanging"; herbage somewhat glutinous-vestite with spreading, jointed, flat- tened hairs 1-1.5 mm long intermingled with shorter ones, many of 1967] Physalis — Waterfall 101 the hairs capitate-glandular, sometimes only a few long ones present; leaf blades ovate, sometimes cordate, larger ones 3-6 cm long and 2-5 em wide, rarely sub-rotund; branches often have only smaller leaves; leaf margins dentate-sinuate, often with 5-10 irregularly shaped teeth on each side, sometimes nearly entire; leaf surfaces often spreading-hairy below and appressed hairy above, sometimes only along the principal veins below and reduced above to a few short hairs or hair-bases; petioles of principal leaves 2-6 cm long; flowers axillary, solitary, large; flowering calyx oblong or ovate- oblong, often 5-8 mm wide at base and 8-12 mm wide at base of calyx lobes, and 10-18 mm long, the upper one-fourth to two-thirds divided into narrowly triangular, sometimes acuminate, lobes 4-11 mm long; corolla campanulate to subcampanulate, 20-35 mm long and 3-4 cm wide when fully expanded, limb not reflexed, not at all 5-lobed, yellowish distally, with 5 large purplish-splotched areas which may cover most of the corolla, these rarely seemingly absent in herbarium specimens; anthers yellow with a slight bluish tinge, 3.5-5 mm long and 1.3-2 mm wide on slender glabrous filaments 10-15 mm long; style 2-3 em long; pedicels 7-20 mm long; fruiting calyx 10-ribbed, hairy, 25-45 mm long and 20-30 mm wide on pedicels 10-23 mm long; berry 12-20 mm in diameter. 10a. var. glutinosa Corolla limb maculate with 5 large purplish, often nearly contig- uous, areas; vestiture rather dense, often giving the plant a slightly greyish appearance. SELECTED SPECIMENS. MEXICO: DURANGO: grasslands with sparse growth of oak scrub, 14 miles w of Durango, June 21, 1950, Maysilles 7014 (MICH); Durango and vicinity in 1896, Palmer 313 (BM, F, GH, NY); mountainside, 5 miles sw of Durango, 3 feet tall, corolla pale maroon inside, Waterfall 15409 (BM, F, GH, HAL, NY, OKLA, P, S, SMU); GUANAJUATO: many stemmed, half-woody, hanging to 1 meter, hills 24 miles w of Xichu, June 14, 1957, McVaugh 14813 (MICH); HIDALGO: Mineral del Monte, Cuesta Blanca, in humosa terra inter rupes porphyriticas ferratas, July 1836, Ehrenberg 585 Type: HAL); Pachuca to Real del Monte, May 1828, Graham 268 (BM); hills above Pachuca, Aug. 5, 1898, Pringle 7576 (F, VT); SAN LUIS POTOSI: in 1878, Parry & Palmer 650 (GH, NY, P); in montibus San Migueliteo in 1876, Schaffner 704 (Type collection of Cacabus mexicanus Watson: GH, NY); ZACATECAS: south slope of La Bufa, Aug. 10, 1948, Dressler 112 (Gu); plant 5 ft. broad and high, hills, Oct. 26, 1888 Pringle 1742 (F, GH, NY, UC, VT). var. eximia (Standley) Waterfall, comb. et stat. nov., P. eximia Standley, Field Mus. Publ. Bot. 27: 273-274. 1937, MEXICO: CHIHUAHUA: Majalca, June 24, 1935, LeSueur 894 (Type: F); LeSueur 115, Aug. 18-20, 1935 (F, GH, UC). This taxon is similar to var. glutinosa, but with the 102 Rhodora [Vol. 69 corolla immaculate, or nearly so, and the plant less densely vestite, appearing greener. 11. Physalis cinerea Waterfall, sp. nov. Planta cinerea; foliis ramealibus ovatis inaequalibus, 2-3 cm longis 15-25 mm latis; corollis immaculatis campanulatis 25-30 mm longis; antheris flavis 3-4 mm longis et 1.5-2 mm latis; filamentis 10-12 mm longis. Physalis cinerea is similar to P. glutinosa, but is smaller and is cinereous with short, stiffly spreading, flattened, jointed hairs, some of which are capitate-glandular; prin- cipal leaf blades (from branches only) are smaller than in P. glutinosa; flowering calyx 10-12 mm long and 5-8 mm wide, divided about one-third to one-fourth way into deltoid, slightly acuminate lobes; no dark spots are evident in the specimens seen, but this characteristic is sometimes lost in drying; mature fruits unknown. MEXICO: SAN LUIS POTOSI: in the region of San Luis Potosí, alt. 6000-8000 ft., Parry & Palmer 649 (Type: Ny, Isotype: GH). 12. Physalis viscosa L., Species Plantarum 183. 1753; P. curas- savica L., ibid. 182, P. pensylvanica L., Sp. PL, ed. 2: 1670. 1762; other synonymy listed under the varieties, Since several of the varieties are extra-limital, no inclu- sive species description is here included. 12a. var cinerascens (Dunal) Waterfall, Rhodora 60: 136. 1958; P. pensylvanica L., var. cinerascens Dunal in D. C. Prodromus 13(1); 435-436. 1852; P. curassavica L., var. integrifolia Dunal, l.c. 438; P. mollis Nutt., var. cinerascens (Dunal) Gray, Proc. Amer. Acad. Arts & Sci. 10: 66. 1875; P. cinerascens (Dunal) Hitchc., Spring Flora of Manhattan 32. 1894; P. mollis Nutt., var. parviflora Rydb., Mem. Torr. Bot. Club 4: 355. 1896; P. saltillensis Fernald, Proc. Amer. Acad. Arts & Sci. 35: 568-569. 1900. Perennial, 5-80 cm tall, erect to decumbent, covered to a greater or lesser degree with stellate, or somewhat dendritic, hairs of varying size; principal leaf blades 1.5-7 cm long, and 1.5-5 em wide on petioles 1-5 em long; leaf-margins dentate, undulate or entire; corolla yellow- ish, dark-maculate, 9-15 mm long, and 8-16 mm wide, reflexed-rotate when fully expanded, on pedicels 7-35 mm long; fruiting calyx 15-35 mm long and 15-28 mm wide, half to three-fourths filled by the berry, on pedicels 8-40 mm long. SELECTED SPECIMENS. MEXICO: cutApas: Escuintla, May 13, 1936, Matuda 239 (MICH, US, VT); Hacienda Monserrate, Sept. 1923, Purpus 9212 (GH, NY, UC); CHIHUAHUA: Aldama, May 15-17, 1967] Physalis — Waterfall 103 1908, Palmer 240 (F, GH, MICH, NY, US); COAHUILA: Sierra de Santa Rosa s of Muzquiz, July 8, 1938, Marsh 1269 (F, GH); near Saltillo, Sept. 1898, Palmer 332 (Type of P. saltillensis: (MICH, NY, US); Parras, July 1910, Purpus 4601 (F, GH, UC, US); along arroyo in desert, 18 miles ne of Saltillo, Aug. 6, 1957, Waterfall 13230 (F, HAL, MICH, OKL, OKLA, S, SMU); DISTRITO FEDERAL: Apr. 10, 1938, Balls 4144 (BM, UC, US); GUANAJUATO: Silao, May 19, 1891, Pringle 5148 (GH, VT); with Prosopis and Acacia, 14 miles se of Leon, Aug. 16, 1957, Waterfall 13880 (OKLA); HIDALGO: valley near Tula, July 16, 1896, Pringle 6362 (BR, F, GH, NY, UC, US, VT); MEXICO: Vallée de Mexico, 1865, Bourgeau 112 (BR, GH, US); MORELIA: Arséne sin. num. (F); NUEVO LEON; w of Linares, Aug. 8, 1930, Bartlett 10831 (F, GH, MICH); Hacienda Pablillo, Galeana, Taylor 44 (F, GH, NY); desert between mountains, 22 miles w of Monterrey, Waterfall 15316 (OKL, OKLA, SMU); OAXACA: San Geronimo, July 1-5, 1895, Nelson 2770 (GH, US); PUEBLA: San Luis Tultitlanapa, Purpus 3583 (UC); QUERETARO: San Juan del Rio, Aug. 18, 1905, Rose et al 9534 (NY, US) ; desert hillside, 2 miles ne of Zimapan, Aug. 20, 1957, Waterfall 14121 (OKL, OKLA, P, SMU); SAN LUIS POTOSI: region of San Luis Potosi, 1878, Parry and Palmer 641 (GH, NY, US); along arroyo run- ning through shrubs, 25 miles ne of San Luis Potosi, Aug. 20, 1959, Waterfall 15709 (OKLA, SMU); TAMAULIPAS: circa Matamoros urbem, 1831, Berlandier 2316 (Lectotype: GH, Isolectotype: F); Tampico, Jan. 1-31, 1910, Palmer 126 (BM, F, GH, NY, US); in canyon 4 km w of Miquihuana, Aug. 4, 1941, Stanford, et al 674 (GH, OKLA, UC) ; ZACATECAS: near Concepcion del Oro, Aug. 11-14, 1904, Palmer 318, (NY); YUCATAN: in 1895, Gaumer 482 (NY); Chichen Itza, June 6, 1932, Steere 1008 (F, MICH). The type collection, Berlandier 2316 from near Mata- moros, has slightly sinuate-dentate leaves. There are col- lections with entire, often relatively broader, leaves which appear to belong to a distinct phase, but there are many in- termediates. Indeed the type itself is intermediate between the phase with entire margins and the one with dentate leaves. Since the two occur from Kansas to central Mexico, no nomenclatural differentiation is made here. 12b. var. spathulaefolia (Torr.) Gray, Proc. Amer. Acad. Arts & Sci. 10: 67. 1875; P. lanceolata Michx. var. spathulaefolia Torr., Bot. Mex. Bound. 153. 1859. Leaf blades ovate to lanceolate or spathulate, basally tapering, petioled. MEXICO: TAMAULIPAS: en las dunas que separan la Laguna Madre y el Golfo, Rancho El Mezquite, Mar. 13, 1963, Gonzales 55 (MEXP). 12c. var. sinuatodentata Schlechtendal, Linnaea 19: 309 1846. 104 Rhodora [Vol. 69 P. curassavica L., var. sinuato-dentata (Schlecht.) Dunal in D. C., Prodromus 13 (1): 438. 1852. This is very similar to var. cinerascens, but appears to be a weakly differentiated local variety from coastal sand dunes near Vera Cruz. It consists of small plants having dendritic or fasciculate hairs in addition to the stellate ones, coarsely dentate leaves varying to entire, and elongate cord-like rhizomes, SPECIMENS EXAMINED. MEXICO: vERACRUZ: Mexico, Ehren- berg (Type: HAL); VERA CRUZ, 1853, Muller 175 (NY) in collibus arenosis pr. Veracruz, July 1828, Schiede 190 (HAL); sand dunes, Veracruz, June 24, 1901, Pringle 8517 (F, NY, OKLA, P, UC, US, VT). 12d. var. yucatanensis Waterfall, var. nov. Planta vestita, trichomatibus dendriticis vel stipato-stellatis ad 2 mm longis. This variety is characterized primarily by the spreading dendritic trichomes of the stem, petioles, pedicels and caly- ces, the longer ones 0.5-2 mm long, some of them stipitate, the basal one-third to two-thirds unbranched ; similar hairs present to a greater or lesser extent on the leaf blades where they are often smaller. SPECIMENS SEEN. MEXICO: vUCATAN: Chichankanab, Gawmer 1798 (Type: F, Isotypes: BM, NY, US); Gaumer 482 (BM, F, GH, US) not all duplicates are this species), Chichankanab, Gawmer 1518 (F, US); Izamal, Gawmer 2427 (F); 2 ft. high, Izamal, Gaumer 23166 (£); Mérida, Rivas 16 (F); Tirul, Mar. 6, 1903, Seler 3911 (F, GH); Chichen Itza, June 6 & 29, 1932, Steere 1008 & 1647 (F) Muna, July 22, 1932, Steere 2109 (F). 13. Physalis Mevaughii Waterfall, sp. nov. Frutex, 1.5-2.5 m alta; trichomatibus paucis brevibus; foliis ovatis vel rhombico-ovatis, attenuatis, majoribus 7-13 cm longis et 4-9 cm latis, integerrimis vel paucidentatis, petiolis 1.5-5.5 em longis; caly- cibus floriferis 8-9 mm longis et 10-15 mm latis ad basim loborum; calycis lobis deltoidibus 3-5 mm longis; pedicellis 25-30 mm longis; corollis maculatis campanulato-rotatis vel reflexo-rotatis, 11-13 mm longis et 20-25 mm latis; antheris violaceis, 2.7-3.5 mm longis, fila- mentis 3-4 mm longis; calycibus fructiferis decagonatis, 25-45 mm longis et 25-35 mm latis; pedicellis fructiferis 4-5.5 cm longis; baccis 15-20 mm latis. This nearly glabrous species of Physalis is characterized by its large fruiting calyces, 10-angled, but with the pri- mary angles more prominent than the secondary ones, 1967] Physalis — Waterfall 105 borne on long pedicels 4-5.5 cm in length. The hardened stems appear to be of a persistent herbaceous type becom- ing woody. SPECIMENS SEEN. MEXICO: JALISCO: shrub 2 m high, precip- itous mountainsides with tall wet mixed forest of firs and deciduous trees, 3-5 miles nw of San Miguel de la Sierra, Nov. 8, 1962, Mc- Vaugh 22034 (Type: MICH) ; steep mountain slopes in moist fir forest, Nov. 12-13, 1952, McVaugh 14064 (MICH); single-stemmed shrub 1.5-2.5 m high, fir forest on steep mountainsides 11-12 miles s of El Rincon, McVaugh 21512 (MICH). 14. Physalis stapelioides (Regel) Bitter, in Fedde, Spec. Nov. 18: 5-7. 1922; Saracha stapelioides Regel, Supplm. Hort. Bot. Petropol. 18. 1864; Saracha stapeliaflora Dcne., Catal. Mus. d'Hist. Nat. Paris 7. 1863 (nomen nudum); Physalis acuminata Greenman, Amer. Acad. Arts & Sci. 35: 311. 1900. Suffrutescent, or “soft-shrubby”, the collected branch-ends some- times herbaceous, 1-3 m high, stem as much as 3 cm thick; herbage with jointed, spreading hairs, some as much as 2-3 mm long, others much shorter, some capitate-glandular, the glands often reddish or brownish; leaf blades ca. ovate, acuminate, in two ranges of contrast- ing sizes, the larger ones, especially on the main stem, 10-20 cm long and 8-15 cm wide on petioles 4-13 cm long, smaller ones often 6-8 cm long and 4-5 cm wide on petioles 2-3 cm long; branch-ends with only smaller leaves sometimes collected; leaf surfaces usually with a few appressed hairs; flowers single; flower buds large immediately prior to anthesis, as much as 10 mm long and 7-8 mm wide; calyx at anthesis oblong to oblong-flaring, 10-15 mm long, the upper one- third to one-half divided into lanceolate to lanceolate-ovate acumi- nate lobes, 8-9 mm wide basally to sometimes 12-13 mm wide at base of calyx lobes; corolla light yellow, 15-25 mm long and 25-35 mm in diameter when fully expanded, pentangular with slight sinuses be- tween the somewhat acuminate apices, conspicuously maculate with 5 clusters of discrete spots, densely matted-hairy in the tube below the maculations; anthers oblong-ovate, bluish, 4-5 mm long on glab- rous filaments 4-7 mm long; fruiting calyx 10-angled, the sinus-veins not prominent, sometimes 2-branched, covered with long and short hairs, some glandular-capitate, ca. ovate, 25-40 mm long, on pedicels 10-25 mm long; berry oily, 12-20 mm in diameter, sessile, or essen- tially so, on the invaginated calyx-base. SELECTED COLLECTIONS, MEXICO: DISTRITO FEDERAL: loose bushy shrubs to 10 ft. tall and as much through, in open places in forest, Desierte de los Leones, 10000 ft altitude, Mar. 28, 1938, Balls 4060 (BM, GH, MICH, UC, US) ; GUERRERO: by stream in oak and pine forest, Teotepec, May 20, 1939, Hinton 14299 (F, GH, US); top of Sierra Madre near Chilpancingo, Dec. 24, 1894, Nelson 2229 (GH, 106 Rhodora [Vol. 69 US); HIDALGO: between Pachuca and Real del Monte, Rose, et al, 8677 (US); JALISCO: simple shrub to 2 m, abundant in fir zone at head of Barranca de la Rosa, northeastern slopes of Nevado de Colima below Canoa de Leoncito, Oct. 10, 1952, McVaugh 13417 (MICH, OKLA); soft shrub to 8 ft. branches over 1 inch diameter, pine-fir forest, north slope of Nevado de Colima, Aug. 25, 1957, Straw & Gregory 1241 (MICH) ; MEXICO: Sierra de las Cruces, Oct. 23, 1892, Pringle 5315 (P. acuminata, Type: GH, Isotype: F); Under cliffs, Sierra de las Cruces, 10000 ft, Sept. 12, 1904, Pringle 13128 (F, GH, MICH, OKLA, US, VT); MICHOACAN: pine-fir forests, mountains w of Cerro San Andres, 8-10 miles nw of Ciudad Hidalgo, Mar. 19, 1949, MeVaugh 9940 (F, OKLA); VERACRUZ: Orizaba, Botteri 220 (GH). 15. Physalis orizabae Dunal, in D.C. Prodrumus 13 (1): 452. 1852; P. subintegra Fernald, Proc. Amer. Acad. Arts & Sci. 35: 567-568. 1900. Perennial from a woody, sometimes slender elongate, base, 15-45 cm high; vestiture of short antrorsely curled hairs often intermixed with varying amounts of long jointed hairs, 1-2 mm long, the latter often more abundant on the pedicels and calyces; leaf blades broadly to narrowly ca. ovate, principal ones usually 3-5 cm long and 2-3 cm wide their margins usually entire, but sometimes with 1-3 irregularly shaped teeth on each side; blades often with only a few scattered, often appressed, hairs; petioles of principal leaves 1-2 em long; calyx at anthesis oblong to more or less campanulate, 6-10 mm long and 5-10 mm wide at base of lobes, upper one-third to one-half divided into lanceolate to ovate-deltoid, sometimes acuminate, lobes; vestiture of calyx of varying amounts of long flattened jointed hairs, more or less spreading; pedicels 10-15 mm long; corolla yellowish, prominent- ly maculate, 12-22 mm long and 15-30 mm in diameter when fully rotate-opened, tomentose in the throat, and with jointed, appressed hairs on the outside; anthers nearly oblong, bluish or violet, 2.2-3 mm long on glabrous, somewhat thickened, filaments 3-5 mm long; fruit- ing calyx 10-angled or 10-ribbed, often large when fully developed, 2-4 cm long and 18-35 mm wide on pedicels 10-15 mm long; berry 1-2 em in diameter, sessile to subsessile on the barely invaginated calyx. SELECTED COLLECTIONS: MEXICO: AGUASCALIENTES: oak forests near summit of Sierra del Laurel, 10 miles se of Calvillo (somewhat questionably referred to this species), McVaugh 18439 (MICH, OKLA); DURANGO: Tejamen, Aug. 21-27, 1906, Palmer 480 (F, GH, NY, US); in pine woods 3 miles sw of El Salto, Waterfall 16186 (F, GH, OKLA, SMU); DISTRITO FEDERAL: Sierra de Ajusco, 9000 ft, June 6, 1904, Pringle 13277 (F, GH, MICH, OKLA, UC, US, VT); HIDALGO: Real del Monte, May 27, 1827, Berlandier 248 (BM); Sierra de Pachuca, July 21, 1901, Rose & Hay 5576 (GH, NY, US) ; JALISCO: cut-over and grazed mountainsides in oak zone, se slopes of Cerro 1967] Physalis — Waterfall 107 Gordo, 12 miles se of Tepatitlan, Aug. 29, 1958, McVaugh 17508 (MICH); near Huejuquilla, Aug. 25, 1897, Rose 2577 (GH, US); MEXICO: Carboneras, Río Verde, Temascaltepec, June 26, 1935, Hin- ton 7717 (F, GH, MICH, NY, US) ; Sierra de las Cruces, 10000 ft, Sept. 10, 1899, Pringle 8225 (Type of P. subintegra: GH, Isotypes: BR, F, OKLA, UC, US, VT) ; MICHOACAN: w of Tancitaro, July 19, 1940, Leaven- worth 296 (F, MICH, NY); near Hamburg Courts, Lago Patzcuaro, Aug. 20, 1949, White 238 (OKLA, US); MORELOS: in mountains near Cuernavaca, Aug. 10, 1949, White 234 (OKLA, US); NUEVO LEÓN: in open woods, San Francisco Canyon, 15 miles sw of Pueblo Galeana, Mueller 346 (F, GH, MICH); OAXACA: nw slope of Mt. Xempoaltepec, 8000 to 10000 ft, Nelson 681 (GH, US); PUEBLA: in pine woods, San Manuel de la Sierra, Sierra Negra, 10400 ft, Balls 4464 (BM, UC, US); abundant near San Martin, July 14, 1929, Mexia 2654 (BM, F, MICH, NY, UC) ; QUERÉTARO: Zimapan, Cottam 10439 (US) ; mountain- side 34 miles se of San Juan del Rio, Aug. 18, 1957, Waterfall 14020 (F, GH, MICH, OKLA, P, SMU); SAN LUIS POTOSÍ: Alvarez, July 13-23, 1904, Palmer 225 (GH, US); TAMAULIPAS: top of mountains, Santa Rita Ranch, April 7, 1926, Runyon 1051 (US); TEPIC: Sierra Madre between Dolores and Santa Gertrudis, Aug. 1897, Rose 2049 (GH, NY, US); VERACRUZ: between Banderilla and Jalapa, Feb. 15, 1943, Gilly et al 146 (MICH); ad radices montis Orizaba, Sept. 1828, Schiede, sim num (Type: HAL). This variable species shows increasing variation at the lower altitudinal and the northern geographic limits of its range. Even near the center of its range the vestiture, size of fruits and of flowers vary considerably. The TYPE, from the foot of Mt. Orizaba, has abundant spreading hairs on the flowering calyces and pedicels. The type of P. subintegra, from Sierra de Las Cruces, has spreading to antrorsely curled hairs on the flowering calyces and pedi- cels and develops the extreme in large thick fruiting calyces. Some collections show the vestiture of the leaves and young stems varying toward shorter, denser, more appressed hairs, approaching P. chenopodifolia (P. puberula) in that respect. 16. Physalis ingrata Standley, Field Mus. Publ. Bot. 17: 391. 1938. Plant herbaceous, or woody below, from a woody root; stems 15 cm (in Type) to 60 cm tall, vestiture of short, retrorse-spreading to spreading hairs; leaf blades ovate, the principal ones 3-5 cm long and 18-25 mm wide, their surfaces more or less vestite with short, spreading to somewhat appressed hairs; petioles 7-20 mm long; flower-buds prior to anthesis ca 6 mm long and 5-6 mm wide; flower- ing calyx broadly oblong to hemispheric or somewhat campanulate, 108 Rhodora [Vol. 69 6-9 mm long and 5-6 mm wide at base of the lanceolate-ovate to del- toid, slightly acuminate, calyx lobes, which are 3-5 mm long; pedicels 7-12 mm long; corolla campanulate, yellowish, dark-maculate, 10-12 mm long and 7-12 mm wide; anthers nearly oblong, 3-4 mm long, yellow with a bluish or greenish-blue tinge, on glabrous filaments 3-4 mm long; fruiting calyx (immature) 17-20 mm long and 11-12 mm wide; berry ca. 8 mm in diameter, sessile or on a stipe as much as 0.5 mm long on the slightly invaginated calyx-base. HONDURAS: on rocky hillside above Siguatepeque, July 14, 1936, Yuncker, et al 5865 (Type: F, Isotype: GH) ; Yuncker; et al 5794 (F, GH). 17. Physalis aggregata Waterfall, sp. nov. Planta herbacea, trichomatibus articulatis, brevibus, apice capitato- glanduliferis, et etiam paucis longioribus, 0.5-1.5 mm longis; foliis lanceolatis vel ovato-lanceolatis, integerrimis vel paucidentatis, 6-10 mm longis et 3-5.5 mm latis, petiolis 2-4.5 cm longis; floribus aggre- gatis; calycibus floriferis 5-7 mm longis et 5-7 mm latis; calycis lobis ovatis, 3-3.5 mm longis; pedicellis 4-6 mm longis; corollis 9-10 mm longis et 13-15 mm latis; antheris flavis vel violaceo-flavis, 2.5-3 mm longis; filamentis 3-4 mm longis; calycibus fructiferis 15-20 mm longis et 12-13 mm latis; baccis 10-11 mm latis. MEXICO: oaxaca: Rancho de Calderón, 5500 ft, San Juan, L. C. Smith 771 (Type: GH). This peculiar species, known only from the type collec- tion, is characterized by its aggregations of 3-7 flowers to form an inflorescence (apparently by branch-reduction, the axis, and sometimes bracteal leaves, present), although the first-formed flowers are solitary and axillary; the jointed trichomes are of two kinds, a dense layer of short, often capitate-glandular ones, and scattered longer ones 0.5- 1.5 mm long; the rather immature flowering calyx is nearly filled by the berry. 18. Physalis lassa Standley & Steyermark, Field Mus. Publ. Bot. 23: 19-20. 1943. Perennial, stems herbaceous, or woody below, 30-60 cm tall, from a woody base, densely covered with flat, jointed spreading hairs taper- ing apically on the TYPE, but some hairs having small capitate glands in some of the associated material; leaf blades ovate, acute, sometimes subcordate, more or less densely covered with soft spread- ing hairs above and below, lighter colored below, margins entire, or rarely with 1 or 2 repand teeth on each margin; principal blades 3.5-7 cm long and 2-4 cm wide on petioles 1-2 cm long, sometimes slightly winged in the upper part; flowering calyx oblong to flaring- 1967] Physalis — Waterfall 109 oblong, 8-11 mm long and 5-7 mm wide at base of lobes, upper one- third to two-thirds divided into narrowly deltoid or ovate-lanceolate to ovate-deltoid lobes, on pedicels 6-8 mm long; corolla yellowish, dark-maculate, tomentose in the throat, 10-15 mm long and 12-15 mm wide when fully open; anthers bluish, oblong, ca. 4 mm long and 1 mm wide on filaments of about equal length; fruiting calyx 10- ribbed or slightly 10-angled, 25-35 mm long and 17-25 mm wide; berry 10-15 mm in diameter, essentially sessile on the slightly (2.5 mm) invaginated calyx base. SELECTED SPECIMENS. MEXICO: GUERRERO: vicinity of Aca- pulco, Feb. 1895, Palmer 510 (GH, US, NY); GUATEMALA: Sierra de las Minas near San Geronimo, Mar. 3, 1907, Kellerman 6656 (F); hillside, 7200 ft, near Tecpam, July 20, 1933, Skutch 481 (MICH, US) ; damp oak forest nw of Jalapa, Nov. 16, 1940, Standley ?7476 (F); between Jalapa and Montanya Miramundo, 1500-2000 m, Dec. 7, 1939, Steyermark 32868 (Type: F); Cerro Pixpix above San Ildefonso Ixtahuacan, Aug. 15, 1942, Steyermark 50638 (F). 19. Physalis chenopodifolia Lamarck, Tableau Encycl. 2: 28, no. 2401. 1793; P. puberula Fernald, Proc. Amer. Acad. Arts and Sci. 36: 502. 1901. Plant herbaceous from a woody base, more or less cinereous with short, antrorsely curled hairs; leaf blades usually ca. ovate with margins irregularly coarse-dentate to repand-dentate with 1-10 teeth on each margin, or entire; calyx at anthesis oblong to flaring-cam- panulate, 6-10 mm long and 3-6 mm wide at base of lobes, divided one-third to two-thirds into usually narrowly deltoid-lanceolate, grad- ually attenuate lobes; pedicels 3-10 mm long; corolla yellowish, prominently dark-maculate, 12-20 mm long and 15-30 mm wide when fully rotate-expanded, tomentose in the throat; anthers tinged bluish or bluish-green, 2-2.5 mm long, on slender filaments 2-5 mm long; fruiting calyx 10-ribbed or slightly 5-angled with intermediate ribs or angles, 20-30 mm long and 15-20 mm wide on pedicels 1-2 cm long; berry 10-15 mm wide, 19a. var. chenopodifolia Blades of principal leaves 3-7 cm long and 2-5 cm wide on petioles 10-35 mm long; plant rather cinereous with antrorsely curled hairs. SELECTED COLLECTIONS. MEXICO: DISTRITO FEDERAL: peren- nial in clumps, Tlalnepantla, 7300 ft, June 9, 1901, Pringle 8511 (F, GH, NY, UC, US, VT); MEXICO: Nanchititla, Feb. 22, 1933, Hinton 3431 (BM, GH, F); Sacro Monte, Amecameca, Pringle 9147 (Type of P. puberula: GH, Isotypes: F, US, VT); GUANAJUATO: base of hills, 34 miles north of Querétaro, Aug. 23, 1961, Waterfall 16551 (GH, MICH, OKLA, UC, US); Pachuca, July 1905, Rose et al 8768 (F, GH, NY, US); MICHOACAN: Morélia, 2200 m, Sept. 5, 1912, Arséne 8658 (NY); PUEBLA: Rancho Posadas prés de Puebla, July 25, 1909, Nicolas 209 (GH, NY, P); near Calchicomula, July 24, 1901, Rose et al 5655 (GH, 110 Rhodora [Vol. 69 NY, US) ; QUERÉTARO: in dense clumps, stony flats with acacias, mimo- sas and cacti, 8 miles e of Querétaro, Aug. 23, 1961, Waterfall 16521 (BM, F, GH, HAL, MICH, OKLA, UC, US); TLAXCALA: Mt. Malinche, Tecarona, 8500 ft, June 21, 1938, Balls 4861 (uc). The extreme with irregularly toothed leaves is repre- sented by the type collection of P. chenopodifolia (P), while the type of P. puberula has leaves with entire margins. 19b. var. chenopodifolia, forma immaculata Waterfall, f. nov. Corolla immaculata vel subimmaculata. MEXICO: TLAXCALA: sunny banks between cultivation, Mt. Ma- linche, June 21, 1938, Balls 4861 (Type: US). 19c. var. viridis Waterfall, var. nov. Planta paucivestita, laminis plus minusve viridibus, integerrimis vel paucidentatis, 30-50 mm longis et 15-27 mm latis. MEXICO: DURANGO, Otinapa, July 25-Aug. 5, 1906, Palmer 420 (Type: US; Isotypes: F, GH, NY, UC). SELECTED COLLECTIONS. MEXICO: CHIHUAHUA: limestone hill- side 32 miles s of Parral, Aug. 9, 1956, Waterfall 12519 (F, GH, HAL, MICH, OKLA, P, S, SMU, US); COAHUILA: oak and juniper assn. on a mountain 24 km nw of Fraile, July 15, 1941, Stanford, et al 370 (Gu, NY); DURANGO: Palmer 420, cited under "Type"; grassland with scattered pines, 19 miles sw of Durango, Aug. 10, 1957, Waterfall 13434 (F, GH, MICH, NY, OKL, OKLA, SMU, UC, US); TAMAULIPAS: in arroyo 19 km sw of Miquihuana, Aug. 11, 1941, Stanford et al 918 (GH, NY); in pine forest 3 mi n of Miquihuana, July 14, 1949, Stan- ford, et al 2438 (US). This species varies into its northern extreme with vesti- ture less abundant, but present in varying quantities, leaves usually entire, their blades somewhat smaller than in var. chenopodifolia, 19d. var. exigua Waterfall, var. nov. Planta parvior, laminis parvis, inaequaliter paucidentatis vel in- tegerrimis. MEXICO: SAN LUIS POTOSI: Charcas, July-Aug., 1934, Lundell 5398 (Type: MICH; Isotype: US). SELECTED SPECIMENS. MEXICO: DURANGO: open pine-oak woods 5.5 mi w of Durango, Aug. 12, 1957, Waterfall 13617 (OKLA); GUANAJUATO: subhumid oak forest on steep slopes, 14.5 mi from Guan- ajuato on road to Dolores Hidalgo, July 5, 1955, Johnston 2626 (OKLA); valley near Tula, July 16, 1896, Pringle 7334 (GH, US, VT); QUERÉTARO: cliff sides and top, 15 mi se of San Juan del Rio, Aug. 17, 1957, Waterfall 13962 (BM, F, GH, MICH, OKLA, P, SMU); SAN LUIS POTOSÍ: region of San Luis Potosí, in 1878, Parry & Palmer 642 (GH, NY, P, US); Lundell 5398, Type, cited above. This variety is characterized by the small leaf blades, 1967] Physalis — Waterfall 111 the principal ones 10-25 mm long and 5-15 mm wide, with margins coarsely and irregularly few-toothed, or, rarely, subhastately lobed, on petioles 3-10 mm long. In kind of vestiture it tends toward the coarser trichomes of P. orizabae. 19e. var. glandulosa Waterfall, var. nov. Planta viscida, trichomatibus articulatis capitato-glandulosis vel attenuatis. MEXICO: QUERÉTARO: cliffs on steep mountainsides, 22 miles ne of Zimapan, Aug. 20, 1957, Waterfall 14170 (Type: OKLA) ; Isotypes (MICH, US). 20. Physalis hastatula Waterfall, sp. nov. Planta herbacea perennis; caulibus 9-10 dm longis, procumbentibus vel suberectis et ramosis, subglabris vel trichomatibus paucis brevibus antrorsis; foliis lanceolatis vel angusto-lanceolatis, subhastatis vel integerrimis, principalibus 4-6 cm longis et 3-4 mm latis, petiolis 12- 22 mm longis; calycibus floriferis 4-6 mm longis et 3-4 mm latis ad basim loborum; calycis lobis lanceolatis, 3-4 mm longis; pedicellis 4-7 mm longis; corollis maculatis, 8-10 mm longis et 3-4 mm latis; antheris violaceis, 1.5-2 mm longis, filamentis 2.5-4 mm longis; calyci- bus fructiferis decagonatis subtilibus, 17-23 mm longis et 12-15 mm latis, baccis 7-10 mm latis. SPECIMENS SEEN. MEXICO: JALISCO: prostrate, vine-like, a few plants in shade of north-facing rocks, Cerro La Campana, near km 36 sw of Ojuelos on road to Aguascalientes, Aug. 9-12, 1958, Mc- Vaugh 16778 (Type: MICH, Isotype: OKLA); bushy, near km 31 sw of Ojuelos on road to Aguascalientes, Aug. 13, 1958, McVaugh. 16889 (MICH, OKLA). This species is characterized by being nearly glabrous, having narrow leaves with a large subbasal tooth or small lobe on each side, sometimes 2 teeth, sometimes entire, and by the 10-angled fruiting calyx with the primary angles more prominent than the secondary ones. 21. Physalis gracilis Miers, Annals and Magazine of Natural His- tory, ser. 2, 4 : 37. 1849; P. Schiedeana Dunal in DC., Prodromus 13 (1): 452. 1852. Herbaceous, 60-120 cm tall, erect or reclining on bushes to pro- cumbent and, sometimes, repent, more or less vestite with flat, jointed, usually spreading hairs ca. 1-1.5 mm long, sometimes glabrate or nearly glabrous; leaf blades ovate, sometimes broadly so, principal ones 3-10 em long and 2-8 em wide, surfaces spreading-hairy, or ap- pressed-hairy or glabrous, entire to irregularly few-toothed, petioles 2-6 cm long; flowering calyx usually spreading-hairy, sometimes 112 Rhodora [Vol. 69 glabrous, 5-10 mm long! and 4-7 mm wide at base of ‘calyx lobes which tend to be ovate-deltoid and are 2-3 mm long; flowering pedi- cels 7-15(-20) mm long; corolla maculate, 10-13 mm long and 13-16 mm wide when open-reflexed, hairy pads below the five spots; an- thers yellowish, sometimes with a slight bluish tinge which may be more pronounced in fresh material, 3-4 mm long on filaments 2-5 mm long; fruiting calyx 10-ribbed to nearly terete, 22-30 mm long and 14-25 wide on pedicels 10-20 mm long; berry nearly spheric, 8-15 mm in diameter, sometimes slightly stipitate on the invaginated calyx base. SELECTED COLLECTIONS. MEXICO: cAMPECHE: Yohaltun, Jan. 4, 1901, Goldman 538 (US): moist soil along river, Champoton, July 7-15, 1932, Steere 1767 (MICH); DISTRITO FEDERAL: Canada de Contreras, June 1, 1963, Galvan s.n. (OKLA); HIDALGO: Real del Monte, Aug. 1840, Galeotti 1199 (BR, P); MORELOS: perennial, climb- ing among shrubs to 12 ft, wet banks, barranca near Cuernavaca, June 1895, Pringle 6319 (BM, BR, F, GH, NY, UC, US, VT); NAYARIT: Sierra Madre near Santa Teresa, Aug. 10, 1897, Rose s.m. (US); OAXACA: Zacuapan, Nov.-April, 1840, Galeotti 1190 (BR); PUEBLA: near Tehuacan, Dec. 22, 1895, Pringle 7034, (GH, VT); SAN LUIS POTOSÍ: Tamazunchale, Nov. 26, 1937, Kenoyer 826 (F); 2 km nw of Tampacan, Mar. 27, 1959, Rzedowski 10268 (OKLA); TABASCO: along Rio Grijalva n of Villahermosa, Sept. 24, 1944, Gilly et al 330 (MICH); in second growth, La Palma, Balancan, June 1-6, 1939, Matuda 3312 (F, GH, MICH, NY); VERACRUZ: Maltrata, May 12, 1937, Matuda 1359 (GH, MICH, US); stems creeping, barranca of Texolo, 3000 ft, April 30, 1899, Pringle 8142 (BM, BR, F, NY, P, UC, US, VT). BRITISH HONDURAS: at edge of forest, Mountain Pine Ridge, San Agustin, July-Aug. 1936, Lundell 6836 (GH, MICH, NY, OKLA, us). COSTA RICA: Los Diamontes, Dec. 2, 1953, Heiser 3781 (F). EL SALVADOR: vicinity of San Salvador, Jan. 4, 1922, Standley 19142 (GH, NY, US); procumbent, moist thicket, vicinity of Metapan, Jan. 29-Feb. 1, 1947, Standley & Padilla 3130 (F). GUATEMALA: weedy stream side, Quezaltenango, Oct. 8, 1934, Skutch 1399 (GH); Esquintla, Donnell-Smith 2089 (US); procumbent in damp thickets 43 km ne of Coban, 1200 m, Standley 70076 (F, US) ; prostrate in semi- chaparral, across river from Sayaxche, May 5, 1942, Steyermark 46312 (F, MICH, NY, US) ; Sierra de los Cuchumatanes, Huehuetenango, July 24, 1942, Steyermark 49464 (F, US). HONDURAS: vicinity of San Marcos de Colon, Jan. 12-22, 1949, Standley 15763 (F): elongate herb on shrubs in wet thicket, near Yuscaran, Dec. 11, 1946, Standley et al 1157 (F); weed along road, Montanya Santa Barbara near Lake Yojoa, Aug. 7, 1948, Williams & Molina 14476 (F, GH); along rail- road near Ceiba, July 14, 1938, Y uncker, et al 8414 (BM, F, GH, NY, US). NICARAGUA: path in rain forest, Jinotega, Cerro la Florida, Dec. 5, 1958, Hawkes, et al 2166 (F). P. Schiedeana Dunal is typified by (HAL) by a slightly 1967] Physalis — Waterfall 113 more hairy phase than is the type of P. gracilis. The spe- cies varies from this to nearly glabrous material. P. Galeottiana Walpers, Repert. Bot. Syst. Lipsiae 574. 1842-1848, nom. nov, for P. glabra Mart. & Gal., Acad. Roy. des Sci. ... Belgique, Bull. 12: 132-133. 1848, may be con- specific. The Type, H. Galeotti 1197 (BR), Cordillera, Zacua- pan, Vera Cruz, June-Oct. 1840, consists of two short vegetative branches, one with two “pairs” of leaves, and one with a single leaf. It is possible that they represent the glabrous extreme of P. gracilis, in which case this name has priority for the species sens. lat. However, it hardly seems desirable to base a species name on the slight evi- dence presented by inadequate vegetative material, so, un- less a flowering and/or fruiting isotype should be found, and shown to belong to this species, it is proposed that the name P. gracilis Miers, based on adequate type material, be retained. The nearly glabrous phase may resemble P. philadel- phica, but may be separated from it by the longer pedicels of P. gracilis. 22. Physalis luteoanthera Waterfall, sp. nov. Herbacea, ad 1 m alta; caulibus tenuiter erectis vel procumbenti- bus; pilis longis, articulatis; foliis ovatis integerrimis vel undulato- dentatis, principalibus 6-10 em longis et 4-6 cm latis; petiolis 15-45 mm longis; calycibus floriferis 7-10 mm longis et 6-10 mm latis ad basim loborum; calycis lobis late ovatis, 2-3 mm longis; corollis luteis, maculatis, 12-13 mm longis et 17-23 mm latis; antheris luteis, 3-4 mm longis; filamentis 2-3 mm longis; calycibus fructiferis ca. 25 mm longis et 19-22 mm latis; pedicellis fructiferis 10-13 mm longis; baccis 8-10 mm latis. TYPE: Steyermark 33726 (F), shaded thickets, lower south-facing slopes of Volcan Santa Maria between Finca Pirineos and Los Positos, Dept. of Quezaltenango, Guatemala. COLLECTIONS SEEN. EL SALVADOR: vicinity of Ahuachapan in wet thicket near stream, Jan. 16-25, 1947, Standley & Padilla 2774 (F). GUATEMALA: Steyermark 33726, Jan. 8, 1940, the Type, referred to above. 23. Physalis peruviana L., Species Plantarum ed. 2: 1670. 1762; P. latifolia Lamarck, Tableau Encyclopédique et Méthodique . . . Bot. 2: 29. 1793; P. peruviana L., var. latifolia (Lam.) Dunal in DC., Prodromus 13(1): 440. 1852. 114 Rhodora [Vol. 69 Perennial, 0.3-1 m tall, tomentose to villous, hairs jointed, often diverging at right angles to the stem; leaf blades ovate, 5-10 cm long and 4-7 em wide, their apices often somewhat lunate-attenuate, their bases often more or less cordate, margins entire, or with few irregu- lar teeth; leaf surfaces with jointed hairs, somewhat appressed, more abundant on lower surfaces and along the principal veins; petioles 1-4 cm long; flowering calyx oblong to oblong-campanulate, 8-9 mm long and 4-6 mm wide at base of the triangular to ovate-attenuate calyx lobes into which the upper half of the calyx is divided; flower- ing pedicels 6-8 mm long; corolla 10-14 mm long and 12-15 mm wide, maculate; anthers bluish, 3.5-4 mm long, on filaments 2-4 mm long; fruiting calyx densely soft-hairy, 10-ribbed or slightly 10-angled, 3-4 cm long and 2.5-3 em wide; berry 12-20 mm long and 10-15 mm wide on the invaginated (3-5 mm) calyx base. SELECTED COLLECTIONS. BERMUDA: Dec. 1915, Collins 442 (GH, NY); herbaceous shrub, Berry Hill Road, Paget, Sept. 12, 1963, Manuel 281 (GH). HAITI: shrub or woody herb, Oriana Road, Holdridge 824 (NY). JAMAICA: New Castle, Sept. 9, 1908, Britton 3311 (NY); above Neweastle, alt. 4000 ft, Feb. 4, 1956, Stearn 191 (GH). 24. Physalis angustior Waterfall, sp. nov. Frutex vel herbaceo-frutex, 1-2 m altus; pilis densis, articulatis, partim glanduliferis; foliis ovatis, apice lunato-attenuatis, principali- bus 6-10 em longis et 4-8 cm latis; petiolis 3-5 em longis; floribus singulis vel 2-3-aggregatis; calycis lobis lanceolato-attenuatis, 5-9 mm longis; pedicellis 12-20 mm longis; corollis luteis, maculatis, 15- 20 mm longis et 20-29 mm latis; antheris violaceis, 3-4 mm longis; calycibus fructiferis 25-30 mm longis et 15-20 mm latis; pedicellis fructiferis 15-30 mm longis; baccis 14-16 mm longis et 12-15 mm latis. TYPE: Pringle 7731, on cliffs, Sierra de Tepoxtlan, 7500 ft, Mar. 14, 1889, (VT); Isotype: (US). COLLECTIONS SEEN. MEXICO: MORELOS: 7731; shaded ledges, Sierra de Tepoxtlan, May 5, 1900, Pringle 8275 (F, GH, MEXP, NY, UC, US, VT); on cliffs of Sierra de Tepoxtlan, 7500 ft, Nov. 20, 1902, Pringle 11056 (F, GH, US, NY). Physalis angustior differs from P. peruviana in its ag- gregations of flowers, its smaller, narrower fruiting calyx and its more woody habit, 25. Physalis Sancti-josephi Dunal in DC., Prodromus 13(1): 451. 1852. Herbaceous, probably coarse and tall, collected tops 40-50 em long, covered with varying densities of jointed hairs of varying lengths, the longer 1-1.5 mm long, many of them tipped with small reddish- brown glands; leaf blades ovate, slightly attenuate apically, margins usually with l-several irregularly shaped and variable-sized teeth or 1967] Physalis — Waterfall 115 undulations on each side, sometimes entire; principal blades 6-10 cm long and 4-8 cm wide, on petioles 3-8 cm long; flowering calyces 6-10 mm long and 6-9 mm wide at base of lobes, upper 2.3-5 mm divided into lanceolate or ovate-lanceolate, somewhat acuminate lobes; flower- ing pedicels 5-10 mm long; corolla yellowish, maculate, 10-20 mm long and 15-22 mm wide when fully expanded; anthers bluish, 2-3 mm long, on filiform filaments 3-5 mm long; fruiting calyx with 5 prom- inent angles and 5 intermediate angles or ribs, 25-35 mm long and 18-25 mm wide, prominently reticulate, the principal veins usually dark-colored; fruiting pedicels 8-15 mm long; berry 8-15 mm in diameter. COLLECTIONS EXAMINED. MEXICO: SAN LUIS POTOSI: Al- varez, July 15-23, 1904, Palmer 220 (F, GH, MICH, NY, UC, US) ; herb in oak grove, 2200-2400 m, Sierra de Alvarez, July 30-31, 1934, Pennell 17841 (NY, US); orilla de camino, km 33 de la carretera San Luis Potosí-Río Verde, July 29, 1954, Rzedowski 3574 (MEXP); LO- CALITY UNDETERMINED: pr. San Jose del Oro, reg. frig., June 1831, Schiede sin num (Type: HAL). 26. Physalis Muelleri Waterfall, sp. nov. Herba perennis 25-45 cm alta; trichomatibus brevibus, ad 0.5 mm longis, partim capitato-glanduliferis; foliis ovatis, inaequaliter mag- nodentatis, principalibus 4-8 cm longis et 2.8-5.5 cm latis; petiolis 15-35 mm longis; calycibus floriferis 6-8 mm longis et ca. 5 mm latis ad basim loborum; calycibus lobis ovato-deltoideis vel lanceolato- attenuatis 2-4 mm longis; pedicellis 5-6 mm longis; corollis luteis, maculatis, 10-12 mm longis et 12-15 mm latis; antheris luteis 3.5-4 mm longis, filamentis 3-5 mm longis; calycibus fructiferis decacostatis, trichomatibus brevibus, partim capitato-glanduliferis; calycibus fructiferis 17-22 mm longis et 15-18 mm latis; pedicellis fructiferis 10-15. mm longis; baccis 10-14 mm latis. TYPE: Diente Canyon, mountains near Monterrey, Nuevo León, July 6, 1933, Mueller 129 (Type: GH, Isotype: F). This species, known only from the type locality, is char- acterized by its large yellow anthers, dark-maculate corolla, 10-angled or 10-ribbed fruiting calyx covered with short spreading hairs, many of them capitate-glandular, open apically, and nearly filled by the berry; also by the ovate leaves, coarsely and irregularly salient-dentate, with much of the vestiture capitate-glandular. 27. Physalis sordida Fernald, Proc. Amer. Acad. Arts & Sci. 35: 568. 1900. Stems herbaceous, 10-50 cm tall from a woody base growing from a horizontal rootstock 3-5 cm below the surface of the earth; herbage 116 Rhodora [Vol. 69 covered with usually brownish, flat, jointed, spreading hairs, up to ca. 1 mm long, some glandular-capitate; leaf blades suborbicular to ovate or lanceolate-ovate, the principal ones 3-7 cm long and 2.5-6 cm wide with vestiture similar to the stem, but less abundant and, sometimes, more appressed, margins coarsely and irregularly salient- dentate to repand-dentate to subentire, on pedicels 1.5-5 em long; corolla yellowish, with 5 darker, but not strikingly contrasting, maculations, 10-13 mm long and 10-15 mm wide when fully rotate or rotate-reflexed, nearly pentagonal in outline, with shallow sinuses; anthers yellow, or tinged with blue or green, 3-4 mm long on some- what thickened filaments of about equal length; fruiting calyx 10- angled or 10-ribbed, 15-30 mm long and 13-20 mm wide; berry spheric to ovoid, 10-15 mm in diameter, sessile, or essentially so, on the in- vaginated calyx base. SELECTED COLLECTIONS. MEXICO: DISTRITO FEDERAL: hill- side near Guadalupe, 7400 ft, July 25, 1904, Pringle 8976 (F, GH, MEXP, NY, UC, VT); DURANGO: Inde, 2000 m. Aug. 1927, Reko 5272 (US); GUANAJUATO: Aug. 1899, Lugis 417 (US); HIDALGO: valley near Tula, 6800 ft, Sept. 22, 1901, Pringle 9526 (F, GH, MEXP, NY, US, VT); MEXICO: near Guadelupe, Sept. 1903, Rose and Painter 6841 (NY, US); OAXACA: Boca de Leon, Telixtlahuaca, 7500 ft, Smith 637 (Type: GH); Mt. Alban, near Oaxaca, Aug. 7, 1949, White 225 (OKLA, US) ; PUEBLA: vicinity of San Luis Tultitlanapa, July 17, 1908, Purpus 3580 (BM, F, GH, NY, UC, US); SAN LUIS POTOSÍ: in montibus San Miguelito in 1876, Schaffner 699 (GH); TAMAULIPAS: in narrow, deep, moist arroyo in hills 19 km se of Miquihuana on road to Palmillas, Aug. 11, 1941, Stanford et al 839 (GH, NY); ZACATECAS: south slope of La Bufa, Aug. 10, 1948, Dressler 101 (GH). HONDURAS: Sel- guapa, Dept. Comayagua, Mar. 22, 1945, Rodriguez 2551 (F). The collection from Honduras is somewhat dubiously referred to this concept, yet it seems closer to it than to any other. In the northern part of its range this species approaches P. hederaefolia var. puberula, but here its bluish-tinged anthers tend to distinguish it. 28. Physalis Pennellii Waterfall, sp. nov. Herbacea, perennis, 5-20 cm alta; trichomatibus articulatis, longis et brevibus ad 1.5 mm longis, partim glanduliferis; foliis ovatis vel lato-ovatis vel subrotundis ad basim, principalibus 13-32(-40) mm longis et 12-25 (-30) mm latis, inaequaliter paucidentatis vel undulatis vel integerrimis; petiolis 5-11(-20) mm longis; calycibus floriferis hemiglobosis, dense longipilosis, 4-5 mm longis et 4-5 mm latis ad basim loborum; calycis lobis deltoideis vel ovato-deltoideis, 2-3 mm longis; pedicellis 3-6 mm longis; corollis luteis, maculatis, 7-13 mm longis et 10-14 mm latis; antheris violaceis, ca. 2.5 mm longis; caly- 1967] Physalis — Waterfall 117 cibus fructiferis 10-costatis, 15-20 mm longis et 11-13 mm latis; pedicellis 9-15 mm longis; baccis 8-12 mm latis. TYPE: Pennell 17523 (US); rocky limestone west of Santa Ana, above Potrero, Sierra de Catorce, July 24, 1934; Isotype: (NY). SPECIMENS EXAMINED: MEXICO: GUANAJUATO: Aug. 1947, Kenoyer 2476 (GH); SAN LUIS POTOSÍ: Charcas, July-Aug. 1934, Lun- dell 5399 (MICH, US); Pennell 17523 cited under "Type". Physalis Pennellii is most nearly related to P. sordida Fern., but is a more dwarf plant with a mixture of longer hairs on the herbage, and with a more densely long-hairy flowering calyx. 29. Physalis latecorollata Waterfall, sp. nov. Herba perennis ca. 30 cm alta; caulibus villosis, pilis articulatis, ad 2-3 mm longis; foliis ovalibus vel ovatis, integerrimis, principali- bus 50-60 mm latis et 38-42 mm latis, petiolis 15-25 mm longis; caly- cibus floriferis 8-9 mm longis et 7-8 mm latis; pedicellis 5-7 mm longis; corollis maculatis, 33 mm latis; antheris violaceis, 3 mm longis; filamentis filiformibus, 5-7 mm longis; calycibus fructiferis ignotis. TYPE: Garcia sin num (MEXP) MEXICO: oaxaca: orilla de ar- royo, campamento Rio Molino cerca de San Miguel Suchistepec, Oaxaca, alt. 2350 m, Sept. 20, 1965, 30. Physalis philippensis Fernald, Proc. Amer. Acad, Arts & Sci. 35: 568. 1900. Stems 15-20 cm long, several-branched, herbaceous from a woody root; vestiture of flattened, spreading, jointed hairs of varying sizes, the longer stem-hairs 2-3 mm long; leaf blades ovate to rhombic- ovate to hastate-ovate, 15-20 mm long and 15-20 mm wide, usually with 1-3 irregularly shaped teeth, spreading-hairy to appressed-hairy on both sides, on petioles 6-12 mm long; calyx at anthesis hemis- pheric or more widely spreading, 5-8 mm long and 8-10 mm wide at base of the ovate-lanceolate to deltoid calyx lobes, 3-4 mm long; pedicels 5-8 mm long; corolla 12-15 mm long and 20-25 mm wide, the limb with 5 prominent violet or purplish markings, the broad, flaring tube tomentose within above the point of stamen-insertion; corolla slightly lobed, the sinuses 3-5 mm deep; anthers bluish or violet, ob- long, 3-4 mm long, on slender, glabrous, violet or bluish filaments 4-6 mm long; fruit unknown. SPECIMENS EXAMINED. MEXICO: oAxAca: Sierra de San Filipe, 9000 ft, June 1, 1894, Pringle 5621 (Type: GH, Isotype: VT). 31. Physalis caudella Standley, Field Mus. Publ. Bot. 17: 273. 1937. Perennial, apparently from a deep, not-collected rhizome, indument 118 Rhodora [Vol. 69 of varying amounts of long jointed hairs (1-) 1.5-3 mm long, or of long and short hairs mixed in varying proportions; principal leaf blades 4-8 cm long and (6-) 15-40 mm wide, ovate-lanceolate to linear- lanceolate, on petioles 5-20 mm long; margins of blades entire to irregularly undulate to saliently few-toothed; flowering calyx 6-12 mm long and 4-6 mm wide at base of lobes, the upper 4-7 mm divided into more or less elongate lobes; flowering pedicels 5-10 mm long; corolla prominently reddish-blue or purplish maculate, 11-18 mm long; anthers bluish, reddish-blue or blue-green, ca. 3 mm long; fruit- ing calyx (2-) 2.5-5 em long and (2-) 2.5-3 em wide, divided above into lobes (6-) 10-15 (-17) mm long, its pedicel ca. 1 em long. 31a. var. caudella Leaf blades ovate to oblanceolate-ovate or lanceolate-falcate, often basally inequalateral, principal ones 4-8 em long and 13-40 mm wide; fruiting calyx 3-5 em long and 2.5-3 em wide. SPECIMENS EXAMINED. MEXICO: CHIHUAHUA: pine-oak slope, Cajurichi, Río Mayo, Sept. 13, 1936, Gentry 2710 (Type: F, Isotypes: GH, US); Mojarachic, July 28, 1940, Knobloch 8013 (Us); near San Juanito, July 26, 1937, 7800 ft, Shreve 8023 (F); Salto de Babicora, July 18, 1937, LeSueur 1448 (F). 31b. var. parva Waterfall, var. nov. Planta nana, foliis angustis, principalibus 4-6 cm longis et 6-12 (-15) mm latis; calycibus fructiferis 18-20 mm longis et 10-15 mm latis. TYPE: Townsend and Barber 122 (F), Isotypes: (BM, GH, NY, OKLA, P, UC, US, VT), MEXICO near Colonia Garcia in the Sierra Madres, 7200 ft, July 11, 1899. Representatives of this taxon bear a striking resem- blance to the type of P. virginiana var. polyphylla (Greene) Waterfall from Piedra in southern Colorado. However, the latter has yellow anthers, moderately darkened areas on the corolla, and vestiture consisting wholly of short, an- trorse hairs, whereas the newly-described taxon includes plants with bluish anthers, purplish maculations, and vesti- ture of spreading, jointed hairs, long and short intermixed. 32. Physalis virginiana Miller, Gardener's Dictionary, ed. 8, no. 4, 1768. Since this is a complex species, with its distribution prin- cipally to the north, in the United States, only the synonymy pertaining to the taxa studied here is given, and also only descriptions of these taxa. The nearest relationship of the Mexican varieties of P. virginiana to other Mexican taxa appears to be with P. 1967] Physalis — Waterfall 119 caudella. The following key may be used to separate the varieties of these two species. a. Anthers blue or violet; long villous hairs more or less abundantly intermixed with shorter ones. b. Fruiting calyces large, 3-5 cm long & 25-30 mm wide; leaves broad, principal ones 4-8 cm long & 13-40 mm wide. P. caudella var. caudella. b. Fruiting calyces smaller, 18-20 mm long & 10-15 mm wide; leaves narrower, principal ones 4-6 cm long & 6-15 mm wide. P. caudella var. parva. a. Anthers yellow, or with a slight bluish tinge; vestiture short and appressed, or plant glabrous, or with a few long stiff hairs. c. Vestiture short and appressed or plant glabrous. d. Plant large, usually 50-60 cm tall, single-stemmed. P. virginiana var. sonorae. d. Plant small, usually 10-30 cm tall, several-stemmed. P. virginiana var. nana. c. Vestiture of few long rigid jointed hairs, short ones usually present on sepal tips only. P. virginiana var. longiseta. 32a. Physalis virginiana Miller, var. sonorae (Torr.) Waterfall, Rhodora 60: 154. 1958; P. pumila Nutt., var. sonorae Torr., Bot. Mex. Bound. 153. 1859; P. longifolia Nutt., Trans. Am. Phil. Soc. (n.s.) 5: 193-194. 1836; P. lanceolata Michx., var. laevigata Gray, Proc. Am. Acad. Arts & Sci. 10: 68. 1875; P. lanceolata Michx., var. longifolia (Nutt.) Trel., Rep. Ark. Geol. Surv. 4: 207. 1891; P. rigida Pollard & Ball, Proc. Biol. Soc. Wash. 13: 134-135. 1900. Herbaceous perennial, 30-60 cm tall, single-stemmed, often branch- ing above, glabrous, or with short, antrorse hairs, especially on the calyces and younger parts; leaf blades lanceolate to narrowly lanceo- late, principal ones 4-7 cm long and 10-16 mm wide on petioles 15-25 mm long; corolla yellow or greenish-yellow, maculate, but not as dark and prominently so as in P. caudella; anthers yellow. COLLECTIONS EXAMINED. MEXICO: CHIHUAHUA: Oct. 1911, Stearns 14 (F); SAN LUIS POTOSÍ: in 1878 (atypical, bluish anthers) Parry & Palmer 643 (NY); lugares incultos, Zaragosa, July 7, 1954, Rzedowski 3520 (OKLA); SONORA: Santa Cruz, Oct. 22, 1893, Mearns 2620 (US); Fronteras, June 1850-2, Thurber 418 (Type: GH, Isotype: F). 82b. var. nana Waterfall, var. nov. Herba perennis; caulibus ramosis ad basim, 5-30 em longis; tricho- matibus brevibus, antrorsis, paucis vel nullis; foliis lanceolatis vel ovato-lanceolatis, principalibus 3-5 cm longis et 1-2 cm latis; petiolis 1-2 cm latis. 120 Rhodora [Vol. 69 TYPE: Waterfall 16600 (OKLA), Isotypes: (F,GH, US), grassy silty flats 38 miles se of Saltillo in NUEVO LEON, MEXICO Aug. 25, 1961, SELECTED COLLECTIONS. MEXICO: coAHUILA: grassy flat east of Puerto de Aire at southern end of Sierra de la Encantada, Sept. 1, 1941, Stewart 1316 (GH); NUEVO LEÓN: wet depression in calcareous desert 16 miles n of Matehuala, Aug. 21, 1959, Waterfall 15753 (F, MICH, OKLA, SMU) ; Waterfall 1660, Type listed above; SAN LUIS POTOSI: alkaline soil in desert near Km 549, 36 miles s of Mate- huala Sept. 27, 1958, McVaugh 18209 (MICH); calcareous desert, 10 miles s of intersection of Highways 80 & 57, 74 miles north of San Luis Potosi, Aug. 20, 1959, Waterfall 15715 (OKLA, SMU) ; SINALOA: vicinity of Mazatlan, Apr. 6, 1910, Rose et al 14095 (NY, US); ZACATECAS: on road from Concepcion del Oro south of Cardono, 7 miles n of San Tiburcio, Sept. 2-3, 1938, Johnston, I. M. 7360 (GH). One of the above collections, McVaugh 18209, exhibits a vertical, elongate, hardened, corm-like rootstock, 25 mm long and 6 mm wide, dug from 11-13 cm below the surface of the earth. Other collections could have been broken away from above such a structure. 32c. var. longiseta Waterfall, var. nov. Planta 15-45 cm alta; caulibus longisetis, trichomatibus paucis; foliis lanceolatis vel linearo-lanceolatis, principalibus 5-8 cm longis et 1-2 em latis; petiolis 3-10 mm longis; calycibus fructiferis 2-3 em longis et 15-25 mm latis; pedicellis 1-2 cm longis, saepe incrassatis in parte superiore. TYPE: Palmer 53 (US), Isotypes: (F, GH, NY, UC), Río Verde, San Luis Potosí, June 2-8, 1904. COLLECTIONS SEEN. MEXICO: SAN LUIS POTOSI: Type, cited above; region of San Luis Potosí in 1887, Parry & Palmer 647 (GH, NY, US). TO BE CONTINUED Volume 69, No. 777 including pages 1-120, was issued March 31, 1967. EARLOW REFERENCE LIBRARY > UL 11 1967 Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL j STUART KIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER pem Editors ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. - Vol. 69 April-June, 1967 No. 778 CONTENTS: Taxonomic and Genetic Relationships of Eastern North American Species of Elymus with Setaceous Glumes Sp Oca NIE Cle: aooccatidid reins M NE 121 Floral Anatomy of Rhexia virginica (Melastomataceae) Richard H. Eyde and James A. Teeri eee 163 The Persistence of the Double-Flowered Form of Celandine Poppy. Richard A. Howard sessen 179 Palynological Notes on American Species of Helianthemum (Cistaceae). Robert L. Wilbur and James D. Perry ........ 184 A Spurless Form of Aquilegia canadensis L. Hobsrt-B, Living aren isse c 195 (Continued on Inside Cover) Che New England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. CONTENTS: — continued A New Status for an Eastern North American Scirpus Alfred E. Sehuyler eese eee eene enne 198 Physalis in Mexico, Central America and the West Indies U. T. Waterfall ./|............... eeeeeeeee eee eee eene enn ntn nnne nnns 203 Erratum MMMMKMMMMMMMMuu———-—-—-——a-A——————— 240 RHODORA.—A quarterly journal of botany, devoted primarily to the flora of North America and floristically related areas. Price, $6.00 per year, net, postpaid. in funds payable at par in United States currency in Boston; single copies (if available) $1.80. Back vol. umes 1-58, with a few incomplete, can be supplied at $5.00 per volume, Volume 59— available at $6.00. Somewhat reduced rates for complete sets can be obtained upon application. Scientific papers and notes. relating directly or indirectly to the plants of North America, will be considered by the editorial com- mittee for publication. Articles concerned with systematic botany and cytotaxonomy in their broader implications are equally accept- able. All manuscripts should be double-spaced throughout, Please conform to the style of recent issues of the journal. Illustrations can be used only if the cost of engraver's blocks is met through the author or his institution. Forms may be closed five weeks in advance of publication. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Albion R. Hodgdon, Dept. of Botany, Nesmith Hall. University of New Hampshire, Durham, New Hampshire 03824 Subscriptions and orders for back issues (making all remittances payable to RHODORA) should be sent to RHODORA, Botanical Museum, Oxford Street, Cambridge, Mass. 02138. Second Class Postage Paid at Boston, Mass. Manufactured by THE LEXINGTON PRESS, INC. Lexington. Mass, QTRbooora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 69 April-June No. 778 TAXONOMIC AND GENETIC RELATIONSHIPS OF EASTERN NORTH AMERICAN SPECIES OF ELYMUS WITH SETACEOUS GLUMES' GEORGE L. CHURCH In the grass genus Elymus, the origin of populations that exhibit varying degrees of reduction in glume structure is considered to have occurred frequently through natural hybridity between Elymus Hystrix L. (Hystrix patula Moench) and other sympatric species. Although spikes of E. Hystrix usually are recognized by the absence of glumes or the presence of reduced stubs, individuals showing occa- sional bristle-like glumes are not uncommon (fig. 17). More- over, Great Lakes populations may be encountered that consist of spikes with the usual divergent spikelets and also the unusual development of long, filamentous or setaceous pairs of glumes at each rachis node (fig. 25 center). Such populations might have arisen through natural crosses be- tween typical glumeless E. Hystrix and some other species with well developed glumes. These studies were greatly facilitated by generous support from the Section for Systematic Biology of the National Science Founda- tion — grants GB-180 and G-9046. Loans from the herbaria at the University of Minnesota, Michigan, North Dakota and Wisconsin have been much appreciated in addition to courtesies extended by the directors of the herbaria at Harvard University, the New York Botanical Garden, United States National Museum, University of North Carolina at Chapel Hill, the Canadian Department of Agri- cuture at Ottawa and the University of Montreal. For assistance in the breeding program, I wish to thank Bruce W. Fowles, David J. McLaughlin, and Amanda K. Sherrill. 121 122 Rhodora [Vol. 69 Growing sympatrically with these variable populations are some well-established species, previously unrecognized in the Great Lakes area; i.e., E. diversiglumis Scribn. & Ball and E. Wiegandii Fern. Both of these species may have narrow or setaceous glumes and together with the sympatric E. canadensis L. var. canadensis they have been regarded as being involved in natural hybrid species complexes with E. Hystrix. The first part of this treatise deals with taxonomic re- assessments of these species. The second part is a report of experiments in interspecific hybridization that have been useful in interpreting the degree of genetic barriers exist- ing between the species concerned, as well as in indicating patterns of natural hybridity by which new taxa might originate. A second paper records the currently active process of speciation involving E. Hystrix and components of E. virginicus L. in the Pine Hills area of southern Illinois. All the known species of eastern American Elymus are tetraploid (2N— 28) and apparently differ only slightly in genomic structure. It has been demonstrated in previous studies that such usually diverse and widespread species as E. canadensis var. canadensis and E. virginicus var. vir- ginicus will hybridize freely even though the F, is sterile (Church 1954, 1959). Partly fertile populations, interme- diate between these species, have been reported to have arisen through introgressive hybridization in Iowa (Pohl 1959) and Texas (Brown and Pratt 1960). Still further evidence of genetic fluidity in the Hordeae tribe, of which Elymus is a member, is presented in the many cases of intergeneric hybrids, both artificial and natural; i.e, Elymus and Hordeum (Bowden 1958, Pohl 1966), Elymus and Sitanion (Stebbins et al 1954), Elymus and Agropyron (Stebbins 1956 and Dewey 1966), Agropy- ron and Hordeum (Gross 1960), and Agropyron and Sitanion (Boyle 1963 and Dewey 1964). As far as is known to the author, the only report of a probable natural hybrid involving Hystrix is that with Agropyron reported by Dore (1950). There is an adequate basis, then, for suspecting that E. Hystrix (Hystrix patula) has played a role in either 1967] Elymus — Church 123 the interspecific or intergeneric hybrid origin of new taxa — a hypothesis that provides a large part of the framework for the present studies. TAXONOMIC RELATIONSHIPS 1. Elymus diversiglumis Scribn. & Ball In their study of Elymus in Minnesota, Booher and Tryon (1948) suggested the hypothesis that populations in the area of "E. interruptus" with reduced glumes might well be a product of hybridity of E. canadensis var. canadensis and Hystrix patula. In a later paper deading with Elymus (Church 1954), preliminary studies indicated that the commonly recognized E. interruptus of the Great Lakes region of North America is a very different taxon from E. interruptus Buckley of Texas (1862), contrary to treatments in current manuals (Hitchcock and Chase 1950, Fernald 1950, Stevens 1950, Gleason 1952, Frazer and Russell 1954, Scoggan 1957). Wiegand (1918) in his study of eastern North American Elymus, however, clearly understood E. diversiglumis Scribn. & Ball as the species with subulate, unequal glumes in the northern Great Plains. Rydberg (1922) likewise recognized E. diversiglumis in his Flora of the Rocky Mountains and. Adjacent Plains. Bowden (1964), in con- ference with the present author, also recognizes E. diver- siglumis in his extensive treatment of Elymus in Canada. It will be demonstrated here that the misunderstood “inter- ruptus complexes" of the central United States and adjacent Canada undoubtedly are composed of E. diversiglumis and of occasional introgressants of this species with E. cana- densis var. canadensis. The original description of E. diversiglumis published by Scribner and Ball (1901) is clear in stating that the “empty glumes are subulate — varying from a mere point to 1.5 cm in the same spikelet." The unequal pairs of subulate glumes that are so characteristic of specimens from central Minnesota (figs. 2 left, 5) are readily matched by those in the TYPE specimens from Crook County in northeastern Wyoming. Elymus diversiglumis is particularly abundant along clearings and road-cuts in the oak-aspen forests of 124 Rhodora [Vol. 69 the Itasca Park region of central Minnesota where it flourishes in company with E. canadensis var. canadensis. It may be distinguished in late July by well-seeded, deeply arched and silvery-hirsute spikes that have matured two to four weeks in advance of those of var. canadensis (figs. 9 vs. 11). In searching for additional distinguishing features for E. diversiglumis, it has been noted that specimens from any part of the range indicated in the list of citations always possess a palea with an obtuse apex, usually augmented by a median tuft of hairs — quite in contrast to the acute, denticulate or retuse palea apices of E. canadensis var. cana- densis, as indicated in the key to species. The two taxa are similar in exhibiting 6 or 7 nodes and leaves 6-10 mm in width. With very few exceptions, however, the distinct pilosity of the upper surface of diversiglumis leaves separ- ates them from those of var. canadensis. When populations are encountered with the glumes falling below 0.8 mm in width and the striations not continuing clearly to the base, hybridity between the two taxa may be suspected. The fact that diversiglumis and var. canadensis may hybridize to give rise to sterile F, offspring that in turn may produce fertile backcrosses to var. canadensis has been demonstrated by the author (Church 1959) and is more fully documented below. Such fertile introgressants and sterile, integrading forms appear occasionally, as might be expected, in disturbed areas. Populations selected for study were distributed in wood- land areas adjacent to, or within a few miles of, the follow- ing lakes in central Minnesota: — McCraney, Itasca, Gull, Long, Round and Mille Lacs. Along the disturbed aspen and oak woodland of the La Salle Trail and new Route 92 through Itasca Park in Hubbard County, for example, abundant stands of E. diversiglumis are established. Most specimens appear like those in figs. 5, 9 and 2 (left) which breed true from seeds grown in the greenhouse. An infre- quent sample (2340)? may give rise to forms with wider *Citations in the text refer to the author's specimens on file in the Brown University Herbarium. 1967] Elymus — Church 125 glumes approaching those of E. canadensis var. canadensis figs. 6, 11) with which it is a common associate. Still other forms (2327) have paleas as long as those in var. canadensis (9.5 mm) and narrow but striated glumes equal in length (22 mm). This intermediate form is quite sterile. Even some typical diversiglumis forms (2328) may be at least 50% sterile. On the other hand, on a stretch of gravel road- side west from Itasca Park to McCraney Lake (Route 113), sympatrie populations of E. diversiglumis, E. canadensis var. canadensis and E. Hystrix were remarkably distinct. However, on an extensively cut Tilia woodland near the shore of McCraney Lake, an unusually tall, vigorous and aggressive stand of E. diversiglumis (2461) was encount- ered. Although it possessed the usual spikelet characters of E. diversiglumis (cf. key), its wide leaves and strongly drooping spikes simulated these features of E. Wiegandii. Seed grown descendants maintain the same vigor in outdoor plots in Providence. Again, a disturbed terrain has facilitated the spread of apparent hybrid forms (2514) of E. diversiglumis in the extensively bulldozed woodland roads adjacent to Gull Lake in Crow Wing County. The plants have glumes approaching those of var. canadensis in appearance, high percentages of shrivelled pollen and low seed set. In spite of the above instances of apparent interspecific hybridity in Minnesota, on an isolated bluff, EIk Mound (Dunn County) in western Wisconsin, the undisturbed woodland glades harbor both taxa, morphologically distinct, in closely adjacent sites. Further support for the integrity of E. diversiglumis has been presented by Frére Jean-Paul Bernard (of the Institu- tion des Sourds-Muets, Montreal) who has made careful studies of Elymus in the Otterburne region of Manitoba, and to whom the author is greatly indebted for sending an unpublished manuscript (1958). Frére Bernard tested the hypothesis that E. diversiglumis might be a shade adapted form of E. canadensis var. canadensis by studying trans- plants of the former species to open, exposed habitats. No changes in morphological features were noted. Likewise, 126 Rhodora [Vol. 69 the present author, in the course of genetical studies, has observed only increased vigor of growth of typical repre- sentatives of E. diversiglumis transferred to open field plots. The importance of field observations is further em- phasized by Frére Bernard who states, "Sur le terrain, la pubescence des lemmas parait blanche soyeuse à premiére vue chez lE. diversiglumis, tandis qu'elle est plutôt terne chez PE. canadensis (toutefois, cette distinction s'efface plus ou moins durant le séchage de spécimens)." Bernard also notes the aggressiveness of this species: "l'auteur a observé des touffes de cette espéce à glumes sétacées crois- sant en bordure d'un chemin traversant un secteur déboisé depuis quelques années." In recognition of the appropriateness of the Scribner and Ball designation for the taxon with unequal glumes found in the Great Lakes region and with scattered stations westward to Wyoming and Saskatchewan, the following modification and expansion of the original description is presented : Elymus diversiglumis Scribner & Ball (U.S.D.A. Bull. Agrost. 24: 48-49. Fig. 22. 1901.). TYPE: T. A. Williams 2653 (us!), Wyoming, Crook County, Bear Lodge Mountains, rich, open woods at 6000 feet. CULMS robust (70) 120 (160) cm high, glabrous; LEAVES 5-15 mm wide, usually pilose above (rarely, slightly scabrous), sheaths over- lapping except in the tallest specimens; SPIKES flexuous and broadly to deeply arched at maturity, (8) 15 (24) em long; RACHIS flattened, margins ciliate, internodes (3) 5-6 (10) mm long; SPIKELETS usually two per node and containing 2-3 florets and frequently an additional, terminal sterile floret; GLUMES 2-15 (20) mm long, 0.1-0.5 mm wide, very unequal in length, sometimes lacking, setaceous from an in- durated base; LEMMAS (7) 9 (12) mm long, silvery hirsute, rarely hirtellous, the awn 2-3 times the length of the body and widely divergent at maturity; PALEA (8) 8.5 (9) mm long, slightly shorter than the lemma, the apex obtuse, usually with a small, median, spatulate projection covered with a tuft of hairs denser than those of the body; STAMENS (2) 2.5 (3.5) mm long. Range: borders of mixed woodlands, clearings and disturbed areas, including river terraces; frequent in Minnesota with scattered sta- tions in Wisconsin, Iowa, Ontario, Manitoba, Saskatchewan, North Dakota and eastern Wyoming. SPECIMENS EXAMINED: SASKATCHEWAN: Saskatoon, Macoun & Her- riot 42962 (CU, SASK), Fraser 12-77 (SASK) ; Assiniboia, Oxbow, Dore 1967] Elymus — Church 127 & Boivin 13324 (DAO); Montagne a l'Original, Boivin 8362 (DAO). MANITOBA: Turtle Mountain, Burgess 2229 (DAO); Beauséjour, Dore 12800 (DAO); Indian Reserve, Prairie Meadow, A. & D. Love 6229 (DAO, MT); Riding Mountain Park, Rowe 416, Halliday 32-157 (CAN, DAO), Lovaas 61-27 (DAO); Otterburne, Bernard 50-132, 54-572, 58- 441, 536, 568, 607 (MT); Kleefeld, Bernard 56-5422, 58-617 (MT); Stuartburn, Bernard 58-568 (MT.); Clear Lake, \Halliday 157-1932 (CAN). ONTARIO: Rainy River District, Dore 9118 (CAN, DAO); Thunder Bay District, Lindsay 1251 (DAO). WYOMING: Crook County, Williams 2653 TYPE, 2679 (US), NORTH DAKOTA: Dunn County, Swallen 5799 (us), Stevens & Moir (NDA); Bottineau County, Brannon 85150 (us), Stevens 2512 (NDA); Pembina County, Bergman (MIN), Stockbridge 974 (NDA); Cavalier County, Stevens (NDA). MINNESOTA: Beltrami County, 275174 (MIN); Clearwater County, 277388 (MIN), Moore & Huff 18986 (DAO-MIN), Church 2310- 16, 2435 (BRU); Polk County, 64722, 64724 (MIN) ; Mahnomen County, Church 24,61 (BRU); Hubbard County, Grant 2914 (IND, MO, MIN, NY), 2714 (MIN, US), 275776 (MIN), Church 2328, 2840-47 (BRU); Becker County, Grant 3901 (GH, MIN), Church 2354-56, 2470, 2480 (BRU); Ottertail County, 371975 (MIN); Crow Wing County, 64727 (MIN), Church 2360, 2506, 2512 (BRU) ; Stearns County, 64723 (MIN); Hennepin County, Wheeler 1224 (GH, MIN, US); Scott County, 64726 (MIN). IOWA: Emmet County, Wolden 530 (18C). WISCONSIN: Bay- field County, Weber (wis); Chippewa County, Ebert (wis); Dunn County, Myer 132 (wis), Church 2570 (BRU); Juneau County, Tru- man (WIS); Vilas County, Stearns 570 (wis); Polk County, Baird (WIS) ; Burnett County, Shinners 4471 (WIS). SPECIMENS INDICATING PROBABLE INTROGRESSION OF E. diversiglumis INTO E. canadensis var. canadensis: SOUTH DAKOTA: Lawrence County — Martin 477 (US) — This specimen from Spearfish Canyon, roughly 25 miles from the TYPE locality in adjacent Wyoming, has rather thin, slightly pilose leaves and very narrow, setaceous glumes as in E. diversiglumis. The glumes are quite equal in length, however, and the 10.5 mm long paleas have acute, truncate or slightly concave apices — all of which features typify canadensis var. canadensis. WYOMING: Crook County — Whereas the TYPE specimens, Williams 2653 (US), from the Bear Lodge Mountains are easily matched by those from Welcome, Williams 2679 (vs), it is noteworthy that Williams 2681 from the same “rich gulches" is nearer to E. canadensis var. canadensis in the bidentate apices of the paleas and the glabrous leaves. The collections of Griffiths 925 and 942 (Us) from Hulett are likewise regarded as forms of var. canadensis with setaceous glumes. Elymus occidentalis Scribn. (1898) is a name cited by Hitchcock and Chase (1950) as a synonym for what is treated here as E. diver- siglumis. Examination of the TYPE specimen, Nelson 4470 (US) from 128 Rhodora [Vol. 69 Albany County, Wyoming, reveals simply a dry river terrace form of E. canadensis var. canadensis. 2. Elymus Wiegandii Fern. The description of E. Wiegandii published by Fernald (1933) makes quite clear the distinction between this spe- cies and E. canadensis var. canadensis. Hitchcock and Chase (1950), however, regard this species as one of many vari- ants of E. canadensis var. canadensis. Bowden (1964) re- duces E. Wiegandii to a variety of E. canadensis. On the other hand, Voss (unpublished) recognizes the specific dis- tinctness of E. Wiegandii as it occurs sparingly in Michigan. Throughout its range, along shaded streams or alluvial river terraces, E. Wiegandii is readily identified by its lax spikes, hanging pendant from the time of emergence from the uppermost of the glaucous, usually overlapping leaf-sheaths (fig. 10). The culms are vigorous, bearing 9-12 leaves, 10-22 mm wide, dark green and frequently pilose on the upper surface. The glumes are usually less than 1 mm wide and the bases abruptly narrowed, indurated but not terete. Probably because of the fact that the glumes are frequently less than 0.5 mm wide, especially in populations in the Great Lakes area, the species has often been confused with E. diversiglumis or lumped under so-called E. interruptus. Elymus Wiegandii may be distinguished at once from E. diversiglumis (cf. key) by the pendant, rather than deeply arched spikes (fig. 13) (a feature often obscured in mounted specimens), the usually wider and always greater number of leaves, the usually glabrous and overlapping leaf sheaths and the bidentate (rather than obtuse) dense-hairy apices of the paleas. The ranges of the two species overlap geographically in the broadest limits of the Great Lakes area, but E. diversiglumis is a plant of more open and drier terrain than the shaded, wet thickets of E. Wiegandii. In the same areas of Minnesota that yielded so many stations for E. diversiglumis as well as E. canadensis var. canadensis and E. Hystrix (Itasca Park and Mille Lacs), several newly reported sympatric and ecologically distinct 1967] Elymus — Church 129 stands of E. Wiegandii are noteworthy as indicated in the following list: — LAKE ITASCA: drainage in wet, coniferous woods, west of Park Road, 2305; north boundary of Itasca Park, wet, mixed hemlock woods — 2438. TULABY LAKE: drainage, mixed woods near Route 113 — 2493. MILLE LACS LAKE: north of Garrison, shore road through mixed woodlands, wet, shady forest— 2517 (whole plant glaucous); north of Garrison, sandy, wet clearing under oaks — 2522; south of Vineland, shore road, mixed woodland — 2293-94, 2297-99, 2302. ROUND LAKE: north of Brainerd, edges of sandy, wet, dense, mixed woods — 2500, 2502. RICE LAKE: Lum Park, brook-dissected, oak-birch woods, near shore — 2499 (whole plant glaucous). Likewise in New England and eastern Canada, the shaded, alluvial habitat of E. Wiegandii is in contrast to the more open and drier woodland-border habitat charac- teristic of E. canadensis var. canadensis. Pilosity of the adaxial leaf surfaces is a frequent but not too constant fea- ture of E. Wiegandii in New England and Quebec as well as the midwest (Fassett 1951). Two populations collected by the author in closely adjacent stations in southwestern Wisconsin (Castle Rock Woods), illustrate the differences between superficially similar representatives of the two species, Both 2580 (E. Wiegandii) and 2579 (E. canadensis var. canadensis) are robust plants (ca. 120 cm tall) with pilose upper leaf surfaces, mostly overlapping sheaths, and paleas 10-12 mm long with bidenticulate apices. E. Wie- gandii, however, has 10-12 nodes, leaves about 18-20 mm wide, sheaths glaucous, spikes pendant and glaucous, glumes ca. 17 mm long and 0.5 mm wide. This population (2580) grows on the edge of a shady, rich woodland resting on an abandoned alluvial terrace! In contrast, 2579 (E. cana- densis var. canadensis) has 7-9 nodes, leaves 9-12 mm wide, the margins moderately inrolled, sheaths green, glumes 21-30 mm long and 0.8-1.0 mm wide. Although the habitat is within ca. 25 feet of a shady, alluvial thicket, the plants are confined to an open, dry roadside. Furthermore, the author has observed many other stations of E. Wiegandii 130 Rhodora [Vol. 69 in the field and invariably the habitat is a rich, alluvial soil in the shade. Although some forms of E. canadensis var. canadensis may be found on the moist lake shores of Lake Michigan, the contrasting habitat factors to those of E. Wiegandii are sand and full sunlight. On the whole, E. Wiegandii appears to be less capable of standing competi- tion on open areas than either E. canadensis var. canadensis or E. diversiglumis and under greenhouse conditions it becomes quickly reduced in size much more than any other species of Elymus of this study. The greatest concentration in the range of E. Wiegandii lies along the tributaries of the Connecticut and St. Law- rence Rivers, spreading out across New York and northern Pennsylvania and extending northward across northern New England to the Canadian Maritimes. More recently, several stations have been reported in the drainage systems of the Rainy, Mississippi, and Red Rivers in Ontario, Michi- gan, Wisconsin, and Minnesota with scattered stations westward in Iowa, the Dakotas, Wyoming and Manitoba. The most recently reported collections are from the banks of the rivers draining into James Bay on the Ontario side (Dutilly & Lepage 1963). The following citations of specimens examined by the author indicate a much wider geographic range for E. Wie- gandii than was known to Fernald in 1933 : — MANITOBA: Assiniboine, Krivda (DAO); Sans Souci, Bernard 499, 5477 (MT); Otterburne, Bernard 504, 525 (MT). ONTARIO: Rainy tiver County, Dore 9175 (BRU, DAO); Thunder Bay County, Garton 2063 (us), Williamson 2239 (PENN); Carleton County, Rolland 31 (us), Dore 10, 541, (BRU, DAO); Cochran County, Dutilly & Lepage 36350, 36809, 38397 (UCU). QUEBEC: Champlain County, Gagnon (DAO); Huntington County, Dore & Cody (DAO), Bassett & Frankton (DAO); Pontiac County, Dore 9028 (BRU, DAO); Brome County, Dore 6 (GH); Vaudreuil County, Ray 28420 (MT); Gatineau County, Dore 15241 (DAO, MT), Dore 5 (GH); St. Maurice County, Stanislas 142 (MT) ; Stanislas County, Marie-Victorin 7331 (MT); Richelieu County, Marie-Victorin 6367 (MT); Portneuf County, Marie-Victorin 45064 (GH, MT); Rimouski County, Lepage 3260 (GH); Gaspe County, Dore 47879 (US); Deux Montagne County, Louis-Marie 2910 (us); Lac St. Jean County. Kennedy 3 (GH); Terrebonne County, Rolland-Ger- 1967] Elymus — Church 131 main 6061 (MT), Rouleau 2528 (MT), Gray Ex. 1405 (GH, ISC), Church 2528, 2532 (BRU); Bonaventure County, Collins 5116 (GH), Collins et. al., (GH, NY), Rousseau 32528 (Gu), Marie-Victorin 32350, 33763, 48988 (DAO, GH), Church 2533-36 (BRU), Raymond 56069 (DAO); Sherbrook County, Dore 345 (DAO); Drummond County, Ver- ret (DAO), NOVA SCOTIA: Pictou County, Smith 11677 (DAO, GH), Erskine 53430 (DAO), NEW BRUNSWICK: Victoria County, Marie-Vic- torin 46549 (GH, MT); Kings County, Svenson & Fassett 3012 (GH); Madawaska County, Dore 47951 (DAO). MAINE: Aroostock County, Fernald 197 TYPE (GH), Church 2537 (BRU); Piscatiquis County, Parlin 1793 (GH), Church 2538, 2540 (BRU); Kennebec County, Fer- nald & Long 12748 (GH); Oxford County, Parlin 2038 (GH) ; Somerset County, Fernald & Long 490 (US). NEW HAMPSHIRE: Coos County, Pease 10685 (GH), Church 2376 (BRU); Grafton County, Church 2226, 2231 (BRU); Sullivan County, Cowden (NEBC). VERMONT: Orleans County, Church 2408 (BRU); Caledonia County, Congdon (GH); Ben- nington County, Church 2210-12, 2217-19 (BRU); Rutland County, Church 2263-64 (BRU), Carpenter 3 (NERC, US), Kirk 1036 (us); Windham County 2214-16 (BRU), Wheeler (Gu); Chittenden County, Church 2250 (BRU), James (NY). MASSACHUSETTS: Hampden County, Church 2526 (BRU); Hampshire County, Church 2200-04 (BRU); Franklin County, Church 2207 (BRU). CONNECTICUT: New Haven County, Blewitt 35 (GH). NEW YORK: St. Lawrence County, Phelps 134 (GH, NY); Washington County, Dobbin & Burnham (GH); Essex County, House 25976 (GH); Columbia County, McVaugh 2087 (PENN); Monroe County, Hitchcock 10024 (US); Schoharie County, Whitney 1844 (TEX); Tompkins County, Wiegand 9295 (GH, NY, US); Madison County, House 2860 (GH); Chemung County, Axtell & Smith 2687 (cv, US); Onondago County, Lippert 295 (SYR); Rens- selaer County, House 24015 (SYR); Oneida County, 1789 (SYR). PENNSYLVANIA: Warren County, Wahl 9573 (GH, PENN); McKean County, Fogg et. al. 19781 (PENN); Susquehanna County, Glowenka 9399 (PENN); Wayne County, Gress & Honesdale (PENN); Wyoming County, Glowenka 9281 (PENN). MICHIGAN: Houghton County, Far- well 11007 (MICH); Chippewa County, Dodge (MICH); Oakland County, Farwell 1556a (MICH). WISCONSIN:Lincoln County, Seymour 13129-30, 14464 (WIS), Jwen 339 (Wis), Church 2560 (BRU); Polk County, Brenner (Wis); Chippewa County, Davis (wis); Eau Claire County, Kunz 356 (WIS), Fassett 22286-7 (wis); Dane County, Cheney (Wis), Hale (wis); Vernon County, litis & Koeppen 11922 (Wis); Richland County, litis & Koeppen 11841 (wis); Grant County, Fassett 28971-2, 26493 (wis), Church 2580, 2589 (BRU), Iltis et. al. 13847 (WIS). MINNESOTA: Kittson County, Moore 11604 (Isc, MIN); Cook County, Butters & Abbe 1024 (GH, MIN); Beltrami County, Pamel 72, 329 (ISC); Clearwater County, Church 2305 (BRU) ; Carlton County, Moore & Putters 3472 (GH, MIN); Becker County, 132 Rhodora [Vol. 69 2438, 2193 (BRU); Crow Wing County, Church 2499, 2517, 2522 (BRU); Aitkin County, Church 2500 (BRU). 10WA: Dickinson County, Thorne 13223 (Isc); Emmet County, Wolden 566 (Isc). NORTH DAKOTA: Bottineau County, Stevens (NDA); Ward County, Crosby (NDA); Benson County, Hitchcock 5064 (US), Lunell 1020620 (us); Pembina County, Stevens (US); Barnes County, Perrin 1346 (US), Stevens & Moir (us); Ransom County, Stevens (US); Cass County, Stevens (NDA), Stevens 114 (DAO, NDA, US, WIS), Bergman & Stevens (NDA), Stevens 1677 (GH); Sioux County, Stevens 1559 (NDA, US). SOUTH DAKOTA: Roberts County, Ball & Over 2224 (1sc), Johnson 10 (GH, TEX). WYOMING: Crook County, Ownbey 2377 (MIN, US), Grif- fiths 925, 942 (us). 3. Elymus canadensis L. a. var. canadensis In central Minnesota, sympatric populations of var. cana- densis and E. diversiglumis may occasionally give rise to variants with narrow glumes (ca. one half mm in width) that represent introgressants from the latter species. It has been noted further that E. Wiegandii, characterized by nar- row giumes (cf. key), remains distinct from sympatric var. canadensis. In both cases, var. canadensis is distin- guished by the stiff, glabrous leaves with involute margins, the erect or broadly arched spikes and the one mm wide glumes. b. var. brachystachys (Scribn. & Ball) Farwell Plants of this variety with compact spikes and glabrous lemmas were raised from seed collected in Llano County, Texas. They differed from the TYPE material (Palmer 420 US!) in possessing glumes that were frequently approxi- mately only one half mm in width. Since Scribner and Ball (1901) in their original description of this taxon had cited collections from Oklahoma to as far north as Michigan, it was thought by the author that Elymus populations with narrow glumes arising from hybridity with var. brachy- stachys might presently be encountered in the Great Lakes area. The search, particularly at the stations reported by Farwell (1920), proved to be unrewarding. However, since earlier collections of var. brachystachys from Texas had segregated for several variations in reduction of glume width (Church 1954), it appears that a process of specia- 1967] Elymus — Church 133 tion leading to varieties with narrow glumes might be occurring in the far South, involving a component of E. canadensis, but in an obviously distinctly different pattern from that which produces introgressants of diversiglumis into var. canadensis in Minnesota. c. var. interruptus (Buckley) comb. nov. Based on E. interruptus Buckley, Acad. Nat. Sci. Phila. Proc. 1862:99. Llano County, Texas. Since the name of this taxon has been misapplied to very different northern species with setaceous glumes, a revision of Buckley's original description is presented. CULMS, slender, tufted 50-80 (100) cm high; LEAVES 15-20 cm long, 5-8 mm wide, scabrous and often glaucous; SHEATHS not exceeding nodes, margins usually ciliolate; SPIKE erect, (6) 8-15 em long; RACHIS glabrous except for scabridulous margins; INTERVALS between spikelets on rachis (6) 8-14 mm long; GLUMES (15) 20 (30) mm long, equal, glabrous or scabridulous, base indurated and often terete, usually less than 0.5 mm wide at the middle of the body; SPIKELETS of 2, 3 or 4 florets with an additional, terminal, sterile floret; LEM- MAS 8-9 mm long, glabrous and the awns 15-20 mm long, straight at maturity, margins scabridulous; PALEAS (7) 8 (9) mm long with the apices narrow, truncate and ciliolate. Rocky, open woodlands in Llano County, Texas. The TYPE specimens collected by Buckley (PH!) consist of two culms about 60 cm high. The glumes are 20-22 mm long, 0.4-0.5 mm wide and the indurated bases have 2-3 nerves. The ISOTYPE (GH!) has even narrower, i.e., setaceous, glumes. Additional specimens from Llano County: Wolf 3123 (us), Smith (us 1019698), Church and Brown 1041 (BRU), Reverchon (US 64974). Although var. interruptus appears to be restricted to Llano County, it is sympatric with var. brachystachys with which it shares a short (ca. 60 cm) stature and a time of floret maturity two months in advance of northern varieties growing in the same field plots in Providence. The two varieties are undoubtedly closely related, but they are dis- tinguishable on morphological grounds (cf. figs. 3, 4, 7, 8). On the other hand, examination of figs. 9, 15, 1, and 2 will reveal the striking differences between var. interruptus and diversiglumis and clear the confusion that has existed be- tween these taxa. 134 Rhodora [Vol. 69 4. Elymus Svensonii sp. nov. In searching records for further information concerning the range of both E. diversiglumis and E. canadensis var. interruptus, the case of the latter variety reported from an isolated station in Tennessee was noteworthy (Svenson 1941). Examination of the specimen on file in the Brooklyn Botanie Garden Herbarium revealed a plant quite distinct from either of the above taxa. Through the kindness of the collector, Dr. Henry K. Svenson, the author was able to find the exact area on the steep banks of the Cumberland River in Tennessee, the only station known in the state or elsewhere. The plants were scattered rather thickly in the edge of a high, limestone cliff woodland over a distance of not more than fifty feet. Although house-lots were laid in contact with the summit, the plants were spread down the river embankment, rather out of reach. Since the population is unique in taxonomic features, it seems appropriate that it be described as a new species and named in honor of the original collector, Dr. Svenson. Elymus Svensonii sp. nov. Planta tota glauca; 80-100 em alta; culmo basi 2 mm lato; laminis 6-8 mm latis at 15-25 longis, supra minusve villosis, vaginis plerumque non excedentibus, auriculis rubidis —- siccis fusco-brunneis; spicis erectis, arcuantibus, 12-16 cm longis, rachi applanato, sinuosissimo et segmentis racheos 8-12 mm longis; spiculis geminatis, in 3-4 floribus constantibus; glumis setosis, variatim 1-18 mm longis, plerumque mediostriatis, basi induratis subteretibus; geminis saepe in longitudine inaequalibus aut omnino deficientibus; lemmatibus 9-10 mm longis et aristis deinde valde curvatis 12-20 mm longis; palea 8-9 mm longa, apice lata, truncata aut parum convexa et margine breviter ciliata. Entire plant glaucous; CULMS 80-100 em high and the bases 2 mm in diameter; LEAVES 6-8 in number, 6-10 mm wide and 15-25 em long, usually slightly villous above — the margins smooth or scabrellous; LEAF SHEATHS usually not overlapping and the auricles dark-red be- coming reddish-brown on drying; SPIKE at first erect becoming broadly arched, 12-16 cm long; RACHIS, flattened, conspicuously sinu- ous, internodes 8-12 mm long, margins hirtellous; SPIKELETS, 2 at each node, composed of 4-5 florets and closely adpressed to the curves of the rachis internodes, florets not readily disarticulating from the pedicel joints; GLUMES, setiform usually with a median striation, indurated and nearly terete at the base, usually in unequal pairs 1-18 mm long, occasionally lacking; LEMMA 9-10 mm long, the widely 1967] Elymus — Church 135 divergent awn 12-20 mm long; PALEA 8-9 mm long, apex broad, truncate or slightly concave, margins villous; ANTHERS ca. 5 mm long; LODICULES asymmetrically bi-lobed, the larger lobe acute with margin ciliolate. Known only from the TYPE locality: steep limestone bluffs on the Cumberland River at the end of an abandoned road off McGavock Pike and ca. 0.9 mile from the intersection with Lebanon Pike, just north of Donelson, Davidson Co., Tennessee, HOLOTYPE: Church 2527 (BRU) (figs. 12, 1 and 2). ISOTYPES: (GH, US). PARATYPES: Svenson 9452 (BKL, MO). Elymus Svensonii bears some resemblance to E. Hystrix, but differs from the latter species in exhibiting variable glumes, consistently arched rather than erect spikes, closely adpressed rather than divergent spikelets, usually a greater number of florets per spikelet, divergent rather than straight lemma awns (figs. 12, 1, 2). Specimens in the United States National Herbarium col- lected by Gattinger in 1885 and labeled “‘precipices on the Cumberland Riv. near Nashville" (us 1021371) could, of course, have come from a station quite similar to that of E. Svensoni. Asprella Hystrix Willd. has been added to the label at some later date. The plants are not too well preserved, and probably represent E. Hystrix with vestigial glumes similar to another Gattinger collection (US 1021372) designated with no locality other than Nashville. 5. Elymus Hystrix L. (Sp. Pl. 560. 1753, Virginia. Clayton.) There is no more distinctive species in moist, shady wooded areas of the eastern United States than what has for a long time been recognized as Hystrix patula Moench. The widely spreading spikelets, falling quickly at maturity and the usual lack of glumes certainly appear to constitute sufficient bases for generic designation (fig. 17). Yet mor- phological variations of the common form are not unknown. For example, populations with pubescent lemmas have been designated as H. patula var. Bigeloviana (Fern.) Deam. Equally common are forms that exhibit occasional filiform glumes. Most striking of all are populations in which long, filiform glumes are found all along a rather sinuous rachis 136 Rhodora [Vol. 69 of rather closely compacted nodes (fig. 25). The latter variations are known throughout the range but are particu- larly frequent in the Carolinas as is indicated in the cita- tions that follow. SPECIMENS EXAMINED: NEW MEXICO: Colfax County, Silveus 4928 (TEX). OKLAHOMA: Adair County, Wallis 7606 (TEX). ARKANSAS; Petit Jean Mts, Scully 974 (Uus); Baxter County, Demaree 29337 (TEX); Randolph County, Demaree 29190 (vs); Sharp County, Demaree 26890A (US). MISSOURI: Jasper County, Palmer 3781 (MO). ILLINOIS: Knox County, Chase 1840 (US); Peoria County, Chase 12053 (US). WISCONSIN: Burnett County, Shinners 4475 (WIS); Calumet County, Strandberg 376 (wis); Dane County, Cottam 184 (wis); Grant County, Iltis 13789 (wis); Green County, Struik (wis), Fell 584,38 (wis); Lafayette County, Coogan 149 (WIs) ; Rock County, Fell 57938 (wis); Sauk County, Stowell (wis), Church 2576 (BRU); Vernon County, litis 6449 (wis), Marks (WIS). INDIANA: Cass County, Deam 25895 (US) ; Sullivan County, Deam 25683 (US). OHIO: Coshocton County, Moldencke 13124 (US). NOVA SCOTIA: Pictou County, St. John 1387 (GH). MARYLAND: Prince Georges County, Maxon 5988 (GH). WASHINGTON, D. C.: High Island, Pollard 446 (us). VIRGINIA: Bedford County, Curtiss (GH); Culpepper County, Allard 1787 (GH); Greensville County, Fernald & Long 8029 (GH). NORTH CAROLINA: Alamance County, Ramseur 2277 (DUKE), Bell 4149 (DUKE); Anson County, Radford 13146 (NCU); Davidson County, Radford 12828, 7069 (NCU); Durham County, Bloomquist 14526 (UNC), 793, 797 (DUKE); Godfrey & Fox 51163 (DUKE, UNC); Wilbur 6193 (DUKE, GH, US); Granville County, Batson 334 (NCU), Fox 4912 (DUKE); Haywood County, Silveus & Bloomquist 9765 (TEX); Lee County, Beard 1072 (NCU); Macon County, Quaterman 1442 (DUKE) ; Orange County, Bloomquist et. al. 402 (GH), Radford & Stewart 681a, 212a (NCU); Person County, Bell 14652 (NCU); Rockingham County, Radford 13639 (DUKE); Rowan County, Horton 1407 (NCU); Stanley County, Ahles 16222 (NcU); Union County, Ahles 27513 (NCU); Wake County, Boyce 1620 (NCU), Godfrey 48127 (GH), Brown 51207 (TEX). SOUTH CAROLINA: Abbeville County, Radford 26037; Chesterfield County, Radford 12397 (DUKE, NCU); Edgefield County, Radford 26361 (NCU); Kershaw County, Radford 23669 (NCU); Lancaster County, Ahles 31356 (NCU); McCormick County, Radford 22327 (NCU); Newberry County, Bell 9235 (NcU); Saluda County, Radford 23123 (NCU). The author has examined in the Linnean Herbarium in London the Savage indexed sheet 100.8 upon which the name Elymus Hystrix is based. Three specimens are rather 1967] Elymus — Church 137 young since the glabrous spikes are barely out of the ter- minal leaf sheaths. Leaves and sheaths are dark green and slightly pilose. Glumes are entirely lacking. 'The spike rachis is thin and straight with ciliolate margins and the internodal distances are ca. 5 mm. The straight awns are ca. 214 times the length of the 10 mm long lemmas. Paleas are 9 mm long and the apices are acute and slightly notched. Except for the fact that the Linnean specimens are imma- ture and do not exhibit the spreading spikelets, they are quite similar to those selected over the wide eastern United States range. The many variations listed above, however, indicate rather strongly that they must be the product of some pattern of hybridity with other species of Hlymus, perhaps arising pandemically in the eastern United States. The experimental studies reported later will present a further basis for evaluating the phylogenetic position of E. Hystrix in the relationships of Elymus species. It will be demonstrated that employment of the origina] Linnean name (Bowden 1964) is quite justifiable on genetic as well as on morphological grounds. Key to Eastern North American Elymus* 1. Pairs of glumes mostly unequal in length, often setaceous, re- duced to pointed stubs or lacking (if paleas 9-12 mm long and apices bidentate, cf. E. Wiegandi) 2. Spikes erect at maturity; spikelets usually widely divergent from the rachis and readily falling when dry; glumes lacking or occasionally present as filiform bristles at some or all nodes; paleas 8 (9-10) mm long, apices obtuse or truncate, often pilose on the Upper margin. ............... ERI UE E. Hystrix. 2. Spikes slightly to broadly arched at maturity; spikelets more or less adpressed to the rachis; glumes setaceous, 0.1 mm-0.5 mm wide, tapering from an indurated base. 3. Lemmas densely hirsute; paleas (7.5) 8 (9) mm long, apices obtuse with median tuft of hairs; upper leaf surface usually pilose; entire plant green otherwise. ............ E. diversiglumis. 3. Lemmas glabrous; paleas 8-9 mm long, apices narrow and truncate or retuse, cilolate; entire plant glaucous. ................ VON D "EEUU I ^e E. Svensonii. 1. Pairs of glumes essentially equal in length. 4. Spikes abruptly pendant on emergence from the leaf sheaths, the latter usually glaucous and overlapping; leaf blades com- monly pilose above, 12-28 mm wide; glumes (0.2) 0.5 (1.2) mm 138 Rhodora [Vol. 69 wide, often setaceous with base narrowed and indurated; paleas (9) 11 (12) mm long, apices bidentate. ............ E. Wiegandii. 4. Spikes erect or becoming slightly to broadly arched at maturity. 5. Glume bases clearly indurated. 6. Glumes 0.8-2 mm wide, swollen on half or all of the adaxial surface as well as the abaxial base, sometimes bowed and disarticulating along with the spikelets in age; paleas 6-8 (8.5) mm long, apices obtuse, truncate or slightly emargi- nate; spikes erect, ..........eeessseeeeeeeeeneee eene E. virginicus. 7. Glumes 1-2.7 em long, spike at times partly included in sheath. ........... sess eene nennen enne var, virginicus. 7. Glumes 2.7-1.0 em long, spike well exerted. ................. BEEDNMNMNNMNJUUJUlJss L| var. glabriflorus. 8. Spikelets glabrous. ......... eee f. glabriflorus. 8. Spikelets hirsute. ............ eene f. australis. 6. Glumes 0.2-1.0 mm wide, indurated for 1-3 mm adaxially, striated middle area more or less convex. 9. Paleas 5.5-6.5 mm long, apices obtuse; rachis internodes 1.5-3.0 mm long, whole spike villous, lemmas rarely glabrescent. .......sseeeeeeesseee eene e nennen E. villosus. 9. Paleas (7) 8.5 (9) mm long. 10. Apices bidentate; rachis internodes (3) 4-5 (7) mm long; glumes scabrous to hispidulous. ............... sagucssssgesuepecsncveseeeseascaccoseossssooosoesccadacaconasasascascasseass E. riparius. 10. Apices truncate or slightly retuse; rachis internodes (6) 8 (9) mm long; glumes 0.2-0.5 (0.6) mm wide; lemmas glabrous; leaf-sheath margins usually ciliate. ———————— — E. canadensis var. interruptus. 5. Glume bases thin and striated or indurated for only 1.0 mm; body of glume not thickened adaxially. li. Spikes (7) 10-20 (30) em long broadly arched at maturity; spike internodes (4) 5 (7) mm long; paleas (8.5) 9-12 (14) mm long, apex very acute and usually bidentate; lemmas hirsute or hirtellous; glumes (0.7)-1.0 mm wide; maturity in July and August. ............. eee ceeueeeeeeeeeecsseeeecseeeeeceseceuceuceuceaeeus E. canadensis var. canadensis. 12. Plants green. ........eesseeeeenh eene f. canadensis. 12. Plants glaucous. ..........eeeeeeeeeeneenneeen f. glaucifolius. 1i. Spikes (7) 10-12 (15) cm long, essentially straight at maturity; spike internodes (2) 4 (5) mm long; paleas (7) 9-10 mm long, apex acute and truncate or bidenticulate; lemmas glabrous or scabrous; glumes 0.5-1.0 mm wide; maturity in late May and June. ............. m :"—— ———— E. canadensis var. brachystachys. *Leaf measurements are from the third leaf beneath the spike. Spikelet measurements are taken from the mid-spike area. Ex- tremes of measurements are indicated in parentheses. Plate 1342. Fig. 1. (left to right) E. greater magnification. diversiglumis 2476; E. E. canadensis var. interruptus 1041. Fig. 2. Same Scale shown in millimeters Svenson 2 as fig. 1 , Plate 1343. Fig. 3. E. canadensis var. interruptus 1041. Fig. 4. E. canadensis var. brachystachys 1028. Fig. 5. E. diversiglumis 2328. Fig. 6. E. canadensis var. canadensis 2300. Seale shown in millimeters Plate 1344. rm Fig. 7. E. canadensis var. interruptus 1041. Fig. 8. E. canadensis var. brachystachys 1028. Scale shown in millimeters Plate 1345. Fig. 9. E. diversiglumis 2508. Fig. 10. E. Wiegandii 2522. Fig. 11. E. canadensis var. canadensi 2300. Fig. 12. E. Svensonii 2521. Plate 1346. Fig. 13. E. Wiegandii 2522. Fig. 14. E. virginicus var. glabriflorus f. glabriflorus 2660. Fig. 15. E. canadensis var. interruptus 1041. Fig. 16. (left to right) E. virginicus var. virginicus 2557; E. canaden- sis var. canadensis 2554. Plate 1347. 2300 ^ 2300 x 2463 21,63 Q . Fig. 17. (left to right) E. canadensis var. canadensis 2300 (pistil- late); experimental F,; E. Hystrix 2463 (staminate). Fig. 18. Back- cross of 2300 (E. canadensis var. canadensis) staminate to experi- mental hybrid 2300 (E. canadensis var. canadensis) X 2328 (E. di- versiglumis). Scale shown in millimeters Plate 1348. Fig. 19. Experimental cross of 2368 (E. Hystrix) X 2361 (E. diversiglumis.) Fig. 20. Another region of the same spike as fig. 19. Scale shown in millimeters Plate 1349. M EHE M [7 2252 Fig. 21. E. riparius 2367. Fig. 22. Experimental cross of 2367 (E. riparius) X 2361. (E. diversiglumis). Fig. 23. E. diversiglumis 2361. Fig. 24. experimental cross of 2361 (E. diversiglumis) X 2252 (E. canadensis var. canadensis). Scale shown in millimeters à M l if jd Plate 1350. Fig. 25. E. villosus 2700; two spikes of the same population of E. Hystrix 2699. Scale shown in millimeters. Fig. 26. Pollen-mother-cell of E. diversiglumis 2461, late first meiotic anaphase showing bridges. 148 Rhodora [Vol. 69 GENETIC RELATIONSHIPS In testing the hypothesis that E. Hystrix and E. cana- densis var. canadensis might cross to give rise to hybrid populations with setaceous glumes in Minnesota (Booher and Tryon 1948), experimental crosses between representa- tive strains were made in order to learn the extent of the genetic barriers existing between them. However, since populations of Elymus with setaceous glumes may be found also in the following cases, similar experimental crosses were made with the hope of obtaining data indicative of the origin of the taxa involved: — E, canadensis var. inter- ruptus in Texas, E. Svensonii in Tennessee and E. virgini- cus var. glabriflorus in Illinois. METHODS. Since none of the strains employed was self- sterile, pistillate parents were emasculated and pollination was effected by dusting the stigmas liberally with pollen collected in cellophane bags that had covered the staminate spikes. Results of the crosses are indicated by numbers separated by colons and represent in the following order: — F, seed- lings obtained, seeds obtained from the cross, florets emasculated. Numbers in square brackets indicate original seedlings obtained in contrast to preceding number of sur- viving mature plants. In cases where no square brackets are used, all seedlings grew to mature plants. Two numbers alone indicate: — mature plants: seeds planted. Cytological data, when given, are based on the examina- tion of at least 50 pollen-mother-cells (p.m.c.’s) in the stages reported. Chiasma frequencies represent average chiasmata per bivalent chromosome at diakinesis or first meiotic meta- phase. 1. Crosses of E. Hystrix with E. canadensis var. cana- densis. On the edge of mixed woodlands in the Brainerd-Mille Lacs Lake area, fertile specimens have been collected that have the general appearance of var. canadensis but which possess glumes reduced in width or length and yet not iden- tifiable as setaceous. With the thought that such strains 1967] Elymus — Church 149 might be introgressants of E. Hystrix into var. canadensis, crosses between these taxa were attempted. Because of the difficulty in emasculating the readily deciduous florets of Hystrix, it was not often possible to use this species as the pistillate parent. The following strains were crossed, all with no success: 2300 (var. canadensis), Mille Lacs Lake, X 2368 (Hystrix), Nerstrand Woods (0:90) ; 2300 (var. canadensis) X 2306 (Hystrix), Itasca Park (0:104) ; 2669 (Hystrix), Pine Hills — Illinois, X 2300 (var. canadensis) (0:100). In contrast to results with the above strains, a vigorous F, was obtained from the cross of 2300 (var. canadensis) X 2465 (Hystrix) from McCraney Lake woods (14:14:43), even though the reciprocal cross failed (0:46). Figure 17 shows the parents and the F,. It can be noted that the latter plant has the thin, flattened rachis of Hystrix, the hirsute lemmas of var. canadensis and the setaceous, uni- striated, basally indurated glumes that are obviously inter- mediate. Cytological observation of p.m.c.’s of the F, above re- vealed chiasma frequencies of 1.80 as contrasted with 2.00 for 2300 (var. canadensis) and 1.99 for 2463 (Hystrix). Roughly 30% of both first and second meiotic divisions in the hybrid exhibited an average of 4 univalent laggards and 2 bridges and yet the young pollen was remarkably free from cytoplasmic extrusions. Nine vigorous hybrid plants flourished while surrounded by parental strains in an out- door plot for five years but continued to remain sterile. Consequently, formation of introgressants from parental backerossing to the hybrid appears to be precluded. Still further evidence of the existence of strong genetic barriers between populations of the above taxa is revealed in the failure of a cross of strains from the Pine Hills wood- land of southern Illinois: 2669 (E. Hystrix) X 2659 (var. canadensis) (0:84) and the reciprocal (0:91). 2. Crosses of E. diversiglumis with E. Hystrix A cross between 2316 (diversiglumis) from Itasca Park and 2513 (Hystrix) from Gull Lake was a failure (0:5:98) 150 Rhodora [Vol. 69 as was the reciprocal (0:6:104). A similar strain 2317 (diversiglumis) from Itasca Park crossed reciprocally with 2463 (Hystrix) from McCraney Lake likewise was a failure (0:96), (0:32). Employing 2368 (Hystrix) from Ner- strand Woods in a reciprocal cross with 2501 (diversiglu- mis) from Itasca Park still resulted in failure (0:104), (0:112). However, the use of 2361 (diversiglumis) from Gull Lake as the staminate parent in a cross with 2368 (Hystrix) produced a vigorous F, plant (1:3:222). This plant strongly resembled Hystrix with a few filiform glumes (figs. 19 and 20) and revealed a chiasma frequency of 1.85. In the examination of ca. 500 pollen-mother-cells, 10% showed univalent laggards, extrusions and bridges. Six- teen seeds produced an F, of 4 plants, entirely sterile, although still resembling the Hystrix parent. 3. Crosses of E. diversiglumis with E. canadensis var. canadensis Striking success attended a cross of a very vigorous 2461 (diversiglumis) from McCraney Lake with 2523 (var. canadensis) from Mille Lacs Lake (13[20] :20:46). Of the surviving F,, half were fertile diversiglumis and half were sterile var. canadensis. Since 2461 (diversiglumis) showed some meiotic irregularities (fig. 26), it was heterozygous, probably for genes from var. canadensis acquired through previous natural hybridity. In this case the above experi- mental cross represented a backcross, a fact that would explain the segregation of progeny in the F,. The reciprocal cross 2523 X 2461 (6:37:67) produced a fertile F, resem- bling diversiglumis but with equal glumes 0.3 mm wide and ribbed to the indurated base. Still further evidence of the heterozygosity of this same 2461 (diversiglumis) is the fact that another cross with 2300 (var. canadensis) from Mille Lacs Lake used as the pistillate parent resulted in two F, plants (2:3:66), one essentially like var. canadensis and the other a vigorous diversiglumis with intermediate glumes. However, the latter segregate yielded only 6 seeds out of 380 florets. The employment of another strain of diversiglumis, less 1967] Elymus — Church 151 vigorous but free from meiotic irregularities, demonstrated more strikingly how partially fertile strains, intermediate with var. canadensis, might arise through introgression. In a cross of 2300 (var. canadensis) Mille Lacs Lake with 2328 (diversiglumis) Itasca Park, the F, (2[3]:24:80) was defi- nitely intermediate with respect to glumes and paleas, but revealed laggards, bridges and extrusions in 30% of the 100 p.m.c.’s examined and a seed set of only 15%. The backcross made with 2300 (var. canadensis) pollen produced a vigor- ous plant (1:1:32), still possessing intermediate glumes but yielding a higher seed set of 25%. This experimental backcross closely resembles natural populations exhibiting a similar degree of intermediate morphology and partial sterility (2327, 2340, 2514) that have been collected in the Itasca Park and Round Lake areas. The use of an allopatric strain of var. canadensis from Vermont demonstrated a lower level of genetic barriers with respect to the vigorous 2361 (diversiglumis) from Itasca Park. This cross of 2361 with 2252 (var. canadensis) from Vermont (figs. 23 and 24) yielded an F, (20:32:76) that had narrow, but rarely setaceous glumes and was 40% fertile. The backcross made with 2361 (diversiglumis) pollen produced a rather weak, sterile F. 4. Cross of E. diversiglumis with E. riparius Although not frequently encountered in the Great Lakes area, E. riparius is a species with distinctly terete and in- durated bases to the glumes, a character which might con- tribute to setaceousness of the glumes of sympatric species with which it might hybridize. However, the F, (4:26:98) of the cross of 2367 (riparius) from Nerstrand Woods with 2361 (diversiglumis) from Itasca Park (figs. 21 and 22) was completely sterile. 5. Crosses of E. Wiegandii with E. canadensis var. canadensis All attempts at hybridization between strains of both E. Wiegandii and var. canadensis met consistently with failure: — 2522 (Wiegandii) Mille Lacs area with 2519 152 Rhodora [Vol. 69 (var. canadensis) Mille Lacs Lake and reciprocal (0:27), (0:26); 2522 (Wiegandii) with 2317 (var. canadensis) Itasca Park and reciprocal (0:20), (0:29); 2511 (var. canadensis) Round Lake with 2522 (Wiegandii) (0:54) ; 2550 (var. canadensis) Lake Michigan shore with 2522 (Wiegandii) (0[8]:8:57) ; 2519 (var. canadensis) Mille Lacs Lake with 2538 (Wiegandii) from Itasca Park (0:31) ; 2300 (var. canadensis) Mille Lacs Lake with 2517 (Wie- gandii) Mille Laes Lake (0:58) ; 2522 (Wiegandii) Mille Lacs with 2523 (var. canadensis) Mille Lacs (0[4]:20: 148); 2550 (var. canadensis) Lake Michigan shore with 2522 (Wiegandii) Mille Lacs (0[2] :8:57). In the cases listed in which a few F, seedlings developed and then died, the failure was associated with an intense formation of anthocyanin pigment. The above experimental data certainly substantiate the recognition of E. Wiegandii as a species very distinct from E. canadensis var. canadensis. 6. Crosses of E. Wiegandii with E. Wiegandi Certain strains of E. Wiegandii from the Mille Lacs Lake area (2499, 2517, 2522) exhibited lagging univalents, bridges and a chiasma frequency of 1.90 in 300 p.m.c.'s examined. Yet fertility in terms of percentage of seeds was not greatly reduced. Since the above data suggested the existence of genetic barriers between populations in central Minnesota, experimental crosses between the follow- ing sympatric strains were made: — 2502 (Wiegandii) Brainerd crossed with 2522 (Wiegandii) Mille Lacs (32: 32:70) and the reciprocal (36:36:70). The complete fertil- ity of the F, is indicative of the taxonomic identity of the strains. By way of comparison with the sympatric cross, an al- lopatrie eross was made between 2438 (Wiegandii) from Itasca Park and 2536 (Wiegandii) from Mattapedia, Que- bec (20:21:42) and the reciprocal (14:14:28). Examina- tion of the pollen-mother-cells revealed chiasma frequencies ranging from 1.70 to 2.00 and irregularities in meiosis were 1967] Elymus — Church 153 rare. Yet, in contrast to the sympatric crosses, the Minne- sota with Quebec crosses were largely sterile. A similar correlation between geographic range and genetic barriers has been found in crosses of E. canadensis var. canadensis from Vermont with strains from Minnesota. For example, in the cross of 2252 (var. canadensis) from Vermont with 2300 (var. canadensis) from Mille Lacs, no seeds were ob- tained (0:212). With respect to the implied limits of genetie barriers between strains in the above tests, the strains of Wiegandii are more closely related to each other than are the interregional strains of var. canadensis. 7. Crosses of E. Wiegandii with E. diversiglumis In broadly sympatric populations of E. diversiglumis and E. Wiegandii, both species may exhibit characters in com- mon; i.e., very drooping spikes, glaucous, dark green adaxi- ally pilose leaves and varying degrees of setaceous glumes. Although other characters may be found to separate these species (cf. key), experimental crosses were made to test the possibility of natural interspecific hybridity. As can be seen from the following data, however, all such crosses were failures: 2522 (Wiegandii) Mille Lacs Lake with 2328 (diversiglumis) Itasca Park (0:86) and reciprocal (0:40), 2361 (diversiglumis) Itasca Park with 2517 (Wiegandii) Mille Lacs Lake (0:70), 2514 (diversiglumis) Gull Lake with 2517 (Wiegandii) Mille Lacs Lake (0:58), 2438 (Wie- gandii) Itasca Park with 2328 (diversiglumis) Itasca Park (0:32). In a cross of 2355 (diversiglumis) from Cotton Lake with 2522 (Wiegandii), T seedlings were obtained (7:7:40). In the course of a month, however, an intense anthocyanin pigmentation developed in all plants and death ensued after two months. A similar barrier was noted above in a cross of E. Wiegandii and E. canadensis var. canadensis. 8. Crosses of E. Svensonii with E. Hystrix Since the Tennessee endemic E. Svensonii resembles E. Hystrix to some extent, it seemed logical to test the rela- 154 Rhodora [Vol. 69 tionship of these species, again by means of artificial hy- bridization. The vigorous strain 2368 of Hystrix from Nerstrand Woods, Minnesota was employed in making re- ciprocal erosses. The cross with pistillate Hystrix resulted in a vigorous, intermediate, partly fertile F, (3:4:24). These plants were still surviving three years later in out- door plots. The F, of three plants (3:27) segregated as follows: a sterile, glaucous Hystriz, a partially fertile, green Hystrix and a fertile intermediate. The latter segre- gate was still vigorous, though largely sterile, after five years in an outdoor plot. Chiasma frequencies in the inter- mediate member of the F, were 1.90 compared to 2.00 for both parents. Pollen of this hybrid appeared normal. A few seeds from the first year plants produced 6 plants (6:48) all resembling Hystrix. These F, segregates were sterile. The reciprocal cross (i.e., with Svensonii as the pistillate parent) produced a vigorous but largely sterile F, (9:16: 22) which resembled a Hystrix with adpressed instead of spreading spikelets at maturity. The two seeds obtained produced an F, (2:58) that resembled a glumeless Hystrix like the staminate parent but possessed reflexed awns and glaucous foliage like the pistillate parent. Ten F, plants resembled Hystrix very closely, were intensely glaucous and mostly fertile, although they remained in a vegetative state for the first year. 9. Crosses of S. Svensonii with E. canadensis var. in- terruptus Since the endemic Svensonii has been confused with var. interruptus (Hitchcock and Chase 1950), crosses between these taxa were made to determine the degree of possible genetic relationship. The cross of 1041 (var. interruptus) from Texas with staminate 2527 (Svensonii) produced a vigorous, intermediate F, (9[11]:33:90) which was still thriving after five years in the field. The F, consisted of 8 vigorous plants (8[11] :18) : — one fertile var. interrup- tus, two fertile Svensonii and five partly fertile intermedi- 1967] Elymus — Church 155 ates. The latter forms had spikes with slightly sinuous rachises, widely separated nodes (an interruptus char- acter), glaucous foliage and very setaceous glumes (Sven- sonii features). The chiasma frequency of the intermediates was 1.87. More than 50% of the p.m.c.’s examined had bridges and univalent laggards, in contrast to normal meioses in the parents, The four seeds obtained from the F, did not germinate. The cross made with Svensonii as the pistillate parent also produced a vigorous F, (32[37] :96:293) and again intermediate in appearance. The F, (8:36:96) segregated equally to give partly fertile intermediates and sterile, veg- etable plants. The single F, plant (1:8:25) obtained was stunted and sterile. The chiasma frequency of the F, was 1.86, very close to that of the reciprocal cross. In some 800 p.m.c.’s examined, irregularities proved to be relatively infrequent. 10. Crosses of E. Svensonii with E. diversiglumis Crosses with 2428 (diversiglumis) from Itasca Park were failures (0:108), reciprocal (0:125). Similar crosses with 2355 (diversiglumis) from Cotton Lake, Minnesota were likewise failures (0:158), reciprocal (0:110). 11. Crosses of E. Svensonii with E. virginicus var. vir- ginicus Since it has been demonstrated that various strains of E. virginicus var. virginicus have genes in common with many other species (Stebbins and Snyder 1956, Church 1959), reciprocal crosses were made with Svensonii in order to probe further into the genetic background of this Tennes- see endemic. The cross with pistillate 2542 (var. virgini- cus) trom Texas, produced a husky, partly fertile, intermediate F, (27:29:129). The F, (23:38) was entirely glaucous like the staminate 2527 (Svensonii) parent, but segregated into several categories as to width of glumes. Most of the latter forms were sterile except for one with glumes only slightly wider than those of Svensonii and a 156 Rhodora [Vol. 69 seed set of 15%. Cytological studies of the p.m.c.’s in the intermediates of the F, revealed a chiasma frequency ‘of 1.82. Laggards and extrusions in 50% of the division stages appeared to be the cause of much of the shrivelled pollen. The reciprocal cross (18:19:150) produced intermediate plants with glumes much nearer to var. virginicus than to Svensonii in form and width. The few seeds obtained failed to germinate (0:14). Chiasma frequency was 1.84 and laggards, bridges and extrusions were common in all later stages of meiosis. A cross with 2641 (var. virginicus) from Michigan em- ployed as the staminate parent produced only a single, stunted sterile F, plant (1[4] :6:116). 12. Crosses of E. canadensis var. interruptus with E. diversiglumis Although the taxonomic confusion between var. inter- ruptus and diversiglumis has been clarified earlier in this paper, it is of interest to present now the degree of genetic barriers between these taxa as based on the data from ex- perimental hybrids, The cross of 1041 (var. interruptus) from Llano County, Texas with 2461 (diversiglumis) from MeCraney Lake, Minnesota as the pistillate parent was relatively successful (17[24] :41:268) in the production of a rather intermediate F, which, however, was completely sterile. Chiasma frequency was 1.83 and in all stages of meiosis examined, ca. 40% of the p.m.c.'c revealed laggards, bridges and extrusions. The cross made with 1041 (var. interruptus) as the pistillate parent again resulted in a completely sterile F, (3:25:70) with narrow glumes and hirsute lemmas. A cross attempted with 2450 (diversi- glumis) from Itasca Park, Minnesota as the pistillate parent was a complete failure (0:134). 13. Crosses of E. canadensis var. interruptus with E. virginicus var. glabriflorus f. glabriflorus The variety glabriflorus of the widespread E. virginicus is found mostly below the 40 degree parallel in North Amer- 1967] Elymus — Church 157 ica and is distinguished from northern varieties by much longer and thinner glumes. Forms with glabrous or hirsute lemmas are both common (cf. key to species). In Union County, southern Illinois, strains of both forms have been found with very setaceous glumes, the origin of which might be due to introgression from some other southern species. Introgression between E. canadensis var. interrup- tus and strains of E. virginicus var. glabriflorus in Texas has been demonstrated by Brown and Pratt (1960). Hence it was considered appropriate to study the relationships between these two taxa further through experimental hy- bridization. A cross of 2603 (f. glabriflorus) from Illinois as the pistillate parent was made with 1041 (var. inter- ruptus) from Texas. The F, plants were intermediate in morphology (7:11:81) and had a seed-set of ca. 5%. The F, generation of 13 plants segregated as follows:- 3 fertile f. virginicus, 4 fertile var. interruptus and 6 slightly fertile intermediates. However, the few seeds from the intermedi- ates failed to germinate. The reciprocal cross produced a fairly similar F, population of 14 plants that segregated in a similar pattern as follows:- 6 partly fertile f. virgin- icus, one fertile var. interruptus and 7 partly fertile inter- mediates that continued to segregate in the F,! Although a surprising degree of genetic compatibility was demon- strated between the allopatric strains of this experimental cross, none of the experimental hybrids resemble the aber- rant populations of f. glabriflorus with setaceous glumes in Union County, Illinois. 14. Cross of E. villosus with E. Hystrix Although E. villosus is quite distinct from other eastern species of Elymus in its smaller florets and very compact spikes, the narrow glumes with terete bases suggest that its possible role in the ancestry of natural hybrid populations of E. virginicus or E. Hystrix should not be overlooked. Fur- thermore, a natural, bigeneric hybrid between E. villosus and Hordeum jubatum L. has been described recently from Iowa (Pohl 1966). 158 Rhodora [Vol. 69 A cross between 2520 (E. villosus) and sympatric 2521 (E. Hystrix) from moist, woodland borders in central Min- nesota was a failure (0:94 and reciprocal 0:94). 15. Cross of E. villosus with E. virginicus f. glabriflorus Again in Union County, Illinois, E. villosus grows sym- patrically with E. virginicus var. glabriflorus f. glabriflorus or with f. australis. Since the strains of E. villosus came to anthesis 3 to 4 weeks earlier than those of E. virginicus var. glabriflorus, experimental crosses were difficult to arrange. However, a cross of pistillate 2652 (E. virginicus f. glabri- florus) with 2676 (E. villosus) produced only one interme- diate plant that was stunted and sterile (1:8:21). Another cross of pistillate 2647 (f. australis) with 2678 ( E. villosus) produced seedlings that died after 3 months (3:4:60). 16. Cross of E. villosus with E. virginicus var virginicus A cross of 2678 (E. villosus) and 2600 (E. virginicus var. virginicus) produced only 5 stunted and sterile F, plants (11:16:93). The reciprocal cross produced only inviable seeds (0:6:97). Within the limits of these data, then, E. villosus appears to be quite isolated genetically from both E. Hystrix and E. virginicus. DISCUSSION AND SUMMARY It has been demonstrated that the frequently encountered populations of Elymus with very unequal, setaceous glumes in the mixed woodlands of central Minnesota are composed mainly of E. diversiglumis, a recently overlooked species. E. diversiglumis is quite aggressive in the Minnesota seg- ment of its range, but is distributed sparingly with no obvi- ous phenotypic variations in Wisconsin, Wyoming, the Dakotas and Manitoba. In the disturbed habitats of wood- land road embankments, however, forms with wider and longer than typical glumes may be found in vigorous, thick stands that reveal rather low levels of fertility in terms of seed formation (20-50%). Such putative hybrids may ex- hibit also irregular meioses in the pollen-mother-cells (fig. 26). 1967] Elymus — Church 159 Attempts at recreating these aberrant forms of E. diversi- glumis by crossing sympatric E. canadensis var. canadensis and E. Hystrix resulted mostly in failures. However, a sterile, intermediate F, was obtained that resembled E. Hy- strix with abundant, filiform glumes (fig. 17), but it did not duplicate any natural populations encountered in Min- nesota. Furthermore, any close genetic relationship be- tween E. diversiglumis and sympatric E. Hystrix does not exist in the light of the failure of most of the experimental crosses between these species. On the other hand, the rare, nearly sterile F, plant that did appear (figs. 19 and 20) re- sembled forms of E. Hystrix elsewhere in the range. This variant has not yet been found in Minnesota. The F, plants were all typically awnless E. Hystrix and quite sterile. Experimental crosses between E. diversiglumis and the usually genetically isolated E. riparius (figs. 21 and 22) presented intermediate forms whose pattern of origin ap- pears parallel to but not duplicative of that of the aberrant E. diversiglumis forms noted above. It was through crosses and backcrosses of strains of E. canadensis var. canadensis and E. diversiglumis that the production of experimental hybrids that duplicated natural aberrant forms ultimately was made. Populations of E. diversiglumis with unusually wide glumes would appear, then, to exhibit introgression from E. canadensis var. cana- densis (fig. 18). The genetic evidence presented indicates that many of the strains of E. diversiglumis employed in experimental crosses were already heterozygous for genes from E. canadensis var. canadensis and that, in effect, back- crosses had been made, Such a pattern of introgressive hybridization would appear to be continuing in central Min- nesota. Of further interest is the apparent fact that intro- gression operates most effectively only with sympatric strains of E. canadensis var. canadensis, since a strain of the latter from Vermont contributed to the complete sterili- ty of the experimental hybrids (fig. 24). Since it has been noted that E. Wiegandii in some of its characters resembles both E. diversiglumis and E. canaden- 160 Rhodora [Vol. 69 sis var. canadensis, it seemed appropriate to determine the degree of genetic barriers existing between all three sym- patric species. The evidence of failure of experimental crosses indicates clearly that not only is E. Wiegandii not involved in hybrid complexes of E. diversiglumis but that genetic isolation from E. canadensis var. canadensis also is quite strong. The reduction of E. Wiegandii to a variety of E.canadensis (Bowden 1964) is not supported by the genet- ical basis for species integrity presented in this study. Returning to E. Hystrix as a source of genetic control over the appearance of reduced or setaceous glumes in hy- bridizing populations, the discovery of E. Svensonii, with irregular sized, bristly glumes, afforded the opportunity through breeding experiments to probe the genetic rela- tionship of this isolated endemic to not only E. Hystrix but to E. canadensis var. interruptus and E. diversiglumis. While clearly isolated genetically from E. diversiglumis, it is significant that E. Svensonii crossed reciprocally with both E. Hystrix and E. canadensis var. interruptus and that the intermediate F, produced a segregating F, in all cases, even though the F, was sterile. In particular, the fact that the F, of E. Svensonii and E. Hystrix segregated for fertile E. Hystrix and partly fertile intermediates would appear to indicate that E. Hystrix played a role in the origin of E. Svensonii, perhaps in the not too distant geological past. Various components of E. virginicus appear to be able to combine with either E. Svensonii or E. canadensis var. interruptus to produce hybrids with setaceous glumes. When E. virginicus var. virginicus was crossed with E. Svensonii, the F, segregated not only for the parents but intermediates as well. In the cross of E. virginicus var. virginicus f. virginicus with E. canadensis var. interruptus, partly fertile intermediates with more or less setaceous glumes continued to segregate even in the F.. In contrast to the demonstrated genetic affinities between E. Svensonii, E. Hystrix and components of E. virginicus, E. villosus appears to be quite isolated from E. Hystrix and both forms of E. virginicus var. glabriflorus. 1967] Elymus — Church 161 In these studies, it is apparent that the opportunities for the evolution of new taxa with setaceous glumes are more frequent in the relatively southern ranges of species of Elymus. A striking instance of the process of speciation, currently operating, will be presented in a second paper involving E. Hystrix and E. virginicus in southern Illinois. DIVISION OF BIOLOGICAL AND MEDICAL SCIENCES BOTANICAL LABORATORIES BROWN UNIVERSITY, PROVIDENCE, RHODE ISLAND LITERATURE CITED BERNARD, FRERE JEAN-PAUL. Personal Communication. BooHER, L. E. and R. M. Tryon. 1948. A study of Elymus in Min- nesota. Rhodora 50: 80-91. BOWDEN, WRAY M. 1958. Natural and artificial X Elymordeum hy- brids. Can. Jour. Bot. 36: 101-123. 1964. Cytotaxonomy of the species and interspe- cific hybrids of the genus Elymus in Canada and neighboring areas. Can. Jour. Bot. 42: 547-601. BovLE, W. S. 1963. A controlled hybrid between Sitanion hystrix and Agropyrcn trachycaulum, Madroño 17(1): 10-16. BROWN, WALTER V. and GENE A. PRATT. 1960. Hybridization and introgression in the grass genus Elymus. Amer. Jour. Bot. 47(8): 669-676. BuckK.EY, S. B. 1862. Elymus interruptus. Acad. Nat. Sci. Phila. Proc. 1862: 99. CHURCH, G. L. 1954. Interspecific hybridization in eastern Elymus. Rhodora 56: 185-197. 1959. Introgressive hybridization in Minnesota spe- cies of Elymus. Proc. Ninth Int. Bot. Cong., Montreal, 2:73. DEWEY, DouGLAs R. 1964. Natural and synthetic hybrids of Agropy- ron spicatum X Sitanion hystrix. Bull. Torr. Bot. Club 91: 396-405. 1966. Synthetic Agropyron-Elymus hybrids. I. Elymus canadensis X Agropyron subsecundum. Amer. Jour. Bot. 53(1): 87-94. Dore, W. G. 1950. Supposed natural hybrid between Agropyron and Hystrix. Can. Field Naturalist 64(1): 39-40. DuTILLY, ARTHEME and E. LEPAGE. 1963. Contribution à la Flore du Versant Sud de la Baie James, Quebec-Ontario. Contrib. Arctic Inst. 12F. FARWELL, O. A. 1920. Notes on the Michigan flora. Mich. Acad. Sci. Rept. 21: 357. 162 Rhodora [Vol. 69 FASSETT, N. C. 1951. Grasses of Wisconsin, Univ. Wis. Press. FERNALD, M. L. 1933. Types of some American species of Elymus. Rhodora 35: 187-198. 1950. Gray's Manual of Botany, 8th edition. FRAZER, W. P. and R. C. RUSSELL. 1954. List of the flowering plants of Saskatchewan. Univ. of Saskatchewan, Saskatoon. GLEASON, H. A. 1952. Illustrated Flora of the Northeastern United States and Adjacent Canada. GROSS, A. T. H. 1960. Distribution and cytology of Elymus macounii Vasey. Can. Jour. Bot. 38: 63-67. HARRINGTON, H. D. 1954. Manual of the Plants of Colorado. Denver. HITCHCOCK, A. S. and AGNES CHASE. 1950. Manual of the Grasses of the United States, 2nd edition. Pour, RICHARD W. 1959. Morphology and cytology of some hybrids between Elymus canadensis and E. virginicus. Proc. Iowa Acad. Sci. 66: 155-159. 1966. X Elyhordeum iowense, a New Inter- generic Hybrid in the Triticeae. Brittonia 18(3): 250-255. RYDBERG, P. A. 1922. Flora of the Rocky Mountains and Adjacent Plains. New York. SCOGGAN, H. J. 1957. Flora of Manitoba. National Museum of Canada Bull. 140. SCRIBNER, F. LAMSON. 1898. USDA Bull. Agrost. 13: 49. and CARLETON R. BALL. 1901. USDA Div. Agrost. Bull. 24: 47-49. STEBBINS, G. L. and A. VAARAMA. 1954. Artificial and natural hy- brids in the Gramineae, tribe Hordeae. VII. Hybrids and al- lopolyploids between Elymus glaucus and Sitanion spp. Geneties 39: 378-395. and L. A. SNYDER. 1956. Artificial and natural hy- brids in the Gramineae, tribe Hordeae. IX. Hybrids between western and eastern North American species. Amer. Jour. Bot. 43:(4) : 305-312. STEVENS, O. A. 1950. Handbook of North Dakota Plants. North Dakota Agricultural College. SVENSON, H. K. 1941. Notes on the Tennessee Flora. Brooklyn Bot. Gard. Contrib. 93: 115. WIEGAND, K. M. 1918. Some species and varieties of Elymus in eastern North America, Rhodora 20: 81-90. Voss, EDWARD G. Elymus in unpublished notes and keys to the Triticeae in Michigan. FLORAL ANATOMY OF RHEXIA VIRGINICA (MELASTOMATACEAE) RICHARD H. EYDE AND JAMES A. TEERI Contradictory comments concerning the flowers of Rhexia L. are found in the two current manuals of our northeastern flora. The description in Gray's Manual (Fernald, 1950) clearly implies an inferior or partially inferior ovary: "Hy- panthium urceolate, adherent to the ovary below and con- tinued above it." In contrast, Gleason and Cronquist (1963) report "hypanthium free from the ovary." From serial sections of flowers we have learned that this point of con- fusion is caused by a developmental peculiarity; viz., cell layers in the fruit wall separate prior to dehiscence in such a way that the partially inferior ovary can appear superior. Flowers of Rhexia have unilocular anthers, another unusual feature; therefore we have thought it worthwhile to de- scribe in some detail the reproductive structures of one species, R. virginica L. This investigation originated from conversations between the first author and Dr. John J. Wurdack, a specialist in the taxonomy of Melastomataceae at the Smithsonian Institu- tion, and was carried out during the summer of 1966 while the second author was supported by the Smithsonian Insti- tution's Summer Program for Students. Dr. Wurdack as- sisted by supplying specimens and information and by read- ing our manuscript critically. Flowers, flower buds, and developing fruits were obtained from plants of Rhexia virginica var. virginica cultivated in the garden of Dr. and Mrs. J. J. Wurdack in Beltsville, Maryland, these plants having been obtained originally from natural stands in the vicinity. A voucher specimen (Wur- dack 2534) has been deposited in the U.S. National Herbari- um. 163 164 Rhodora [Vol. 69 Serial sections 10-15, thick were prepared with a rotary microtome, after fixing in formalin-acetic acid-ethanol, de- hydrating in a t-butanol series, and embedding in ‘“Para- plast." The sections were stained with tannic acid, ferric chloride, safranin, and fast green. Altogether, more than 50 flowers and developing fruits were prepared in this man- ner. For 3-dimensional examination of the vascular pattern, we cleared about 20 flowers, utilizing both the chloral- lactophenol method of Amann (Bersier & Bocquet, 1960) and the simultaneous clearing and staining method of Fuchs (1963). Fuchs’ procedure, involving prolonged immersion of the material in 10-15% NaOH, had to be modified some- what, for concentrations of this strength tend to macerate the softer floral tissues. We found that a 5% solution of NaOH could be used, thus reducing the maceration to within tolerable limits. Cleared flowers were dehydrated with an ethanol series and transferred to toluol for examination, thereby hardening the tissues enough to permit manipula- tion with camel's-hair brushes. Line drawings were prepared by the second author with the aid of a Wild M5 Stereomicroscope and drawing attach- ment. Photographs are by Jack Scott, Smithsonian photog- rapher. FRhexia virginica is an erect perennial herb (often some- what woody), widely distributed in eastern North Ameri- ca. The flowers, produced throughout the summer in open cymes, are conspicuous for their four showy purple petals and bright yellow stamens. These are borne near the sum- mit of an urceolate hypanthium, which terminates in four short calyx lobes, the lower part of the hypanthium being partially adnate to a four-locular inferior ovary. The locules are in the same radii as the petals — as Eichler (1873) em- phasized, this is an exception for the family — and each contains a massive axillary placenta with many bitegminous ovules. The tabular upper part of the hypanthium sur- rounds an elongate style with truncate stigma. Before an- thesis the petals, like those of other Melastomataceae, are tightly overlapped in the configuration known as "right con- 1967] Rhexia — Eyde and Teeri 165 tort," or “right convolute,” to use Eichler's (1878, p. 480) term." Prominent multicellular trichomes with globose glandular heads occur on the exposed abaxial surfaces of petals, on the exterior of the hypanthium, and on the superior part of the ovary. Although Holm (1907) found trichomes of two kinds on the leaves and stems of Rhexia virginica, all of the floral trichomes are alike. They resemble Holm’s Fig. 6, except that the glandular heads are not papillose. Else- where in the Melastomataceae, a family noted for the great diversity of its trichomes, the Rhexia type has been figured from at least three genera, Dissotis ( Feissly, 1964, Fig. 13), Comolia (Pflaum, 1897, Fig. 17) and Sonerila (Palézieux, 1899, Fig. 13B). Feissly's drawings convey quite accurately the dense cytoplasm of the head. This terminal structure does not usually persist through microtechnical procedures ; almost all of the trichomes in our sections were “decapi- tated." Before the flower opens, the stamens are bent inward in the manner characteristic for the family (Fig. 1), witn their apices inserted in eight pockets around the upper part of the ovary. Ziegler's (1925) illustrations show that the pockets (“Gruben”) in Melastomataceae with dimorphic stamens are of two depths, those ensheathing the long antesepalous stamens being much deeper than those en- sheathing the antepetalous stamens. Adjacent pockets in Rhexia differ slightly in depth, even though all stamens are superficially alike, and cross sections through the tips of the incurved anthers indicate that the antesepalous stamens are actually a bit longer than their neighbors (Fig. 14). "In order to appreciate the right-handedness of the configuration, it is necessary to imagine a “‘bug’s-eye” view from inside the corolla or to hold up page 7 of Eichler’s Bliithendiagramme and view his Fig. 1E through the opposite side by transmitted light. The normal bota- nist's-eye view of a flower bud, from above and outside the corolla, shows each petal overlapping its neighbor to the left. Readers who are interested in the history of this rather puzzling convention should consult Appendix C of B. Daydon Jackson's *A Glossary of Botanic Terms." (For additional confusion, see also Schoute, 1935, p. 25.) 166 Rhodora 11 10 3 2 [Vol. 69 Fig.1. Flower of Rhexia virginica before anthesis; diagram of longitudinal section. Numerals indicate levels of the sections in Fig. (Legend continued to foot of page 169) Rhexia — Eyde and Teeri 167 1967] 168 Rhodora [Vol. 69 VASCULAR SYSTEM Melastomataceae is one of several families with internal phloem. This character can be observed in some of the floral vascular bundles of Rheaia when they are completely differ- entiated, but the position of phloem elements is somewhat variable and difficult to establish in many cross sections; therefore we have illustrated the vascular system in Fig. 1-11 without indicating xylem and phloem. Major features of vasculation in the Rhexia flower can best be understood by following cross sections from base to apex, beginning with the vascular cylinder of the peduncle (Fig. 2). The lowermost vascular strands to separate from this cylinder are eight large bundles that pass through the hypanthium to perianth members and stamens. Our Fig. 3 shows six of the eight separating a bit lower than the other two (which in this case lead to two calyx lobes). We have found this "six-and-two" pattern to be common, but it seems to have no special significance, The next bundles to sepa- rate from the central cylinder are the four dorsal carpellary bundles (indicated by the letter d in Fig. 4-7). The dorsals consist of only a few cells in cross section and are therefore easily overlooked. 'They originate just below the locules and remain immediately adjacent to the locules throughout their length, ultimately ending in the “roof” of the ovary, below the styles. It can be seen in cleared preparations that the central axis of the flower contains an anastomosing complex of longitudinal bundles from which numerous transverse pla- cental bundles branch in ladder-like fashion (Fig. 1). The complex originates from the basal vascular cylinder, and represents the combined ventral bundles of the four carpels ; however, individual ventrals are not clearly distinguishable below the stylar level. Instead, many cross sections show an almost complete vascular cylinder in the axis (Fig. 7) ; others show four patches of vascular tissue in a more or less square arrangement (Fig. 5, 6). In successively higher sections taken above the placentae, the amount of central vascular tissue decreases until only four slender strands 1967] Rhexia — Eyde and Teeri 169 remain in the superior part of the ovary, where they alter- nate with the four pollen transmitting tracts. (We have not attempted to show the distal continuation of the central vas- cular system in Fig. 1, but see Fig. 8-11). In the style the tracts merge in a broad column of transmitting tissue and each of the vascular strands divides, yielding four pairs of ventral carpellary bundles (Fig. 10) that branch irregularly below the stigma, forming an irregular ring in cross section (Fig. 11). Four of the eight hypanthial bundles pass into the calyx lobes as midveins. The other four divide at the perianth level into three strands, a central strand supplying a pet^' and two lateral strands passing into the adjoining calyx lobes. In addition, each hypanthial strand provides a bun- dle to a stamen, this separation also occurring at the peri- anth level. All of these longitudinal bundles are united in the upper part of the hypanthium by a continuous trans- verse vascular complex that underlies the torus. In common with many genera of Melastomataceae, Rhexia has a prominent appendage (Fig. 1) on the abaxial side of each stamen, vascularized exactly as Wilson (1950, Fig. 14) portrayed it. Wilson's interpretation that the vascularized appendages are vestiges of primitive telomes has not con- vinced other botanists (Eames, 1961, p. 133, 134; James, 1956, p. 209, 210; Leinfellner, 1958). Floral vasculature of Rhexia resembles closely that of Aciotis fragilis, which Morley (1953, p. 264 and Fig. 44) chose as an example of a relatively unmodified member of the family. A difference worth noting, however, is that the 2-11. In many flowers, the dorsal appendage (a) of the anther is longer than the example shown here. Trichomes have been omitted from this and succeeding drawings. Fig. 2-11. Cross sections of an unopened flower (cf. Fig. 1). Some of the diagrams include features from more than one section, and Fig. 7-11 are oriented somewhat differently from Fig. 2-6. All dorsal carpellary bundles (d) are labeled in Fig. 4, only one of the four in Fig. 5-7. p — pollen transmitting tissue (Fig. 8; stippled in Fig. 9- 11); v - ventral carpellary bundles. Further explanation in text. 351. ar - m -— ET. MEIN PITT OT NM EK. Fig. 12-14. Flowers of Rhexia virginica before anthesis, showing aerenchyma (s) of ovary wall and hypanthium. Arrows on anthers indicate “Offnungsgewebe” of Ziegler (see text). — Fig. 12. l.s., X85. — Fig. 13. x.s. through inferior part of ovary, X60. — Fig. 14. x.s. through hypanthium and anthers. Section passes through pore in antipetalous anther, but pores of adjoining anthers are below plane of section; this suggests a slight dimorphy in length of stamens, X 90. Plate 1352. Fig. 15- 1T. “Stamens of ; seem virginica. — Fig 15. Empty an- ther, x.s. showing crushed remnants of septa (arrows); vascular tis- sue on right (a) supplies dorsal appendage, X70. — Fig. 16. Tip of anther, median l.s. showing pore, a few remaining pollen grains, X55. — Fig. 17. Tip of anther before anthesis, l.s. roughly paradermal to hypanthium, showing pore (center) and four sporangia. Separa- tion of aerenchyma tissue (s) has been exaggerated in the preparation of sections, X220. 172 Rhodora [Vol. 69 dorsal] carpellary bundles in Aciotis, and in other genera that Morley figures, extend into the styles; whereas in Rhexia the dorsals terminate at the stylar base, and the ventral complex supplies the style. HISTOLOGY OF OVARY WALL In its basal, inferior portion, where carpellary tissues merge indistinguishably with hypanthial tissue, the com- pound ovary wall of an immature flower comprises 8-12 cell layers. The free portion is thinner, about six cell layers in most places. It is thinnest along the midline of each carpel (Fig. 8, 9, 14), where the cells do not enlarge appreciably during growth, a phenomenon related to the presence of the poorly differentiated dorsal bundle. Dehiscence of the ma- ture fruit begins along these dorsal midlines, and is thus loculicidal. During development, druses become abundant in cells of the gynoecium, as elsewhere in the flower. Other types of crystals, tannin cells, and isolated sclereids are absent. Cells of the very young ovary wall are rectangular in outline when examined in transverse or longitudinal sec- tions, and the epidermal layers retain this appearance as the flower develops (except where papillae and multicellular trichomes develop on the outer epidermis). Cells of the lav er subjacent to the outer epidermis also remain rectangular in outline. In the inferior part of the ovary, however, the deeper cell layers develop intercellular spaces (Fig. 12, 13). The result is a tissue much resembling the spongy mesophyll of leaves, the individual cells being very irregular in form. Similar developmental changes take place in the free portion of the hypanthium (Fig. 14), but not in the superior part of the ovary wall. The aerenchymatous zone is one of irreg- ular cell contact, where separation occurs quite easily dur- ing the preparation of sections (see especially Fig. 17). In the same manner, tissues in the inferior part of the ovary can separate — beginning in the radii between the staminal pockets — when flowers and fruits dry under natural condi- tions or in a plant press. To an observer aided only by hand 1967] Rhexia — Eyde and Teeri 173 lens and razor, the ovary may then appear completely superior. ANTHER Unilocular anthers are known to occur in several plant families, but the unilocular condition is brought about in most cases by the ontogenetic union of two locules (Mahe- shwari, 1950, p. 39). The condition found in Rhexia — four locules uniting at maturity to form one — is rare? As in some other Melastomataceae (cf. Centradenia, Heeria; Zie- gler, 1925), sporangia develop in a parallel arrangement. Rhexia is believed to be the only genus of the family, how- ever, in which all three septa regularly deteriorate before anthesis (see James, 1956, p. 204, for comments on this point by Asa Gray and contemporaries). At early stages of development each septum is made up of three or more cell layers, but the cells gradually disintegrate through the ac- tivity of the uninucleate tapeta. Eventually the tapetum of one sporangium is juxtaposed to that of the neighboring sporangium with only fragments of cell walls between them. Rupture of these tenuous partitions occurs more or less simultaneously throughout the length of the anther (Fig. 8-11), and the central septum persists no longer than the others. The breaks occur toward the adaxial side, so that cross sections prepared after pollen has escaped show the flattened remnants of septa on the connective side (Fig. 15). At the time of dehiscence, the anther wall comprises two cell layers, neither of which has the irregular banded wall thickenings of the '"endothecium" or “fibrous layer”? in a typical textbook anther. Some local modification of the epi- "Eames (1961, p. 114) gives as examples Arisaema, Callitriche, Clusia, but cites no references. We have come across published com- ments on anther development only for Callitriche. Anthers of the latter have a very different form from those of Rhexia, and unilocu- larity develops only near the apex; the lower part of the anther is bilocular at maturity (Hegelmaier, 1864, p. 39). 'See Eames (1961, p. 125, 138) for a discussion of these terms. Some Melastomataceae do have a fibrous layer. In lacking a fibrous layer, Rhexia resembles Melastoma malabathricum and Mouriri guia- nensis. (Subramanyam, 1948; Venkatesh, 1955). 174 Rhodora [Vol. 69 dermal cells takes place during development, those at the base of the anther becoming papillose, and those surround- ing the pore becoming transversely elongate (Fig. 12, 14, 16), but overall thickening of the cell walls is negligible. Pollen emerges through a solitary pore located near the apex of the anther on its adaxial side. The pore is circular to elliptical in sectional outline (Fig. 17) and only about .05 mm in diameter.’ It originates as a distinct depression early in the development of the anther. About five layers of cells surrounding the interior of the pore and separating it from the sporogenous tissue remain small and stain densely throughout the development of the anther (Fig. 12, 14). Ziegler (1925) used the term “Offnungsgewebe” to describe an identical tissue in Centradenia, the anthers of which are histologically similar to those of Rhexia. The small size and dense appearance of the cells invite com- parison with the “resorption tissue" by which anthers of certain Ericaceae dehisce. To judge from publications on Ericaceae, however, cells surrounding the Rhexia pore dif- fer in two ways: (1) they lack the calcium oxalate crystals of ericaceous resorption tissue (Matthews & Knox, 1926), and (2) they do not “become converted into an amorphous, apparently fluid mass” (Copeland, 1943). In most families with poricidal anthers, a derivation from septicidal dehiscence is evident upon inspection, for the pores are in pairs, are somewhat elongate along the lateral furrows, and open with the aid of a fibrous layer (Leclere du Sablon, 1885). Complete transitions can be shown in some plant groups from anthers opening longitudinally to anthers opening apically (see, for instance, Venkatesh, 1957). A similar progression has been proposed for Melasto- mataceae (Venkatesh, 1955) but the evidence involves only Memecylon and Mouriri, genera rather remote from Rhexia and its allies. Anthers of Memecylon open the “normal” way, by two complete longitudinal slits; whereas the two slits in Mouriri are restricted toward the apex, supposedly “Pore size is one of the characters used by James (1956) to sub- divide Rhexia into two series of species. 1967] Rhexia — Eyde and Teeri 175 indicating an evolutionary stage through which all poricidal Melastomataceae have passed. As far as we are aware, however, no one has illustrated a series of transitional forms? between the paired, linear openings of Mouriri and the solitary, circular, centrally situated pore of Rhexia. One might therefore propose a somewhat different explana- tion: that the small, rounded pore evolved as a new struc- ture, beginning as a functionless subapical indentation between the outer anther sacs and assuming its present importance when longitudinal dehiscence was lost. Addi- tional comparative studies of melastomataceous anthers would be very desirable. Geotropism in the androecium of Rhexia deserves com- ment. Open flowers are oriented with floral axis away from the vertical and toward the horizontal position, Since the emerging anthers are reflexed and the pore is located on the adaxial side, the apical end of the lowermost anther (nearest the earth's surface) hangs downward with its pore directed outward. All of the other anthers are brought into positions approximately parallel to this one by a coun- terclockwise (as viewed from outside) twisting of the filaments, the filament farthest from the earth twisting 1809. It is interesting that in Monochaetum, a genus placed next to Rhexia in Cogniaux's (1891) monograph of Melasto- mataceae, the pore is located on the abaxial side of the anther. Nevertheless, the outward orientation of all pores becomes the same as in Rhexia, because in Monochaetum the lower stamens twist and the uppermost stamen does not (Troll, 1922; Ziegler, 1925). According to Ziegler, the upper filaments of Tibouchina twist like those of our Rhexia flowers. The result is the same in the three genera: all pores open toward insect visitors. Leggett (1881) reported that the lower part of the Rhexia anther acts as a bellows, causing puffs of pollen to ‘Dr. Wurdack suggests Miconia as a genus to examine for such transitional forms, but he adds that an adequate range of fluid- preserved material would be hard to obtain. 176 Rhodora [Vol. 69 emerge forcefully from the pore when pressed by a bumble- bee's "foot" or a pointed object in a botanist's hand. We have verified this in company with Dr. Wurdack, using a dissecting needle to apply the pressure and observing the resulting puff of pollen through a stereomicroscope. Dr. Wurdack suggests that in nature the pollen gathering be- havior of the bee may be more complex than this simple observation would indicate, for he has noticed an unusually vigorous buzzing of bumblebees while they are visiting his Rhexia plantings. Vibrations by bee visitors are known to play a role in the floral biology of Melastoma malabathri- cum (Pijl, 1954) and of some other plants in which anthers have apical pores (Michener, 1962). The older literature describes additional complexities in the behavior of insects visiting Melastomataceae (for review, see Harris, 1905, p. 229). Accordingly, the floral biology of Rhexia might well be worth the attention of an enterprising graduate student with a flair for cinematography. DEPARTMENT OF BOTANY SMITHSONIAN INSTITUTION, WASHINGTON, D. C. DEPARTMENT OF BOTANY UNIVERSITY OF NEW HAMPSHIRE, DURHAM LITERATURE CITED BERSIER, J. -D. and G. BOCQUET. 1960. Les methodes d'éclaircisse- ment en vascularisation et en morphogénie végétales comparées. Arch. Sci. 13: 555-566. COGNIAUX, A. 1891. Mélastomacées, Vol. 7, 1256 pp. In A. and C. de Candolle, Monographiae Phanerogamarum. Paris, G. Masson. COPELAND, H. F. 1943. A study, anatomical and taxonomic, of the genera of Rhododendroideae. Am. Midland Nat. 30: 533-625. EAMES, A. J. 1961. Morphology of the angiosperms. New York, McGraw-Hill. xiii + 518 pp. EICHLER, A. W. 1875, 1878. Blüthendiagramme construirt und erlautert. Leipzig, W. Engelmann. viii + 575 pp. [Erster Theil, 1875; Zweiter Theil, 1878] 1967] Rhexia — Eyde and Teeri 177 FEISSLY, C. 1964. Sur l'ornementation du tube calicinal de quelques Osbeckiées Africaines. Bull. Soc. Neucháteloise Sci. Nat. 87: 137-170. FERNALD, M. L. 1950. Gray's Manual of Botany. Ed. 8. New York, American Book Co. lxiv + 1632 pp. Fucus, C. 1963. Fuchsin staining with NaOH clearing for lignified elements of whole plants or plant organs. Stain Technol. 38: 141-144. GLEASON, H. A. and A. CRoNQUIST. 1963. Manual of vascular plants of northeastern United States and adjacent Canada. Princeton, N. J., Van Nostrand. li + 810 pp. Harris, J. A. 1905. The dehiscence of anthers by apical pores. Missouri Bot. Gard. Ann. Rep. 16: 167-257. HEGELMAIER. F. 1864. Monographie der Gattung Callitriche. In- auguralabhandl. Naturwiss. Fak. Univ. Tübingen. Stuttgart. 64 pp., 4 pl. HOoLM, T. 1907. Morphological and anatomical studies of the vegeta- tive organs of Rhexia. Bot. Gaz. 44: 22-33; pl. 1, 2. JAMES, C. W. 1956. A revision of Rhexia (Melastomataceae). Brit- tonia 8: 201-230. LECLERC DU SABLON. 1885. Recherches sur la structure et la déhis- cence des anthéres. Ann. Sci. Nat., 7. sér. Bot. 1: 94-134; pl. 1-4. LEGGETT, W. H. 1881. Fertilization of Rhexia virginica L. Bull. Torrey Club 8: 102-104. LEINFELLNER, W. 1958. Zur Morphologie des Melastemaceen-Staub- blattes. Osterr. Bot. Zeitschr. 105: 44-70. MAHESHWARI, P. 1950. An introduction to the embryology of angio- sperms. New York, McGraw-Hill. X + 453 pp. MATTHEWS, J. R. and E. M. KNox. 1926. The comparative mor- phology of the stamen in the Ericaceae. Trans. Proc. Bot. Soc. Edinburgh 29: 243-281. MICHENER, C. D. 1962. An interesting method of pollen collecting by bees from flowers with tubular anthers. Rev. Biol. Trop. 10: 167-175. MoRLEY, T. 1953. The genus Mouriri (Melastomataceae); a sec- tional revision based on anatomy and morphology. Univ. Calif. Publ. Bot. 26: 223-312. PALÉZIEUX, P. DE. 1899. Anatomisch-systematische Untersuchung des Blattes der Melastomaceen mit Ausschluss der Triben: Micro- lieieen, Tibouchineen, Miconieen. Bull. Herb. Boissier 7, Appendix 5. 85 pp, 9 pl PFLAUM, F. 1897. Anatomisch systematische Untersuchung des Blattes der Melastomaceen aus den. Triben: Microlicieen und Tibouchineen. Dissertation. K. Bayr. Ludwigs-Maximilians- Universitat, München. 91 pp., 2 pl. 178 Rhodora [Vol. 69 PIJL, L. VAN DER. 1954. Xylocopa and flowers in the tropics. III — Observations on some Papilionaceae, Melastoma, Calotropis, Cas- sia and some orchids, with general considerations. K. Nederl. Akad. Wetensch. Proc. C. 57: 553-562. SCHOUTE, J. C. 1935. On corolla aestivation and phyllotaxis of floral phyllomes. Verh. K. Akad. Wetensch. Amsterdam, Afd. Natuurkunde (Sect. 2). Deel 34, No. 4, 77 pp. SUBRAMANYAM, K. 1948. An embryological study of Melastoma malabathricum L. Jour. Indian Bot. Soc. 27: 11-19. TROLL, W. 1922. Uber Staubblatt- und Griffelbewegungen und ihre teleologische Deutung. Flora 115: 191-250; Taf. IV-VI. VENKATESH, C. S. 1955. The structure and dehiscence of the anther in Memecylon and Mouriria. Phytomorphology 5: 435-440. 1957. The form, structure, and special ways of dehiscence of anthers of Cassia — III. Subgenus Sewna, Phy- tomorphology 7: 253-273. WILSON, C. L. 1950. Vasculation of the stamen in the Melastomaceae with some phyletic implications. Am. Jour. Bot. 37: 431-444. ZIEGLER, A. 1925. Beiträge zur Kenntnis des Androeceums und der Samenentwicklung einiger Melastomaceen. Bot. Arch. 9: 398-467. THE PERSISTENCE OF THE DOUBLE-FLOWERED FORM OF CELANDINE POPPY. In September of 1917 Karl Sax of the Bussey Institution sent to the editors of Genetics an article entitled “The In- heritance of Doubleness in Chelidonium majus Linn." which was published in that journal the following year (Genetics 3: 300-307. 1918). The origin of the material Sax studied was not specified beyond the fact that “a double- and a single-flowered plant growing near the Bussey Insti- tution were transplanted and reciprocal crosses made." Sax does not make ciear whether the plants he obtained were under cultivation or spontaneous as a portion of the weedy flora. Apparently no herbarium specimens were prepared of the material studied and spontaneous plants were not found, recorded or preserved by Palmer in his study of the spontaneous flora of the Arnold Arboretum grounds (Jour- nal Arnold. Arb. 11: 100. 1930). Across South Street from the headquarters and the greenhouses of the Bussey In- stitution the Arnold Arboretum built a house which was occupied by E. H. Wilson from its construction until his death in 1931 and subsequently by Karl Sax. To the rear of this house is a low marshy area of land which has served as a dumping ground for greenhouse soils, pruning debris and discarded plants from the nurseries and experimental plots. Recently it has been possible to fill this marshy area more rapidly and so a routine examination of the area was made for unusual plants which might be effected by changes of soil level or of drainage or which should be preserved from damage through the filling process. This survey re- vealed a small population of Chelidonium majus with single- and double-flowered plants intermixed. We suspect that these are the descendants of the plants Sax had studied and had discarded in the swampy area. A set of specimens was prepared to record the presence of the double-flowered form in 1966 and in the process of preparing labels for these specimens a problem of nomenclature of the form and a conflict of names was discovered. 179 180 Rhodora [Vol. 69 References to a double-flowered form of Chelidonium majus in the literature are numerous but specimens of such forms are exceedingly rare in herbaria. J. Gilbert in 1798 (Histoire des Plantes de Europe 1: 167. 1798) acknowl- edged that in cultivation a double-flowered form was pro- duced. De Candolle (Regni Vegetabile Systema Naturale 2: 99. 1821) refers to both simple and double forms in his treatment of Chelidonium majus. Nicholson in the first edition of the Dictionary of Gardening (1: 310. 1884) noted that there was “also a double-flowered form." Henslow (British Wild. Flowers in Their Natural Colours and. Form 18. 1910) stated “C. grandiflorum from Dahuria, C. lacinia- tum from South Europe, and the double form of our own wild species are cultivated as garden plants." However, the treatment of the genus in Flora Ewropaea (1: 251. 1964) makes no mention of the double-flowered form, and no name for this appears in the contemporary European floras examined. It has been possible to check the herbaria at Edinburgh, Kew, and the British Museum and no double- flowered specimens were found. The date of the introduction of the double-flowered form into the United States is not recorded and the extent of its distribution is unknown. None of the editions of Gray's Manual of Botany or of the Britton and Brown Illustrated Flora or the local floras examined makes any mention of the existence of a double-flowered Celandine Poppy. There are only two such specimens in the herbarium of the New England Botanical Club: one from Milton, Massachusetts collected in 1917, and one from Fairfield, Connecticut col- lected in 1937. No specimens have been found in herbaria at New York, Washington, or Philadelphia. Bailey's Manual of Cultivated. Plants (p. 428. 1949) calls Chelidonium a monotypic genus in temperate Europe and Asia comprising a biennial or perennial herb, often in old gardens, sometimes double-flowered. An inquiry to the Bailey Hortorium was answered with the information that several European nurseries offered seed of a “Chelidonium 1967] Celandine — Howard 181 majus florepleno” and that a varietal name had been pub- lished in 1955. I am grateful to the staff at the Bailey Hor- torium for the suggestions which prompted a search of various nursery catalogues and seed lists of botanical gar- dens. A seed list from the botanical garden at Antwerp for example (1963 Delectus Sporarum Seminum Fructuum, Hortus Botanicus Antverpiensis, p. 12. 1963) shows that the name “Chelidonium majus L. flore-pleno" is still in use in the 1963 edition, although no valid publication of the name can be found. The description of the double-flowered variety was pub- ished by André Lawalrée in 1955 (Brussels Bull. Jard. Bot. de l'Etat B 25: 409, 410. 1955) with the name Chelidonium majus var. pleniflorum Lawalrée. He indicated the variety had up to 20 petals with 7-14 stamens. The holotype in the Brussels herbarium was collected by J. Tijskens s. n. in May 1953 in the district Campinien, Bonheiden along the route to Malines above the village of Kiwit in Belgium. A search of the scanty European collections in the Gray Herbarium then revealed one double-flowered specimen col- lected by Keinheinz near Potsdam in 1962. This discovery prompted an examination of Parey's Blumengürtnerei where (2nd ed. 1: 679. 1958) the treatment of Chelidonium prepared by C. R. Jelitto used the name C. majus var. plénum Wehrh. a reference we found repeated in the seed list of the Leiden Botanical Garden issued in 1954 (Hortus Botanicus Academicus Lugduno-Batavus, Delectus Semi- num 19. 1954). Finally, the description of the double- flowered form by H. R. Wehrhahn was found in Die Gar- tenstauden (5: 464. 1930) as Chelidonium majus L. “var. plenum hort. Mit gefüllten Blüten. Merkwiirdig ist, dasz alle (auch die hier nicht genannten Formen) vóllig samen- beständig sind." This tortuous search for a name, an authority and a de- scription comprising valid publication was given in detail to indicate the significance and the close relationship of horticultural and botanical literature. Perhaps the double- 182 Rhodora [Vol. 69 flowered Chelidonium majus existing as a cultivated plant should be regarded as a cultivar and named as Chelidonium majus ‘Plenum.’ Its existence as a weed persisting in a dump area indicates it is also a part of the local flora of Massachusetts and as such its correct name had to be sought and found in the literature of cultivated plants in Europe. The correct name for a double-flowered Celandine Poppy is: Chelidonium majus L. var. plenum Wehrhahn, Die Gar- tenstauden 5: 464, 1930. C. majus var. pleniflorum Lawalrée, Brussels Bull. Jard. Bot, de l'Etat 25: 409, 410. 1955. C. majus L. ‘Flore-pleno’ Hort. lacking known valid pub- lication. Specimens seen. United States. Massachusetts: Jamaica Plain, R. A. Howard 16101 (NEBC), Milton, N. T. T. Kidder s. n. (NEBC). Connecticut: Fairfield, E. H. Eames 11829 (NEBC). The persistence of the collection from which Sax may have obtained his material for a period of 50 years led us to re-examine the nature of the double-flowered form. Sax's breeding experiments were not repeated but a simple count was made of floral parts of mature double flowers selected from 190 different plants. These showed petals ranging in number from 12 to 22 in individual flowers, averaging 18. 19, while the stamens of the same flowers ranged in num- ber from 6 to 19 and averaged 11.52. The average total number of petals and stamens in the double forms was 29.71. From his breeding experiments Sax had concluded that the double-flowered form was recessive and the varia- tion in petal number and of stamen number was greater than in the single-flowered forms. However, he found that the sum of the petal number and the stamen number was about the same in all individuals whether single- or double- flowered. The mean sum of the petals and the stamens of the double-flowered forms in Sax's report was 29.49. 1967] Celandine — Howard 183 There appears to have been little change in the population of the double-flowered Chelidonium when left alone for 50 years. RICHARD A. HOWARD ARNOLD ARBORETUM JAMAICA PLAIN, MASS. PALYNOLOGICAL NOTES ON AMERICAN SPECIES OF HELIANTHEMUM (CISTACEAE) ROBERT L. WILBUR AND JAMES D. PERRY" Heydacker (1963) presented palynological evidence pur- portedly demonstrating that certain American species of the cistaceous genus Helianthemum Mill. sens. lat. were actually generically distinct and should be considered in- stead as members of the New World segregate Crocanthe- mum Spach sens. str. Since one of us (Daoud and Wilbur, 1965) recently concluded otherwise in connection with a revision of the North American species, an investigation of the pollen of the twenty species recognized from that area and H. brasiliense Lam. from eastern South America was undertaken in an effort to evaluate Heydacker's contention. Crocanthemum was originally described by Spach (1836) to include those American species which he thought pos- sessed only petaliferous flowers. Those American species which he believed to have both petaliferous and apetalous flowers were assigned to his genus Heteromeris while spe- cies thought to be characterized by only apetalous flowers comprised his genus Taeniostema. These three New World genera were judged to be clearly separable from the Old World genera in that they supposedly possessed the combi- nation of alternate, estipulate leaves and short, erect styles. Many authors have treated all the American representa- tives of Helianthemum sens. lat. as members of the segre- gate Crocanthemum sens. lat. characterized by the above mentioned alternate, estipulate leaves and short, straight styles. Britton (Ill. Fl. N. U.S. 2: 539-541. 1913), Bicknell (Bull. Torrey Club 40: 613-616. 1913), Janchen (Österr. Bot. Zeitschr. 71: 266-270. 1920 and Engler and Prantl’s Nat. Pflanzenfam. 2 Aufl. 21: 289-313. 1925) and Barnhart (Small’s Man. Se. Fl. 878-880. 1933) have all employed Crocanthemum in this broadened concept rather than in *Work on this paper was carried out with the support of the Na- tional Science Foundation Grant G18799. 184 1967] Palynological Notes — Wilbur & Perry 185 the original restricted usage for species supposedly charac- terized by the presence of only petaliferous flowers. Fernald (Rhodora 43: 609-615. 1941) has presented the most careful appraisal of the generic status of Crocanthe- mum known to us. He concluded that there was no justifica- tion for the generic segregation of the American species since the morphological features examined supposedly char- acterizing the genera prove highly inconstant. Such a failing would also cast doubt upon the naturalness of the same groups treated as subgenera or sections. The generic and infrageneric classification of the Cistaceae is not pres- ently in a satisfactory state and doubtless can only be im- proved by the acquisition of comparative data from numerous botanical subdisciplines. Cytological data on this family are accumulating from the Mediterranean area and Europe and offers promise of providing another useful guide for supraspecific classification within the Cistaceae when a larger percentage of the species have been examined. Only one count has been reported for the approximately forty-five species of the family found in this hemisphere. It seems likely, judging from the limited data already ob- tained, that systematically meaningful clues to relationship may be derived from detailed palynological studies but in our opinion reclassification is unwarranted until a much broader survey of the family has been undertaken. Brizicky (1964), like Fernald, concluded that there was not yet sufficient evidence to support the generic segrega- tion of the American species but noted Heydacker's paly- nological evidence supposedly favoring the retention of apparently all American species within a special section ["tribu"] of Helianthemum except for H. carolinianum (Walt.) Michx. and H. brasiliense (Lam.) Pers. The sup- posedly very distinctive pollen of these two species con- vinced her of the validity of Crocanthemum's claim for generic recognition. Heydacker (1963) studied the pollen of 55 species of the Cistaceae and concluded that the palynological evidence supported the recognition of the following seven genera: 186 Rhodora [Vol. 69 Cistus L., Crocanthemum Spach, Fumana (Dun.) Spach, Halimium (Dun.) Spach, Helianthemum Mill., Hudsonia L. and Lechea L. Unfortunately, Heydacker did not name more than a quarter of the species examined and it is not possible to learn from her account how many species in each genus or how many specimens of each species were exam- ined, Surprisingly enough, Heydacker concluded from paly- nological evidence that the small Old World genus Tubera- ria, Which is often included within Helianthemum itself, is not worthy of generic recognition and on the basis of its pollen might better be included within Cistus rather than Helianthemum. This conclusion is startlingly in contradic- tion to the evidence provided by the morphology of the flower and fruit. In contrast Jean and Pons (1963) con- cluded that palynology supported generic status for Tubera- ria but demonstrated it to be more similar to Helianthemum than to Cistus. The pollen of Helianthemum was said by Heydacker to be more elongate than that of the other cistaceous genera and also to be the largest in the family (approximately 80 p long). She found that the American representatives of Helianthemum examined did not conform with their Euro- pean congeners, for grains of the former often were equi- axial or even shorter than wide. The ratio of polar to equatorial measurements (P/E) of the European repre- sentatives was reported to be from 1.2-1.6. The exine proved variable but in the pictured example was striate and rather finely ridged at the base of the elevated striations. The pollen of Crocanthemum, of which both C. brasiliense (Lam.) Spach and C. carolinianum (Walt.) Spach were examined, was said to be very distinctive. She reported the pollen to be subspherical or slightly short-axial with a P/E ratio of about 0.8. The thickened exine was reportedly reticulate with rather thick or enlarged bases to the elevated network. The pollen for our study was prepared for examination by the acetolysis method of Erdtman (1960) from anthers carefully excised from chasmogamous flowers. The prepa- 1967] Palynological Notes — Wilbur & Perry 187 Plate 1353. Figs. 1-4. Fig. 1. Helianthemum brasiliense. Fig. 2. H. corymbosum. Fig. 3. H. arenicola. Fig. 4. H. carolinianum. [Vol. 69 Rhodora 188 Plate 1354. CUM fU e wi -— E] n ianthemum. concolor. Hel se S. . T. H. nutan o Fi canadense. 6. H. ig. Fig. . greenei, H 8 1967] Palynological Notes — Wilbur & Perry 189 ration of the slides was made by Mrs. Shirlee Cavaliere and Mrs. Myra Stewart of Professor D. A. Livingstone's labora- tory. We are most appreciative of their assistance. The advice and aid of Dr. Livingstone were most helpful but any mistakes or errors contained in this paper are to be attributed to the authors alone. A Spencer microscope with 10 X oculars and a 44 X objective was used for measurements which were made to the nearest optical line (1 line = 1.52 a). Data for Table 1 were obtained by measuring at least 40 fully expanded pol- len grains from each collection. In the case of Helianthe- mum chihuahuense, grains allowing a polar diameter measurement were scarce; therefore fewer polar than equatorial diameter measurements were taken. Pollen-wall ornamentations were compared by photographs at 1500 X Collectors and collection numbers are given in the table. The information presented in Table 1 in our opinion offers no palynological support for those who contend that the American species are divisible into two or more genera. The ratio of polar/equatorial measurements of H. caro- linianum and H. brasiliense are matched by the P/E ratio of numerous other American species. Variability of the P/E ratio of H. carolinianum and H. brasiliense was great- er than that indicated by Heydacker, ranging from 0.7-1.2 in contrast to the 0.8 reported in her paper. All of the American species studied with a P/E ratio of 0.6-1.2 were lower than or equalled only the lowermost range of that reported for the Old World species of 1.2-1.6. These data, if substantiated with a more meaningful sample, would give only slight additional support to those who have argued that the American species as a group are generically distinct from their Old World relatives. The sculpturing of the exine of H. brasiliense and H. carolinianum fails also to support the separation of those two species from the re- maining American species. Although H. brasiliense pos- sessed the coarsest sculpturing of any American species, the sculpturing of H. carolinianum was matched by that of several other American species. We cannot confirm Hey- TILO |roeorp, $6 |ccop TT-8'0 (|greeep| 6T | rgr LLFS0 roeezp 0$ | BOF OTLO (LLES) ge | Zor ELLO |LLCCSP eg | gIC TTS0 |97?€980 TZ LTS UTELO 899919 ZP | w9c TT-L0 — (LL6-e6p | Te | 9'0€ TT-L0 — (€80618| 87 | HLE CULO Vesorr Le | rer asuey ('Ae( IS uto] OLLV3 4/d, (7) NOISNAWIG 'IVDIIOLV 21653 | OFE-F'9E 9'8*-0 TF | 0 L-0 TY OTT-OIP | 9'rc-0'8€ | 9'ye-e' Gy 9'8f-Y'68 Y 98-708 9 TS-F'6E © asuRy EZ 6 OF 6T OVP 8I eV 80 GOP LAS SLE 0% 0'0€ GS 6'06 | VS | O'P? 6T THE | se Tro Aad 2g [ULIN 9je[norjod A[our,q 9je[norje1 Ápour,q 9je[noT39.1 AjasrBog IJe MIJ A[939.19pO]X 9je[noI39.I Á[9s.1€0^) 9]9LI3S A[93€10qv]^q peuesxoru, Ssodpri 'oje[norjo.r A[9S1907) a}B[NoYar Á[9urq 9j9[norjo1 ÁA[out,q 91e[nor3a.r K[es.1807) (7) NOISNAWIG HV'IlOd. ƏNIYALATAOS | 2199 PSU 'ODNIXƏW :'S2€AA "S L123[noO ELIZ tediey “ey “09 WELE :'xuorpgy wnsoquié.100 096cI uUojurH 'oorxo]y 'uvuroo[€0;) :€3oj1() (Əy) xo[oouoo €$66 ƏSOY 'oo-Xo]q foS|eprH US)EM "S asuenyenyiys 180Z Jodie ‘ery “OD uoAed0q “XU CULM) unuerur[o.reo JAeq[Iyy SSEW ‘PAY ould H'XYPIW (“I) əsuəpreuvə LSS 1949H “Avnsnig 'ooptAeQuo]q 9p ojueurjrede([ *o.r197) SIJ (uv) osuer[mse4q uoj[aMOu ‘SSEW J1938q9AA UI NPU 909 TƏUYLYIS 'OINXƏW ‘1soj0g SINJ ueg :'[suoH wnəzuəJre OLZ8 SIA `L “OD Uurpp[ueiqp sudeyg v[ooruo. ie ALITVOOT ANV SWIOSdS ‘H 'H t W'IHVIL TT-8°0 @T-L'0 TT-9°0 Urso ETSO e T-L'0 T'T-8°0 e r80 6 r80 0'T-6'0 TT8°0 9 TS-GOP 98P-O'TV DIST 9'T67-G'6F 9'8y-Y' 66 ree OLT 9'TG87-qQ'6T 0'Iv-6' VE OTP-6PE 9° 1S-9'°8P 9'v9-q'Gy 8c VV OPV 88r Ly S'er T8F L'9F y'88 p88 L’6¥ roc ecp-r'66 9'8v-Y' 96 0'LY-Y 9E 9'8y-Y' 66 cCer-Y 96 9'sr-V'98 9'sr-v' 65 GOR vss Col Y $8 9°8P-0' TY 9'8v-0' TP 9j1e[nor39. K[93€.I9po]A 9jv[norjo.1 Aout, q 9]e[norjo. AJOL IPON əƏyemə17ə1 ÁA[eur,q aye[Notyat ABUL aye[Noyet Ajour 9je[n2rjo.r ÁA[939.t9po]Ay suorje[nortja. pesue.rrg Ajasuep A194 ‘PULA 9je[not39.t ÁA[91919pO]A 91e[not39.t ÁK[939I9pO]A 9j?[nor394 Á[939 19po]A €*10uOS “WN 67891 zni) Burg yoynys U8 xe '3eT] ( TROOT dieu ^D “a 'sexaL “OD Sie :'udeu) umuttd.109d L8ZST peuiog P pieapooM 'seyW “OD e[qejsuriegq 8056 Sutiqy ‘JED fejuopy pa :nw unrriedoos ‘H H 8889 SStmj) “BD *O2IX9]IA DUS)EA CS Te[2u rd YSsnVAdA, ‘ONIN *oosi[|ef' :"[sutay suezed GESOT UBIO “BIULOFITBD) e(egq :‘sepuelg 'S ‘J, Ssugjnu 0399 SSUnj) Bld “OD exe] Aug Wyse "uvungoH Jue) ‘puv[s] ssuIqoy Touoold 6TIT "e[gurojenz) : [Bund Ser) wnye19w0[s ("uxorg) umnsounp J9AI[O :YySing urmi[OoJIULIEUISOIX * p|uxorg 'ssepp 'pue[s] joxonjueN :'uxorg wnnsuidoid . STOE IUM 'H 'H rH H H 192 Rhodora [Vol. 69 dacker's claim that Crocanthemum is generically character- ized by enlarged bases of the exine columns since those of H. brasiliense were indistinctly, if at all, swollen and those of H. carolinianum were, in our material, clearly not en- larged basally. Swollen bases to the exine columns of a more pronounced degree were noted in H. mutans, H. greenei and H. georgianum. Electron microscopy, perhaps of the surface scanner type, would doubtless provide more convincing evidence but from what we have observed it would not appear that palynology offers much promise for guidance in the reclassification of the American species of Helianthemum s.l. We would conclude therefore that there is no palynological basis yet suggested which would support the separation of H. brasiliense and H. carolinianum from the other American species or for the maintenance of the generic segregate Crocanthemum. The pollen of the species of Helianthemum illustrated by Jean and Pons (1963) in their study of the Cistaceae of France was clearly striate and unlike the reticulate pattern exhibited by all American species examined in this study except H. canadense. The pollen of H. canadense was also conspicuously striate but unlike the European species illus- trated exhibits a very broken surface since the grooves and ridges in one area tend to run in a different direction than those in another. The pollen of H. canadense examined by us was by far the most distinctive of the species examined but should doubtless be reinvestigated to determine whether its apparent uniqueness is consistent or an artifact in preparation. In passing we would like to comment upon Brizicky's (1964) transfer of Grosser's Halimium section Spartioides to Helianthemum apparently with the same circumscription and the same three species as it originally contained. It is not believed that the following statement quoted from Abrams (1951) constitutes valid publication of the transfer of the section from either Halimium (Pflanzenreich IV. 193 (Heft 14) 33. 1903) or Crocanthemum (Nat. Pflanzenfam. ed. 2. 21: 305. 1925): “The Pacific States species belong 1967] Palynological Notes — Wilbur & Perry 193 to the section Spartioides, characterized by the broom-like habit and the absence of cleistogamous flowers." To attri- bute this as constituting a new combination by Abrams would be in our opinion contrary to Article 33 of the Code. The two sections of the exclusively American subgenus Lechioides (Dunal) Reichenb. were contrasted as follows: Flowers usually dimorphic; eastern N. and S. America, the West Indies, Mexico, Central America and Chile ...... Sect. Lechioides Flowers homomorphic, broomlike subshrubs; western North America (California and Baja California) and Chile ..............- Sect. Spartioides (Grosser) Brizicky desstosesosesosossosssosossscsosseccsoso toten This infrageneric classification does not seem satisfactory since there actually are no characters mentioned by Brizicky or known to us which bind the species together. The alleged broomlike habit of the species of Section Spartioides actually does not even cover all of the extremely diverse growth forms that are included within H. scoparium. Homomorphic species which apparently would be included in section Lechioides are the Mexican and non-spartioid H. nutans T.S. Brandeg. and H. patens Hemsl. Fewer than one per cent of the specimens of H. carolinianum (Walt.) Michx. of section Lechioides seen had any cleistogamous flowers. It seems to us most unlikely that H. scoparium and H. greenei will be separated as a group from all the other North American species when enough is known to establish a biologically sound infrageneric classification. In any event the formation of an infrageneric classification might better wait for the discovery of features which bind species into definable groups. DEPARTMENT OF BOTANY DUKE UNIVERSITY DURHAM, NORTH CAROLINA REFERENCES CITED ABRAMS, LEROY. 1951. ‘Helianthemum in Illustrated Flora of the Pacific States 3: 122-123. BrizicKy, GEORGE K. 1964. The genera of Cistaceae in the south- eastern United States. Jour. Arn, Arb. 45: 346-357. 194 Rhodora [Vol. 69 Daoup, H. S. and Rorert L. WILBUR. 1965. A revision of the North American species of Helianthemum (Cistaceae). Rhodora 67: 63-82, 201-216, 255-312. ERDTMAN, G. 1960. The acetolysis method. A revised description. Svensk Bot. Tidskr. 54: 561-564. HEYDACKER, FRANCOISE. 1963. Les types polliniques dans la famille des Cistaceae. Pollen et Spore 5: 41-49. JEAN, MARIE-THERESE and ARMAND Pons. 1963. Contribution a étude palynologique des Cistacées de la flore de France. Ann. Sci. Nat. Bot. ser. 12. 4: 159-204. A SPURLESS FORM OF AQUILEGIA CANADENSIS L. ROBERT B. LIVINGSTON An area with great diversity of habitat and supporting a large population of columbines has been noted in an exten- sive juniper pasture in East Deerfield, Massachusetts. In May, 1962, I observed Aquilegia canadensis forma flaviflora Britton ex House in the pasture, and in 1963 I collected the salmon colored, forma Phippenii Munz. Further it was found that columbines from the area attained markedly different heights when transplanted into a relatively uni- form home garden. In May, 1964, I again visited the Deerfield pasture and found a plant whose flowers were completely spurless. Munz (1946) reported that: “Plants with spurless flowers have been found in Boylston, Massachusetts, by school chil- dren, May 14, 1913, and at St. Johnsbury, Vermont, May 20, 1896, Francis P. Perry.” Since spurless forms of the red columbine have been noted at least twice previously in New England it appears reasonable to designate this plant as a new form. Aquilegia canadensis var. canadensis forma ecalcarata, Livingston forma nov. Similar to Aquilegia canadensis L. proper with the follow- ing exceptions: Flowers pendant; petals yellow, 7-9 mm. long, 4-5 mm. wide, oblong-obtuse, spurless, spur replaced with small cupule near base, cupule 1.5-2 mm. deep, interior red; stamens about 1 cm. long, exceeding the twisted styles; basal leaves mostly biternate, deep green and glabrous above, light green and glaucous beneath, usually with node-like swellings at or below mid petiole. Forma a varietate typica differens his sequentibus: Flores penden- tes; petala lutea, 7-9 mm long., 4-5 mm. lat., oblonga-obtusa, sine calcari, pro quo cupula parva prope basim substituta est, cupula 1.5- 2 mm. profundite, intra rubra; stamina c. 1 cm. long., stylos tortos excedentia; folia basalia plerumque biternata, maxime viridia gla- braque supra, dilute viridia glaucaque infra, inflatione nodiformi ad aut infra partem petiolo mediam plerumque praedita. 195 196 Rhodora [Vol. 69 cy es i ve |i M l | . | | Figure 1. Aquilegia canadensis f. ecalcarata. Habit, flower, and all parts except staminodia 2/3 X; staminodia 1 1/3 X. 1967] Aquilegia — Livingston 197 TYPE in the Herbarium of the University of Massachu- setts, collected in East Deerfield, Hampshire County, Massa- chusetts, adjacent to exposed field stone in rocky pasture, about 14, mile west of River Road, 114 miles north of the Sunderland bridge where it crosses the Connecticut River, Livingstone 3291. Original plant collected May 20, 1964, and propagated vegetatively in greenhouse from vegetative offshoots. Type is original plant. One clonal offshoot trans- planted back to original site, June, 1965, another retained in home garden as source of seed and to provide additional clonal material. DEPARTMENT OF BOTANY UNIVERSITY OF MASSACHUSETTS, AMHERST LITERATURE CITED Munz, PHILIP A. 1946. Aquilegia The cultivated and wild colum- bines. Gentes Herbarum Vol. 7, Fasc. 1: 1-150. A NEW STATUS FOR AN EASTERN NORTH AMERICAN SCIRPUS ALFRED E. SCHUYLER! Among plants that Fernald originally described from southeastern Virginia were those he named Scirpus atro- virens var. flaccidifolius. Although he pointed out most of their unique characteristics in his original description (1938), other authors (Beetle, 1947; Gleason, 1952) have not given these plants any taxonomic recognition. They resemble plants of S. atrovirens Willd., S. hattorianus Mak., and S. georgianus Harp., but differ consistently from them in several characteristics and are recognized here as: Scirpus flaccidifolius (Fern.) Schuyl., stat. et comb, nov. Scirpus atrovirens var. flaccidifolius Fern., Rhodora 40: 396. 1938. (Fernald & Long 8109, GH; ISOTYPES, PH, US). Scirpus flaccidifolius has lax and reclining mature culms with inilorescences that lop over in contrast to the nearly upright mature culms of S. atrovirens, S. hattorianus, and S. georgianus (collectively referred to as related species in this discussion). The primary inflorescence rays become relatively long at maturity (about 30 cm in Fernald & Long 10544, GH) in comparison with those of related species. The lower leaf sheaths and blades are nearly smooth or some- what nodulose-septate, but not conspicuously so as is usually the case with S. atrovirens. The small glomerules of spike- lets (fig. 1) rarely contain more than 15 spikelets, and individual spikelets (fig. 2) are relatively wide (mostly 2-3 mm). Related species frequently have more than 15 spikelets per glomerule and consistently narrower spikelets. ‘I am grateful to Hans Wilkens for help in the field, maintaining transplants of our collections, and reading the manuscript; to Ruth Patrick and H. Radclyffe Roberts for reading the manuscript; and to Morton Gilbert and Jay Sacks for their photographic services. *The distinctions of this species from S. atrovirens and S. georgianus were recently pointed out by myself (1967). 198 1967] Scirpus — Schuyler 199 we Figure 1. Inflorescence of S. flaccidifolius (Fernald & Long 10141, PH). Figure 2. Spikelet of S. flaccidifolius (Fernald & Long 10140, GH). Figure 3. Achene and bristles of S. flaccidifolius (Fernald & Long 10140, GH). 200 Rhodora [Vol. 69 The blackish scale color is similar to the scale color of S. hattoriamus but differs from the usually brownish scale color of S. atrovirens and S, georgianus. Many of the scales are slightly mucronate, particularly those from lower por- tions of the spikelets, whereas those of related species are all mucronate. The bristles (fig. 3) are frequently longer than the achenes, resembling S. atrovirens in this respect, but differing from S. hattorianus which usually has bristles shorter than to about equaling the achenes. The presence of 6 bristles distinguishes S. flaccidifolius from S. georgia- nus, which either lacks bristles or has up to 3 short ones. The achenes (fig. 3) are mostly 1-1.2 mm long and compare closely in length with those of S. atrovirens but are gen- erally longer than those of S. hattorianus and S. georgianus. The chromosome number of n = 27 was comparatively easy to determine from meiotic material of S. flaccidifolius collected in Northampton Co., North Carolina (Schuyler & Wilkens 3858, PH). This number differs from the 28 units observed in meiotic figures from plants of S. hattorianus collected in Berks Co. (Wilkens 10270, PH) and Northamp- ton Co., Pennsylvania (my no, 3735, PH), and Cattaraugus Co., New York (my no. 3443, PH?). It has been difficult to obtain results from meiotic material of S. atrovirens (my experience being comparable to that reported by Hicks, 1928) and no definite number can be reported here. In meiotie material of S. georgianus, three different numbers, 25 (New Jersey: Burlington Co., my no. 3863, PH), 26 (New Jersey: Burlington Co., my no. 3846, PH), and 27 (Pennsylvania: Lehigh Co., my no. 3852, PH), have been observed and obviously indicate the need for further cyto- logical investigation of this species. While collecting in southeastern Virginia and northeast- ern North Carolina during the spring of 1966, Hans Wil- kens and I found Scirpus flaccidifolius growing in or near previously reported localities (Fernald, 1938 & 1940), and also found it at other places near Fontaine Creek farther 'Although previously cited as S. atrovirens (Schuyler, 1964), sub- sequent study has shown that the plant is S. hattorianus. 1967] Scirpus — Schuyler 201 downstream from where it had been collected previously. Although all of the localities we found were in the vicinity of large streams (e.g., Nottoway River, Meherrin River, and Fontaine Creek), none were in the low areas close to the margins of these streams; instead they were in some- what higher areas, though still in the wet, swampy bottom- lands associated with these streams. We further noticed that S. flaccidifolius grew abundantly where there was evidence of human disturbance such as clearings for power lines near the Nottoway River and roadside clearings in bottomlands near Fontaine Creek. However it was also observed growing in a more natural area on the east side of the Meherrin River near Haley's Bridge along a small meandering stream through bottomland woods south of highway no. 730, It is likely that further collecting will extend the range of S. flaccidifolius, but at the present time only the following collections are known: VIRGINIA: sussEX CO. SW of Homeville, bottomland swamp, Nottoway River, Fernald & Long 10141 (GH, PH) ; E of the Nottoway River, in wet clearing for power lines along highway no. 40, Schuyler 3855 (PH). SOUTHAMPTON CO.: near Haley’s Bridge, wooded alluvial bottomland of Meherrin River, Fernald & Long 8109 (GH, PH, US). GREENSVILLE CO.: SE of Taylor's Millpond, wooded bottomland of Fontaine Creek, Fernald & Long 10140, 10544 (GH, PH); SE of Taylor’s Millpond, along road near bridge over Fontaine Creek, Schuyler & Wilkens 3859 (PH). NORTH CAROLINA: NORTHAMPTON co.: 1.7 mi WNW of Margarettsville, along road S of Fontaine Creek, Schuyler & Wilkens 3858 (PH); Rich Square, along stream, Blom- quist (PH). ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA PHILADELPHIA, PENNSYLVANIA LITERATURE CITED BEETLE, A. A. 1947. Scirpus. North American Flora 18: 481-504. FERNALD, M. L. 1938. Noteworthy Plants of Southeastern Virginia. Rhodora 40: 364-424, 1940. A Century of Additions to the Flora of Vir- ginia. Rhodora 42: 419-498. GLEASON, H. A. 1952. The New Britton and Brown Illustrated Flora of the Northeastern United States and Adjacent Canada. The New York Botanical Garden, New York. 3 v. 202 Rhodora [Vol. 69 Hicks, G. C. 1928. Chromosome Studies in the Cyperaceae, with Special Reference to Scirpus. Bot. Gaz. 86: 295-317. SCHUYLER, A. E. 1964. Notes on Five Species of Scirpus in Eastern North America. Bartonia No. 33: 1-6. 1967. Scirpus hattorianus in North America. Not. Nat. No. 398: 1-5. PHYSALIS IN MEXICO, CENTRAL AMERICA AND THE WEST INDIES U. T. WATERFALL Continued from Vol. 69, No. 777 33. Physalis tehuacanensis Waterfall, sp. nov. Herba, perennis, 22-27 cm alta; trichomatibus articulatis, partim capitato-glanduliferis, ad 1.5-2 mm longis; foliis ovatis vel deltoideo- ovatis, inaequaliter magno-dentatis vel sinuato-dentatis, principalibus 12-30 mm longis et 12-30 mm latis, petiolis 12-35 mm longis; calycibus floriferis 5-6 mm longis et 4-5 mm latis, lobis lanceolatis vel ovato- lanceolatis 2-2.5 mm longis; pedicellis 4-6 mm longis; corollis luteis, immaculatis, 8-9 mm longis et 10-11 mm latis; antheris luteis, 2.7- 3.5 mm longis; filamentis 1.5-3 mm longis; calycibus fructiferis 15-18 mm longis et 14-16 mm latis; pedicellis 8-10 mm longis; baccis 10-11 mm latis. TYPE: MEXICO: PUEBLA: Smith, Peterson & Tejeda 3992 (F), Isotype (US), Tehuacan city dump, July 20, 1961. This species is characterized by the long, coarse, multi- cellular hairs, intermixed with shorter ones, many of the trichomes tipped with reddish-brown glands, by the coarse- ly and irregularly dentate leaves, with the hairs of the blade appressed, and especially abundant on the veins, those on the petioles spreading, by the yellow anthers and the yellowish, immaculate (but possibly faded) corolla. It is known only from the type locality in the state of Puebla. 34. Physalis microphysa Gray, Proc. Amer. Acad. Arts & Sci. 21: 402. 1886; P. campanulata T. S. Brandegee, Plantae Mexicanae Pur- pusianae, IV. Univ. Calif. Publ. Bot. 4: 278. 1912. Plant herbaceous, several-stemmed from a woody base, the stems 15-60 em long; herbage rather densely covered with varying lengths of flat, jointed, spreading, tapered hairs, and some capitate-glandu- lar ones, up to 2 mm long, the longer ones often, but not always, the least abundant; leaf blades reniform to ovate or subhastate, the principal ones 12-20 mm long and 10-23 mm wide, rarely smaller, the widest, reniform ones on the lower part of the stems; margins entire, or each side with 1-3 teeth, irregular in shape and size; vesti- ture of blades more or less appressed, seldom as long as that of the stems; petioles usually 5-12 mm long, but lower ones up to 6 cm long, abruptly expanding onto the blade; flowering calyx bristly hairy, 4-6 mm long and 4-6 mm wide at base of lobes, divided one-half 203 204 Rhodora [Vol. 69 to two-thirds of its length into lanceolate, slightly acuminate, lobes; corolla white with a slight yellowish or greenish tinge, apparently immaculate, or slightly light-green maculate, 8-10 mm long and 12-20 mm wide when rotate-expanded, more or less pentangular in outline, on pedicels ca. 2 mm long; anthers yellow, ovate to ovate-oblong, 1-1.5 mm long, on slender glabrous filaments 2-3 mm long; fruiting calyx teretish, ovate-oblong to campanulate, 9-12 mm long and 8-12 mm wide, open to open-flaring apically; fruit with thin, dry pericarp, 3-7 seeds in each locule, sessile in the calyx. SELECTED COLLECTIONS. MEXICO: CHIHUAHUA: limestone ledges, Santa Eulalia Mts, Aug. 14, 1885, Pringle 317 (Type: GH, Isotypes: BR, F, OKLA, UC, US, VT) ; limestone, Canyon del Rayo, north- ern end of Sierra del Diablo, July 25-29, 1941, Stewart 917 (GH); COAHUILA: near Saltillo, June 1, 1947, Hinton 16703 (US); arroyo banks, Canyon del Pajarito, Sierra de la Madera, Cuatro Cienegas, Sept. 6, 1939, Muller 3158 (UC); SAN LUIS POTOSÍ: Purpus 53123 (Type of P. campanulata: UC, Isotypes: F, GH, US). 35. Physalis parvianthera Waterfall, sp. nov. Planta herbacea, 25-40 cm alta; caulibus villosis, pilis articulatis plus minusve viscidis; foliis deltoideo-ovatis vel ovatis vel lanceolato- ovatis, plus minusve villosis, integerrimis vel paucidentatis, majori- bus 25-30 mm longis et 20-30 mm latis; petiolis 10-14 mm longis; calycibus floriferis villosis, 4-5 mm longis et 5-6 mm latis, dentibus ca. 1-1.5 mm longis, lanceolato-attenuatis; pedicellis 3-5 mm longis; corollis luteis, immaculatis, 7-10 mm longis et 15-20 mm latis; an- theris violaceis 1.5-1.8 mm longis, filamentis filiformibus; calycibus fructiferis teretibus, parvis, 8-9 mm longis et 5-7 mm latis; pedicellis fructiferis 6-8 mm longis. TYPE: Paray 1614 (MEXP) MORELOS: Sierra de Chalchi, cerca de Tepcztlan, Aug. 1955. The small nearly terete fruiting calyx, immaculate yel- low corollas, small violet anthers and soft vestiture char- acterize P. parvianthera. Its nearest relative, P. microphysa, has larger yellow anthers, longer calyx lobes, a somewhat sparser, harsher vestiture and usually smaller leaves. The "small-anthered" reference is made in comparison with P. microphysa. 36. P. hederaefolia Gray, Proc. Amer, Acad. Arts & Sci. 10: 65. 1875. Herbaceous perennial, 1-5 dm tall, with a varying mixture of long jointed hairs and short trichomes, or with short hairs only, vestiture viscid or not, glandular-capitate or not, rarely partly stellate (a distinguishing characteristic of var. cordifolia), antrorse or spread- ing; leaf blades subreniform to ovate, rarely ovate-lanceolate, mar- 1967] Physalis — Waterfall 205 gins coarsely and irregularly salient-dentate to shallowly few-lobed, to undulate or entire, principal ones 20-50 (-80) mm long and 15-40 (-60) mm wide on petioles 10-45 mm long; flowering calyces 5-9 mm long and 4-6 mm wide at base of lobes, upper 2-4 mm divided into deltoid to lanceolate lobes, on pedicels 2-7 (-15) mm long; corolla maculate, obviously to obscurely so, 7-10 mm long and 10-17 mm wide, the limb reflexed when fully expanded; anthers usually yellow, sometimes with a violet tinge, 2-4 mm long, on slightly thickened filaments, 1-5 mm long; fruiting calyx 10-angled or 10-ribbed, vary- ingly vestite, 12-25 mm long and 9-17 mm wide on pedicels 5-10 mm long; berry 9-15 mm wide. 36a. var. hederaefolia; P. Palmeri Gray, Synoptic Flora 2(1): 235. 1888. Vestiture including long, jointed hairs. SELECTED COLLECTIONS. MEXICO: CHIHUAHUA: plowed field, Sitenapuchi, July 7, 1955, Pennington 510 (OKLA); COAHUILA: Hermanas, Apr. 20, 1939, Marsh 1629 (F, GH, OKLA); among shrubs on toboso-flat, 4 km w of San Juan, sw of Sierra de las Cruces, July 11, 1941, Stewart 810 (GH); along shallow watercourse in desert, 70 miles w of Saltillo, Aug. 7, 1957, Waterfall 13289 (F, MICH, OKL, OKLA, SMU, US); DURANGO: desert, 68 miles sw of Torreon, Aug. 11, 1959, Waterfall 15389 (OKLA, SMU, US); NUEVO LEON: rolling hills, 75 miles s of Laredo, Sept. 17, 1942, Gentry 6743 (GH, UC, US); SAN LUIS POTOSI: ex convalli San Luis Potosi, in 1877, Schaffner 401 (NY); SONORA: vicinity of Hermosillo, Mar. 8, 1910, Rose et al 12537 (GH, NY, US); canyon de la Bellota, Sierra de la Cabellera, Oct. 7-10, 1941, White 4693 (MICH); TAMAULIPAS: in canyon 4 km w of Miquihuana, Aug. 4, 1941, Stanford et al 674 (OKLA, US, not the GH duplicate). 36b. var. puberula Gray, Proc. Amer. Acad. Arts & Sci. 10: 65. 1875. Short-vestite; with short antrorse hairs only, or with varying amounts of glandular-capitate trichomes. SELECTED COLLECTIONS. MEXICO: CHIHUAHUA: rocky hills near Chihuahua, May 21, 1885, Pringle 15, (BM, BR, F, GH, MICH, NY, UC, US, VT); gravel desert with Larrea & Fouquieria, 30 miles s of Chihuahua, Aug. 8, 1956, Waterfall 12486 (F, MICH, OKLA, S, SMU, US); calcareous stony hillside, 11 miles n of Parral, Waterfall 12509 (F, MICH, OKLA, SMU, US); stony grassland, 148 miles s of Ciudad Juarez, Aug. 8, 1961, Waterfall 16092 (BM, F, GH, MICH, NY, OKLA, P, S, UC, US, VT); COAHUILA: Torreon, Oct. 13-20, 1898, Palmer 463 (BM, F, GH, MICH, NY, UC, US) ; limestone hillside, 10 miles n of Saltillo, Aug. 25, 1961, Waterfall 16625 (OKLA); DURANGO: Durango and vi- cinity, April to Nov. 1896, Palmer 165 (BM, F, GH, NY, UC, US) ; desert, 68 miles sw of Torreon, Aug. 11, 1959, Waterfall 15388 (F, OKLA, SMU, UC); stony hillside with grass, cacti and yuccas, 19 miles sw of Durango, Aug. 15, 1959, Waterfall 15523 (F, MICH, OKLA, SMU, US) ; 206 Rhodora [Vol. 69 NUEVO LEÓN: arroyo in desert between mountains, 22 miles w of Mon- terrey, Aug. 10, 1959, Waterfall 15317 (OKLA); SONORA: Puerto de los Aserraderos, Aug. 4-9, 1940, White 3224 (GH, MICH) ; ZACATECAS: near Concepcion del Oro, Aug. 11-14, 1904, Palmer 318 (GH, NY, US) ; stony mountainside, 2.2 miles se of Sombrerete, Aug. 15, 1959, Water- fall 15569 (OKLA, SMU). 36c. var. cordifolia (Gray) Waterfall, Rhodora 60: 158. 1958; P. Fendleri Gray var. cordifolia Gray, Synopt. Fl. N. Amer. 2(1): 395. 1878; P. Fendleri Gray, Proc. Amer. Acad. Arts & Sci. 10: 66. 1875. Varying quantities of stellate hairs, or flat branched trichomes, present, at least on the calyx lobes. COLLECTIONS SEEN. MEXICO: BAJA CALIFORNIA: 64 miles se of Ensenada, Aug. 26, 1960, Broder 350 (US). 37. Physalis glabra Bentham, Botany of the Voyage of the Sulphur, 39-40. 1844; P. hastata Rydberg, Mem. Torr. Bot. Cl. 4:363. 1895. Several-stemmed from a woody base, usually herbaceous, sometimes woody, 10-100 em long, glabrous or nearly so; leaf blades narrowly lanceolate to ovate, sometimes basally lobed or hastate, sometimes un- dulate-margined or few-toothed, rarely with 5-7 teeth on each margin; blades of principal leaves 20-40 mm long and 7-30 mm wide on petioles 12-25 mm long; flowering calyx 3-5 mm long and 3-5 mm wide on pedicels 10-20 mm long; corollas 7-15 mm long and 9-20 mm wide, re- flexed-rotate when fully open, usually yellowish, rarely drying with a bluish tinge (then resembling P. crassifolia var. versicolor), some- times with a tendency toward slightly darker spots near base of limb; pedicels 15-30 mm long; anthers yellow, 2-2.5 mm long on rather thick filaments 2-4 mm long; fruiting calyx 10-ribbed, 10-23 mm long; 11- 20 mm wide, half-filled by the berry; fruiting pedicels 17-30 mm long. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: islands off the coast and on mainland, San José del Cabo, March-June 1897, Anthony 353 (F, GH, OKLA, UC) ; Todos Santos, Jan. 28, 1890, Brande- gee 422 (Type of P. hastata Rydb.: uc, Isotype: GH); San José del Cabo, Sept. 14, 1930, Jones 27313 (BM, GH, NY, UC, US). 38. Physalis crassifolia Bentham, Botany of the Voyage of the Sulphur, 40. 1844; P. pedunculata Greene, Pittonia 1:268-269. 1889, non P. pedunculata Mart. & Gal. Bull. Acad. . . . Belgique 12:132. 1845; P. Greenei Vasey & Rose, Contr. U.S. Natl. Herb. 1:18. 1890; other synonymy under subordinate taxa. Perennial, usually herbaceous, sometimes woody-stemmed, some- times flowering and fruiting the first year of growth, hence apparent- ly annual, 10 cm to 4 dm tall; plant variously vestite with few to abundant, usually short hairs often tipped with viscid or oily glands; leaf blades ovate, margins entire or toothed, principal ones (1-) 2-6 (-10) em long and 1-5 (-7) em wide on petioles 1-4 (-6) cm long; flowering calyx rarely oblong, 5-9 mm long and 3-5 mm wide, the 1967] Physalis — Waterfall 207 deltoid to broadly lanceolate teeth 1-3 mm long; flowering pedicels usually 1-3 em long; corolla yellow and immaculate, or with slightly darkened spots, rarely bluish, or becoming bluish when dry, 8-30 mm long and 8-20 mm wide when fully expanded, at which time it is reflexed-rotate to funnelform, slightly sinuate-margined; few to num- erous trichomes within near top of corolla tube; anthers yellow, 2.5- 3.5 mm long, on filaments 4-14 mm long; fruiting calyx 10-ribbed, the ribs usually smooth, sometimes muriculate; fruiting calyx glab- rous to somewhat hairy, 1-6 cm long and 8-35 mm wide, on pedicels 5-40 mm long, usually much inflated around the berry. The disposition of individual collections in this complex, and the concept and delimitation of taxa, present problems not easily solved. P. crassifolia, sens. str., is not far re- moved from P. glabra Benth., here retained as a separate species. In the broad sense, some of the collections — those with pedicels shorter than usual — may approach P. heder- aefolia Gray. Much of the material considered here is perennial, some- times even becoming shrubby. Some of it reproduces as an annual, or produces flowers and fruit during its first year of growth. Since this difference has been used by some authors as a major one of value for the separation of spe- cies, it is not surprising to find annual material described as species under the names P. Greenei Vasey & Rose, P. flava Wiggins, and P. filipendula Brandegee. However, other tropical, or subtropical, plants are known to flower and fruit the first year, with herbaceous stems, then pro- ceed to become more or less woody, and persist as peren- nials. The author postulates the possibility of such an occurrence in this complex population of Baja California and the adjacent mainland of Mexico. Under such an in- terpretation, P. Greenei would largely be referred to var. crassifolia of the following treatment, but some of the ma- terial of var. versicolor and some of var. infundibularis, flowering the first year, would be similar. Therefore the name is not placed under varietal synonymy. Since P. Greenei Vasey & Rose is a substitute name for P. pedunculata Greene, non Mart. & Gal., 1845, l.c. supra, it might be pertinent here to comment that the type of P. pedunculata Mart. & Gal. (BR, specimen in flower only) ap- 208 Rhodora [Vol. 69 pears to belong to the concept of Brachistus Pringlei S. Wats., 1890. Other notable variation within the species-concept in- cludes the size, shape and color of the corolla, the amount and kind of vestiture, the length of the pedicels, and the size of the fruiting calyx. In much of the material the corolla is reflexed-rotate when fully expanded, although most of the corollas found on collections are not in this condition. I. M. Johnston pro- posed his var. infundibularis for plants with large, usually funnelform corollas, the vestiture varying greatly, Wiggins proposed the name P. flava for somewhat similar material with large yellow corollas and muriculate fruiting calyces. Some of the material referred to var. infundibularis may have slightly darkened maculations, only slightly differen- tiated from the surrounding yellow corolla, but the type material seems to have wholly yellow corollas, as is char- acteristic of the species-complex. Since P. flava seems to represent an extreme with a brighter yellow corolla and less glandular herbage than the type of var. infundibularis, but not of the whole population, it is referred to synonymy under that taxon. Many sheets of annual material, with smaller, rotate corollas, have been referred to var. crassi- folia, Rydberg proposed the name P. versicolor for material with the corolla drying bluish, the leaves thin and toothed, and with extra long pedicels and small calyces. Sometimes material with thick, entire leaves has the corolla drying bluish, and there is intergradation in pedicel-length and calyx-size. Nevertheless, the taxon is retained here in the varietal category. To its synonymy are referred P. versi- color var. microphylla Rydberg, based on second-growth material probably produced during unfavorable growing conditions, and P. sonorensis Standley, a phase usually with a yellow corolla, which has persisted until it has become shrubby. At first consideration P. filipendula Brandegee would ap- pear quite distinct by virtue of its large fruiting calyces. However, it does have the yellow anthers of the P. crassi- 1967] Physalis — Waterfall 209 folia complex. The corolla exhibits a tendency toward the development of darker maculations, but that occurs else- where in the complex. In other species-complexes (P. vir- giniana, sens. lat., for example) there have developed extremes with unusually large fruiting calyces. Nomen- clatural recognition has been given to one such development as P. virginiana forma macrophysa (Rydberg) Waterfall. It is postulated that a similar situation occurs in P. crassi- folia, sens. lat., the representatives being proposed later in this paper for nomenclatural recognition under the name var. crassifolia subvar. amplifolliatus (follis: bag, bellows), based on a type obviously related to subvar. crassifolia, forma muriculata by virtue of the muriculate ribs of the fruiting calyx but including material with smooth ribs such as is present in the type of P. crassifolia. 38a. var. crassifolia. P. cardiophylla Torrey, Bot. Mex. Bound. 153. 1859; P. crassifolia var. cardiophylla (Torr.) Gray, Synoptic Flora 2(1):235. 1878. Corolla 8-12 (15) mm long, reflexed-rotate when fully open; flower- ing calyces usually 4-6 mm long on pedicels from somewhat longer than, to 6-7 times the length of the flowering calyx. 38b. var. crassifolia forma crassifolia Ribs of fruiting calyx smooth. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: Cedros Island, March-June 1897, Anthony 306 (BM, F, GH, NY, UC, US); vol- canic rocks, San Quintin, April 7, 1936, Epling & Stewart sin. num. (F, NY, US); Valley of Palms, Apr. 8, 1882, Jones 3726 (BM, MICH, NY, UC, US); Magdalena Bay, Revillagigedo Islands, May 30, 1925, Mason 1952 (US); abundant on dunes, Bahia Magdalena, Solis 41 (US) ; sandy hills and flats 25 miles north of Punta Prieta, April 15, 1931, Wiggins 5362 ( GH, MICH, NY, UC, US); 9 miles east of San Ignacio, Nov. 3, 1946, Wiggins 11358 (GH, UC, US). SONORA: arroyos and milpas, San Bernardo, Rio Mayo, Feb. 23, 1935. Gentry 1344 (F, GH, UC); Papago Tanks, Nov. 17, 1907, MacDougal sin. num. (US); gravelly bed of arroyo de San Francisco, El Travelo, north of Alamos, Pennell 19503 (US). 38c. var. crassifolia forma muriculata (Greene) Waterfall, comb. et stat. nov., P. muriculata Greene, Bull. Calif. Acad. 1:209, 1885. Ribs of fruiting calyx more or less muriculate. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: Cape San Quentin, May 10, 1885, Greene 5109 (Type of P. muriculata: UC, Isotype: US) ; Cedros Island, April 1, 1897, Brandegee, sin. num. (UC) ; vicinity of Loreto, Nov. 14, 1962, Carter 4453 (vc); San Quintin, April 7, 1936, Epling & Stewart, sin num. (F, US); sandy wash 5 210 Rhodora [Vol. 69 miles south of Miller's Landing, April 17, 1931, Wiggins 5401 (GH, MICH, NY, UC, US); sandy arroyo 1.5 miles sw of La Paz, alt. 15 ft., Dec. 2, 1959, Wiggins 15723 (au, UC). 38d. var. crassifolia subvar. amplifolliata Waterfall, subvar. nov.; incl. P. filipendula Brandegee. Calycibus fructiferis magnis, inflatis, 30-35 mm longis et 25-30 mm latis, costis muriculatis vel levibus. The principal characteristic differentiating subvar. ampli- folliata from subvar. crassifolia is the large, inflated fruit- ing calyx, about one-fourth filled by the berry ; its ribs may be muriculate or not; the corollas may be immaculate, or with slightly contrasting maculations. TYPE: E. W. Nelson and E. A. Goldman 7405, Santa Anita, Jan. 11, 1906, (GH); Isotype (US). SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: Nelson and Goldman 7405, type cited above; San José del Cabo, Oct. 4, 1890, Brandegee sin num. (type of P. filipendula) (uc); San José del Cabo, Jan. 20, 1923, Jones 24393 (F, GH). 38e. var. versicolor (Rydberg) Waterfall, Rhod. 60:160. 1958; P. versicolor Rydb., Bull. Torr Bot. Club 22:307. 1895; P. versicolor var. microphylla Rydb., Bull. Torr. Bot. Club 22:307. 1895; P. genuicaulis A. Nels., Bot. Gaz. 47:430. 1909; P. sonorensis Standley, Field Mus. Publ. Bot. 22:102. 1940. Flowering calyces usually 3-4 mm long on pedicels 5-10 times their length; corollas similar to those of var. crassifolia, in shape and size, yellow to greenish-yellow, sometimes whitish toward their periphery, often drying with a bluish tinge; leaves usually dentate and thin. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: among loose rocks on beach, San Francisco Island, Gulf of California, May 30, 1921, Johnston 3952 ( GH, NY, UC, US) ; Los Angeles Bay, Nov. 22-Dec. 20, 1887, Palmer 561 (BM, GH, NY, UC, US) ; canyon about a mile from beach, 35 miles s of Punta Prieta, Oct. 29, 1946, Wiggins 11302 (GH, UC, US) ; CHIHUAHUA: in cultivated field in valley, Guasa- remos, Río Mayo, Aug. 5, 1936, Gentry 2358 (F); SINALOA: annual with thickened corm, wet wash with jungle growth, Imala, Nov. 29, 1939, Gentry 5090 (GH, MICH, NY); annual on rocky volcanic slope with coastal thorn forest, Cerros de Navachiste about Bahia Topolo- bampo, Sept. 26-30, 1954, Gentry 14305 (US); vicinity of Culiacán, April 21, 1910, Rose et al 14959 (US) ; SONORA: on malpais hill, Lower Sonoran thorn forest, Bachoco, 12 miles e of Cajeme, Feb. 20, 1937 (Type of P. sonorensis Standley), Gentry 3011 (F); Guaymas, July 1887, Palmer 94 (Type of P. versicolor var. microphylla Rydberg: NY; Isolectotypes: GH, UC, US); Guaymas in 1887, Palmer 622 (Type of P. versicolor Rydberg: Us, Isotype: GH); embankment in low area near river, 9 miles s of Navajoa, Aug. 17, 1956, Waterfall 12818 (F, 1967] Physalis — Waterfall 211 GH, MICH, NY, OKL, OKLA, P, SMU, US); mesquite-grassland, 9 miles w of La Angostura, Aug. 19, 1941, White 4018 (GH, MICH, NY, US); grassy Olneya-Prosopis-Cercidium plain, 11 miles e of Willard, be- tween Hermosillo and Colorado, Sept. 5, 1941, Wiggins & Rollins 298 (GH, MICH, NY, UC, US). 38f. var. infundibularis I. M. Johnston, Proc. Cal. Acad. Sci. ser. 4, 12:1156-1157. 1924; P. flava Wiggins, Contr. Dudley Herb. 3:73. 1940. Corolla usually funnelform, sometimes the limb more expanded, usually 15-20 mm long, uniformly yellowish, or with slightly con- trasting maculations. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: com- mon on silty flat forming dense oily clumps one to two ft. high, Angel de La Guardia Island, Gulf of California, June 30, 1921, Johnston 4203 (Type: GH, Isotype: US) ; Puerto Refugio, Jan. 26, 1940, Daw- son 1026 (MICH); southeast side of Cedros Island, Aug. 16, 1932, Howell 10692 ( GH, US); desert washes and slopes, 5-10 miles n of Catavina, Mar. 7, 1930 (GH, MICH, NY, UC, US). 39. Physalis Mimulus Waterfall, sp. nov. Planta herbacea, glabra; foliis lateovatis vel reniformibus, majori- bus 12-40 mm longis et 16-25 mm latis, paucis undulato-dentatis, petiolis 10-32 mm longis; calycibus floriferis 2-2.5 mm longis et ca. 2.5 mm latis; calycis lobis deltoideo-lanceolatis, 0.6-0.9 mm longis; pedicellis 3-6 mm longis; corollis maculatis, 3-4 mm longis et 4-6 mm latis; antheris violaceis, 1-1.3 mm longis, filamentis filiformibus; calycibus fructiferis 10 mm longis et 8-9 mm latis. TYPE: MEXICO: cor1iMA: Socorro Island, March 1889, C. H. Town- send, s.n. (US). The specific name alludes to the resemblance of post-an- thesis calyces, and of the broad leaf blades with few, ir- regular, low, undulate dentations, to certain species of Mimulus, 40. Physalis purpurea Wiggins, Contr. Dudley Herb. 3:74, pl. 19, figs. 9-10. 1940. Herbaceous, perennial or persisting annual, erect, branched, 25-75 cm tall, spreading-puberulent, especially on the stems, petioles and pedicels; leaf blades ovate to ovate-deltoid, bases sometimes oblique, rarely cordate, margins coarsely and irregularly dentate to sinuate- dentate; principal blades 2.5-5 cm long and 2-5 cm wide on petioles 2-5 cm long; flowering calyx 4-5.5 mm long and 3-4 mm wide at base of lobes, upper 1-2 mm divided into broadly lanceolate to deltoid lobes, on peduncles 17-25 mm long; corolla violet or purplish, the throat yellowish, hairy within near the top of the tube, reflexed-rotate when fully expanded, 7-11 mm long and 9-13 mm wide; anthers yellow, 212 Rhodora [Vol. 69 1.5-3 mm long, on filaments 4-6 mm long; fruiting-calyx 10-ribbed, 23-30 mm long and 17-24 mm wide, on pedicels 2-4 cm long; berry nearly spheric, 10-18 mm wide. This species is related to the P. crassifolia complex, ap- proaching it through P. crassifolia var. versicolor. It is probably as distinct in its direction of divergence as is P. glabra in a different direction. COLLECTIONS SEEN. MEXICO: sonora: Bahia San Carlos, Feb. 8, 1940, Dawson 1063 (F, MICH) ; bed of creek, 5 km w of Pilares de Nacozari, Nov. 19, 1939, Drouet et al 3680 (F); bases of cliffs e of Guaymas, Dec. 3, 1939, Drouet et al 3841 (F, US); bases of cliffs e of Guaymas, Dec. 3, 1939, Drouet et al 3854 (F, MICH, NY, UC,US) ; hills w of tannery in shade of shrubs on basaltic slopes, Oct. 21, 1939, Gentry 4671 (F, GH, MICH, NY, OKLA, UC, US). 41. Physalis longicaulis Waterfall, sp. nov. Planta herbacea, e radice lignosa fusiforme; caulibus procumbenti- bus glabris vel trichomatibus paucis antrorse adpressis; caulibus 3- 9 dm longis; foliis ovatis, principalibus 17-25 (-35) mm longis et 17-24 mm latis, magne et inaequaliter paucidentatis; foliis oppositis; internodis 6-9 cm longis; petiolis 5-15 mm longis; corollis maculatis, 7-10 mm longis et 8-12 mm latis; pedicellis 15-25 mm longis; antheris luteis, ca 2.5 mm longis; filamentis filiformibus, 3-5 mm longis; calyci- bus fructiferis 17 mm longis et 11 mm latis, pedicellis 30 mm longis; baccis ca. 6 mm latis. TYPE: C. H. & M. T. Mueller 135, Diente Canyon, mountains near Monterrey, NUEVO LEÓN, MEXICO July 1933 (F). This species is peculiar in its small, opposite, or seeming- ly opposite, leaves, the pairs separated by long internodes, and in the small fruiting calyx. 42. Physalis Rydbergii Fernald, Proc. Amer. Acad. Arts & Sci. 35:569. 1900. Stems herbaceous or woody below, several from a woody root, 15- 30 em tall, covered with short, fine, divergent multicellular hairs, some of them capitate-glandular; leaf blades lanceolate to ovate-lanceolate or rhombic-lanceolate, principal ones 10-15 mm long and 3-7 mm wide, tapering to a petiole 3-8 mm long, both surfaces covered with hairs similar to the stem-hairs, margins subentire, or with 1 to several teeth of varying size and shape; flowering calyx oblong to oblong- campanulate, 4-5 mm long and 2-2.5 mm wide at base of lobes, the upper 1-1.5 mm divided into ovate to deltoid lobes; pedicels 5-7 mm long; corolla yellowish, immaculate, or slightly maculate and fading in drying, 7-9 mm long, 6-8 mm wide, probably wider if fully expand- ed; anthers yellow, ca. 2 mm long on filaments 3-4 mm long; fruiting 1967] Physalis — Waterfall 218 calyx 8-12 mm long and 8-10 mm wide, oblong-ovate, open at the apex, nearly filled by the berry. COLLECTIONS EXAMINED. MEXICO: sINALOA: Ymala, Sept. 25 to Oct. 8, 1891, Palmer 1713 (Type: GH, Isotypes: GH, NY, US). This unique species is still known only from the type col- lection. 43. Physalis philadelphica Lamarck, Encycl. Méthod. Botanique 2: 101. 1786; P. chenopodifolia Wildenow in L., Species Plantarum, ed 4: 1023-1024, 1797, non P. chenopodifolia Lamarck, Tabl. Encycl... 28. 1793; P. ixocarpa Brotero ex Hornemann, Hortus Regius Botani- cus Hafniensis, Suppl. 26. 1819; P. aequata Jacq. f. ex Nees, Linnaea 6: 470. 1831; P. laevigata Mart. & Gal., Bull. Acad... . Belgique 12: 131. 1845; P. philadelphica Lam. var. minor Dunal in DC., Prodro- mus 13(I): 450. 1852. Annual, 15-60 cm tall, glabrous to sparsely vestite with short ap- pressed hairs, and, sometimes, long pilose ones; leaf blades 2-7 cm long, ovate to ovate-lanceolate, margins dentate to sinuate-dentate to entire; petioles one-half the length of the blade to equalling it; corolla yellowish, maculate, 10-18 mm wide, rotate-reflexed when fully extended; anthers blue, or yellowish with bluish margins, ca. 3 mm long, usually strongly twisted after dehiscence; fruiting calyx 10- ribbed, 2-3 cm long, often well-filled by the fruit which may be slightly oily at maturity, and is sessile in the calyx; fruiting ped- uncles 3-8 mm long. The type of P. philadelphica Lamarck, in the Lamarck collections (P), is the same as P. ixocarpa. It has a few short hairs on young parts, but does not have long ones. Rydberg's reference (1896:337) of perennial material to this taxon, which he attempted to differentiate from P. ixocarpa, pertains to collections of the variable P. virgin- iana, complex of eastern North America, Some of them the author has seen and has referred to var. subglabrata (Mac. & Bush) Waterfall. SELECTED COLLECTIONS. MEXICO: cuiAPAS: Los Lagos, 34 miles se of Comitan, Jan. 17-20, 1952, Carlson 2228 (MICH, NY, UC); 3 miles e of Teopisca, June 12, 1960, King 2866 (MICH, NY, UC); CHIHUAHUA: canyon, La Cienegita, Rio Mayo, Sept. 10, 1936, Gentry 2633 (F, GH, UC, US) ; 3 miles w of Cd. Camargo, 4000 ft., Aug. 2-5, 1939, White 2274 (GH, MICH); COAHUILA: La Novia, between Laguna de Leche and Magueyal, Aug. 30, 1941, I. M. Johnston 8635 (GH); Saltillo, June 1898, Palmer 325 (GH, NY, US); DISTRITO FEDERAL: Lago de Texcoco, Oct. 21, 1962, Rzedowski 16273 (OKLA); DURANGO: Durango and vicinity, April to Nov. 1896, Palmer 288 (BM, F, GH, NY, UC, US); GUERRERO: Rio Balsas, Aug. 26, 1910, Orcutt 4387 (F); 214 Rhodora [Vol. 69 HIDALGO: Real del Monte, Berlandier 225 (P); Ixmiquilpan, July 1905, Rose et al 9068 (NY, US); JALISCO: along lumber road, 11 miles from Atenquique, April 3, 1951, MeVaugh 11743 (MICH); barranca near Guadalajara, June 1886, Palmer 1 (BM, GH, NY, US) ; grassland with desert shrubs, 11 miles se of Lagos de Mareno, Waterfall 13872 (OKLA); MEXICO: in 1865, Bourgeau 872 (NY, P, US); MICHOACAN: Ario in 1840, Galeotti 1188, Type collection of P. laevigata Mart. & Gal. (BR, P); Balsas, Huetamo, Mar. 23, 1934, Hinton 5823 (MICH, NY, US); MORELOS: valley below Cuernavaca, 4000 ft, Sept. 16, 1900, Pringle 8446 (F, GH, NY, UC, US, VT); NUEVO LEÓN: Rio Santa Cata- rina, Monterrey, June 17, 1934, Pennell 16790 (NY, US); wet depres- sion in caleareous desert, 16 miles n of Matehuala, Aug. 21, 1959, Waterfall 15750 (OKLA, SMU); OAXACA: valley about Cuicatlan, Nov. 3, 1894, Nelson 1874 (GH, US); Tomellin Canyon, 3500 ft, Dec. 9, 1895, Pringle 6263 (BM, F, GH, NY, UC, US, VT); PUEBLA: vicinity of San Luis Tultitlanapa, July 1908, Purpus 3581 (UC); QUERETARO: vicinity of Querétaro, in 1912, Basile 203 (US); SAN LUIS POTOSÍ: region of San Luis Potosí, in 1878, Parry & Palmer 640 (GH, NY, US); calcareous desert, 74 miles n of San Luis Potosí, Aug. 20, 1959, Waterfall 15717 (OKLA, SMU); SINALOA: Fuerte, Mar. 27, 1910, Rose et al 13575 (US) ; Mazatlán, Jan. 1889, Wright 1253 (F, GH, UC, US) ; TABASCO: San Juan Bautista, Jan. 28, 1907, Collins & Doyle 267 (us); TAMAULIPAS: valley near Miquihuana, Aug. 8, 1941, Stanford et al $01 (GH, NY, UC); TLAXCALA: Tlaxcala, June 23, 1938, Balls 4902 (BM, NY, UC, US); VERACRUZ: Orizaba, July 24, 1866, Bourgeaw 2694 (BR, GH, OKLA, P); Cordoba, Aug. 1936, Matuda 261 (MICH); Jalapa, in 1894, Smith sin. num. (MICH). COSTA RICA: Cartago, cultivated, Oct. 6, 1953, Heiser 3608 (F); EL SALVADOR: in sand along stream, Ahuachapan, Jan. 9-27, 1922, Standley 20244 (GH, NY, US). GUATEMALA: open slopes in pine-oak forest, e of Quiche, Nov. 20- Dec. 4, 1940, Grant 657 (F, GH); near Amatitlan, Dec. 29, 1938, Standley 61335 (F, US); damp thicket, sea level, San Jose, Jan. 30, 1939, Standley 64015 (F); Coban, April 19, 1941, Standley 90935 (F). CUBA: Havana, May 7, 1914, Ekman 734 (NY); HAITI: Massif de la Selle, Port-au-Prince, Feb. 2, 1926, Ekman 5495 (NY). JAMAICA: Hope Road, Jan. 13, 1898, Harris 6983 (MICH). NEVIS: Charlestown, Mar. 9, 1959, Proctor 19451 (GH). The above citations all refer to var. philadelphica forma philadelphica. The following is a forma containing plants which develop long jointed hairs in varying abundance, especially about the sepals and pedicels. Not all duplicates of the same collection will exhibit the same degree of dens- ity of long hairs, but most collections do not have them in any quantity. 43b. forma pilosa Waterfall, forma nov. 1967] Physalis — Waterfall 215 Forma philadelphica simillima, sed calycibus et pedicellis plus minusve pilosis. TYPE: Paul C. Standley 79899 Finca la Alameda near Chimalten- ango, GUATEMALA, Dec. 11-22, 1940 (F). SELECTED COLLECTIONS. MEXICO: CHIHUAHUA: arroyo, Sierra Bajura, Río Mayo, Dec. 28, 1934, Gentry 1207 (F). GUATE- MALA: in field near Antigua, Melhus & Goodman 3489 (F); Standley 79899, type, cited above; Jalapa, Nov. 28, 1939, Steyermark 32159 (F). 43c. var. parviflora Waterfall, var. nov. Var. philadelphica simillima, sed flores minores, corollis 3-5 mm longis et 6-7 mm latis; antheris violaceis, 1.2-1.5 mm longis. TYPE: partial shade nearly south of Cerro Sanganguey, 12 miles se of Tepic, NAYARIT, Aug. 16-18, 1959, Feddema 582 (MICH, Isotype: OKLA). COLLECTIONS EXAMINED. MEXICO: JALISCO: low marshy area about 1 mile w of Ayo el Chico, Aug. 23, 1958, McVaugh 17219 (MICH); NAYARIT: Feddema 582, cited as Type. 43d. var. immaculata Waterfall, var. nov. Var. parviflora simillima, sed corollis immaculatis vel subimmacula- tis. TYPE: U. T. Waterfall 16115 silty flat in desert, 14 miles ne of Parral, CHIHUAHUA, Aug. 10, 1961 (OKLA, Isotypes: GH, US). COLLECTIONS EXAMINED. MEXICO: CHIHUAHUA: flood plain of Río Conchos near Valle de Zaragosa, May 19, 1955, Pennington 328a (OKLA); Waterfall 16115, Type, cited above; GUANAJUATO: ditch in cornland, 11 miles se of Salamanca, Aug. 16, 1957, Waterfall 13907a (OKLA); JALISCO: stony hilltop in desert scrub, 7 miles se of Lagos de Mareno, Aug. 15, 1957, Waterfall 13860 (OKLA, SMU) ; SONORA: Rio Bavispe, Colonia Oaxaca, July 25, 1938, White 672 (GH, MICH); Rio Bavispe, June 27-29, 1940, White 2888 (GH, MICH). 44. Physalis michoacanensis Waterfall, sp. nov. Planta herbacea; caulibus multirameis, 1 m altis, puberulis; foliis ovatis vel rhomboideo-ovatis, basi inaequalibus, marginibus integer- rimus, majoribus 6-11 cm longis et 3-5 latis, petiolis 3-6 cm longis; calycibus floriferis 2-4 mm longis et 3-4 mm latis ad basim loborum; calycis lobis ovato-lanceolatis vel deltoideo-lanceolatis, 1.5-1.8 mm longis; pedicellis 3-5 mm longis; corollis maculatis vel parvimaculatis, reflexo-rotatis, 8-10 mm longis et 8-9 mm latis, ostiis dense longitricho- matibus; antheris violaceis, 2-2.2 mm longis; calycibus fructiferis subteretibus, 30-25 mm longis et 20-25 mm latis; pedicellis 5-8 mm longis; baccis 8-10 mm latis. TYPE. Rogers McVaugh 17931 (MICH, Isotype: OKLA) abundant in partial shade, old lava flows 4 miles nw of Apatzingan, in sparse woodland of Cordia, Amphipterygium, Apoplanesia, elev. 300 m, MICHOACÁN: MEXICO, Sept. 16, 1958. 216 Rhodora [Vol. 69 45. Physalis angulata L., Species Plantarum 1: 183. 1753. Annual, 25-100 em tall, branched, glabrous or with a few short, an- trorsely appressed hairs; leaf blades usually ovate to ovate-lanceolate, narrower in var. lanceifolia, usually deeply and irregularly incised- toothed, or undulate-toothed, sometimes entire, principal ones 5-11 cm long and 3.5-8 cm wide, on petioles 4-8 cm long; flowering calyx usually (3-) 4-7 mm long, divided above into deltoid to broadly lance- olate, sometimes slightly acuminate lobes (1.5-) 2-3 mm long, width 2-4 mm at base of lobes, on pedicels 5-10 mm long; corolla immacu- late, or with indistinct spots, somewhat hairy on the inside of the tube, 6-12 mm long and 7-12 mm wide when fully expanded; anthers bluish or violet, (1-) 2-2.5 mm long on slender filaments 3-5 mm long; fruiting calyces 10-angled or 10-ribbed, 20-35 mm long and 17-22 mm wide, on pedicels (7-) 10-25 mm long; berry 10-12 mm in diameter on a gynobase ca. 1 mm long. The concept of P, angulata has contained different kinds of plants at different times. Some are incompatible with the current delimitation of the species, while others, based on obvious, but inconsequential variation, have resulted in the description, or reinterpretation, of taxa not retained here. The specific name itself seems to have resulted in misin- terpretation. Some authors have apparently thought it re- ferred to the fruiting calyx, and have separated material with 10-ribbed fruiting calyces. If one considers only the evidence of Linnaeus' intention, as indicated in Species Plantarum itself, it is noted that the specific names are generally taken from a species-characterization, either that of Linnaeus himself, or of one of the cited references. In this case we find “Physalis ramosissima, ramis angulatis glabris", indicating that the name refers to the branches, not the fruiting calyx. The first-cited reference is Hortus Cliffortianus which reads ‘5. Physalis annua ramosissima, ramis angulosis glabris, foliis dentato-serratis . . . Crescit in Carolina". In the herbarium of the Linnaean Society in London (the author is grateful to its director, Mr. O'Grady, for his cour- tesies) is to be found, as a part of the Linnaean Herbarium, a sheet of P. angulata (labelled 247.9). The specimen is nearly entire-leaved, anthers are light blue, fruiting calyx 10-ribbed with the primary ribs larger, flowering pedicels 1967] Physalis — Waterfall 217 are 6-12 mm long and fruiting pedicels 15-20 mm long; the fruiting calyx is ca. 30 mm long and 23 mm wide. The Riksmuseum in Stockholm has in its Linnaean Her- barium a similar specimen, one with entire leaves or with leaves having 1 or 2 teeth on each margin. A notation on the back of the sheet indicates that it was raised in the botanical gardens in Uppsala. Link in 1821 (Enumeratio Plantarum Horti . .. Berolin- ensis) described P. dubia ... "calycibus demum angulatis pulverulento-subtomentosis . . ." from Brasil. Nees (Lin- naea 6: 471. 1831) renamed the species P. Linkiana based on P. dubia Link of which he says “Nomen mutandum erat, ob Physalidem Gm. antiquiorem", Of the calyx he only says, "sub lente scabriusculus". Asa Gray (Proc. Amer. Acad. Arts and Sci. 10: 64. 1875) transferred the name “Linki- ana" to the varietal status. In 1891 Otto Kuntze (Rev. Gen. Plant....2: 452) transferred Link’s “dubia” to the varietal status under P. angulata. In 1962 Stehlé (Bull. Soc. Bot. France 109: 28) reduced "Linkiana" to the status of a forma under P. angulata. There does not seem to be a type extant to represent P. dubia Link. It seems doubtful if a plant described as having calyces ‘“pulverulento-subtomentosis” could belong to the concept of P. angulata as characterized either by the two extant collections from the Linnean Herbarium, or by rep- resentatives of the population in the eastern United States on which the citation in Hortus Cliffortianus was based. The flowering calyces of P. angulata vary from glabrous to having varying amounts of short, appressed hairs, but usually these are sparse. It seems possible that P. dubia could have referred to material now called P. Lagascae. In any event, in the absence of a type to which the name P. dubia can be tied, it seems dubious, if a pun is permitted, that it should be assigned to P. angulata. Following this course of action, none of the names listed in the preceding paragraph can be used, since they all spring from P. dubia. It should be noted, however, that in the varietal and formal category, the name has pertained to the extreme of P. an- 218 Rhodora [Vol. 69 gulata with large-toothed leaves. In the following treatment both this and the entire-leafed extreme are considered as var. angulata. 45a. P. angulata L., var. angulata; P. ramosissima Miller, Gar- dener's Dictionary: 12. 1768; P. capsicifolia Dunal, in D. C. Prodr. 13(1): 449. 1852; P. angulata var. capsicifolia (Dunal) Griseb., Kar. 96: 1857; P. angulata var. ramosissima (Miller) O. E. Schulz, in Urban, Symb. Ant. 6: 143. 1909; P. angulata forma ramosissima (Miller) Stehlé, Bull. Soc. Bot. France 109: 28. 1962. Leaves ovate to ovate-lanceolate or oblongish, coarsely and irreg- ularly toothed to entire; flowering calyx usually 4-7 mm long, with lobes 2-3 mm long; corolla usually immaculate, sometimes with in- distinct spots, usually 6-12 mm long; anthers usually 2-2.5 mm long. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: San José del Cabo, April 1899, Grabendorfer sim mum (UC); JALISCO: near Chapala, Oct. 5, 1903, Rose & Painter 7628 (US); MICHOACÁN: Zarzamora, Hinton 15035 (F, GH; the US duplicate is P. philadel- phica); TABASCO: alrededores del Barrio de Santa Cruz, San Juan Bautista, Aug. 13, 1888, Rovirosa 263 (NY); VERACRUZ: Cordoba, Aug. 1938, Matuda 224,0 (mMIcH). BRITISH HONDURAS: Jones Bank, Belize River, March 1933, Lundell 4002 (F, MICH). COSTA RICA: Guapiles, Mar. 12, 1924, Standley 37388 (us). EL SALVA- DOR: dry hillside, Laguna de Olomega, Dept. de San Miguel, Feb. 20, 1922, Standley 21054 (GH, US). NICARAGUA: Bahia de Blue- fields Río Escondido, Mar. 26, 1949, Molina 1990 (F). PANAMA: Changuinola, Jan. 25, 1921, Carleton 74 (Us); San José Island, Jan. 1, 1946, I. M. Johnston 993 (GH); around Culebra, Nov. 1911, Pittier 4782 (F, GH, US). ANAGDA: rocky plain, Britton & Fishlock 1044 (NY). ANTIGUA: near St. John, Feb. 4, 1913, Rose et al 3257 (NY). BAHAMAS: creek, Andros, June 6, 1890, Northrop 616 (NY). BARBADOS: Waterford in 1900, Bovell sin num (NY). BERMUDA: south road, Aug. 16, 1913, Collins 280 (GH, NY). CUBA: sand dunes, Laguna de Cortes, Mar. 11, 1911, Britton & Cowell 9863 (GH, NY); Bayate, Oriente Province, Aug. 1, 1914, Ekman 2348 (NY); near river s of Havana, June 23, 1914, Leon 4325 (NY). DOMINICAN REPUBLIC: near Puerto Libertador, Manzanilla Bay, Howard 9635 (GH, NY). GRENADA: wooded hillside near Victorial, Proctor 17222 (GH). GUADELOUPE: in 1898, Duss 2590 (Ny). HAITI: Gonave Island, Mar. 3-14, 1920, Leonard 3065 (GH); Bassin Blue, Apr. 16, 1929, Leonard 14705 (GH, NY). JAMAICA: Saint Ann’s Bay, Mar. 27-30, 1908, Britton 2484 (NY); Windsor estate, Aug. 22, 1956, Proc- tor 10503 (GH). PUERTO RICO: edge of marsh, Mayaguez, March 31, 1918, Britton 2353 (NY); Cerro Ventana, Vieques Island, Feb. 20, 1914, Shafer 2983 (NY). ST. CROIX: Catherine’s Rest, April 9, 1896, Ricksecker 361 (NY, UC). ST. KITTS: cane fields, Britton & 1967] Physalis — Waterfall 219 Cowell 755 (NY). ST. THOMAS: Eggers 295 (GH, NY). ST. VIN- CENT: G. W. Smith 268 (NY). MARTINIQUE: in 1879, Duss 1930 (NY). TORTOLA: Sea Cow Bay, Britton 683 (NY). TRINIDAD: Britton et al 1057 (G). In attempting to account for another name, it may be of interest to note that P. Halicacabum Crantz, Inst. Rei Herb. 367. 1766, is described as “P. ramosissima ramis angulatis glabris, foliis ovatis dentatis", and that under it is “Beta, Alkekengi indicum glabrum capsicifolia". The first phrase seems only slightly modified from Linneaus' in Species Plantarum, and the appended variety is exactly the same. Therefore it seems probable that the name should be included in the synonymy of P. angulata, despite its re- ferral to P. Alkekengi in Index Kewensis. 45b. Physalis angulata L., var. lanceifolia (Nees) Waterfall, Rho- dora 60: 163-164. 1958; P. lanceifolia Nees, Linnaea 6: 473. 1831. Leaves lanceolate to linear-lanceolate; corolla usually 4-6 mm long; anthers usually 1-2 mm long. SELECTED COLLECTIONS. MEXICO: CAMPECHE: Palizada, July 25-28, 1939, Matuda 3867 (GH, MICH, OKLS, US); SINALOA; vicinity of Labradas, Sept. 9, 1925, Ferris & Mexia 5149 (GH); Mazatlan, Sept. 1932, Ortega 7074 (F, NS); SONORA: bed of Rio de Sonora, vicinity of Hermosillo, Mar. 5, 1910, Rose et al. 12421 (US). 46. Physalis minimaculata Waterfall, sp. nov. Planta annua, ramosa; caulibus pubescentibus; foliis ovatis, basi inaequalibus, margine inaequaliter magnodentatis, principalibus 3-5 em longis et 2-2.5 mm latis, petiolis 2-4 mm longis; calycibus floriferis 3.5-4 mm longis et 3-3.5 mm latis ad basim loborum; calycis lobis 1.5-2 mm longis; pedicellis 4-6 mm longis; corollis minimaculatis, rotatis vel reflexo-rotatis, 7-10 mm longis et 8-9 mm latis; antheris violaceis, oblongis, 2-2.5 mm longis; filamentis crassis; calycibus fructiferis decangulatis, subtilibus, ca. 18 mm longis et 12 mm latis. TYPE: Rogers McVaugh 17902 (MICH). MEXICO: MICHOACÁN: bushy annual to 60 cm high, flowers white, old lava flows 4 miles nw of Apatzingan in sparse woodland of Cordia, Amphiteryrium and Apoplanesia, elev. ca. 300 m, Sept. 16, 1958. 47. Physalis ampla Waterfall, sp. nov. Planta herbacea; caulibus glabris vel antrorse paucitrichomatibus; folis ovatis, integerrimis vel undulato-dentatis; principalibus 4-6 cm longis et 3-4 cm latis; petiolis 3-5 cm longis; calycibus floriferis 4-5 mm longis et 3-4 mm latis ad basim loborum; calycis lobis late ovatis vel deltoideo-ovatis, 1-1.5 mm longis; pedicellis 3-8 mm longis; 220 Rhodora [Vol. 69 calycibus vestitis, pilis antrorsis, articulatis, ca. 1 mm longis; corol- lis immaculatis, 5-6 mm longis et 4-6 mm latis; antheris violaceis, ca. 1 mm longis; calycibus fructiferis decacostatis, 30-35 mm longis et 25-30 mm latis, glabris vel paucitrichomatibus; pedicellis 5-10 mm longis; baccis 10-12 mm latis. TYPE: T. S. Brandegee sin. num. MEXICO: SINALOA: vicinity of Culiacán, Sept. 12, 1904 (uc); also examined: Brandegee, Oct. 10, 1904, vicinity of Culiacán. This species is characterized by the small corollas, hairy calyces, short pedicels and large, inflated fruiting calyces which are glabrous or have a few hairs, these often growing from the more or less deltoid lamellations on the calyx ribs. 48. Physalis Lagascae Roemer & Schultes, in Linnaeus Systema Vegetabilium 4: 679. 1819. Annual, 10-90 cm tall, stems vestite in varying degrees with jointed hairs up to 2-4 mm long; leaf blades ovate to ovate-lanceolate, often somewhat rhombic and inequilateral at bases, margins entire to ir- regularly dentate or sinuate; principal blades 2-4 cm long on petioles 1-3 (-4) em long; flowering calyx divergently long-hairy, 3-4 mm long and 2.5-3 mm wide at base of calyx lobes, which are nearly deltoid and 0.7-1.5 mm long, pedicels 2-5 mm long; corolla maculate, but the spots usually not strongly contrasting and sometimes ap- parently immaculate, 5-7 mm long and 5-6 mm wide when fully open, hairy in the tube; anthers bluish or violet, 1.2-1.5 mm long, on filaments ca. 1.5 mm long; fruiting calyx 10-ribbed, the ribs sometimes with projections that appear to be lamellate hair-bases, or with hairs having such bases, sometimes glabrous, 13-20 mm long and 12-20 mm wide on pedicels 3-6 mm long; berry nearly spheric, 6-10 mm wide. 48a. var. Lagascae; P. micrantha Link, Enumeratio Plantarum Horti . .. Berolinensis 1: 181. 1821-22; P. parviculea Blake, Contrib. U. S. Natl. Herb. 24: 20. 1922. Plant varyingly hairy, often abundantly so on the stem, the hairs persistent. SELECTED COLLECTIONS. MEXICO: DURANGO: Durango and vicinity, Sept. 1896, Palmer 622 BM, F, GH, UC, US) ; JALISCO: steep rocky hills 2 miles nw of Tequila, Sept. 3, 1960, McVaugh 18629 (MICH); precipitous mountainsides with arborescent Ipomoea above the west end of Lake Chapala, Nov. 8, 1959, MeVaugh 377 (MICH); near farmland, 42 miles e of Guadalajara, Aug. 18, 1959, Waterfall 15634 (BM, F, MICH, NY, SMU); MICHOACAN: on mountainside above Lake Chapala, near Km 543, Sept. 23, 1958, McVaugh 18181 (MICH); MORELOS: valleys below Cuernavaca, Sept. 1900, Pringle 8447 (F, GH, NY, OKLA, UC, US, VT) ; NAYARIT: steep mountainsides 2 miles ne 1967] Physalis — Waterfall 221 of Santa Maria del Oro, Sept. 15, 1960, McVaugh 19035 (MICH) ; QUERETARO: silty flats with shrubs and cacti, 8 miles e of Querétaro, Aug. 23, 1961, Waterfall 16548 (OKLA); SINALOA: stony hillside, 57 miles ne of Mazatlán, Aug. 16, 1956, Waterfall 12770 (OKLA); ZACATECAS: hills 5 miles sw of Jalpa, Aug. 30, 1960, McVaugh. 18513 (micH). EL SALVADOR: Cerro de San Jacinto near San Salvador, Feb. 8, 1922, Standley 20611 (GH, NY, US). GUATEMALA: dry riverbed, Los Amate, Dept. of Izabel, May 9, 1919; Type of P. parviculea Blake, Blake 7318 (GH, US); grassy slopes around Ipala, Oct. 23, 1939, Steyermark 30334 (F). HONDURAS: vicinity of El Zamorano, April 14, 1949, Standley 19002 (F). PANAMA: Sabanas, Dec. 2, 1921, Heriberto 281 (Us). MARTINIQUE: in 1899, Duss 4050 (NY). 48b. var. glabrescens Schulz, in Urban Symbolae Antillanae 6: 147. 1909. Similar to var. Lagascae, but less hairy, glabrate, the stems with few hairs or none. SELECTED COLLECTIONS. MEXICO: CHIAPAS: in grassy field, Nov. 27, 1949, Matuda 17225 (F); GUERRERO: Coyuco-Ziranderangio, May 1, 1934, Hinton 5928 (NY); MEXICO: Cumbre-Tejupilco, Sept. 6, 1935, Hinton 8418 (MICH, US); MICHOACAN: Tacupa, Mar. 23, 1934, Hinton 5821 (BM); VERACRUZ: région d'Orizaba, Bourgeau 2940 (P); YUCATAN: Chichankanab, Gaumer 9426 (F, GH, US). COSTA RICA: San Ramon, April 10-16, 1938, Brenes 22847 (NY). EL SALVADOR: San Salvador, Dec. 20, 1921, Standley 19662 (GH, NY, US); old corn field, San Vicente, Mar. 2-11, 1922, Standley 21724 (GH, NY, US); wet forest, Sonsonate, Mar. 18, 1922, Standley 22363 (GH, US). GUATEMALA: damp field, Zacapa, Oct. 7, 1940, Staadley 73608 (F); damp thicket, Jutiapa, Oct. 24, 1940, Standley 75631 (F). HONDURAS: gravel stream bed, near El Zamorano, Aug. 13, 1947, Standley 12177 (F). NICARAGUA: Managua, Aug. 1932, Garnier 1164 (US); moist thicket, Chichigalpa, July 12, 1947, Standley 11275 (F). PANAMA: near Matias, Dec. 30, 1925, Standley 28920 (US). CUBA: Type collection of var. glabrescens, Wright 3636 (GH, US, NY, P); Havana, Sept. 1916, Leon 6806 (NY); on crotch of tree, Baracoa, ORIENTE, Nov. 30, 1910, Shafer 7691 (NY). The usage of the name P. Lagascae introduces the ques- tion of the identity of P. minima L. In Species Plantarum, page 183-184, this species follows P. pubescens as “8. Phy- salis ramosissima, pedunculis fructiferus folio longioribus. Hort. Cliff. 62. Roy. lugdb. 427. Solanum vesicarium in- dicum minimum. Herm. ludgb. 569. t. 571. Pee-inota-in- odien. Rheed. mal. 10. t. 140. f. 71. Habitat in Indiae aridis sordidis”. 222 Rhodora [Vol. 69 There is no P. minima in the Linnaean material at Stock- holm, Neither is there any in the Hortus Cliffortianus her- barium at the British Museum. Paul Hermann, Horti Academici lugduno-Batavia Catalogus . . . 1679-1686 ("Herm. lugdb.") describes material from Malabar and Ceylon, and has an accompanying illustration showing calyces with spreading hairs. Heinrich van Rhede tot Drakestein, Hortus Malabaricus . . . (“Rheed. mal.") in his tab. 71 shows elongate apiculate fruiting calyces, the upper one-fourth to one-third divided into free, divergent calyx-tips. No hairs are shown on the drawing, but then neither are they for P. pubescens, so it is probable that this characteristic simply was not illustrated. It also shows corolla lobes rather deeply incised, but perhaps this is ar- tistic liberty. In any event, it does not resemble anything I have seen. One might assume that P. minima is a nomen con fusum and that it should be dropped altogether. Or one might con- clude that, whatever it is, it refers to Asiatic material, not West Indian and New World. After all, Linnaeus did say "habitat in India" not “in India utraque" which would have intentionally included the West Indies. Furthermore his references are to Asiatic works. True, New World material has been identified as P. minima, but it was for this material that P. Lagascae was established. Attempting to form a concept of P. minima by examining Asiatic material cur- rently so referred is fruitless. In the rather extensive col- lections at the Kew Herbarium there is under this name material of both P. angulata and P. pubescens, as well as material with small 10-ribbed fruiting calyces, some of which is very similar to that referred herein to P. Lagascae. The elucidation of this material awaits further study. 49. Physalis microcarpa Urban & Ekman, in Urban Plantae Haitien- sis, Arkiv for Botanik 21A(5): 59. 1927. Annual 15-30 (-60) cm tall, slightly branched to much-branched, erect to nearly prostrate, hairs few, short, antrorsely curled; leaf blades ovate or rhombic-ovate to lanceolate, principal ones 25-35 (-45) mm long and 11-30 (-40) mm wide, on petioles 12-45 mm long, their margins usually entire, sometimes slightly and irregularly repand- 1967] Physalis — Waterfall 223 dentate, bases sometimes inequilateral, apices often somewhat at- tenuate; flowering calyces 1.2-2.5 mm long and 1-2 mm wide, divided above into lobes 0.4-0.8 mm long, pedicels 0.5-2 mm long; corolla 2.8-3.5 mm long, immaculate, tube slightly hairy within; anthers 0.3-0.7 mm long, bluish or violet, sometimes yellowish (perhaps due to fading), on filaments 0.5-1.0 mm wide, nearly terete, slightly 10-ribbed, sometimes slightly angled, on pedicels 1.5-4 mm long; fruit 4-6 mm wide. SELECTED COLLECTIONS. CENTRAL AMERICA: moist sand- bar, Metepan, Jan. 29, 1947, Standley and Padilla 3337 (F). GUATE- MALA: brushy rocky slope, Jutiapa, Oct. 24, 1940, Standley 75090 (F); damp thicket, above Chiquimula, Oct. 14, 1940, Standley 74334 (F). HONDURAS: weed in field along Rio Yeguare e of E] Zamorano, Dec. 10, 1946, Standley 1465 (F); Tegucigalpa, Nov. 29, 1949, Stand- ley 24792 (F). HAITI: Jean Rabel, Jan. 27, 1929, Leonard 21614 (GH, NY). SANTO DOMINGO: Llanura de Vega prope Monte-Cristi, April 3, 1926, Ekman 5841 (Type: S); common weed in field, Valle del Ciabo, Nov. 4, 1930, Ekman 16125 (GH, NY). 50. Physalis acutifolia (Miers emend. Sandwith) Sandwith, Kew Bull. 14: 232. 1960; Saracha acutifolia Miers, Ann. & Mag. Nat. Hist., ser. 2, 3: 449. 1849; P. Wrightii Gray, Proc. Amer. Acad. Arts & Sci. 10: 63. 1875. Annual, nearly glabrous, the few hairs short, stiff, flattish and appressed; leaf blades ovate-lanceolate to linear-lanceolate, principal ones usually 4-12 cm long, on petioles 1.5-7 cm long; leaf margins usually irregularly and often coarsely dentate, sometimes regularly and saliently dentate; flowering calyx usually 4-5 mm long on pedicels 5-12 times its length; corolla light yellow, sometimes with a greenish tinge, rotate with short tube, 15-23 mm wide when fully open, with 5 hairy pads near base of limb alternating with the stamens; anthers usually ca. 3 mm long, yellow with a blue or a blue-green tinge; filaments slender, slightly longer than the anthers; fruiting calyx ca. 2-2.5 cm long and 1.7-2 cm wide, on pedicels 2.5-6 cm long some- times nearly filled by the fruit. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: east base of Cerro de la Giganta, Oct. 7, 1951, Carter & Kellogg 3129 (UC) ; CHIHUAHUA: Candelaria, Oct. 24, 1911, Stearns, sin. num. (US) ; Janos, Aug. 26, 1939, White 2576 (GH, MICH); SINALOA: Culiacan, Sept. 23, 1904, Brandegee sin num (UC); rocky volcanic slopes with coastal plain forest, 24 miles n of Los Mochis, Oct. 3, 1954, Gentry 14409 (Us); Elota, Nov. 1925, Ortega 5633 (GH, US) ; SONORA: thorn forest 3 miles s of Navajoa, Sept. 16, 1959, Gentry 17824 (US); Guaymas, in 1887, Palmer 175 (GH, NY, US) ; 37 miles w of Hermosillo, Aug. 28, 1941, Wiggins & Rollins 146 (GH, MICH, NY, UC, US); Horconcitos, Río Bavispe, Sept. 5, 1940, White 3714 (MICH). 224 Rhodora [Vol. 69 It is quite possible that P. dentata Dunal in DC. Prodro- mus 13(1) : 441-442. 1852, should be referred to this taxon. A photograph, Negative 31435 of the Field Museum of Natural History, shows leaves, large corollas, and post-an- thesis nodding calyces, as well as the long pedicels, of P. acutifolia. The label only says “Nueva Espana". It was probably collected by Mocino in 1791 on his trip from Tepic to Los Alamos to Aguascalientes. 51. Physalis sulphurea (Fernald) Waterfall, comb. nov., Margaran- thus sulphureus Fernald, Proc. Amer. Acad. Arts & Sci. 35: 566-567. 1900. Annual, 10-35 cm tall, little-branched above, but sometimes several- stemmed from the base, stems becoming thick, hollow, up to 5-10 mm wide when pressed flat, glabrous, or with a few antrorsely ap- pressed hairs on younger parts, sepals and pedicels; leaf blades ovate to elliptic-lanceolate, principal ones 1.5-6 em long and 1-4 cm wide, often basally inequilateral and sometimes decurrent into a slightly alate petiole 5-25 mm long, margins entire or irregularly repand-dentate; flowering calyx quadrate to oblong-campanulate, 2-3 mm long and 2-4 mm wide, divided above into deltoid lobes 0.7-1.5 mm long, on pedicels 4-10 mm long; corolla immaculate, or with barely discernible spots, somewhat hairy within the tube, 5-7.5 mm long and 6-8 mm wide, possibly wider if fully expanded; anthers bluish or violet, oblong to ovate, 1-1.5 mm long on filaments 1-2 mm long; fruiting calyces 7-13 mm long, nearly terete or 10-ribbed, thin, the veins obscure, on pedicels (5) 10-15 mm long; berry 5-9 mm in diameter. In describing Margaranthus sulphureus, Fernald cited two collections, Bourgeau 111 and Pringle 8215. The latter, valley of Mexico, alt. 2250 m, Federal District, Oct. 4, 1889 is selected as LECTOTYPE (GH), with à second sheet as Isolec- totype (GH). Most of the corollas are unexpanded, but one or two do have the limb expanded. They definitely are not the urceo- late corollas of Margaranthus, but correspond well to those of Physalis. At one time in the course of this study, the author was of the opinion that P. Eggersii O. E. Schulz in Urban, Sym- bolae Antillanae was synonymous with Margaranthus sul- phureus Fernald, and consquently annotated a few sheets 1967] Physalis — Waterfall 225 of this taxon in several European herbaria as P. Eggersii. However, a sheet of the type collection of P. Eggersii, Eg- gers 1057, Water Island prope St. Thomas (P), proves to be not P. sulphureus, but appears to be an extreme of P. Lagascae var. glabrescens. COLLECTIONS SEEN. MEXICO: DISTRITO FEDERAL: Pringle 8215 cited above as type; wet soil, Valley of Mexico, 7300 ft, suly í, 1901, Pringle 9426 (GH); JALISCO: west end of Lake Chapala, July 14, 1940, Hitchcock & Stanford 7167 (GH, US); MEXICO: orilla de Lagunas de Zumpangos, May 5, 1963, Cruz-Cisneros 603 (MEXP) ; San Juan Citlaltepec, Orillas de la Laguna de Zumpango, 2250 m, May 5, 1963, Rzedowski 16000 (OKLA); MICHOACÁN: southeast shore of Lake Patzcuaro, 2200 m, July 15, 1941, Schery 130 (MICH, US); LOCALITY UNDETERMINED: Bourgeau 111 (BR, OKLA, P, U:), see dis- cussion above; Schaffner 362 (P). 52. Physalis lobata Torrey, Ann. Lyc. Nat. Hist. New York 2: 226- 227. 1828; Quincula lobata (Torr.) Raf. Atlantic Journal 1: 145. 1832; P. sabaena Buckley, Proc. Acad, Sci Phil. 14: 6. 1863; Chamae- saracha physaloides Greene, Bull. Torr. Bot. Club 9: 122. 1882; Quincula lepidota A. Nels. Bot. Gaz. 47: 450. 1909. Perennial, branching from the base, spreading or procumbent; indument of varying amounts of crystalline vesicles, flattening when dried, sparse or abundant enough to give the plant a scurfy ap- pearance; leaf blades ovate-lanceolate to linear-lanceolate, cuneate- ly narrowing to a winged petiole, principal ones usually 4-10 cm long and 0.5-3 em wide, usually pinnatifid, sometimes sinuate or entire; corollas blue or violet, rotate, 1.5-2 cm broad, with 5 hairy pads alternating with the bases of the filaments; anthers 1.5-2 mm long, yellow, on slender filaments; style twisted and bent to one side; flowering calyx 3-4 mm long, divided into triangular lobes 1.5-2 mm long, on pedicels 1-3 (-5) mm long; fruiting calyx 1.5-2 cm long, pentagonal-ovoid ,inflated; fruiting pedicel 1-2.5 (-3) em long; berry spheric-ovoid, 4-7 mm in diameter, sessile, or essentially so, on the invaginated base of the calyx. SELECTED COLLECTIONS. MEXICO: CHIHUAHUA: 28 miles s of Camargo, June 21, 1950, Dressler 1125 (GH); silty flat in desert, Aug. 9, 1961, Waterfall 16110 (OKLA, MICH); 3 miles n of Jiminez, Aug. 1, 1939, White 2176 (GH, MICH); COAHUILA: Sabinas, May 21, 1902, Nelson. 5776 (F, GH, US); Torreon, Oct. 13, 1898, Palmer 461 (BM, F, NY, UC, US); SONORA: playa, 2 miles w of Cerro Colorado, n of Sierra Pinacate, Apr. 28, 1951, Kamb 2016 (Uc); Charco, Feb. 213, 1910, Lumholtz 19 (GH); TAMAULIPAS: Nuevo Laredo, June 1935, Clark 6623 (NY). 226 Rhodora [Vol. 69 P. lobata reaches the southern limit of its distribution in northern Mexico, extending from southwestern United States, where it is abundant in western Oklahoma, western Texas, New Mexico and parts of Arizona. 53. Physalis Hintonii Waterfall, sp. nov. Herba perennis, 30-55 cm alta; trichomatibus dendriticis, ad 1 mm longis; foliis ovatis, pauci-undulatis, principalibus 5-9 em longis et 3-6 latis, petiolis 10-35 mm longis; calycibus floriferis 10-16 mm longis et 9-14 mm latis ad basim loborum; calycis lobis deltoideo- ovatis, 5-10 mm longis; corollis maculatis, 13-18 mm longis et 17-20 mm latis; antheris violaceis, 3-4 mm longis; filamentis 3-5 mm longis; calycibus fructiferis pentangulatis, ad apicem apertis, 25-35 mm longis et 18-22 mm latis; baccis ca. 1 cm latis. P. Hintonii is easily recognized by its spreading dendritic (several-branched) hairs, its large flowering calyx with broad sepals, and its oblongish fruiting calyx usually little constricted, and apically open. TYPE: George B. Hinton 8457 (NY), Tejupileo, Dist. of Tema- scaltepec, Estado Mexico, Sept. 17, 1935; Isotypes (NY, US). COLLECTIONS SEEN. MEXICO: mexico: Hinton 8457 cited above as type; Tejupilco, Temascaltepec, 1340 meters, Sept. 2, 1933, Hinton 4686 (NY, US); same location, Aug. 21, 1935, Hinton 8198 (us); Cumbre, Tejupilco, Temascaltepec, Hinton 8200 (F, US). 54. Physalis Greenmanii Waterfall, sp. nov. Planta herbacea, 1-2 m alta; trichomatibus articulatis, ad 1-3 mm longis; foliis ovatis, acuminatis vel lunato-acuminatis, cordatis, in- tegerrimis, principalibus 6-10 cm longis et 3.5-6.5 cm latis, petiolis 1.3-5 em longis; pilis laminis paucis, adpressis; calycibus floriferis 8-10 mm longis et ca. 7 mm latis ad basim loborum; calycis lobis ovatis vel ovato-lanceolatis, attenuatis, 5-7 mm longis; pedicellis 5-10 mm longis; corollis maculatis, 10-14 mm longis et 15-18 mm latis; antheris violaceis, 3-4 mm longis, filamentis filiformibus; calycibus fructiferis pentangulatis, 25-40 mm longis et 25-30 mm latis, calycis lobis longe acuminatis; pedicellis fructiferis 7-10 mm longis; baccis 10-12 mm latis. P. Greenmanii is characterized by its cordate leaves with few, long and short, strongly appressed hairs, the stem with spreading, more or less intertwined jointed hairs of vary- ing lengths, the longest 1-3 mm long, by the large light — yellow, prominently mottled-maculate corolla, and by its rather large, 5-angled fruiting calyx abruptly narrowed to long-acuminate lobes. 1967] Physalis — Waterfall 227 TYPE: C. G. Pringle 8104 (US), about thicket near Jalapa, 3-6 ft tall, 4000 ft altitude, Veracruz, Mexico, April 27, 1899; Isotypes: (F, GH, MEXP, NY, UC, US, VT). This must be the collection referred to by Greenman (1900) as P. glutinosa when, after describing P. Pringlei and citing specimens, he said “From P. glutinosa on the other hand, P. Pringlei differs in the leaves . . . not cordate.” The collection cited above, presumably available to Green- man, was originally labelled “P. glutinosa Schlecht.", the identification probably having been based on a reading of Schlechtendal’s original description, but the plant not con- specific with the type of P. glutinosa and other material so referred in this paper, noting particularly the large, non- reflexed corolla, 20-35 mm long, of the latter species. It is interesting to note that no material in American herbaria has been heretofore correctly referred to P. glutinosa, since that species is the same as the one we have called Cacabus mexicanus and Physalis eximia. Greenman's comparison to his newly described species was pertinent, but since the specimens he thought to be P. glutinosa were not that spe- cies, they require another name. 55. Physalis Pringlei Greenman, Proc. Amer. Acad, Arts & Sci. 35: 311-312. 1900. Herbs, (45 cm-) 1-1.7 m tall; vestiture of glutinous jointed spread- ing hairs of varying lengths, some with brownish capitate glands; leaf blades ovate, somewhat narrowed, sometimes unequally so, to the petiole, not cordate, usually 4-8 cm long, each margin with 1-3 irregular teeth or small lobes, or entire, sparsely appressed-hairy, often with more abundant hairs along the veins; petioles usually 1-3 cm long; calyx at anthesis oblong or oblong-flaring, 7-10 mm long and 5-8 mm wide at base of lobes, the upper one-third to one-half divided into ovate to lanceolate, acuminate lobes; corolla yellow, marked with 5 somewhat confluent, strongly darkened areas, with tomentum between these and the point of stamen insertion, usually 15-20 mm long and 18-22 mm wide when fully expanded, slightly rotate-pentangular in outline; anthers bluish, oblong, 2.2-3 mm long on glabrous filaments 3-4 mm long; fruiting calyx ovate to oblong- ovate, 5-angled with intermediate ribs or slight angles, with vestiture similar to stem but more sparse, 2.5-4 cm long and 18-25 mm long and 10-15 mm wide, sessile, or essentially so, on the invaginated base of the calyx. 228 Rhodora [Vol. 69 55a. var. Pringlei Calyx lobes acuminate, often prominently so; leaf margins ir- regularly sinuate, or with few irregular teeth or small lobes. SELECTED COLLECTIONS. MEXICO: DISTRITO FEDERAL: Sierra de Ajusco, 8500 ft, Oct. 2, 1895, Pringle 6216 (BR, F, GH, NY, OKLA, P, UC, US, VT); GUANAJUATO: cliffs, Acambaro, Oct. 6, 1904, Pringle 13448 (GH, US) ; JALISCO: rocky outcrop in mountains e of Mamantlan, about 15 miles sse of Autlan, July 30, 1949, Wilbur 1988 (MICH); MEXICO: near Cima, Sept. 1903, Rose & Painter 7172 (GH, NY, US); An Angel, al sur del Xitle, Sept. 15, 1952, Rzedowski 1759 (MEXP) ; NUEVO LEÓN: upper San Francisco Canyon, 15 miles sw of Pueblo Galena, May 17, 1934, Mueller 432 (F, GH, MICH); OAXACA: Sierra de Clavellinas, 3-5 ft., 9000 ft alt., Oct. 18, 1894, Pringle 6001 (Type: GH, Isotypes: BM, BR, NY, UC, US, VT). 55b. var. curtiloba Waterfall, var. nov. A var. Pringlei differt calycibus curtioribus, laminis integerrimis. TYPE: Edward Palmer 144 (US) San Ramon, DURANGO, Apr. 21 to May 9, 1906, Isotypes: (F, GH, NY, UC). COLLECTIONS EXAMINED. MEXICO: CHIHUAHUA: canyon slope in pine-oak forest, Arroyo Hondo, Sierra Charuco, Apr. 16-30, 1948, Gentry 8074 (MICH, UC, US); DURANGO: Palmer 144 cited above as type; SONORA: vicinity of Alamos, Mar. 19, 1910, Rose et al 13104 (GH, NY, US). 56. P. angustiphysa Waterfall, sp. nov. Planta herbacea; trichomatibus longis, articulatis, mollibus, divari- catis; foliis ovatis vel late ovatis, inaequaliter paucidentatis, princi- palibus 4-5 (-10) em longis et 3-4 (-6) cm latis; calycibus floriferis 3-4 mm longis et 3-4 mm latis ad basim loborum; calycis lobis lanceo- latis vel late lanceolatis, 2-2.5 mm longis; corollis luteis, maculatis, 8-9 mm longis et 12-14 mm latis; anthers violaceis 2-2.5 mm longis, filamentis filiformibus, 3-4. mm longis; calycibus fructiferis pent- angulatis, 23-27 mm longis et 10-18 mm latis; baccis 10-13 mm longis et 6-8 mm latis. TYPE: Julian A. Steyermark 51977 (us). GUATEMALA: lower brushy slopes, La Sierra (Tujimach) across river from San Juan Atitlan, Dept. Huehuetenango, Sept. 8, 1952; Isotypes: (F, US). COLLECTIONS EXAMINED. MEXICO: MICHOACAN: by stone fence in llano, 1 meter high, Zitacuaro-Coyota, Aug.. 25, 1938, Hinton 13156 (F, MICH, NY); NAYARIT: precipitous mountainsides, Carpinus- Magnolia forest, in ravines, 8 miles w of Tepic, Sept. 10, 1960, Mc- Vaugh 18872 (MicCH). GUATEMALA: Steyermark 95177 cited above as Type. The fruiting calyx is often twice as long as wide. The type collection has long, soft, somewhat tangled, jointed, 1967] Physalis — Waterfall 229 spreading hairs. However, McVaugh’s collection from Nay- arit has larger leaves and less dense vestiture than the type. 57. Physalis subrepens Waterfall, sp. nov. Planta herbacea, 50-60 cm longa; caulibus tenuibus adscendentibus vel procumbentibus, interdum repentibus ad nodis inferioribus; trichomatibus articulatis ad 2 mm longis; foliis ovatis vel late ovatis, acuminatis, interdum subcordatis, paucidentatis vel undulatis vel in- tegerrimis, principalibus 3-8 cm longis et 1.8-6 cm latis; petiolis 22-45 mm longis; calycibus floriferis 7-10 mm longis et 5-7 mm latis ad basim loborum; calycis lobis lanceolatis vel triangularibus, acumi- natis, 4-5 mm longis; corollis luteis, maculatis, 9-14 mm longis et 15-20 mm latis; antheris violaceis, 2.7-3 mm longis, filamentis flli- formibus, 2.5-3 mm longis; calycibus fructiferis pentangulatis et pentacostatis, 20-28 mm longis et 12-18 mm latis; pedicellis fructi- feris 10-20 mm longis; baccis 8-12 mm latis. TYPE: C. G. Pringle 13591 (VT) barranca below Trinidad Iron Works, 5500 ft, Hidalgo, Aug. 21, 1905; Isotypes: (F, GH, MICH, OKLA, UC, US). COLLECTIONS EXAMINED. MEXICO: DISTRITO FEDERAL: Con- treras, June 1, 1963, Galvan s. n. (MEXP); HIDALGO: Barranca near Honey Station, July 16, 1904, Pringle 13440 (GH, MICH, US, VT); Pringle 13591, Type, cited above; VERACRUZ: Lomogrande, Mount Orizaba, 9300 ft, April 29, 1938, Balls 4404 (UC). 58. Physalis volubilis Waterfall, sp. nov. Herba perennis; caulibus adscendentibus vel procumbentibus, in- terdum ad basim lignosis, e radice lignose fusiforme; pilis articulatis ad 1 mm longis; foliis ovatis vel late ovatis, acutis vel acuminatis, saepe cordatis, integerrimis, principalibus 25-50 mm longis et (15-) 20-40 mm latis, petiolis 10-30 mm longis; calycibus floriferis 8-10 mm longis et ca. 5 mm latis ad basim lobis; calycis lobis anguste linearibus, 4-6 mm longis; pedicellis 15-40 mm longis; corollis luteis, maculatis, 12-15 mm longis et 15-25 mm latis; antheris violaceis, 2.5-3.5 mm longis; calycibus fructiferis anguste ovatis, 25 mm longis et 10 mm latis vel 28 mm longis et 15 mm latis; baccis 8-11 m latis. TYPE: Wm. C. Leavenworth 519 (GH), vine, common locally in open pine woods, large tubers on roots, 2 miles above Tancitaroi, 7000 ft, MICHOACÁN, Aug. 10, 1940; Isotypes: (F, MICH, NY). SELECTED SPECIMENS. MEXICO: MEXICO: Meson Vieja, Dist. Temascaltepec, Apr. 29, 1932, Hinton 570 (US); MICHOACAN: procum- bent in clearing in oak-pine forest, Zitacuaro-Cacique, 2600 meters, June 6, 1938, Hinton 11928 (MICH, NY, US) ; Leavenworth 519, Type, cited above; prostrate from a small tuber, runners rooting at nodes, black voleanie ash in pine forest, 3.5 miles n of volcano Paricutin, June 14, 1950, Turner 1895 (MICH); vine, very common, about 5 230 Rhodora [Vol. 69 miles from Uruapan, Km 63-64 on Mexico 241, Aug. 22, 1949, White 239 (OKLA). Physalis volubilis differs from P. subrepens by having narrower flowering calyces, longer flowering pedicels (15- 40 mm long in volubilis, but only 5-12 (-15) mm long in subrepens, and usually entire leaves as compared with the usually toothed leaves of the former. It is possible that J. Rzedowski 10021 (OKLA) belongs here. It is somewhat atypical, and, collected from "orilla de camino, Potrerillos, Municipio de Xilitla, San Luis Potosí, Feb. 28, 1959," it is well out of the range indicated by the cited collections. However, it may be that it awaits collecting in the intermediate areas. 59. Physalis viridoflava Waterfall, sp. nov. Planta perennis e radice lignose fusiforme; caulibus herbaceis, prostratis, breve antrorse vestitis, trichomatibus articulatis; foliis ovatis vel late ovatis, interdum inaequalibus, integerrimis, principali- bus 25-35 mm longis et 20-28 mm latis, petiolis 10-24 mm longis; calycibus floriferis 5-7 mm longis et 3-4 mm latis ad basim loborum calycis lobis anguste lanceolatis, attenuatis, 3-5 mm longis, pedicellis 20-35 mm longis; corollis viridoflavis, immaculatis, 10-12 mm longis et 18-20 mm latis; antheris flavis vel viridoflavis, 2.5-3 mm longis, filamentis filiformibus 2.5-3 mm longis; calycibus fructiferis im- maturis pentangulatis. TYPE: David P. Gregory and George Eiton 99 (MICH), in rela- tively undisturbed pine forest, on route 110 at km 59-60, 20 miles due wsw of Jiqyilpan, and several miles s of Mazamitla, JALISCO, MEXICO, June 18, 1966; Isotype: (NY). This species, growing from a fusiform, or thicker, woody root, has long pedicels, greenish-yellow corollas with no evident maculations (collector says “with chocolate-brown cireular areas toward center"), anthers yellow or green, (collector says *green") and matted hairs in the throat of the corolla. Only immature fruiting calyces are known, but they are 5-angular. 60. Physalis longipedicellata Waterfall, sp. nov. Planta perennis, herbacea, 1 m alta, ramosa; trichomatibus mol- libus, articulatis ad 2 mm longis, majoribus 0.3-0.6 mm longis; foliis late ovatis vel lanceolatis, interdum inaequilateris, ad apicem attenu- atis; marginis pauci-dentatis vel undulatis vel integerrimis; foliis principalibus 6-10 cm longis et 3-6 cm latis, petiolis 3-6 cm longis, 1967] Physalis — Waterfall 231 angustissime alatis; corollis ignotis; calycibus fructiferis pentangu- latis, vestitis, 22-26 mm longis et 20-24 mm latis lobis lanceolatis, attenuatis, 6-8 mm longis; pedicellis fructiferis 20-45 mm longis; baccis l em latis. TYPE: Rogers McVaugh 13828 (MICH), in pine-oak-fir forests, steep slopes near summits, Sierra de Manantlan near Aserradero el Cuarton, 15-20 miles se of Autlan, JALISCO MEXICO, Nov. 2, 1952; Isotype: (OKLA). The name refers to the long pedicels as compared with other species having a 5-angled, hairy fruiting calyx. This large, branched, soft-hairy plant has large leaves, entire, with few shallow undulations, or a few coarse teeth, sometimes only 1 or Z per leaf; the fruiting calyx is abun- dantly covered with short hairs intermixed with a few long jointed hairs up to 2 mm long. 61. Physalis lignescens Waterfall, sp. nov. Frutex; trichomatibus brevibus, divaricatis vel antrorse crispis; foliis ovatis vel deltoideo-ovatis, integerrimis vel pauce undulatoden- tatis, principalibus 3.5-6 cm longis et 2.5-4 cm latis, petiolis 1-2 cm longis; calycibus floriferis 4-6 mm longis et 3.5-6 mm latis ad basim loborum; calycis lobis lanceolatis vel ovato-lanceolatis, 2.5-9 mm longis; pedicellis 8-15 mm longis; corollis luteis, maculatis, 13-16 mm longis et 20-22 mm latis; antheris violaceis, 2-2.5 mm longis, filamentis 4-6 mm longis; calycibus fructiferis (immaturis) 28 mm longis et 15 mm latis; baccis 8-9 mm latis. TYPE: Rogers McVaugh 13454 (OKLA), steep mountainsides in pine-oak forest (zone of broad-leaved trees in barrancas), ne slopes of the Nevado de Colima, below Canoa de Leoncito, 2250-2550 meters, JALISCO, Oct. 12, 1952. SPECIMENS EXAMINED. MEXICO: JALIsco: MeVaugh 13454, Type, cited above; open pine-covered ridges and slopes in mountains e of Mamantlan about 15 miles sse of Autlan by way of Chante, July 22, 1949, Wilbur 1776 (MICH). 62. Physalis jaliscensis Waterfall, sp. nov. Frutex; caulibus 1.5-2.5 m. altis, glabris vel pauce brevitrichomati- bus ex partis junioribus; foliis ovatis vel lanceolato-ovatis, at- tenuatis vel lunato-attenuatis, integerrimis vel pauce undulatis, principalibus 6-9 cm longis et 2-5 cm latis, petiolis 1-3 cm longis; calycibus floriferis 8-11 mm longis et 7-10 mm latis ad basim loborum; calycis lobis deltoideis vel lanceolato-deltoideis, ad apicem attenuatis, 4-6 mm longis, pedicellis 20-35 mm longis; corollis luteis, maculatis, 13-15 mm longis et 18-25 mm latis, campanulato-rotatis; antheris violaceis, 3-3.5 mm longis, filamentis 3-3.5 mm longis; calycibus 232 Rhodora [Vol. 69 fructiferis pentangulatis, 25-30 mm longis et 20-26 mm latis, pedicel- lis 35-45 mm longis; baccis 10-15 mm longis. TYPE: Rogers McVaugh 21512 (MiCH), fir forest, steep mountain- sides and barrancas 1-2 miles n of sawmill “La Cumbre" on the divide above headwaters of Río Mascoto, 25-30 km se of Talpa de Allende, Nov. 29, 1960, Isotype: (OKLA). SPECIMENS SEEN. MEXICO: JALIsco: McVaugh 21512, type, cited above; abundant along lumber road, steep slopes in pine forest, Sierra del Halo, near a lumber road leaving the Colima highway 7 miles ssw of Tecalitlan, 3 miles from highway, Aug. 14, 1957, Mc- Vaugh 16172 (MICH, OKLA). 63. Physalis maxima Miller, Gardener’s Dictionary, ed 8; no. 15. 1768. Annual, 0.6-1.5 meter tall; vestiture of short, spreading, more or less viscid hairs, and some capitate- glandular ones, often with larger, longer hairs 1.5-4 mm long on petioles, pedicels and stems; leaf blades ovate, coarsely and irregularly porrect-dentate to sinuate- dentate to entire or with 1 or 2 teeth on each margin; principal blades 6-10 (22) cm long and 4-9 (16) cm wide, bases slightly to strongly unequal, petioles 2-12 cm long; flowering calyces 5-8 mm long, divided into subulate, or subulate-attenuate lobes 3-6 mm long, these often long pilose; flowering pedicels (3-) 10-25 (-75) mm long, often pilose; corollas white or whitish, slightly to moderately macu- late, 6-8 mm long and 10-15 mm wide when fully open, tube hairy within; anthers slightly bluish-tinged or wholly yellow, 2.5-3 mm long, on slender filaments 2.5-4 mm long; fruiting calyx strongly 5-angled, 40-60 cm long and 25-40 cm wide, vestite in varying degree, but tending to be glabrate, lobes subulate-acuminate, 10-15 mm long; slender fruiting pedicels (10-) 20-30 (-80) mm long, usually re- flexed; berry 1-2 cm in diameter, spheric to spheric-oblong. Physalis maxima is characterized by its large fruiting calyces, long pedicels, white corollas, long, narrow, attenu- ate calyx lobes, and long hairs, especially on calyces, pedicels, petioles and upper stem. However, all these char- acteristics are variable, most particularly length of pedicels and calyx lobes, and the presence of long hairs. As the calyx lobes and pedicels become shorter, and the long hairs disappear, the plants approach those referred to P. nican- droides var. attenuata. The material in the Sloane Herbarium of the British Museum has only a few scattered long hairs. HS 146 15c (lower right) has one fruiting calyx 55 mm long and 40 mm 1967] Physalis — Waterfall 233 wide and another 55 mm long and 38 mm wide; flowering pedicels are 20-25 mm long; a calyx, just after the corolla has fallen, is 7 mm long with very narrow calyx lobes 5 mm long; fruiting pedicels are 22 mm long. Hs 294 53 has flowering pedicels 25-40 mm long; a young fruit has pedi- cels 50 mm long ; anthers yellow, 9 mm long. Neither of these, nor another, HS 296 7, apears to be the same as the one photographed as The Bailey Hortorium negative no. 5097, labelled “Type” with the annotation that the pasted strip below the specimen is in Miller's hand, and that on the reverse Solander pencilled “E. Vera Cruz, Houstoun, 1730". P. maxima has not been recognized as a part of the Mexi- can Flora since Miller originally described it from the Houstoun Vera Cruz collections. Most of the material seen has either been unnamed, or has been referred to P. nican- droides, SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: Cape Region, Oct. 18, 1902, Brandegee s. n. (UC) ; cHIAPAS: Hacienda Mon- serrate, Sept. 1926, Purpus 9262 (UC); CHIHUAHUA: milpa margin, LaCienegita, Río Mayo, Sept. 10, 1936, Gentry 2632 (F, GH, UC, US); COLIMA: subshrub over 1 meter high, alluvial plain, 18 km s of Manzanillo, Sept. 9, 1935, West 3531 (GH, UC); GUERRERO: llano, Santa Barbara, Coyuca, Sept. 30, 1935, Hinton 8503 (BM, F, GH, MIVB, NY, UC); Acapulco, Aug. 1866, Thiebaut 1098 (P); JALISCO: along dry ditch in forest, 10-16 miles ne of Autlan, Oct. 2, 1960, McVaugh 19767 (MICH); MEXICO: Rincon del Carmen, Sept. 16, 1932, Hinton 1740 (F); MICHOACÁN: Barranca de Río Cancita, 9 miles se of Apatzingan, Sept. 18, 1958, McVaugh 17981 (MICH, OKLA); MORELOS: near Yautepec, Aug. 27, 1903, Rose & Painter 6583 (US); NAYARIT: Maria Madre, Tres Marias Islands, Oct. 21, 1925, Ferris 5582 (US); roadside in semi-deciduous tropical forest on seaward — facing mountainsides, 12-13 miles s of Las Varas, Sept. 20, 1960, MecVaugh 19226 (MICH, OKLA); OAXACA: valley of Oaxaca, July 14, 1897, Pringle 7474 (GH, US, VT) ; SINALOA: Culiacán, Sept. 19, 1904, Brandegee s. n. (UC); sandy soil along river near Fuerte, Mar. 7, 1910, Rose et al 13576 (US); SONORA: Sierra de Alamos, Mar. 18, 1910, Rose et al 13059 (NY, US); VERACRUZ, Houston s. n. (Photo: MICH, NY); YUCATAN: 1917-1921, Gawmer 24310 (F, GH); ZACATECAS: pastured hills 5 miles s of Jalpa, Aug. 30, 1960, MeVaugh 18508 (MICH). COSTA RICA: Cartago, Nov. 1887, Cooper 6002 (vs). GUATEMALA: dry slopes on hill bordering Laga Retana, Dept. 234 Rhodora [Vol. 69 Jutiapa, Nov. 26, 1939, Steyermark 31975 (F). HONDURAS: sand- bar along Río Chiquito, Comayagua, Mar. 12-23, 1947, Standley et al 5956 (F); Quebrada de Santa Clara ene of El Zamorano, Aug. 6, 1949, Standley 22280 (F); moist thicket, vicinity of Tegucigalpa, Oct. 7, 1949 (F). NICARAGUA: near shore of Lake Nejapa, near Managua, Maxon et al 7560 (us); vicinity of Chichigalpa, July 12, 1947, Standley 11346 (F). 64. Physalis nicandroides Schlechtendal, Linnaea 19: 311. 1846. Annual, 1-2 meters tall, branched above; stems, petioles and pedi- cels viscid-pubescent with short, spreading, jointed hairs 0.3-0.8 mm long, some of which may be capitate-glandular; leaf blades ovate, often broadly so, more hairy below than above, veins of the lower surface often appressed-hairy, margins ccarsely and irregularly angulate-sinuate-dentate, bases unequal, sometimes joining the petiole 1 cm higher on one side than on the other; principal leaves 6-12 em long and 5-10 cm wide on petioles 2-4 em long; flowering calyx 5-7 mm long, divided into subulate-acuminate lobes 3-4 mm long, pedicels 2-3 mm long; corolla white, greenish-white or slightly yellowish, maculate, but often faded, or indistinct, 4-8 mm long and 6-8 mm wide, tube hairy within; anthers blue tinged or faded to yellowish, 1.5-2 mm long, on filaments 2-2.5 mm long; fruiting calyx strongly 5-angled, 30-45 mm long and 23-40 mm wide, the body nearly as broad as long, somewhat hairy, but tending to be glabrate, subulate- acuminate lobes 10-15 mm long; fruiting pedicels 5-25 mm long, reflexed; the dry, thin-walled berry nearly spheric, 12-22 mm in diameter, sessile on the invaginated base of the calyx. 64a. var. nicandroides Fruiting pedicels short and thick, 5-10 (-13) mm long and (1-) 1.5-2 mm thick. SELECTED COLLECTIONS. MEXICO: DURANGO: Durango, Nov. 1896, Palmer 914, (GH, UC, US) ; GUANAJUATO: near Dolores Hidalgo, Aug. 10, 1947, Kenoyer 1925 (GH); JALISCO: abundant, Cuesta de San Marcos, 15 km sse of Actlan de Juarez, Nov. 7, 1959, McVaugh & Koelz 342 (MICH, OKLA); low ground near Lake Chapala, 2 miles w of Tizapan, Aug. 18, 1961, Waterfall 16394 (F, OKLA, S, US); MEXICO: Luvianos, Temascaltepec, Aug. 26, 1932, Hinton 4529 (US); MICHOACAN: Morélia, Nov. 1912, Arsène 8576 (BM, F, GH, NY, US); NAYARIT: Tepic, Jan. 5 to Feb. 6, 1892, Palmer 2035 (GH, US); OAXACA: suburbs of Oaxaca, Oct. 1, 1894, Smith 708 (F, UC, US) ; SAN LUIS POTOSI: hill 20 miles s of intersection of Highways 80 & 57, 64 miles ne of San Luis Potosí, Aug. 20, 1959, Waterfall 15712 (BM, F, GH, MICH, NY, OKLA, P, SMU, UC); SINALOA: in 1927, Ortega 6694 (F); VERACRUZ: in 1855, Muller 564 (NY); YUCATAN: Common, Izamal, Gaumer 1446 (F, GH, US); LOCALITY UNDETERMINED: Hacienda de la Laguna, Sept. 1828, Schiede 606 (HAL: Type). COSTA RICA: forest, 1967] Physalis — Waterfall 235 Santa Maria de Dota, Dec. 26, 1925, Standley & Valerio 43321 (vs). GUATEMALA: Guatemala, in 1939, Aguilar 141 (F); near Lake Retana, Oct. 10, 1944, Melhus & Goodman 3784 (F). HONDURAS: brushy bank, Tegucigalpa, Sept. 6, 1951, Standley 28697a (F); El Banco, Comayagua, Sept. 29, 1951, Williams 18335 (F, US). 64b. var. attenuata Waterfall, var. nov. Pedicellis fructiferis tenuibus, 10-25 mm longis, plerumque 0.5- 1.0 mm latis. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: Todos Santos, Oct. 4, 1889, Brandegee s.n. (UC); CHIHUAHUA: southwestern Chihuahua, in 1885, Palmer 226 (GH,NY,US) ; GUERRERO: Ilano, Coyuco, July 10, 1935, Hinton 8035 (US) ; MORELOS: near Cuernavaca, July 29, 1906, Pringle 13776 (GH, MICH, OKLA, US, VT) ; NAYARIT: precipitous mountainside in forest, 7-8 miles w of Mazatan, Sept. 14, 1960, Mc- Vaugh 19027 (MICH); OAXACA: Coyula, Aug. 13, 1895, Smith 616 (GH); TAMAULIPAS: Victoria, in 1907, Palmer 24,0 (GH). COSTA RICA: La Balsa & Cateratas de San Ramon, Alajuela, Oct. 25, 1925, Brenes 4550 (NY); San Jose, Aug. 1932, Valerio 225 (F). GUATE- MALA: brushy rocky slope, between Zacapa and Chiquimula, Oct. 9, 1940, Standley 73840 (F); wet thicket, e of Cuilapa, Nov. 25, 1940, Standley 78317 (F). HONDURAS: Geguare, Morazan, Aug. 1934, Rodriguez 319 (F); moist thicket near Tegucigalpa, Oct. 7, 1949, Standley 24120 (F, US). Representatives of this taxon are not frequently collected in Mexico, the cited collections being the only ones seen from each of the Mexican states. It is more frequent in Central America. 65. Physalis cordata Miller, Gardener’s Dictionary, ed 8, Physalis 14. 1768. Annual, erect to sometimes decumbent, 15-80 cm long, appearing glabrous, but sparse, very short, antrorse trichomes usually present, rarely with a few long hairs, leaf blades ovate, or rarely broader, often acuminate apically and inequilateral basally, margins several- toothed; principal blades 3-8 em long and 2.5-8 cm wide on petioles 3-9 cm long; flowering calyx 5-6 mm long, divided about half way into narrow acuminate segments, pedicels 4-10 em long; corolla yellowish, maculate, hairy in the throat, 5-10 mm long and 9-22 mm wide when fully expanded; anthers bluish or greenish-blue, 1.8-3 mm long, on slightly to densely hairy filaments of about the same length; fruiting calyx 5-angled, often somewhat turbinate, cordate-based, glabrous, 25-43 mm long and 20-25 mm wide, the teeth narrow, acuminate, 7-10 mm long, sometimes somewhat porrect; berry nearly spheric, 7-15 mm in diameter; fruiting pedicels 10-25 mm long. P. cordata is another of Miller’s species not recognized in 236 Rhodora [Vol. 69 the American flora since its description based on Houstoun's Vera Cruz collections. Much of this material has been labelled P. turbinata in herbaria, but it is not that species as here interpreted, See discussion under P. turbinata. A few sheets belonging to this species-concept were seen by the author in the preparation of his earlier study of Phy- salis (Waterfall, 1958), and in making subsequent identifi- cations. They were referred by him to P. pubescens var. glabra (Michx.) Waterfall, based on P. obscura Michx. var. glabra Michx. The TYPE, studied in Paris in August, 1965, is on a small, pasted-on square of paper in the lower left hand side of a sheet, the remainder of whis is var. viscido- pubescens Michx., i.e, P. pubescens L. The type of var. glabra represents a few-haired extreme of P. pubescens, differing only in its degree of hairiness. It has an obviously hairy, short-toothed fruiting calyx on pedicels 5-8 mm long not a glabrous, long-toothed fruiting calyx on pedicels usually 15-25 mm long, or longer, as does P. cordata. In the Sloane Herbarium of the British Museum of Natu- ral History in London is preserved the Houstoun material from Vera Cruz on which Philip Miller based his P. cordata. It has the characteristic fruiting calyces, long fruiting peduncles, and long petioles of this species-concept, as is well-illustrated by a photograph, from negative 5099, in the herbarium of the Bailey Hortorium. SELECTED COLLECTIONS. MEXICO: CHIAPAS: Esperanza, Esquintla, Apr. 13, 1947, Matuda 16489 (F); GUERRERO: Acapulco, March 1895, Palmer 510 (GH); JALISCO: coastal plain, 4 miles n of Bahia Navidad, Nov. 1960, McVaugh 20843 (MICH, OKLA); NAYARIT: roadside and thickets between banana plantations, 15 km se of San Blas, Aug. 25, 1959, Feddema 958 (OKLA); OAXACA: Galeotti 1236 (P); SINALOA: Sante Fe, Dec. 1921, Ortega 4384 (US); VERACRUZ: negative 5089 of the Bailey Hortorium photographs of Philip Miller colleetions, is probably of one of the Houstoun collections, as is the specimen in the Sloane Herbarium of the British Museum labelled ns 294 54, listed in Ray’s Hist. 356 as “Alkekengi americanum anuum lamiifolia fructu cordato, Houst."; YUCATAN: Izamal, Gaumer 23170 (F). COSTA RICA: near Port Limon, May 27, 1911, Pittier 3637 (NY, US); near Turrialba, Cartago, Nov. 16, 1953, Heiser 3735 (F). 1967] Physalis — Waterfall 237 EL SALVADOR: Laguna de Maquigue, La Union, Feb. 18, 1922, Standley 20958 (GH, US) ; between San Martin and Laguna de Ilopan- go, San Salvador, Apr. 1, 1922, Standley 22560 (GH, US). GUATE- MALA: Quiriga, Izabal, May 15, 1922, Standley 24073 (GH, US) ; lower slopes of Volcán de Zunil, Feb. 3, 1941, Standley 85842 (F). HONDURAS: wooded swamp, near Juticalpa, Olancho, Mar. 5, 1949, Standley 17709 (F); Lancetilla Valley near Tela, Atlantida, Standley 55683 (F, US). NICARAGUA: moist thicket, Chichigalpa, July 12, 1947, Standley 11410 (F). PANAMA: wooded hillside near Catival, Colon, Jan. 9, 1924, Standley 30236. BAHAMAS: Spring Point, Ack- lin’s Island, Jan. 8, 1906, Brace 4256 (NY). BERMUDA: Brown & Britton 386 (GH, NY, UC). CUBA: south of Providencia, Aug. 19, 1930, Leon 14653 (GH, NY); near Cienfuegos, Jan. 25, 1903, Pringle 40 (vr). CURACAO: Patrick, Mar. 20, 1913, Britton & Shafer 3072 (NY). DOMINICAN REPUBLIC: Santo Domingo (Ciudad Trujillo), Feb. 24, 1930, Ekman 14329 (GH). GRENADA: Aug. 16, 1905, Broad- way s.n. (GH). GUADELOUPE: in 1898, Duss 3984 (NY). HAITI: south of Jean Rabel, Jan. 27, 1929, Leonard 12695 (NY). JAMAICA: Windsor Estate, Aug. 26, 1956, Proctor 15689 (GH). PUERTO RICO: near Catano, Feb. 14, 1927, Britton 8832 (NY). TORTOLA: Fishlock 299 (GH). VIRGIN GORDA: Jan. 6, 1919, Fishlock 200 (NY). 66. Physalis porrecta Waterfall, sp. nov. Planta annua, 0.4-1.3 m alta; caulibus sparse et inaequale vestitis; trichomatibus articulatis, plus minusve planis; foliis ovatis, margini- bus inaequaliter magnodentatis vel integerrimis, principalibus 5-11 em longis et 4-7 cm latis, petiolis (2-) 3-7 em longis; calycibus flori- feris 5-7 mm longis, lobis 1.5-4 mm longis anguste attenuatis, pedicel- lis 4-5 mm longis; corollis 7-10 mm longis et 12-15 mm latis, maculis pallido-fulvis, marginibus diffusis; antheris luteis vel violaceis, 1.8-2.3 mm longis, filamentis 2-4 mm longis; calycibus fructiferis glabris, pentangulatis, 3-5 em longis et 2.2-3 cm latis, lobis porrectis, rostris 7-13 mm longis; pedicellis fructiferis 6-15 mm longis; baccis 12-15 mm latis. TYPE: Alexander F. Skutch 2931 (us), edge of forest, vicinity of El General, Prov. San Jose, alt. 1160 meters, COSTA RICA, Nov. 1936; Isotypes: (GH, NY). SELECTED COLLECTIONS. MEXICO: oaxaca: vicinity of Cafetal Concordia, April 1, 1933, Morton & Makrinius 2556 (US). COSTA RICA: Skutch 2931, type, cited above; moist forest, El Munceo, Río Navarro, Provincia de Cartago, Mar. 6, 1926, Standley & Valerio 51093 (Us). GUATEMALA: clearing in forest, Volcan Zunil, Dept. Quezaltenango, Aug. 7, 1934, Skutch 970 (US); barranca above Duenas, Dept. Sacatepequez, Jan. 21, 1939, Standley 63252 (F, US); bushy rocky slope, near divide on road from Zacapa to Chiqui- mula, Oct. 9, 1940, Standley 73738 (F). 238 Rhodora [Vol. 69 P. porrecta is characterized by its usually abruptly beaked fruiting calyx, its yellowish corollas varying from immacu- late to obviously marked, but usually inconspicuously so, with brownish spots, their margins often diffused, anthers varying from yellow to bluish or greenish-blue, and with vestiture usually more abundant on one side of the stem than on the other, branches and petioles sometimes with the inner side vestite, and the outer side glabrous, or nearly so. 67. Physalis latiphysa Waterfall, Rhodora 60:169. 1958, Annual 15-45 cm tall, branched, more or less villous or glandular- villous, leaves ovatish, thin, entire to few-toothed, acuminate, princi- pal ones 5-7 cm long and 3-5 cm wide on petioles 1.5-7 em long; flowering calyces 3-4 mm long, divided half way into lanceolate or narrowly lanceolate lobes, on pedicels 3-8 mm long; corollas yellowish with small, dark maculations, 3-4 mm long; fruiting calyces strongly 5-angled, sparsely appressed-hairy, 2.5-4 em long and (2.5-) 3-4 em wide with linear-subulate lobes 7-10 mm long extending 5-7 mm be- yond the calyx-body; fruiting pedicels 1-1.5 cm long; berry 13-17 mm in diameter. SPECIMENS EXAMINED. MEXICO: soNORA: between Tepapa and Batuc, Sept. 28, 1934, Wiggins 7503 (MICH, US). The principal range of this species is in southern Arizona (Waterfall, 1958), but more material should be found in northern Sonora. 68. Physalis ignota Britton, Mem. Torr. Bot. Club 16:100-101. 1920; P. pentagona Blake, Contr. U.S. Natl. Herb. 24:20-21. 1922. Erect annual, 10 cm to 1 meter tall, usually densely covered with short, more or less glutinous, nearly straight to antrorsely curled, multicellular hairs; leaf blades ovate to rhombic-ovate, sometimes acuminate, margins entire to repand-denate, surfaces with vesti- ture similar to that of the stem, but more sparse, or nearly glabrous with concentrations along the midrib and principal veins, principal blades 4-12 cm long and 3-8 cm wide on petioles 2-7 cm long; flower- ing calyx oblongish to campanulate, 5-8 mm long and 3-6 mm wide at base of lobes, upper one-fourth to two-thirds divided into ovate- lanceolate to narrowly lanceolate-attenuate lobes; flowering pedicels 3-7 mm long; corollas yellowish, immaculate, campanulate, 6-10 mm long and 4-10 mm wide, sparsely hairy in the throat; anthers yellow or bluish tinged, oblong, 2-2.5 mm long, on filaments of about equal length; fruiting calyces strongly 5-angled, usually densely and evenly covered with short, erect, jointed hairs; fruiting pedicels 7-15 mm long; berry spheric to ovoid, 9-15 mm long, capitate-glandular, as is the inside of the fruiting calyx; gynophore 1-2 mm long. 1967] Physalis — Waterfall 239 SELECTED COLLECTIONS. COSTA RICA: Las Canas, Guana- caste, Nov. 12, 1953, Heiser 3722 (F). EL SALVADOR: sand along river, San Miguel, Feb. 24, 1922, Standley 21092 (GH, NY, US); sand along stream, San Salvador, Mar. 30, 1922, Standley 23156 (GH, US, NY). GUATEMALA: dry riverbed, Los Amates, Izabel, May 9, 1919, Blake 7313 (Type collection of P. pentagona Blake: US); sand bars, Los Amates, May 24, 1922, Standley 24418 (GH, NY, US); gravelly slope, Jutiapa, Oct. 24, 1940, Standley 75054 (F); Rio Selguapa, Comayagua, Mar. 21, 1945, Rodriguez 2530 (F); weedy field, El Za- morano, Morazan, July 26, 1949, Standley 21740 (F). NICARAGUA: summit of Sierra de Managua, May 14, 1947, Standley 8682 (F); wet thicket, Chichigalpa, July 12, 1947, Standley 11155 (F). PANAMA: San Carlos, Dec. 5, 1938, Allen 1136 (GH, NY, US). SAN SALVA- DOR: Aug. 1922, Calderon 1134 (Us). CUBA: Cienfuegos, Oct. 27, 1928, Jack 6566 (NY); Playa de Mariano Habana, Nov. 28, 1915, Leon 6864 (NY); Nuevitas, Mar. 31, 1909, Shafer 1127 (NY). TO BE CONTINUED Volume 69, No. 778, including pages 121-240, was issued June 30, 1967. 240 Rhodora [Vol. 69 Sar * sg Lapidario ELLE S , 2*8 à H ` a Sida Rai Hofmelstería Gentryi Wircins Contr. Dalley Herb, 4: BS, 1350 sep. Mofzatateria siurtsata nauciesta io 3742 Jis. Marca St{gante abore Sto ROO ikio avs m, Tame,” itr magum inea to otirra Oe Joo", tatouttve Leafy budh wits Werrr eme. tle leaves. Plate 1338. Pleurocoronis gentryi Holotype (Ds) Gentry 3472. The above plate replaces the one incorrectly shown in Rhodora Vol. 69 on p. 37. FARLOW REFERENCE LIBRARY OCT 9 4 1967 Dodova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL 3 STUART EIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER > Associate Editors ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. Vol. 69 June-September, 1967 No. 779 CONTENTS: A Preliminary Revision of Tragia (Euphorbiaceae) in the United States Kim I. Miller and Grady L. Webster ...... 241 Seventeenth Report of the Committee on Plant Distribution R. C. Bean, A. F. Hill, R. J. Eaton and S. K. Harris ...... 306 Correction in the 16th Report R. C. Bean, A. F. Hill, R. J. Eaton and S, K. Harris s. 818 Physalis in Mexico, Central America and the West Indies DU. T. Waterfall moraro ivesi ienis ient exccenas cem eee estan niae 319 Pine Hills Elymus George L: Church ..................... eren 330 (Continued on Inside Cover) The Nem England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. CONTENTS: — continued Studies in the Eupatorieae (Compositae) IV R. M. King crececccccccsscccsssessseesseceseeeseceeseececsesesecesseceeeeeeseeneeenseesens 352 Further Changes and Additions to the Flora of Porto Rico and the Virgin Islands Brother Alain H. Liogier .......... 372 Notes from the Pringle Herbarium, II Frank C. Seymowr ....... essent tenant nnne nnne 377 Halesia carolina L. in Kentucky, Indiana and Ohio Edward W. Chester ...ccccsccccsecssccesssssseecceesssecceessscseeccsecsseceeseaeaees 380 Petalostemon foliosus in Alabama Jerry M. Baskin and Carol Caudle .................. 1 383 Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 69 July-September No. 779 A PRELIMINARY REVISION OF TRAGIA (EUPHORBIACEAE) IN THE UNITED STATES! KIM I. MILLER AND GRADY L. WEBSTER The genus Tragia (Crotonoideae — Acalypheae) com- prises about 150 species of nettle-like plants which are widely distributed in tropical and subtropical or warm-tem- perate regions of both the New World and Old World. It is by far the largest and most diverse of the 19 genera of subtribe Plukenetiinae recognized by Pax and Hoffman (1919), and various generic segregates have been proposed during the past 150 years. In the New World, the center of diversity of Tragia is in Brazil, and only about 20 species have been recorded from North America. These include some taxa of particular phylogenetic interest, however, for T. bailloniana from Mexico and Central America may be one of the most primi- tive species in the genus. The present study has been de- liberately restricted in scope to include only those taxa represented in the United States, because it seems to us that the collections available from Mexico are at present inadequate to serve as the basis of even a preliminary treat- 1. This study incorporates part of a doctoral dissertation submitted by the senior author in partial fulfillment of the requirements for the Doctor of Philosophy degree at Purdue University. Investigations were supported by grants from the Purdue Research Foundation and (for the junior author) the National Science Foundation and the John Simon Guggenheim Memorial Foundation while at the Gray Herba- rium, Harvard University. The assistance of Dr. Lillian Miller and Miss Jeanie Taylor is gratefully acknowledged. 241 242 Rhodora [Vol. 69 ment (at least in the absence of field work in critical areas of the Mexican plateau). Mexican specimens of the T. nepetifolia complex have been examined in order to relate them to those from the United States, and it is hoped that a revision of the Mexican species can eventually be added to the present work in order to account for all of the North American taxa. NOMENCLATURAL HISTORY The genus Tragia was originally established by Plumier (1703) on the basis of the American species designated by Linnaeus as T. volubilis. The first comprehensive revision of the group was that of Baillon (1858), who recognized no less than 10 genera within the concept of Tragia held by A. Jussieu (1824). Doubtless Baillon's disposition was ex- treme, but some of the taxa such as Leptorachis, Cteno- meria, Bia, and Zuckertia are very distinctive and have been regarded as worthy of generic status by a few workers. The first detailed enumeration of all the taxa of Tragia in anything approaching a modern sense was that of Muel- ler Argoviensis (1865, 1866), who accepted Leptorachis as a distinct genus and recognized 11 sections to accommodate the 49 species then known for Tragia. In the ‘Flora Brasil- iensis' (1874), Mueller reduced Leptorachis to a section of Tragia, a disposition which was followed by Bentham (1880) and Pax (1890). In the last general revision of Tragia, Pax and Hoffmann (1919) reduced the number of sections to 9 while accepting 123 species, 58 of these American and 65 Old World; only one species, T. volubilis, was recorded for both the Eastern and Western Hemispheres, and it is presumed to be Amer- ican in origin. Pax and Hoffmann rather slavishly followed Mueller in placing the U. S. taxa of Tragia into two sec- tions: Leucandra and Eutragia; they thus perpetuated Mueller's error in putting specimens of T. ramosa in two different sections under two species names (T. ramosa and T. nepetifolia). However, by recognizing T. teucriifolia [T. brevispica] and T. amblyodonta as distinct, they raised the number of species occurring in the United States to 11. 1967] Tragia — Miller and Webster 243 Johnston (1962) has reviewed some of the systematic problems presented by the taxa in the Southwest, which he refers to as ‘noseburns’ (from the effects of the stinging foliage on grazing stock). He has correctly pointed out that the distinction between sections Leucandra and Tragia, at least as construed by Pax and Hoffmann, is very weak. It seems clear that this supposed sectional distinction, at least as applied to taxa in the United States, cannot be main- tained. On the basis of his inability to find stable, geographically correlated morphological characters in the Southwestern noseburns, Johnston has placed all of the plants in the com- plex under a single species name, T. nepetifolia Cav., and relegated no less than 18 names to synonymy. It is indeed evident that the noseburns are bewilderingly variable in habit and leaf shape as well as in number of flower parts. The plasticity of individual plants in responding to environ- mental variations and the large amount of intra- and inter- population diversity, together with the often fragmentary nature of available herbarium specimens, results in a baf- fling pattern of variation which frankly cannot always be untangled even after the most intensive study. Nevertheless, a detailed analysis of specimens from throughout the southern United States, combined with field studies in the Gulf States and Texas (Miller 1964), has shown that there are morphologically recognizable taxa among the noseburns which do have distinctive geographi- cal ranges. Consequently, we are unable to concur with Johnston's conclusion that the Southwestern noseburns all belong to a single species. Amalgamation of all these popu- lations into one species appears to be contradicted by the geographical correlations noted above as well as by (inter alia) the following facts: (1) the chromosome complement in at least one taxon, T. amblyodonta, appears to be de- caploid (n — 55), whereas the other noseburns investigated cytologically are tetraploids (n = 22) ; (2) peculiar horned fruits similar to those of the widespread tropical vine T. volubilis occur only on plants with the leaf shape and flow- 244 Rhodora [Vol. 69 ers of T. brevispica; (3) the female calyx-lobes of plants assigned to T. betonicifolia are on the average much longer than those of related species; (4) typical specimens of T. ramosa have styles which are slender and much smoother (scarcely or not papillate) than those of related taxa, in- cluding T. nepetifolia. On the basis of all of the evidence, we have concluded that 5 species of noseburns should be distinguished where Johnston has recognized only one; this results in a total of 13 species for the United States instead of the 9 or 10 which would be recorded in accordance with his concept. It must be admitted that the system presented here is not an especially tidy one, as a considerable number (ca. 5-10% ) of the specimens examined remain dubious or unassignable, and the disposition of some of the taxa (in particular T. leptophylla) is still doubtful. However, the degree of dif- ficulty encountered in annotating the often fragmentary herbarium specimens or in identifying them by means of a key should not be confused with the problem of whether there are really any distinctions between the wild popula- tions. Our conclusion that there are 5 U. S. species of nose- burns in the T. nepetifolia complex is not founded on irrefutable proofs, but it is at least a reasonable hypothesis which is deserving of criticism and verification by syste- matists in the Southwest. MORPHOLOGY GROWTH FORM. All of the U. S. species of Tragia are perennial herbs with a woody taproot and alternate leaves. The stems may be either erect, decumbent, or twining; within a single species both twining and non-twining types may occur. At one end of the spectrum, such species as T. nigricans and T, saxicola have never been observed to twine, while on the other hand T. glanduligera and T. cordata are always twining. In T. ramosa and some related species there appears to be considerable amplitude of habit, as otherwise similar plants may be either erect, decumbent or twining. These differences, which may reflect ecotypic 1967] Tragia — Miller and Webster 245 variation in some instances and simple adaptability in others, need to be more carefully studied. LEAVES. The leaves of Tragia are quite variable in both size and shape, but within a broad range are sufficiently dis- tinctive to present useful taxonomic characters. The peti- oles, which vary in length from less than 1 mm in T. urens to over 8 em in T. cordata, have 5 free vascular traces. The leaf-blades vary from linear and entire in some forms of T. urens to serrate or dentate in most species, while in the rare T. laciniata the blades are tripartite with coarsely toothed lobes. Stomata occur on both upper and lower sur- faces, but are more numerous below (they may be restricted to areas along the major veins on the upper surface). Most of the stomata are of the ‘rubiaceous’ type as defined by Metcalfe and Chalk (1950), but sometimes they may ap- proach the ‘cruciferous’ type (Miller 1964). The stomata of T. amblyodonta are outstanding in being strikingly larger than those of related species: mean values are 26.6, for the upper epidermis and 22.0, for the lower, as opposed to 20.71 and 18.3, in T. nigricans, the species with the next largest stomata. It seems probable that this larger stomatal size of T. amblyodonta is correlated with its high degree of polyploidy. The most striking feature of the leaves of Tragia is the presence of characteristic stinging hairs, which have been described by Rittershausen (1892) and Knoll (1905). These hairs consist of an elongated central cell, which according to Knoll is of subepidermal origin, surrounded by 3-5 ‘jack- et-cells'. An extremely sharp-pointed crystal at the distal end of the central cell acts as the piercing agent when the end of the hair is touched. The stinging hairs are much alike in all U. S. species of Tragia, although there is a tre- mendous variation in the painrulness of the sting; T. am- blyodonta appears to be the worst offender, whereas T. urens, despite its suggestive epithet, stings little or not at all. Possibly the most important variable is the quantity and nature of the proteinaceous poisoning element which is released when the end wall of the cell pulls back and per- 246 Rhodora [Vol. 69 mits the sharp crystal tip to penetrate the skin. Stinging hairs are also found on the stems, inflorescence axes, caly- ces, and ovaries, and may be very prominent on the ripened fruit. Unicellular hairs are also present on the stems and leaves, usually intermixed with the stinging ones. Long-stalked multicellular glandular hairs occur on the inflorescences of T. glanduligera intermixed with both other types; this con- dition is common in various tropical taxa but is not known in other U. S. species, although a few minute short-stalked glandular hairs may be found on the inflorescence. INFLORESCENCE. In all of the U. S. taxa of Tragia, the inflorescence is an androgynous raceme (fig. 1) with one female flower (rarely two) at the base and 3-20 (or more) male flowers at the distal nodes. Our representatives have all flowers solitary in the axil of the subtending bract, but some tropical species may have several male flowers per bract. The staminate bracts are entire and more or less cucullate, while the female may be either entire or 3-lobed. The position of the racemes of the U. S. species is con- sistently opposite the leaves, as pointed out by Johnston (1962), so that the racemes are actually terminal and the main stem axis is thus a sympodium. In some tropical spe- cies the situation is more complicated because the racemes terminate lateral branches (e.g., in T. mexicana), and the inflorescence pattern in temperate species is apparently de- rived by reduction. Tragia urens shows an approach to the condition in some of the tropical species in that the racemes are borne terminally on lateral leafy branches as well as at the upper nodes of the main stem; in this species, there- fore, the lower part of the stem appears to be monopodial while the upper part is sympodial. FLOWERS. The staminate pedicel is articulate, and the basal portion persists after the fall of the male calyx. The male pedicel is supplied with 2 vascular traces and the fe- male with 5 (Miller 1964). The flowers are apetalous, the perianth being represented by a gamophyllous calyx; the union of the calyx-lobes may be slight, but it is always 1967] Tragia — Miller and Webster 247 Plate 1356. Fig. 1. Longitudinal section of inflorescence, Tragia nigricans, showing androgynous arrangement of flowers, X 20 (K. & L. Miller 1173). Fig. 2. Longitudinal section of staminate flower, Tragia nigricans, showing connate filaments, X 60 (K. & L. Miller 1173). Fig. 3. Longitudinal section of pistillate flower, Tragia urtict- folia, showing papillate stigmatic surface, stigmatoid tissue in stylar column and single anatropous ovule, X 30 (K. & L. Miller 848). Fig. 4. Longitudinal section of ovary, Tragia urticifolia, showing anatropous ovule and obturator, X 100 (K. & L. Miller 848). 248 Rhodora [Vol. 69 Fig. 5. Terminal staminate flower, Tragia ramosa (K. & L. Miller 1341). Fig. 6. Staminate flower, Tragia urticifolia, from near the mid- dle of the staminate portion of the inflorescence (K. & L. Miller 848). 1967] Tragia — Miller and Webster 249 evident, and the lobes never disarticulate separately from the receptacle. The staminate flowers (fig. 2) have mostly 3 or 4 valvate calyx-lobes per flower, but the number may fluctuate within a single inflorescence, depending on the species. The pistillate calyx nearly always consists of 2 whorls of 3 lobes each, which unlike the male are imbricate in the bud. Both male and female calyx-lobes usually possess 3 veins which are however more elaborately branched in the female. ANDROECIUM. In the vast majority of plants of Tragia in the United States, the androecium consists of 3 to 5 stamens (fig. 6), the number varying over this range within a single inflorescence in many species. However, T. smallii and T. urens ordinarily have only 2 stamens (very rarely 3). The filaments, which may be thickened and fleshy or rather slender, are supplied with a single vascular trace; in most species, they are connate only at the base, but in T. nigri- cans they may be fused for half or more their length. There is usually a small pistillode in the center of the flower be- tween the connate bases of the stamens. The stamens and calyx-lobes of the termina] male flower are commonly more numerous than those of the axillary flowers of the inflorescence, there being usually 4 to 6 calyx- lobes and 5 to 8 stamens (fig. 5). Johnston (1962) described this terminal flower as “probably representing the mon- strous joining of 2 or more flowers". However, such termi- nal flowers were found to occur more or less regularly in most of the U. S. species, although of course they might not be evident on partially-grown racemes. It is rather difficult, therefore, to see how the terminal flowers are any more “abnormal” than are e.g. the first-produced cyathia of Eu- phorbia, which may often have an increased number of floral parts (Croizat, 1942). It is possible, as Johnston suggests, that the larger number of parts of terminal male flowers may have added to the taxonomic confusion, since both Mueller (1866) and Pax and Hoffmann (1919) sepa- rated T. ramosa (as T. stylaris) into a separate section 50 Fig. 7. Fig. 8. Rhodora [Vol. 69 Pollen grain, Tragia ramosa (K. & L. Miller 1341). Pollen grain, Tragia smallii (K. & L. Miller 865). 1967] Tragia — Miller and Webster 251 from T. nepetifolia on a supposed difference in stamen num- ber. POLLEN GRAINS. 'The most extensive previously pub- lished treatment of the microspores of Tragia is that of Punt (1962), who described the variation among 8 species, and also contributed some interesting speculations about taxonomic relationships in the subtribe Plukenetiinae. Punt distributed the 8 species of Tragia studied into 4 mor- phological groups: (1) inaperturate, intectate, pilate (T. fallax, T. sellowiana): the ‘Tragia fallax type’; (2) trico:- pate, intectate, pilate (T. geraniifolia, T. ramosa, T. tristis, T. volubilis) : the ‘Tragia tristis subtype’ (groups 2, 3, and 4 all being placed in the ‘Plukenetia type’); (3) tricolpo- rate, intectate, reticulate (T. stolziana): the ‘Plukenetia verrucosa subtype’; and (4) tricolpate, tectate, psilate (T. capensis): the ‘Plukenetia volubilis subtype’. TABLE 1. POLLEN SIZE IN NORTH AMERICAN SPECIES OF Tragia” Diameter in Section TRAGIA Voucher polar view (u) range mean T. amblyodonta Miller & Miller 1178 21.3-26.4 23.4 T. betonicifolia Miller & Miller 1350 19.6-23.0 21:2 T. brevispica Cory 54778 18.7-22.1 20.7 T. cordata Demaree 15878 21.83-25.5 24.0 T. glanduligera Lundell & Lundell 10665 18.7-25.2 20.1 T. laciniata Peebles et al. 5641 17.0-20.4 19.0 T. nepetifolia Tucker 2449 17.0-20.4 18.4 T. nigricans Miller & Miller 1173 23.0-25.5 24.5 T. ramosa Miller & Miller 1167 17.9-22.1 19.9 T. saxicola Curtiss 2517 22.1-26.4 24.8 T. urticifolia Miller & Miller 848 29:1:282 24.6 Section LEPTOBOTRYS T. small Miller & Miller 865 25.5-28.1 26.9 T. wrens Miller & Miller 776 20.4-26.4 22.9 Miller (1964) investigated the pollen of the 13 taxa occurring in the United States, using the glycerine jelly method of Wodehouse (1935) rather than the acetolysis method of Punt (Table 1). This led to the discovery that 2. [Table 1] Parameter estimates are based on measurements of 30 grains per specimen. 252 Rhodora [Vol. 69 the tricolpate grains of the U. S. taxa have operculate colpi, a feature not mentioned by Punt, whose acetolyzed preparations evidently did not show any traces of the op- ercula. Because of the distinctness of the pila at outermost focus, most of the U. S. species (fig. 7) would fall into Punt's Tragia tristis subtype. However, it is actually extremely difficult to decide if these grains are truly intectate, as the pila may be proximally fused. Two U. S. species, Tragia smallii and T. urens, were found to differ from any of those described by Punt. In these two plants of the southeastern United States the pol- len grains (fig. 8) are rounded-triangular in optical section and have very broad pore-like colpi with net-like incrusta- tions apparently representing the operculum of other spe- cies. The pila are furthermore partially united into an extremely fine reticulum, so that in some respects the grains might fit into Punt's Plukenetia volubilis subtype. Examination of other neotropical species of Tragia in- dicates that most of them clearly fall into one or another of the groups listed above. Particularly interesting is T. bailloniana, which alone among the American taxa observed is distinctly reticulate, with even-margined colpi. It would therefore go into the Plukenetia verrucosa subtype adjacent to such Atrican species as T. siolziana and T. pungens. Punt (1962) has suggested that sect. Bia might perhaps be best treated as a distinct genus because of the inaper- turate grains. However, in view of the rather wide spec- trum of morphological variation which has already come to light, it would appear that the pollen of this section is not much more aberrant than that of some sections with colpate grains. Furthermore, such a genus would be ex- tremely diffieult to distinguish on the basis of floral char- acters. Because of the imperfect correlation between pollen characters and taxonomic boundaries, it still seems best to retain all of the sections of Tragia within a single genus. GYNOECIUM. The gynoecium in Tragia is of a type rather widespread in the subfamily Crotonoideae, with 3 united carpels and a single anatropous ovule in each locule (fig. 3). 1967] Tragia — Miller and Webster 253 The styles are elongated, often with a conspicuously papil- late stigmatic surface, and may be united up to half or more their length. The conducting tissue in the stylar canal is prolonged into each locule as a cap-like obturator ap- pressed to the micropylar end of the ovule. An extension of the obturator, which is made up of closely packed fila- mentous cells, projects into the micropyle and is in contact with the nucellus. As in most genera of Euphorbiaceae, the obturator of Tragia is transient and atrophies during de- velopment of the seed. The solitary pendent anatropous ovule (fig. 4) is typical of that found in many genera of the Crotonoideae, tribe Acalypheae: it is crassinucellate, with two thick integu- ments, both of which contribute to the formation of the micropyle. A peculiarity which the Tragia ovule shares with that of many Acalypheae is the very massive inner integument and the expanded chalazal region which is supplied with distinct vascular traces. The nucellus in Eu- phorbiaceous ovules of this kind lacks a beak, and as if in compensation the obturator often provides a direct contact between the tip of the nucellus and the transmitting tissue in the style. FRUIT. The fertilized gynoecium of Tragia develops into a dry thin-walled capsule which dehisces more or less ex- plosively at maturity. At dehiscence the cocci first separate septicidally and from the columella; the lateral walls split open along the line where the carpel and columella were attached; and the cocci also split on the abaxial side (lo- culicidally), allowing the seed to be thrown from the sepa- rated coccus. The vasculated exocarp of the coccus is in most species formidably armed with a dense array of sting- ing hairs; the non-vasculated endocarp is composed of radially elongated sclerified cells. SEEDS. At the time of formation of the proembryo, the nucellus has been reduced to a plug adjacent to the mi- cropyle, and the inner integument and chalaza are still massive. However, as the embryo develops the inner in- tegument becomes partially replaced by endosperm; in the 254 Rhodora [Vol. 69 mature seed its outer epidermis has become the hard pali- sade layer. The collapsed layers of the outer integument become flattened on the seed surface. The mature seeds are nearly spherical, with a smooth or slightly roughened more or less mottled coat. CYTOLOGY. Miller (1963) first published counts of chro- mosome numbers in the genus Tragia for 5 U. S. species. With the addition of one species and correction of identifi- cations, verified counts for 6 species are presented here (Table 2). It is obvious that the results presented here are barely a start, since only one population has been counted for each species, and 7 of the 13 U. S. species are still entirely unstudied. However, the chromosome studies do warrant the following observations. TABLE 2. CHROMOSOME NUMBERS IN Tragia (Miller, 1963)’ Section TRAGIA Voucher 2n T. amblyodonta Miller & Miller 1178 ca. 110 T. brevispica Miller & Miller 1076 44 T. ramosa Miller & Miller 1167 44 T. urticifolia Miller & Miller 848 44 Section LEPTOBOTRYS T. smallü Miller & Miller 882 44 T. urens Miller & Miller 776 44 The six counts reported clearly establish that the base number in Tragia sect. Tragia and sect. Leptobotrys is x = 11. No counts are yet reported from species belonging to the other seven sections, and such information is particu- larly to be desired since the results might help to evaluate the status of taxa such as sect. Bia. Since no chromosome counts have been reported for any of the other 18 genera of Plukenetiinae recognized by Pax and Hoffmann, knowledge of the base number is not of any assistance in determining intergeneric relationships. The finding of n = 55 in T. amblyodonta is interesting, since the other five U. S. taxa counted all agree in having 3. [Table 2] Two of the chromosome numbers reported in Miller (1963) were based on misdetermined specimens. The count reported for T. glanduligera was actually made on material of T. brevispica, and that reported for T. ramosa represented T. amblyodonta. 1967] Tragia — Miller and Webster 255 n = 22. This provides some support for our decision to recognize T. amblyodonta as a distinct species, rather than sinking it in T. nepetifolia as was done by Johnston (1962). The greater stinging capacity of T. amblyodonta would ap- pear to be correlated with this in some way (larger cell size?). RELATIONSHIPS In the treatment of Pax and Hoffmann (1919), all of the U. S. species of Tragia were placed in sect. ‘Eutragia’ ex- cept for T. ramosa and T. brevispica, which went into sect. Leucandra. As Johnston (1962) has pointed out, this dis- position of T. ramosa, and T. brevispica was based on an untenable supposed distinction in stamen number; since stamen numbers of 3-6 occur in all the noseburns, they cannot be separated into two sections by means of this character. Probably sect. Leucandra, if it is a valid taxon at all, will have to be characterized on some other basis; and in any event it is not represented by any species in the United States. On the other hand, two U. S. species, T. urens and T. smallii, do seem to be rather sharply set off from their con- geners by male flowers with only Z stamens that produce à different sort of pollen; Although Mueller (1866) recog- nized a sect. Leptobotrys for T. urens, Pax and Hoffman (1919) merged it with their sect. Eutragia. We believe that this group originally recognized by Baillon and Mueller should be resuscitated in order to emphasize the consider- able morphological gap between T. urens, T. smallii and the Southwestern noseburns. The remaining 11 U. S. species must all be referred to sect. Tragia, but they do not seem to represent a cohesive phyletic unit, as at least 2 species may have their closest relationships outside our area (in Mexico). The nearest species to T. cordata may be T. affinis of western and cen- tral Mexico; the latter is strikingly similar in habit and leaf-shape, even though its male flowers (with 13-18 stamens) are very different. Tragia glanduligera is a pri- marily Mexican species which only enters extreme southern 256 Rhodora [Vol. 69 Texas, and its immediate relationships must be with some taxon to the south. The core of 9 species left after peeling off the elements mentioned does present the picture of a probably mono- phyletic — though certainly not a homogeneous — group. It may be analyzed into 1 cluster and 3 isolated species: (1) the noseburns proper, comprising T. amblyodonta, T. betonicifolia, T. brevispica, T. nepetifolia, T. ramosa, and T. saxicola; (2) T. nigricans; (3) T. laciniata; and (4) T. urticifolia. Relationships in the noseburn complex remain to be exactly elucidated, partly because of its tremendous unexplained variability. Hybridization between the nose- burns — all of which are at least partly sympatric except for T. saxicola — may possibly account for some of the dif- ficulty, although we have found it impossible to distinguish morphological aberrancy due to interspecific crossing from the protean plasticity of unhybridized populations. Culti- vation under controlled conditions and experimental crosses will be necessary to have even a chance of resolving this problem, It does appear that ecological preferences may act to separate the taxa to some extent: for example, in Texas T. betonicifolia and T. urticifolia tend to occur on sandy or granitic soils, whereas T. amblyodonta and T. brevispica are usually found on limestone. A further complication in assessing relationships among the noseburns is that several species also extend into Mex- ico. At least T. amblyodonta, T. brevispica, and T. ramosa are known to go south of the Rio Grande; and the popula- tion of T. nepetifolia in Arizona represents only a small northernmost extension of a primarily Mexican population. Although we have examined a considerable number of Mexican specimens, we have not attempted to account for the Mexican representatives of the noseburn complex in this paper, beyond establishing the typical element associ- ated with the name T. nepetifolia. It is possible that thor- ough study of the Mexican taxa may reveal the name T. nepetifolia to be camouflaging as many different species as has proved to be the case in Texas. 1967] Tragia — Miller and Webster 257 Pax and Hoffmann (1919) remarked on the parallelism evident between the temperate or subtropical representa- tives of sect. Tragia in North and South America, pointing out 4 pairs of vicarious species: T. pinnata — T. laciniata, T. bahiensis — T. glanduligera, T. geraniifolia — T. nepeti- folia, and T. tenella — T. amblyodonta (southern taxon listed first in each pair). There is also a certain resemblance between T. paxii of Argentina (Lourteig & O'Donell, 1941) and T. brevispica of Texas, particularly with regard to the crested fruits. It would appear that all these cases are indeed parallelisms, in the sense that the taxa involved are in each instance more closely related to species in their own area than to the antipodal vicariant; on the other hand, the overall resemblance between the taxa of Texas — north- ern Mexico with those of Argentina probably does reflect a genetic continuity in the past analogous with the bipolar distribution of Prosopis (Johnston, 1940) and a number of other subtropical taxa of the New World. SYSTEMATIC TREATMENT Approximately 2,500 specimens of Tragia were examined during this study. We wish to express our appreciation for the courtesies extended by the directors of the institutions listed below in allowing us to borrow or examine their ma- terial. ARIZ University of Arizona, Tucson; BM British Museum (Natural History), London; FLAS University of Florida, Gainesville; G Conserv- atoire et Jardin Botaniques, Geneva; GA University of Georgia, Athens; GH Gray Herbarium, Harvard University, Cambridge; IA State University of Iowa, Iowa City; 1sc Iowa State College, Ames; K Royal Botanic Gardens, Kew; KANU University of Kansas, Law- rence; KY University of Kentucky, Lexington; LL Texas Research Foundation, Renner; M Botanische Staatssammlung, Munich; MA Instituto Cavaniles, Madrid; MICH University of Michigan, Ann Arbor; MO Missouri Botanical Garden, St. Louis; Ncsc North Caro- lina State University, Raleigh; NMC State University of New Mex- ico, University Park; Ny New York Botanical Garden, New York; P Muséum National d'Histoire Naturelle, Paris; PH Academy of Natural Sciences, Philadelphia; PUL Purdue University, Lafayette; SMU Southern Methodist University, Dallas; TEX University of Texas, 258 Rhodora [Vol. 69 Austin; UARK University of Arkansas, Fayetteville; UNM University of New Mexico, Albuquerque; UsF University of South Florida, Tampa. Future collectors of Tragia are hereby admonished to take careful notes on variation seen within natural popula- tions, with particular reference to the following characters: (1) proportion of plants showing twining habit; (2) viru- lence of stinging hairs (this demands a certain amount of noblesse oblige on the part of the investigator) ; (3) color of male flowers; (4) presence of winged fruits (in popula- tions from Texas). A more complete list of specimen citations has been pro- vided by Miller (1964), together with statistical analyses of some of the data. A tabulation of the historically im- portant Lindheimer collections is presented here as an ap- pendix. Measurements of specimens have been made on dried material except for the flowers, which were examined after being boiled. Voucher specimens for pollen and chro- mosome studies are deposited at the Kriebel Herbarium, Purdue University (PUL). Synonyms and descriptions refer primarily to the taxa as represented in the U. S., and do not necessarily apply to the genus as a whole. TRAGIA L. Sp. Pl. 2: 980. 1753; Gen. PI. ed. 5, 421. 1754. Perennial herbs, sometimes becoming suffrutescent, erect or de- cumbent to trailing or twining. Stems solitary to many from the woody crown of the taproot, alternately branched; pubescence of stiff stinging hairs intermingled with soft spreading hairs. Leaves alternate, petiolate or sessile; margins entire to serrate, coarsely toothed to divided; stipules lanceolate, acute, entire, ciliate, usually persistent. Plants monoecious; inflorescences [in most U. S. taxa] opposite the leaves at the upper nodes (actually terminal, but soon surpassed by the branch from the axil of the subtending leaf) ; racemes androgynous, the lower 1 (2) flowers pistillate, the remaining 2-20+ staminate; individual flowers bracteolate. Flowers apetalous; calyx- lobes 3-6; disk absent in the the flowers of both sexes. Staminate flowers pedicellate, abscission zone below the middle of the pedicel, the basal portion persistent; stamens 2-6 (8) [more in some extra- limital taxa]; filaments connate at least at the base, anthers free; pistillode small. Pistillate flowers pedicellate, abscission zone below the middle of the pedicel, the entire structure with the calyx and columella usually persistent; staminodia absent; ovary usually 1967] Tragia — Miller and Webster 259 of 3 carpels, subglobose, hispid to densely hispid with stiff stinging hairs; ovules solitary in each locule, anatropous, with two integu- ments, obturator penetrating into the exostome in contact with the nucellus; styles 3, spreading or coiling outward at anthesis, united at the base and up to 1/2 or more their length, stigmas smooth or papillate. Fruit an explosively dehiscent capsule, the cocci separat- ing from a persistent columella; columella with 3 interlocular points apically. Seeds one per locule; seed coat dry, crustaceous, smooth or slightly rough, brownish black or with a tawny mottling, ecaruncu- late; endosperm whitish; embryo straight, cotyledons foliaceous and considerably broader than the terete radicle. KEY TO THE TAXA A. Stamens 3-4 (rarely 5); calyx-lobes of staminate flowers usually 3, sometimes 4 (rarely 5) ......MHHHHeH Section Tragia B. Stamens 2; calyx-lobes of staminate flowers usually 4 or 5 (rarely 6) ....cccccecsccsosssrsscssccerssensssnscnsssnseaasonnees Section Leptobotrys Section TRAGIA 1. Leaf blades undivided .................. ee y (2) 1. Leaf blades 3-divided ........................ e 10. T. laciniata 2. Inflorescence without conspicuous stalked glandular hairs .... (3) 2. Inflorescence with conspicuous stalked glandular hairs .......- wu restarts thet —— 1. T. glanduligera 3. Filaments of the stamens connate only at the base «+++ (4) 3. Filaments of stamens connate 1/3 — 1/2 or more their length; leaf blades deeply and sharply toothed, cuneate at the base .... ecu T 11. T. migricans 4. Persistent base of the staminate pedicel usually conspicuously shorter than the subtending bract ................ eH (5) 4. Persistent base of the staminate pedicel usually almost equal- ling in length or exceeding the subtending bract; stigmatic surfaces conspicuously papillate; ovary densely pubescent with stinging hairs .......-. Be T. urticifolia 5. Calyx-lobes of the pistillate flower shorter than the gynoe- "o BENE. (6) [^1 Calyx-lobes of the pistillate flower longer than the gynoecium; bracts of the staminate flowers strikingly long, (1-) 2 mm; staminate flowers 14-75 per inflorescence .... 4. T. betonicifolia 6. Leaf blades rounded to acute, scarcely acuminate, not broadly ovate with a deeply cordate base; seeds less than 4 mm long; plants twining or not ........ neret enne nnne (7) 6. Leaf blades broadly ovate, more than half as long as broad, acuminate, deeply cordate at the base; seeds more than 4 mm long; plants always twining ........- HH 2. T. cordata 7. Leaf blades ovate to triangular or lanceolate, the broader ones more or less cordate to truncate at base ........... (8) 260 Rhodora [Vol. 69 7. Leaf blades suborbicular to oblong, truncate or rounded at the base; stigmatic surfaces only slightly papillate; plants never twining (restricted to extreme southern Florida) .................- ————— ——————ÉÁÉUAARERRRREERRERERREEREEE 9. T. saxicola 8. Stigmatie surfaces distinctly papillate, styles connate only at the base or slightly higher; leaf blades broadly triangular to OVATE RRRRRRRRRRENMMMM (9) 8. Stigmatic surfaces not or scarcely papillate (slightly so in aber- rant specimens with narrowly lanceolate leaf-blades); styles connate 1/3 to 1/2 their length and usually terminating in slender, recurved tips; leaf blades usually 2-4 times as long AS broad ......... sese eene eere serre tiris 7. T. ramosa 9. Plants not densely pubescent on the stem and leaves ............ (10) 9. Plants copiously to densely pubescent, especially on young stems, the entire plant having a greyish-green cast; leaf blades cre- nate to serrate or dentate, bases truncate to sagittate; ovary very densely pubescent with stinging hairs .. 8. T. amblyodonta 10. Leaf blades coarsely serrate-dentate; staminate calyx often red- ish; staminate flowers 8-35 per inflorescence; ovary pubescent with stinging hairs (but usually not very densely); fruits monomorphic (all wingless and 3-seeded); plants erect to decumbent ......... seen nette 6. T. nepetifolia 10. Leaf blades serrate; staminate calyx never reddish; staminate flowers 3-6 (-10) per inflorescence; ovary densely pubescent with stinging hairs; fruits usually dimorphic, either 3-seeded and wingless or l-seeded and crested or winged; plants very often twining .........sseeseseeeeeeeen eene tnnt 5. T. brevispica Section LEPTOBOTRYS l. Leaf blades blunt-toothed or lobed, bases of lower leaves trun- cate; seeds 4.0-4.5 mm long; inflorescences opposite the leaves throughout. RMMMMMMMM 12. T. smallii 1. Leaf blades entire, blunt-toothed, or lobed, cuneate at the base; seeds 3.0-4.0 mm long; inflorescences terminating leafy lateral branches, opposite the leaves only toward the tip of the main LS ci.sercorssvenssserensesenenssssnecssstensenssconsescesoussensessenenseseparsacs 13. T. urens Section 1. TRAGIA Sect. Eutragia Muell. Arg. Linnaea 34: 182. 1865; DC. Prodr. 15 (2): 932. 1866. Perennial herbs, erect, decumbent, or twining; stems and foliage with more or less severely stinging hairs; leaves serrate or dentate, or sometimes deeply lobed or divided; inflorescences opposite the leaves; female flowers 1 or 2 per raceme; staminate flower with 3 or 4 (rarely 5) calyx-lobes; stamens. 3-6, filaments more or less cylindri- cal, somewhat thickened, free or connate; pollen grains distinctly colpate; pistillode subglobose; pistillate flower with 6 (very rarely 5) 1967] Tragia — Miller and Webster 261 calyx-lobes; styles free or connate below, smooth or papillate; seeds 2-5.5 mm long. LECTOTYPE: Tragia volubilis L. (designated by Britton and Wil- son, Sci. Surv. Porto Rico 5: 491-492. 1924). This group of about 30 species almost equally divided between North America and South America is here con- strued nearly as defined by Pax and Hoffmann (1919), except that 7. wrens and T. smallii are removed to sect. Leptobotrys, while T. ramosa and T. brevispica are trans- ferred here from sect. Leucandra. It seems probable that additional neotropical species of sect. Leucandra (those with a low stamen number) will be transferred to sect. Tragia when a more thorough study is made. The more or less weedy ‘noseburn’ species placed to- gether in T. nepetifolia by Johnston (1962) are here dis- tributed into species 5-8; T. betonicifolia, although treated by Shinners (1958) as a variety of T. urticifolia, has often been mistaken for T. ramosa and may be regarded as à peripheral member of the 'noseburn' group. At least the first two species, T. glanduligera and T. cordata, appear to be more closely related to extra-limital taxa than to any U. S. species; in an eventual revision of all of the taxa of sect. Tragia they would therefore probably go into sep- arate subsections or series from the remainder (species 3-11). 1. Tragia glanduligera Pax & Hoffm. Pflanzenreich IV. 147. IX [Heft 68]: 55. 1919. (Fig. 17) LECTOTYPE: Yucatan, Gawmer 731 (GH!; isotypes F, MO!). Stems few to many from the crown of the woody taproot, 1-2 mm in diameter near the base, up to a meter or more long, green to purplish-green, many-branched, trailing and twining; upper internodes 12-85 mm long, lower internodes 20-35 mm long. Leaf blades elliptic to ovate, 2.5-4.0 em long, 1.5-2.0 em broad, apically acute to acuminate, basally truncate to cordate, margins crenate to serrate, thin, green, pubescent, ciliate; petioles 6-22 mm long; stipules lanceolate to nar- rowly ovate, entire, 1.5-4.5 mm long, 0.5-1.2 mm broad at the base, persistent, ciliate, abaxially pubescent. Racemes with the lowermost 1 (2) nodes pistillate, the remaining 10-30 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 0.9-2.0 mm long, ciliate, acute, 262 Rhodora [Vol. 69 Ramen NSS be SS M v. Pa LAN ^N Ks TUA (NES NS yi MC zs. Der e o re GN NDA AS: OPA] s £g E7) [LY © Fig. 17. Habit, Tragia glanduligera (M. C. Johnston 542235). 1967] Tragia — Miller and Webster 263 T. NEPETIFOLIA T. URTICIFOLIA Map 1. Distribution of Tragia glanduligera and Tragia mepeti- folia. Map 2. Distribution of Tragia cordata. Map 3. Distribution of Tragia urticifolia. 264 Rhodora [Vol. 69 entire or three-lobed, abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 0.5-1.5 mm long, pubes- cent, entire. Glandular hairs present over entire inflorescence. Stami- nate flowers: pedicels slender, 1-2.0 mm long, lower persistent part ca. 0.3-0.7 mm long; calyx-lobes 3, broadly oblanceolate to obo- vate, 0.7-1.2 mm long, abaxially pubescent, entire, spreading to re- flexed at anthesis; stamens 3; filaments thickened and fleshy, 0.2-0.4 mm long, connate basally. Pistillate flowers: pedicels 1.0-2.3 mm long, becoming 3-7 mm in fruit; calyx-lobes 6, lanceolate to narrowly ovate, 0.7-1.5 mm long at anthesis, 1.2-2.3 mm long in fruit, acute, entire, ciliate, abaxially sparsely pubescent; styles connate about 1/3 their length, stigmatic surfaces papillate. Fruit 2-2.5 mm long, 4-5 mm broad; columella 1.5-2 mm long; seeds nearly spherical, 1.9-2.2 mm long, brownish-black when mature. DISTRIBUTION: restricted in the U. S. to extreme south- ern Texas in dry sandy soil (Map 1). REPRESENTATIVE SPECIMENS: — TEXAS: CAMERON CO., Las Palmas Plantation ca. 4 mi SW of Brownsville in a palm grove, Correll 14,848 (LL); Resaca del Rancho Viejo, Cory 51454 (NY, SMU); 2.4 mi N of Lozano, Johnston 2758 (SMU); 3 mi east of Harlingen near Arroyo Colorado, Johnston 3640 (TEX); Camp Perry (Boy Scout Camp), Johnston 542235 (TEX); palm grove S of Brownsville, Lun- dell & Lundell 10002 (smu); N of Los Fresnos in mesquite brush, Lundell & Lundell 10665 (LL, SMU, TEX); Brownsville, Tharp 1870 (TEX). HIDALGO CO., La Joya, Walker s.n. (2/11/1942) (TEX). KENEDY CO., Norias Division of King Ranch a few miles W of La Calandria, Johnston 541069 (SMU). WILLACY CO., Tharp s.n. (26/VI/1941) (TEX). Pax and Hoffmann (op. cit. 56) cited collections of Ber- landier, Gaumer, and Valdez when describing this species. The Berlandier specimens are somewhat atypical and it is not certain that they are conspecific. The Valdez collection is perhaps equally representative, but as the Gaumer col- lection is more widely distributed it has been chosen as the lectotype. Tragia glanduligera is a primarily Mexican species of the Gulf coastal lowlands between Yucatan and the Rio Grande, crossing into Texas only in four of the southern- most counties. Specimens from further north which have been reported as T. glanduligera (Jones, Rowell, & Johnston 1967] Tragia — Miller and Webster 265 Fig. 18. Habit, Tragia cordata (M. E. Wharton 9369). 266 Rhodora [Vol. 69 1961) actually belong to T. brevispica, a species which clearly differs in its non-glandular inflorescence, more sharply toothed leaves, and larger dimorphic fruits. The glandular inflorescence and diminutive seeds (smaller than those of any other U. S. species) easily distinguish T. glanduligera from other taxa occurring in Texas, 2. Tragia cordata Michx. Fl. Bor. Amer. 2: 176. 1803. (Fig. 18) TYPE: Kentucky, ‘entre Danville et Beards’ town’, Michaux (P!). Tragia macrocarpa Willd. Spec. Pl. 4: 323, 1805; based on the same type. Tragia Michauxii Baill. Étud. Gén. Euphorb. 460. 1858; based on the same type. Stems solitary or few from the crown of the woody taproot, 1.5-3 mm in diameter near the base, up to 1.5 m or more long, green to yellowish-green, few to many-branched, trailing and twining; upper internodes 2-7.5 cm long, lower internodes 3-10 cm long. Leaf blades broadly ovate, 4.5-13 cm long, 3.6-10 em broad, apically acuminate, basally cordate, margins coarsely serrate, thick, green, pubescent, ciliate; petioles 1.5-8.5 em long; stipules lanceolate to narrowly ovate, entire, 1.8-10 mm long, 0.4-2.1 mm broad at the base, persistent, ciliate, abaxially sparsely pubescent. Racemes with the single lower- most node pistillate, the remaining 20-60 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1.6-2 mm long, ciliate, acute, entire or three-lobed, abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 1.5-2 mm long, pubes- cent, entire. Staminate flowers: pedicels slender, 1.5-2.2 mm long, lower persistent part ca. 0.7-1 mm long; calyx-lobes 3(91%) (4), oblanceolate to narrowly obovate, 0.7-1.5 mm long, abaxially pubes- cent, entire, spreading to reflexed at anthesis; stamens 3 (96%) (4); filaments thickened and fleshy, 0.2-0.5 mm long, connate basally. Pistillate flowers: pedicels 1-1.5 mm long, becoming 2.5-3 mm in fruit; calyx-lobes 6 (96%) (-7), elliptic to ovate, 1.5-2 mm long at anthesis, 2-3 mm long in fruit, acuminate, entire, ciliate, abaxially pubescent; styles connate 1/4-1/3 their length, stigmatic surfaces papillate. Fruit 5-7 mm long, 11-13 mm broad; columella 2.2-3.4 mm long; seeds nearly spherical, 4.3-5.3 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: Indiana to Missouri and east Texas more or less following the eastern deciduous forests (Map 2). REPRESENTATIVE SPECIMENS: — ALABAMA: JEFFERSON Co., Bir- mingham, Earle 2074 (NY). ARKANSAS: HOT SPRINGS CO., Magnet Cove, Demaree 224,07 (SMU). FLORIDA: JACKSON CO. Florida 1967] Tragia — Miller and Webster 267 Caverns State Park, Beck 225 (FLAS, GA). GEORGIA: LEE Co., along Fowlton Creek near Armena, Thorne & Muenscher 8413 (GA). ILLI- NOIS: POPE CO. Golconda, Forbes (F). INDIANA: CRAWFORD CO., roadside 1/4 mi W of Leavenworth, Deam 18583 (NY). KENTUCKY: LYON CO. Skinframe Creek, Eggleston s.n. (14/VI/1909) (NY). MISSOURI: MCDONALD CO. rocky ground, uncommon, Bush s.n. (24/VII/1893) (NY). OKLAHOMA: DELAWARE CO., Little Kansas, DEMAREE 22407 (SMU). TENNESSEE: DAVIDSON CO., woods near Nashville, Gattinger 2518 (KANU, NY, PH, SMU, UARK). TEXAS: BOWIE C0., 3 mi N of Texarkana, Correll 15249 (LL). Tragia cordata, which attains the northernmost latitude of any species in the genus, is not easily confused with any of the other U. S. taxa: it is set apart by the vining habit, large deeply cordate leaves, and large fruits and seeds. Its closest affinities would appear to be with Mexican species such as T. affinis Rob. & Greenm. 3. Tragia urticifolia Michx. Fl. Bor. Amer. 2: 176. 1803. (Figs. 6, 14, 19) Type: Georgia [ex Michaux, loc. cit.], Michaux (P!). Stems solitary or few from the crown of the woody taproot, 3-4 mm in diameter near the base, 25-65 cm tall, green to yellowish- green, simple to few-branched, erect to decumbent; upper internodes 8-30 mm long, lower internodes 12-60 mm long. Leaf blades triangu- lar-lanceolate to narrowly ovate, 2.7-6.7 em long, 0.9-3.2 em broad, apically acute, basally cordate to truncate, petioles 3-11(-17) mm long; stipules lanceolate to narrowly cordate, entire, 2-5 mm long, 0.6-1.6 mm broad at the base, persistent, ciliate, abaxially pubescent. Racemes with the lowermost 1 (2) nodes pistillate, the remaining 11-40 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1-1.5 mm long, ciliate, acute, entire or three-lobed, glabrous or abaxial- ly sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 0.7-1.5 mm long, pubescent, entire. Staminate flow- ers: pedicels slender, 1.5-2 mm long, lower persistent part ca. 1-1.8 mm long, as long as or exceeding the bract; calyx-lobes 3 (93%) (4), oblanceolate to narrowly ovate, 1.2-2.1 mm long, abaxially pubescent, entire, spreading to reflexed at anthesis; stamens 3; fila- ments thickened and fleshy, 0.3-0.8 mm long, connate basally. Pistil- late flowers: pedicels 0.5-1 mm long, becoming 8-4 mm in fruit; calyx-lobes (5) 6 (86%), lanceolate to ovate, 1.3-2.2 mm long at anthesis, 2-3 mm long in fruit, acute, entire, ciliate, rarely abaxially pubescent; styles connate 1/3-1/2 their length, stigmatic surfaces 268 Rhodora [Vol. 69 Figs. 9 and 10. Early stage in development of winged fruit, Tragia brevispica (K. & L. Miller 1076). Fig. 11. Mature one-seeded winged fruit, Tragia brevispica (K. & L. Miller 1076). Fig. Fig. 1178). Fig. Fig. Fig. 12. 13. 14. 15. 16. Pistillate flower, Tragia ramosa (K. & L. Miller 1341). Pistillate flower, Tragia amblyodonta (K. & L. Miller Pistillate flower, Tragia urticifolia (K. & L. Miller 848). Pistillate flower, Tragia smallii (K. & L. Miller 865). Pistillate flower, Tragia urens (K. & L. Miller 776). Tragia — Miller and Webster 269 1967] SN c — rs v z Habit, Tragia urticifolia (K. & L. Miller 848). Fig. 19. 270 Rhodora [Vol. 69 very papillate. Fruit 3.5-4.5 mm long, 7-8 mm wide; columella 2-2.8 mm long; seeds nearly spherical, 3-4 mm long, brownish-black when mature. DISTRIBUTION: North Carolina south to northern Florida and west to Texas and Arkansas, typically in dry sandy soil in open woods and fields (Map 3). REPRESENTATIVE SPECIMENS: — ALABAMA: TALLADEGA CO., Child- ersberg, Hood 4680 (FLAS). ARKANSAS: GARLAND CO. near Hot Springs, Palmer 29069 (UARK). FLORIDA: OKALOOSA Co., Crest- view, West s.n. (27/IX/1950) (FLAS). GEORGIA: BAKER CO., bank of Cooleewahee Creek at Newton, Thorne 5695 (G^, 14). LOUISIANA: WINN PARISH, 10 mi W of Winnfield, Webster & Wilbur 3266 (NCSC, SMU). MISSISSIPPI: LAUDERDALE CO. 18 mi SW of Quitman on Rt. 18, Miller & Miller 872 (PUL). NORTH CAROLINA: GRANVILLE co., 0.6 mi W and 2 mi N of Hester, Ahles & Leisner 17499 (TEX). SOUTH CAROLINA: EDGEFIELD CO., ca. 3 mi. W of Owdoms, Miller & Miller 844 (PUL). TEXAS: BASTROP CO., Bastrop State Park, Cor- rell & Johnston 17433 (LL): SALINE CO., Redhills Lake, Miller & Miller 915 (PUL). Tragia urticifolia is a very distinct species although it has been misinterpreted by some writers of floras. Gleason (1952), in 'Britton & Brown's New Illustrated Flora', re- versed the distinction between T. urticifolia and T. nepeti- folia when he stated that ‘Tragia nepetifolia is similar in habit but the persistent pedicel-base of the staminate flow- ers is as long as the bract, the short filaments are as wide as the anthers, and the styles are scarcely connate.’ These statements pertain to T. urticifolia and not to T. nepeti- folia, as he intended. Gleason also attributed a much wider distribution to T. urticifolia than actually exists. The only species which is at all close to T. urticifolia is T. betonicifolia, which is largely allopatric, overlapping with T. urticifolia only in eastern Texas and in Arkansas. Although T. betonicifolia seems sufficiently distinct to be regarded as a separate species (as discussed below), it doubtless may be regarded as the western vicariant of T. urticifolia. 1967] Tragia — Miller and Webster 271 = Sy NN LAN ZA Fig. 20. Habit, Tragia betonicifolia (J. Hancin 1311). 272 Rhodora [Vol. 69 4. Tragia betonicifolia Nutt. Trans. Amer, Phil. Soc. (n.s.) 5: 178. 1837. (Fig. 20) TYPE: ‘Arkansas’, Red River, Nuttall (BM!). Tragia urticifolia var. texana Shinners, Field & Lab. 19: 183. 1951. TYPE: 6 mi NW of Grapevine, Tarrant Co., Texas, Shinners 11133 (SMU!) Stems few to many from the crown of the woody taproot, 1-2.5 mm in diameter near the base, 20-50 cm tall, green to yellowish-green, few-branched to many-branched, erect or decumbent to trailing; upper internodes 10-30 mm long, lower internodes 15-45 mm long. Leaf blades ovate, ovate-lanceolate to triangular-lanceolate, 1.5-5.5 cm long, 0.9-3.5 em broad, apically acute, basally cordate to truncate, margins sharply serrate, thick, green, pubescent; petioles 0.8-3.5 em long; stipules ovate to ovate-lanceolate, attenuate, entire 2.5-6 mm long, 0.5-2.5 mm broad at the base, persistent, ciliate, abaxially sparsely pubescent. Racemes with the lowermost 1 (2) nodes pistil- late, the remaining 14-75 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1.5-2 mm long, ciliate, acute, entire or three- lobed, abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 1-2 mm long, pubescent, entire. Stami- nate flowers: pedicels slender, 0.7-1 mm long, lower persistent part ca. 0.3-0.6 mm long; calyx-lobes 3 (68%)-4 (26%) (5), ob- lanceolate to narrowly obovate, 1.2-2.3 mm long, abaxially pubescent, entire, spreading to reflexed at anthesis; stamens 3 (71% )-4 (27%) (5); filaments thickened, 0.4-1 mm long, connate basally, Pistillate flowers: pedicels 0.7-1 mm long, becoming 3-4 mm in fruit; calyx- lobes 6, lanceolate, 1.8-3.0 mm long at anthesis, 3-5 mm long in fruit, acute to attenuate, entire, ciliate, abaxially sparsely pubescent; styles connate only at the base or up to 1/3 their length, stigmatic surfaces papillate. Fruit 4-5 mm long, 7-9 mm broad; columella 2.5-3 mm long; seeds nearly spherical, 3-4 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: open woods, fields, and roadsides, usually on dry sandy soil, Texas north through Oklahoma and Arkansas to Kansas and Missouri (Map 4). REPRESENTATIVE SPECIMENS: — ARKANSAS: FRANKLIN CO., Horse- shoe Mountain (Little Short Mountain), Iltis & Iltis 5400 (MICH, SMU). RANDOLPH CO., Ravenden Springs, clay thickets, Demaree 29217 (SMU). KANSAS: MCPHERSON co., McPherson State Park, McGreg- gor 10739 (KANU). Ellsworth Co., 1 mi E of Terra Cotta, McGreg- gor 12396 (KANU). MISSOURI: JASPER CO., near old mines, Webb City, Demaree 40340 (SMU); 2 mi NE of Webb City, Miller & Miller 1350 (PUL). OKLAHOMA: ATOKA CoO., 10 mi S of Kiowa, Matlock . ee . o S e . "e M e ° * D e . le e. . M ee e e ê t. 7 RS . ° ° : m" Srt ur A T"*.* * 6 di e | pete 4 be jest G *. «8s T RAMOSA ero . Map 4. Distribution of Tragia betonicifolia. 273 Map 5. Distribution of Tragia brevispica and Tragia saxicola. Map 6. Distribution of Tragia ramosa. 274 Rhodora [Vol. 69 114 (SMU). JOHNSTON CO., 10 mi S of Tishomingo along Highway 99, Robbins 3066 (SMU). TEXAS: LEE CO., 2.5 mi SW of Giddings, Cory 55763 (SMU). TARRANT CO, E of Euless on Bear Creek N of High- way 183, Whitehouse 16138 (SMU). Although Shinners described this species as a variety of T. urticifolia, it is readily separable from that plant by the length of the bracts subtending the staminate flowers and by the pistillate calyx-lobes. In T. betonicifolia the persis- tent base of the staminate pedicel is greatly exceeded by the subtending bract, whereas in T. urticifolia the pedicel base usually equals in length or exceeds the subtending bract. The calyx-lobes of the pistillate flower in T. betonici- folia are usually more lanceolate and always longer than those of T. urticifolia (and of most related taxa) ; the en- tire gynoecium of T. betonicifolia is enveloped by the calyx at anthesis, but in T. urticifolia and most other U. S. spe- cies of Tragia the top of the ovary and the styles are pro- truding at this stage. Both Fernald (1950) and Gleason (1952) referred the Missouri populations of T. betonicifolia and T. ramosa to T. urticifolia, but the latter species is not known to enter Missouri. The specimens from the New York Botanical Garden designated as having been used for the illustrations in Gleason's manual actually represent two Missouri col- lections of T, betonicifolia. 5. Tragia brevispica Engelm. & Gray, Jour. Boston Soc. Nat. Hist. 5: 262. 1845. (Figs. 9-11, 21)* TYPE: Texas, ‘Black, clayey soil, in the prairies west of the Brazos. 4. The typification of Tragia brevispica presents unusual intricacies even for this genus, and it is with some reluctance that we have adopted it in place of the name T. teucriifolia. Engelmann and Gray, in their original publication, based the name on Lindheimer 307, which according to Blankinship (1907) belongs to Fascicle II of the Lindheimer collections (see the appendix to this paper), made in 1844. The only specimen examined with a label corresponding to the locality data cited by Engelmann and Gray is in the Paris Herbarium, and it seems possible that this sheet was sent on exchange by Gray, who did not bother to retain the original label. The specimens at 1967] Tragia — Miller and Webster 275 ra, M. ANNs: Fig. 21. Habit, Tragia brevispica (E. Whitehouse 16302). 276 Rhodora [Vol. 69 May — July’ [1844], Lindheimer 307 [Fascicle II] (Gu, lectotype!; MO, NY, P, SMU, isotypes!) Tragia teucriifolia Scheele, Linnaea 25: 586. 1852. TYPE: Texas, ‘am Rande von Gebiisch und im Pflanzengestrüpp an sonnigen Stellen bei Neubraunfels, Julio — Septbr., 1846’, Lindheimer 522 [Fascicle III] (GH, Mo, TEX!). Tragia nepetifolia Steucriifolia (Scheele) Muell. Arg. in DC. Prodr. 15(2): 934, 1866. Stems solitary or few from the crown of the woody taproot, 1-2 mm in diameter near the base, 20 cm tall up to a meter or more long, green to yellowish-green, few to many-branched, erect to trailing and twining; upper internodes 10-20 mm long, lower internodes 20-35 mm long. Leaf blades broadly triangular, 1.8-5 cm long, 1.3-3 em broad, apically acute, basally truncate to shallowly cordate, margins finely and shallowly serrate, thin, green, pubescent, ciliate; petioles 7-30 mm long; stipules lanceolate to ovate, entire, 1.4-3 mm long, 0.7-1.4 mm broad at the base, persistent, ciliate. Racemes with the lower- most node pistillate, the remaining 3-6(10) nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1-1.4 mm long, ciliate, acute, entire or three-lobed; bracts of the staminate flowers lanceolate, sub- cucullate, ca. 1-1.8 mm long, ciliate, entire. Staminate flowers: pedi- cels slender, 1-1.5 mm long, lower persistent part ca. 0.4-0.7 mm long; calyx-lobes 3 (25%)-4 (70%) (5), ovate to obovate, 1-1.5 mm long, abaxially pubescent, entire, spreading to reflexed at anthesis; stamens 3 (3196), 4 (54%)-5 (15%); filaments slender, thickened and fleshy, 0.3-0.6 mm long, connate basally. Pistillate flowers: pedicels 0.5-1 mm long, becoming 2-4 mm in fruit; calyx-lobes 6, ovate, 1.3-2.0 mm long at anthesis, 1.8-3.5 mm long in fruit, acute, entire, ciliate or abaxially pubescent; styles connate only at the base or up to 1/3 their length; stigmatic surfaces papillate. Fruits often of two kinds on the same plant; some 3-seeded, typically dehiscing, 3.5-4 mm long, 6.5-7.0 mm broad; other fruits with 1-3 distinct wings, more or less indehiscent; columella 1.8-2.8 mm long; seeds nearly spherical, 2.8-3.8 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: limestone regions of central Texas, ex- tending into northern Mexico (Map 5). Paris, Harvard, Missouri Botanical Garden, and New York herbaria all seem to belong to the same gathering, representing a plant which is more narrow-leaved than other individuals collected by Lindheimer in later years, There are no fruits on any of the sheets of Lindheimer 307, and the flowers are not sufficiently characteristic so that we can be absolutely positive of the specific identity of T. brevispica with T. teucriifolia. 1967] Tragia — Miller and Webster DIN REPRESENTATIVE SPECIMENS: — TEXAS: COLLIN CO., 6 mi S of Fris- co on McDonald Farm, Wagner 114 (SMU). ELLIS CO. along White Rock Creek, 11 mi S of Italy, Correll & Johnston 17366 (LL). MC- LENNAN CO., Smith 240 (TEX). REAL CO. 4 mi N of Leakey, Miller & Miller 1165 (PUL). SAN PATRICIO CO., Welder Wildlife Refuge, Miller & Miller 1076 (PUL). TARRANT CO., Fort Worth, Bluebird Avenue, Cory 54778 (KANU, LL, SMU). TRAVIS CO. Austin, Albers’ herb garden, Warnock W1076 (TEX); Pecan Springs, E of Austin, Lundell & Lundell 8939 (LL). Mueller (1866) treated T. teucriifolia as a variety of his all-inclusive T. nepetifolia, but may not have examined any specimens of the type collection, as he cited Lindheimer 307 only under his var. ramosa. Johnston (1962) does not even recognize the taxon at the subspecific level. Our own concept, in contrast, is essentially in agreement with that of Pax and Hoffmann (1919), who accepted T. teucriifolia as a distinct species and placed it closest to T. ramosa. The commonest forms of T. brevispica appear very different from T. ramosa because of their twining stems, broader rather bluntly toothed leaves, papillate styles, and dimor- phic fruits. However, occasional specimens are encountered which are difficult to place in one species or the other. Be- cause of the extraordinary plasticity of T. ramosa, it is not easy to exclude the possibility that such ‘aberrant’ or 'in- termediate' specimens may simply be imperfect herbarium samples of that widespread species. On the other hand, T. brevispica and T. ramosa are widely sympatric over much of central Texas, and it would not be surprising if occasional hybridization occurred. Such specimens as e.g. Tharp et al. 51-953 from NW Travis County (TEX 89561) might pos- sibly represent interspecific crosses. Without careful field observations, however, we would not want to positively specify a hybrid origin for any specimens examined in this study. Despite the confusion of T. brevispica with T. ramosa, it may actually be more closely related to the Mexican 7. nepetifolia, which in our area is known only from southern Arizona. The Arizona race of T. nepetifolia has leaves and 278 Rhodora [Vol. 69 female flowers rather similar to those of T. brevispica, but clearly differs in its reddish male flowers, non-twining hab- it, and monomorphic fruits. The south Texas and Mexican twining species T. glanduligera is not sympatric with T. brevispica as far as can be determined from available rec- ords, but their ranges approach closely in the Corpus Christi area and it seems possible that they may overlap slightly. However, 7. glanduligera is certainly not closely related to T. brevispica, and is easily distinguished by its glandular inflorescence, more finely toothed and less deeply cordate leaves, and smaller seeds. The growth habit of T. brevispica appears to show con- siderable lability in response to environmental conditions. Plants growing in open areas are more or less erect (the type collection being of this kind), while those in shade are normally trailing with much longer internodes and broader leaves. Both the erect and trailing forms develop the characteristic winged fruits. 6. Tragia nepetifolia Cav. Icones Pl. 6: 37, pl. 557 fig. 1. 1801. (Fig. 22) TYPE: Mexico, Hidalgo, ‘inter Ixmiquilpan et Zimapan', Née (MA, photographs!). Tragia nepetifolia X genuina. Muell, Arg. in DC. Prodr. 15(2): 934. 1866. Stems few to many from the crown of the woody taproot, 1-2 mm in diameter near the base, 15-40 cm tall, green to purplish-green, few to many-branched, erect, decumbent or trailing; upper internodes 10-25 mm long, lower internodes 15-50 mm long. Leaf blades trian- gular to ovate, 1.8-4 cm long, 0.9-2.7 cm broad, apically acute, basally truncate to cordate, margins irregularly coarsely toothed, thin, green, pubescent, ciliate; petioles 3-25 mm long; stipules lanceolate to nar- rowly ovate, entire, 2-5 mm long, 0.7-2 mm broad at the base, persistent, ciliate, abaxially pubescent. Racemes with the single lower- most node pistillate, the remaining 8-35 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1.3-1.6 mm long, ciliate, acute, entire or three-lobed, abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 1.3-1.6 mm long, pubes- cent, entire. Staminate flowers: pedicels slender, 1.4-1.7 mm long, lower persistent part ca. 0.5-0.7 mm long; calyx-lobes 3(79%) -4 (21%), elliptic to ovate, 1-2 mm long, glabrous or abaxially pubescent, entire, spreading to reflexed at anthesis; stamens (2) 3(88%)-4 219 4505 Wey trat TEA O SeN tia A LIZZ as Tragia — Miller and Webster 1967] Habit, Tragia mepetifolia (T. H. Kearney & R. H. Pig: 22. Peebles 10320). 280 Rhodora [Vol. 69 (11%); filaments thickened and fleshy, 0.3-0.6 mm long, connate basally. Pistillate flowers: pedicels 1.3-1.6 mm long, becoming 2.9- 3.3 mm in fruit; calyx-lobes 6, lanceolate to narrowly ovate, 1.4-2.3 mm long at anthesis, 2-3.5 mm long in fruit, acute, entire, ciliate, abaxially pubescent; styles united only at the base or up to 1/3 their leneth, stigmatic surfaces papillate. Fruit 3-4.5 mm long, 6-8 mm broad; columella 2-2.7 mm long; seeds nearly spherical, 3-4 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: southern Arizona and possibly New Mex- ico south into central and western Mexico (Map 1). REPRESENTATIVE SPECIMENS: — ARIZONA: GRAHAM CO., Turkey Creek, 6 mi W of Point of Pines, 70 mi E of San Carlos, Bohrer 439 (ARIZ). MARICOPA CO., Camp Creek, Harrison & Hastings 6615 (ARIZ). PIMA CO. Sabino Canyon, Santa Catalina Mountains, Kearney & Peebles 10320 (ARIZ); Baboquivari Canyon, Kearney & Peebles 14982 (ARIZ). SANTA CRUZ CO., Patagonia Mountains, cliffs E of the old Flux Mine, ca. 5000 feet, Niles, Mason & Williams 172 (ARIZ). YUMA co., Castle Dome Mountains, Arnold s.n. (16/1X/1937) (ARIZ). MEXICO. CHIHUAHUA: above Tinaja Canyon, W of Vieja Casas Grandes, Tucker 2449 (ARIZ). HIDALGO: hills above El Salto, Pringle 11303 (MICH). JALISCO: 18 mi S of Valparaiso, McVaugh 17685 (MICH). As pointed out by McVaugh (1961), the name T. nepeti- folia. has been ‘loosely interpreted by recent botanists... to apply to a wide variety of herbaceous plants with the stems erect or nearly so... from Arizona and eastern United States to Guatemala.’ The responsibility for this state of affairs rests mainly on Mueller (1866), who rather ineptly combined most of the taxa in the noseburn complex under the name T. nepetifolia while at the same time plac- ing part of the specimens of T. ramosa under a different Species name in another section of the genus. Johnston (1962) rectified the confusion with regard to T. ramosa but by combining all the noseburn taxa arrived at a con- cept of T. nepetifolia even more inclusive than that of Mueller. Part of the difficulty in defining the limits of T. nepeti- folia has been due to uncertainty with regard to its typifi- cation. Fortunately, through the courtesy of Dr. F. Bellot Rodriguez, director of the Instituto Cavanilles at Madrid, 1967] Tragia — Miller and Webster 281 we have obtained photographs of the two sheets which presumably served as the basis of Cavanilles’ description. Johnston (1962) cited 'Sessé and/or Mocino’ as the col- lectors of the type, but as indicated by Mueller (1866), the specimens were actually collected by Née between Ixmi- quilpan and Zimapan. These specimens of Née, which are rather accurately (though somewhat crudely) depicted in Cavanilles’ plate, have a leaf shape suggestive of T. betonicifolia or T. urtici- folia, and rather closely match the Mexican specimens cited above. The Arizona population referred by us to T. nepeti- folia agrees with the Mexican plants fairly well, although it differs in having generally more bluntly toothed leaves and a tendency to reddish male flowers. It is possible that these Arizona plants represent a distinct subspecies of T. nepetifolia or even a distinct species, but until the Mexican populations have been thoroughly studied there seems little utility in creating a special taxon for the small outlying race in Arizona, As noted under T. brevispica, that species is rather sim- ilar to T. nepetifolia in some respects but differs in its twining habit and dimorphic fruits. The two species are allopatric in the United States, although further explora- tion may show their ranges to overlap in Chihuahua or northern Nuevo León. The bulk of the U. S. specimens which have passed as T. nepetifolia really belong to T. ramosa, which typically differs in its narrower more finely toothed leaves, greenish male flowers, and slender smooth styles. Both species are allopatric over much of the state of Arizona, and detailed studies of their habitat preference would probably be rewarding. 7. Tragia ramosa Torr. Ann. Lyc. N. Y. 2: 245. 1828. (Figs. 5, 12, 23) TYPE: ‘sources of the Canadian’ [evidently in SE Colorado], Long’s Expedition, James (NY!). Tragia angustifolia Nutt. Trans, Amer. Phil. Soc. (n.s.) 5: 172. 1837. 282 Rhodora [Vol. 69 DO cce SS (3) Fic. 2! , ig. 23. Habit, Tragia ramosa (C. L. & G. York 55234) 1967] Tragia — Miller and Webster 283 TYPE: ‘on the prairies of the Red River, in arid situations’, Nut- tall (x, lectotype!; NY, isotype!) .* Tragia scutelariifolia Scheele, Linnaea 25: 587. 1852. TyPE: New Braunfels, Texas, July 1846, Lindheimer 521 [Fascicle III] (B, not seen, perhaps destroyed). Tragia ramosa var. ? leptophylla Torr. Bot. Mex. Bound. 201. 1859. LECTOTYPE: Texas, ‘near Howard’s Springs’ [SW Crockett Co.], Bigelow (NY!). PARATYPE: ‘New Mexico’ [probably also from W Texas], Wright 1796 (G, NY!). Tragia stylaris Muell. Arg. Linnaea 34: 180. 1865. Tragia stylaris a latifolia Muell. Arg. loc. cit. Type: New Mexico, Fendler 776 (G!). Tragia stylaris B angustifolia Muell. Arg. loc. cit. TYPE: New Braun- fels, Texas, July 1846, Lindheimer 521 [Fascicle III] (G!). Tragia stylaris y leptophylla (Torr.) Muell. Arg. op. cit. 181. Tragia nepetifolia £ ramosa (Torr.) Muell. Arg. in DC. Prodr. 15 (2): 934. 1866. Tragia nepetifolia B latifolia Muell. Arg. loc. cit. TYPE: ‘Novo Mex- ico’ [probably W. Texas], Wright 1794 (G, GH, syntypes!)- 5. Johnston (1962) suggested that the type of this taxon might be at BM. No sheets labelled as such were seen there, but specimens matching those at Kew and New York were mingled on one of the sheets of T. betonicifolia. In view of the labelling confusion of the BM specimens, choice of the Kew collection as lectotype seems desirable. 6. The two sheets of Wright 1794 at the Gray Herbarium apparent- ly represent several different collections made during 1851 and 1852, as no less than 4 different species are included: T. amblyodonta, T. brevispica, T. nepetifolia and T. ramosa. The specimens could have come from widely separated areas between eastern Sonora and the Pecos River, as indicated in the tabulation of localities by Wooton (1906). Unless Wright's original labels are discovered, which now seems unlikely, it may be impossible to clarify the origin of this melange of taxa. The presence of material of T. amblyodonta re- sembling that of Wright 1793 suggests that perhaps part at least of Wright 1794 came from the Pecos River Valley, but this is only a presumption, and cannot be true for the fragments representing T. nepetifolia, which does not enter Texas. Mueller (1866) also cited Mexican specimens of Aschenborn, Hart- weg, and Virlet d'Aoust with his description of var. latifolia. It is possible that one of the elements contained in Wright 1794 may repre- sent the same subspecific taxon as the Mexican collections, but until the Mexican variants of T. nepetifolia are straightened out, it seems best to regard all of these collections as syntypes. 284 Rhodora [Vol. 69 Tragia nepetifolia à angustifolia Muell. Arg. loc. cit. TYPE: Texas, Berlandier 2542 (G!). Tragia nepetifolia ņn scutellariifolia (Scheele) Muell. Arg. loc, cit. Tragia ramosa a latifolia (Muell. Arg.) Pax & Hoffm. Pflanzenr. IV. 147. IX. [Heft 68]: 40. 1919. Tragia nepetifolia var. leptophylla (Torr.) Shinners, Southw. Nat. 6: 101. 1961. Stems solitary or few to many from the crown of the woody tap- root, 1-2 mm in diameter near the base, 12-50 em tall, green, purplish- green or yellowish-green, simple to many-branched, erect, decumbent, trailing, sometimes with a tendency to twine; upper internodes 3-30 mm long, lower internodes 7-40(60) mm long. Leaf blades linear- lanceolate to ovate or basal leaves sometimes reniform, 1-4 cm long, 0.4-2 cm broad, apically acute, basally truncate to cordate, margins serrate, thin, green, pubescent, ciliate; petioles 2-20 mm long; stipules lanceolate to ovate, entire, 1.2-4.5 mm long, 0.5-1.8 mm broad at the base, persistent, ciliate, rarely abaxially sparsely pubescent. Racemes with the lowermost 1(2) nodes pistillate, the remaining 2-20 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1.5-2 mm long, ciliate, acute, entire or three-lobed, sometimes abaxially sparse- ly pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca, 1.2-1.8 mm long, pubescent, entire. Staminate flowers: pedicels slender, 0.7-2 mm long, lower persistent part ca. 0.4-1.5 mm long; calyx-lobes 3(42%), 4(49%)-5(7%) (6), oblanceolate to narrowly obovate-oblong, 1-2.2 mm long, abaxially pubescent, entire, reflexed or spreading at anthesis; stamens (2) 3(20%) 4(85%) 5(81%) 6(11%) (7-10), filaments thickened and fleshy, 0.8-1 mm long, con- nate basally. Pistillate flowers: pedicels 1-1.5 mm long, becoming 2-2.5 mm in fruit; calyx-lobes (5) 6(97%) (7), lanceolate to elliptic- lanceolate, 0.8-2.5 mm long at anthesis, 1.5-3 mm long in fruit, acute, entire, ciliate, abaxially pubescent; styles connate up to 1/3 their length; stigmatic surfaces not papillate (except slightly so in aber- rant specimens). Fruit 3-4 mm long, 6-8 mm broad; columella 1.5-2.6 mm long; seeds nearly spherical, 2.5-3.5 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: open areas, often in disturbed habitats, Missouri and Arkansas west to California, Nebraska and Colorado south into Mexico (Map 6). REPRESENTATIVE SPECIMENS: — ARIZONA: GILA CO, 10 mi S of Pine, Peebles & Fulton 9460 (ARIZ). ARKANSAS: BENTON CO., Glade, Demaree 6911 (UARK). CALIFORNIA: SAN BERNARDINO CO, New York Mountains, Alexander & Kellogg 1326 (ARIZ). COLORADO: BOULDER CO., Steamboat Mountain, 2 mi NW of Lyons, Weber 3906 1967] Tragia — Miller and Webster 285 (TEX). KANSAS: MEADE Co., 11.5 mi E and 18 mi S of Meade, Horr & MeGreggor 3870 (LL, KANU). MISSOURI: STE. GENEVIEVE CO., 1 mi N of Bloomsdale, Hubricht B1526 (Ariz). NEBRASKA: FRANKLIN co., Look-out Mountain, Tolstead 411275 (TEX). NEVADA: CLARK CO. Wilson's Ranch, Clokey 8013 (GA, NCSC, NY, SMU). NEW MEX- ICO: GRANT CO, Mangas Springs, 18 mi NW of Silver City, Metcalfe 65 (ARIZ). OKLAHOMA: Kiowa co., Entrance to Quartz Mountain State Park, Miller & Miller 1341 (PUL). TEXAS: TRAVIS co., Knob Hill, 12 mi W of Austin, Miller & Miller 922 (PUL). UTAH: WASHINGTON CO., 5 mi E of Virgin, Hitchcock 3019 (PH). When the inflated concept of T. nepetifolia held by Muel- ler (1866) and Johnston (1962) is dissolved into its com- ponents, probably the majority of the Tragia collections west of the Mississippi River are referable to T. ramosa. Although Torrey originally described the species in 1828, he later (in the Botany of the Mexican Boundary Survey) confused it with T. urticifolia. Mueller (1866) partially resolved the confusion by pointing out that T. urticifolia differed in its more coarsely papillate styles but introduced several new misunderstandings which have obscured re- lationships among the noseburn taxa up to the present time: (1) he applied the epithet ramosa, when transferring it to varietal status under T. nepetifolia, not to Torrey's plant but rather to specimens of T. brevispica; (2) he in- cluded specimens of T. ramosa in 3 different varietal names under T. nepetifolia; and (3) he placed other specimens of T. ramosa. in the 3 different varieties of his T. stylaris in a different section. He thus succeeded in divorcing the epi- thet ramosa from any of the plants to which it should prop- erly be applied, and as a result the application of the name T. ramosa has remained highly confused. Pax and Hoff- mann (1919) actually managed to improve on Mueller's malpractice by using Torrey's epithet ramosa twice: once for a species placed in sect. Leucandra, and again for a variety of T. nepetifolia, Judging from their citations of specimens, Pax and Hoffmann applied the name T. nepeti- folia, ramosa. mainly to the typical subspecies or race of T. nepetifolia, and not to the population here associated with Torrey's name. Shinners (1958) on the other hand did 286 Rhodora [Vol. 69 apply the name T. nepetifolia ramosa to the plants here accepted as T. ramosa. Shinners has more recently (1961) proposed to call the common Texas noseburns T. nepetifolia var. leptophylla. This is correct on strictly nomenclatural grounds, but it would be rather unfortunate if the epithet leptophylla were generally taken up, as the collections of Bigelow and Wright represent very aberrant plants of rather dubious status. No similar specimens to the original collections of lepto- phylla have ever been found, and it is possible that the small scraps gathered by Bigelow and Wright might be hybrids or even unusual growth forms of some species other than T. ramosa. Fortunately, we believe that the question as to the status of var. leptophylla is rather academic, since T. ramosa is certainly not conspecific with T. nepetifolia. In distinct contrast to that species, T. ramosa has long, thin, nearly smooth styles, greenish male flowers, a more densely hispid ovary, and typically narrower more sharply toothed leaves. Because of its broad spectrum of variability, T. ramosa does include forms which may be difficult to distinguish from any of the other taxa of noseburns, especially in the absence of good flowers and fruits. Furthermore, occasional interspecifie crosses may occur which obscure the specific lines; a possible cross with T. brevispica has already been cited, and the discovery of crosses with T. betonicifolia and T. amblyodonta would not be surprising. Johnston (1962) was correct in exposing as worthless the supposed difference in stamen number used by Mueller and by Pax and Hoffmann to segregate T. stylaris into a different section. Actually the number of stamens in non- terminal male flowers varies from 3-5, while in terminal flowers it may reach up to 10 (a figure larger than that recorded by any previous workers). Future quantitative studies on taxa of the noseburns should clearly specify the position of male flowers measured in order to insure an accurate estimate of the range of variation jn stamen num- ber. 1967] Tragia — Miller and Webster 287 [8 o N^! 1 Ü T A »- v3, ATUS Om ge RV, ae v 5 LJ A B ` C å D E Fig. 24. Habit, Tragia amblyodonta (K. & L. Miller 1178), A-E. Variation in leaf shape. A, T. L. Steiger 421; B, B. L. Turner 1023; C, B. C. Tharp s.n. (10/VII/41); D, B. H. Warnock 46805; E, L. H. Shinners 19583. 288 Rhodora [Vol. 69 8. Tragia amblyodonta (Muell. Arg.) Pax & Hoffm. Pflanzenr. IV. 147. IX. [Heft 68]: 51. 1919. (Figs. 13, 24) Tragia nepetifolia y amblyodonta Muell. Arg. in DC. Prodr. 15 (2): 934. 1866. TYPE: Texas, ‘alluvial soil, bottom of the San Pedro R. [Devil's River], July 6’ [1852], Wright 1793 (G, holotype!; data from isotype, GH!). Stems few to many from the crown of the woody taproot, 1-1.6 mm in diameter near the base, 12-50 cm tall, green to grayish- green, few to many-branched, erect, decumbent or trailing; upper internodes 10-28 mm long, lower internodes 15-35 mm long, Leaf blades ovate, triangular or hastate, 1-4.5 cm long, 0.8-3 cm broad, apically acute to obtuse, basally truncate, cordate to hastate, margins crenate to serrate, thick, green, pubescent, ciliate; petioles 4-18(30) mm long; stipules ovate to rhombic-lanceolate, entire, 1-3.5(4.5) mm long, 0.6- 1(1.6) mm broad at the base, persistent, ciliate, abaxially pubescent. Racemes with the lowermost 1 (2) nodes pistillate, the remaining 3-30 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 0.9-1.3 mm long, ciliate, acute, entire or three-lobed, abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, sub- cucullate, ca. 0.9-1.5 mm long, pubescent, entire. Staminate flowers: pedicels slender, 0.7-1 mm long, lower persistent part ca. 0.2-0.4 mm long; calyx-lobes 3(55%)-4(41%) (5), oblanceolate to obovate, 0.9-1.9 mm long, abaxially pubescent, entire, spreading to reflexed at anthesis; stamens (2) 3(67%)-4(28%) (5); filaments thickened and fleshy, 0.2-0.7 mm long, connate basally. Pistillate flowers: pedicels 0.3-0.4 mm long, becoming 1.5-2 mm in fruit; calyx-lobes (5) 6 (92%), lanceolate to narrowly elliptic, 1-2.3 mm long at anthesis, 2-3.5 mm long in fruit, acute, entire, ciliate, abaxially pubescent; styles connate only at the base or up to 1/3 their length, stigmatic surfaces papillate. Fruit 3-4.5 mm long, 5.5-8 mm broad; columella 1.8-2.5 mm long; seeds nearly spherical, 2.6-3.6 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: rocky areas, usually on limestone, south- west Texas to New Mexico, Arizona and Nevada (Map 7). REPRESENTATIVE SPECIMENS: — ARIZONA: COCHISE co., Tombstone, Loomis & Peebles 5348 (ARIZ). NEW MEXICO: CATRON CO., Beaver- head, Castetter 7104 (ARIZ), CHAVES CO., 20 mi S of Roswell, Earle & Earle 291 (NY). DONA ANA CO., Dunn 8463 (UNM). QUAY CO., San Juan, Hershey 4010 (UNM). TEXAS: EL PASO CO, Scenic Drive above El Paso on Mt. Franklin, Warnock 5752 (TEX). JIM HOGG CO., 7 mi N of Petroleum, Tharp, Follansbee & Thompson 51-1533 (TEX). JIM WELLS CO., Alice-San Diego, Tharp s.n. (19/1V/1931) (TEX). T. SMALLII ef (P^ 3 TA ^ | | aD SS i T ie on p—— — T o | $ e 9 4 | i 3° "4 L + ^ . eee | wi (d A" E J E e > t ° c ^ / e 9" e ef LÀ a 8, ( x PINE $ LP ^ ^W ( w e. 1^», e CAS - es 2? X x T URENS ^ $ Map 7. Distribution of Tragia amblyodonta. Map 8. Distribution of Tragia laciniata and Tragia nigricans. Map 9. Distribution of Tragia smallii. Map 10. Distribution of Tragia urens. 289 290 Rhodora [Vol. 69 KLEBERG CO., 12.5 mi S of Kingsville a few miles W of Naval Air Station, Johnston 541558 (TEX). LIVE OAK CO., Three Rivers, just N of town, Thompson & Turner 24 (TEX). PECOS CO., 2 mi E of Ft. Stockton, Warnock 46134 (TEX). TERRELL CO., 20 mi E of Dryden, Parks, Turner & Warnock 72 (TEX). VAL VERDE CO., 1 mi N of the bridge over the Pecos River on Rt. 90, Miller & Miller 1178 (PUL). This plant was given specific status by Pax and Hoff- mann under dubious auspices, since they did not see the type collection and misapplied the name to specimens of T. glanduligera. Shinners (1958), perhaps misled by this, suggested that var. amblyodonta might be the same as T. teucriifolia [T. brevispica]. In fact, T. amblyodonta has little in common with T. brevispica; it is occasionally viny but scarcely ever really twining. Typical specimens of T. amblyodonta are distinctive because of the greyish cast (due to denser pubescence) and broad shallowly few- toothed leaves. Sometimes one may encounter specimens with the dense pubescence of T. amblyodonta combined with the leaf shape of T. ramosa. We have no convincing evidence that such collections are interspecific hybrids, but it would not be surprising in view of the fact that T. ramosa appears to be the most closely related species. As pointed out earlier, T. amblyodonta appears to differ from other U. S. species of Tragia in having a higher chro- mosome number: 2n — 110 instead of 44. Correlated with this are larger stomata and pollen grains than are found in the other species. Additional cytological investigations on populations of T. amblyodonta would be desirable, in order to determine whether the apparent chromosomal difference is really a constant one. 9. Tragia saxicola Smali, Fl. Southeast. U. S. 702. 1908. (Fig. 25) Type: Florida, Dade County, rocky pine woods between the Ever- glades and Biscayne Bay, A. H. Curtiss 2517 (NY!). Stems few from the crown of the woody taproot, 0.7-1.2 mm in diameter near the base, 12-35 cm tall, green or yellowish-green, few to many-branched, erect; upper internodes 8-18 mm long, lower in- ternodes 10-38 mm long. Leaf blades suborbicular, oblong to broadly ovate, 1.2-3 cm long, 1-2.2 cm broad, apically rounded to acute, Tragia — Miller and Webster 291 1967] QI rm D Sen Du. i E AA Woz N Did» m NY Habit, Tragia saxicola (A. H. Curtiss 2517). 292 Rhodora [Vol. 69 basally subcordate to cordate, margins crenate-dentate to serrate, thin, green, pubescent, ciliate; petioles 5-13 mm long; stipules lanceo- late to narrowly ovate, entire, 1.5-4 mm long, 0.6-1.5 mm broad at the base, persistent, ciliate, abaxially pubescent. Racemes with the single lowermost node pistillate, the remaining 12-20 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 0.8-1.4 mm long, ciliate, acute, entire or three-lobed, abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 0.8-1.2 mm long, pubescent, entire. Staminate flowers: pedicels slender, 1.5-1.9 mm long, lower persistent part ca. 0.5-0.7 mm long; calyx-lobes 3 (68% )-4(32%), obovate, 1.0-1.5 mm long, abaxially pubescent, en- tire, spreading to reflexed at anthesis; stamens 3(81%)-4(19%) ; fila- ments slender, thickened and fleshy, 0.4-0.6 mm long, connate basally. Pistillate flowers: pedicels 0.9-1.2 mm long, becoming 3.2-3.7 mm in fruit; calyx-lobes 6, lanceolate to ovate-lanceolate, 1.5-2.0 mm long at anthesis, 2.0-3.0 mm long in fruit, acute, entire, ciliate, abaxially pubescent; styles connate only at the base or up to 1/3 their length, stigmatic surfaces slightly papillate. Fruit 3.0-4.0 mm long, 6.0-7.0 mm broad; columella 1.7-2.1 mm long; seeds nearly spherical, 2.4-3.0 mm long, brownish-black with a tawny mottling when mature. DISTRIBUTION: restricted to extreme southern Florida and the Keys typically in pine woods on limestone (Map 5). REPRESENTATIVE SPECIMENS — FLORIDA: DADE CO., Lone Pine Key, Atwater M-160 (GA); Biscayne Bay, Curtiss 2517 (FLAS, GA, KANU, NY, PH, UARK). MONROE CO. Big Pine Key, Eyles & Eyles 8192 (NCSC), Killip 41089 (NO), Killip 42441 (1A), Small 3782, 3955 (NY). This geographically isolated member of the noseburn complex is distinctive in habit because of its slender wiry stems and broad sharply toothed leaves, but in reproduc- tive characters it is not very different from the Texas nose- burns. Possibly T. betonicifolia is the closest species vegetatively, but the pistillate calyx of T. saxicola is more like that of T. ramosa. A relationship with T. urticifolia also seems evident, but that plant has narrower leaves and longer persistent staminate pedicels. 10. Tragia laciniata (Torr.) Mull. Arg. Linnaea 34: 182. 1865; in DC. Prodr. 15(2): 933. 1866. (Fig. 26) Tragia urticifolia var. laciniata Torr. Bot. U. S. Mex. Bound. 200. 1967] Tragia — Miller and Webster 293 í N A i j Y P " y 4 y RA s DA a i Ro "al NY px SS M ' RB s e zl rts d! ps soU. 4 > A ty a he, f hi SENE AY Y A u M 7 k^ Wt AW a A e A ! "7 LJ WA Wurm wl iM e AN aN in = c e (8) Fig. 26. Habit, Tragia laciniata (R. H. Peebles et al. 5641). 294 Rhodora [Vol. 69 1858. LECTOTYPE: ‘New Mexico’, Wright 1795 (NY!). Stems few to many from the crown of the woody taproot, 1.3-1.5 mm in diameter near the base, 25-50 cm tall, green to purplish-green, few to many-branched, erect to decumbent; upper internodes 15-20 mm long, lower internodes 20-30 mm long. Leaf blades 3-divided, 2.5-4 cm long, the segments acute, their margins irregularly coarsely toothed, thin, green, pubescent, ciliate; petioles 7-18 mm long; stipules lanceolate, acute, entire, 1-2.8 mm long, 0.6-1.2 mm broad at the base, persistent, ciliate or abaxially sparsely pubescent. Racemes with single lowermost node pistillate, the remaining 10-20 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 0.5-1.5 mm long, ciliate, acute, entire or 3-lobed; bracts of the staminate flowers lanceolate, subcucullate, ca. 0.5-1.5 mm long, pubescent, entire. Stam- inate flower: pedicels slender, 0.8-1.6 mm long, lower persistent part ca. 0.3-0.7 mm long; calyx-lobes 3(90%) (4), obovate to ob- lanceolate, 1-1.4 mm long, abaxially pubescent, entire, spreading to reflexed at anthesis; stamens 3; filaments slender, 3.5-4 mm long, connate basally. Pistillate flower: pedicels 0.6-1 mm long, becoming 2.5-3 mm in fruit; calyx lobes 6, lanceolate, 2-2.4 mm long at anthesis, 2.5-3 mm long in fruit, acute, entire, ciliate; styles connate 1/4-1/2 their length, stigmatic surfaces papillate. Fruit 3.5-4 mm long, 6-7 mm broad; columella 2-2.3 mm long; seeds nearly spherical, 3-3.2 mm long, brownish-black when mature. DISTRIBUTION: Arizona and Sonora, Mexico, perhaps ex- tending into New Mexico. All material examined was from Arizona, and the provenance of the Wright specimen is uncertain, as the label ‘New Mexico’ on his collections was used very loosely (Map 8). REPRESENTATIVE SPECIMENS: — ARIZONA: Clark 12540 (UNM). SANTA CRUZ CO., Sycamore Canyon, Kearney & Peebles 14455 (ARIZ) ; Nogales to Ruby, Kearney & Peebles 14916 (ARIZ, NY); between Nogales and Arivaca, Peebles, Harrison & Kearney 5641 (ARIZ). (NEW MEXICO?): Wright 1795 (PH). Tragia laciniata is vegetatively the most distinctive of the U. S. species; the three-parted leaves are quite unmis- takable. Pax and Hoffmann (1919) note its similarity to the South American T. pinnata, but the resemblance is probably due to parallelism. In reproductive characters T. laciniata agrees with members of the noseburn complex and is perhaps closest overall to some of the forms of T. nepeti- folia. 1967] Tragia — Miller and Webster 295 11. Tragia nigricans Bush in Small, Fl. Southeast. U. S. 702. 1903; Missouri Bot. Gard. Report. 17: 122. 1906. (Fig. 27) TvPE: Texas, upper Hondo, common in woods, June 1885, Rever- chon 1594 (NY, syntype!). Stems solitary or few from the crown of the woody taproot, 1-2 mm in diameter near the base, 15-55 cm tall, purplish-green to grayish- green, turning black in drying, simple or few-branched, erect; upper internodes 9-19 mm long, lower internodes 13-45 mm long. Leaf blades oblong to oblong-lanceolate, 3-7 em long, 1-2.5 cm bread, apical- ly and basally acute, margins deeply dentate, thick, green, pubescent, S m eo A wo 1 "i UE ri D A A OA [ Y n7 | o yy 7.3 « V NA EL. AN y / URN VA ix yv 4f) "t oe ny d á p^ T. TN Warren. 5 » GY EES AN Sry P p E ADM = E oY Faeroe ON MOSS. ed CLL JIE 2 CM ©) Fig. 27. Habit, Tragia nigricans (K. & L. Miller 1173). 296 Rhodora [Vol. 69 ciliate; petioles 1-5 mm long; stipules lanceolate, attenuate, entire, 2.5-6 mm long, 0.5-1 mm broad at the base, persistent, ciliate. Racemes with the single lowermost node pistillate, the remaining 2-5 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 2-2.6 mm long, ciliate, acute, entire or three-lobed; bracts of the staminate flowers lanceolate, ca. 1-2 mm long, pubescent, entire. Staminate flowers: pedicels slender, 1.3-1.6 mm long, lower persistent part ca. 0.2-0.4 mm long; calyx-lobes 3(6496)-4(3296) (5), oblanceolate to narrowly ovate-oblong, 1.5-2.5 mm long, abaxially pubescent, entire, reflexed at anthesis; stamens (3) 4(48%)-5(49%) (6); filaments slender, 0.7-1.3 mm long, connate 1/3-1/2 or more their length. Pistillate flowers: pedicels 0.7-1 mm long, becoming 2-3 mm in fruit; calyx-lobes (5) 6(96%), narrowly rhombic-lanceolate, 1-2.3 mm long at anthesis, 2.4-4 mm long in fruit, acute, entire, ciliate; styles connate only at the base, stigmatic surfaces very slightly papillate. Fruit 3-3.5 mm long, 6.5-7 mm broad; columella 1.7-3.2 mm long; seeds nearly spherical, 2.5-3.2 mm long, brownish-black when mature. DISTRIBUTION: restricted to a small area in south cen- tral Texas, in rocky soil of open woods and fields (Map 8). REPRESENTATIVE SPECIMENS: — TEXAS: coMar co., Fischer's Store, Palmer 12194, (TEX). KERR CO., limestone river bluffs near Kerrville, Palmer 33811 (NY). MEDINA CO., Upper Hondo, Reverchon 1594 (SMU). REAL CO., ca. 4 mi N of Leakey on Ranch Road 336, Miller & Miller 1164 (PUL); Leakey, Palmer 10162 (M). UVALDE CO. Garner State Park, Miller & Miller 1173 (PUL), Turner 3855 (TEX). Tragia nigricans is one of the most distinctive species of the genus of those occurring in the U. S. The paucity of existing herbarium material may be misleading insofar as indicating the range and relative abundance of the species ; additional localities on the Edwards Plateau may be ex- pected. The general form of the leaves and the unique way in which the filaments in the staminate flowers are united one-half or more their length to form a cylinder set T. nigricans apart from the other U. S. species. Section 2. LEPTOBOTRYS (Baill) Muell. Arg. Linnaea 34: 183, 1865; DC. Prodr. 15(2): 946. 1866. Allosandra Raf. Aut. Bot. 51. 1840. LECTOTYPE: Allosandra lanceo- 1967] Tragia — Miller and Webster 297 lata Raf. (superfluous name for Tragia innocua Walt. — T. urens L.). Leptobotrys Baill. Etud. Gén. Euphorb. 478. Type: Leptobotrys dis- color Baill. (= T. urens L.). Low perennial herbs, not twining; stem with stinging hairs but plants not severely stinging; leaves entire or rather coarsely and irregularly toothed or lobed; inflorescences opposite leaves or terminal on lateral leafy branches; staminate flower with 4 or 5 (rarely 6) calyx-lobes; stamens 2 (rarely 3); filaments thickened and fleshy, connate basally; pollen grains ‘pseudocolpate’, the operculum diffuse; pistillode flattened between the filaments, laminar; pistillate flower with 6 calyx-lobes; styles connate below, ventrally more or less papil- late; seeds 3-4.5 mm long. This section is here delimited so as to include only the two species of the Southeastern United States: T. smallii and T. urens. Although Mueller (1866) and at first Pax (1890) ac- cepted Leptobotrys as a section of Tragia, primarily on the basis of the androecium of only 2 stamens, Pax and Hoff- mann (1919) submerged it as a synonym of their sect. Eutragia. The discovery of the characteristic pollen con- figuration in both T. urens and T. smallii, together with the androecial difference, would seem to justify reviving the section again. Tragia urens, with its peculiarly shaped leaves and tendency to monopodial stems, is furthermore very different from the other U. S. species in habit. How- ever, T. smallii closes the gap to some extent, since it agrees better with the noseburn Tragias in leaf shape and inflores- cence. Although the taxa of sect. Leptobotrys do not appear to have any close ties with the North American representa- tives of sect. Tragia, it is not obvious which of the neotropi- cal taxa of Tragia is closest. The similarity in pollen in sect. Bia is suggestive, but the male flowers of that group are very different by virtue of their many stamens and well- developed disk glands. Perhaps some of the South Ameri- can species of sect. Leucandra (sensu Pax and Hoffmann, 1919) are most nearly related, but this is hardly more than a guess in our present state of knowledge. 298 Rhodora [Vol. 69 12. Tragia smallii Shinners, Field & Lab. 24: 37. 1956. (Figs. 15, 28) TYPE: Louisiana, Vernon Parish, 1 mi S of Mayo. McVaugh 8461 (SMU, holotype!; TEX, isotype!). Tragia betonicifolia sensu Small, Fl. S. E. U. S. ed. 2, 702. 1913; and Pax & Hoffm. Pflanzen. IV. 147. IX. [Heft 68]: 62, 1919; non Nutt. (1837). Stems solitary to few from the crown of the woody taproot, 1-2 mm in diameter near the base, 12-25 cm tall, purplish-green to grayish- green, simple to few-branched, erect; upper internodes 6-35 mm long, lower internodes 7-16 mm long. Leaf blades orbicular to elliptic, 2-4 cm long, 0.8-2.5 em broad, apically rounded to acute, basally rounded to obtuse, margins coarsely serrate or crenate to lobed, Fig. 28. Habit, Tragia smallu (K. & L. Miller 865). 1967] Tragia — Miller and Webster 299 thick, green, pubescent, ciliate; petioles 1-4 mm long; stipules lanceo- late to narrowly ovate, entire, 1.5-3.8 mm long, 0.7-1.5 mm broad at the base, persistent, ciliate, pubescent. Racemes with the lowermost 1 (2) nodes pistillate, the remaining 4-11 nodes staminate; bracts of the pistillate flowers lanceolate, ca. 1.5-2.5 mm long, ciliate, acute, entire or three-lobed, sometimes abaxially sparsely pubescent; bracts of the staminate flowers lanceolate, subcucullate, ca. 0.8-1.2 mm long, pubescent, entire. Staminate flowers: pedicels slender, 1.5-1.9 mm long, lower persistent part ca. 0.4-0.6 mm long; calyx-lobes 4(50%)-5 (40%) (6), lanceolate to narrowly ovate, 0.9-1.5 mm long, abaxially pubescent, entire, spreading at anthesis; stamens 2(95%) (3); fila- ments thickened and fleshy, flattened laterally, 0.2-0.5 mm long, con- nate basally. Pistillate flowers: pedicels 0.5-1 mm long, becoming 2.8-3.4 mm in fruit; calyx-lobes 6, linear, 1.3-1.7 mm long at anthesis, 1.5-2.3 mm long in fruit, acute, entire, ciliate, abaxially pubescent; styles connate about 1/4 their length, stigmatic surfaces roughened but not papillate. Fruit 5-7 mm long, 9-13 mm broad; columella 2.2.3.2 mm long; seeds slightly elongated, 4-4.5 mm long, brownish- black with tawny mottling when mature. DISTRIBUTION: Florida to east Texas along the gulf coast typically in dry sandy soil at the edge of woods and along roadsides (Map 9). REPRESENTATIVE SPECIMENS: — ALABAMA: BALDWIN CO. 3 mi N of Foley, Whitehouse 24262 (sMv). FLORIDA: HIGHLANDS CO., Sebring, Garrett s.n. (10/V/1948) (FLAS). GEORGIA: DECATUR CO., 1.8 mi SSW of Faceville, Thorne & Davidson 17122 (1A). LOUISI- ANA: Vernon Parish, 2 mi W of Leander, Webster & Wilbur 3249 (SMU). MISSISSIPPI: LAUDERDALE CO. 18 mi SW of Quitman on Route 18, Miller & Miller 865 (PUL). TEXAS: HARDIN CO., 7 mi S of Kountze, Whitehouse 23253 (SMU). NEWTON CO. 3 mi N of the junction of Route 63 and Farm Road 692, Thompson & Turner 151 (TEX). Shinners was correct in giving a new name to the plant which Small (1913) and Pax and Hoffmann (1919) had confounded with T. betonicifolia of Nuttall; the latter oc- curs only to the west of the range of T. smallii. Vegetatively, T. smallii resembles T. saxicola to some extent, but its pol- len grains agree closely with those of T. wrens. Since it furthermore has staminate flowers with only 2 stamens and a flattened pistillode, it would appear to belong closest to T. urens. 300 Rhodora [Vol. 69 13. Tragia urens L. Spec. Pl. ed. 2: 1391. 1763. (Figs. 16, 29) Ricinus parvus urens, foliis quercinis, virginianus Plukenet, Alm. bot. 320, t. 107. f. 5. 1696. (No type present in Herb. Sloane, BM). Tragia innocua Walt. Fl. Carol. 229. 1788. (No type present in Herb. Walter, BM). Tragia urens a subovalis Michx. Fl. Bor. Amer. 2: 175. 1803. Type: Herb. Michaux (P!) Tragia linearifolia Ell. Sketch. 2: 563. 1824. TYPE: southern Georgia (not seen). Tragia urens B lanceolata Michx. loc. cit. (based on T. innocua Walt.). Tragia wrens à linearis Michx. loc. cit. TYPE: Herb. Michaux (P!). Allosandra lanceolata Raf. Aut. Bot. 51. 1840. Allosandra verbenifolia Raf. loc. cit. Leptobotrys discolor Baill. Étud. gén Euphorb, 479, t. 2, f. 17, 18. 1858. SvNTYPES: Michaux, Leconte, Noisette (P!). Tragia discolor (Baill.) Muell. Arg. in DC. Prodr. 15 (2): 946. 1866. Tragia discolor a subovalis (Michx.) Muell. Arg. loc. cit. Tragia discolor B cuneata [f.] a, latifolia Muell. Arg. loc. cit. TYPE: Carolina, Bose (not seen). Tragia discolor B cuneata [f.] b, lanceolata (Michx.) Muell. Arg. loc. cit. Tragia discolor 8 linearis (Michx.) Muell. Arg. in DC. Prodr. 15 (2): 947. 1866. Tragia urens var. a innocua (Walt.) Pax and K. Hoffm. Pflanzenr. IV 147. IX [Heft 68]: 58. 1919. Stems solitary or few from the crown of the woody taproot, 1-2.2 mm in diameter near the base, 20-50 cm tall, purplish-green to grayish-green, few to many-branched, erect; upper internodes 7-15 mm long, lower internodes 10-25 mm long. Leaf blades broadly ob- lanceolate to linear, 2-8 em long, 2-14 mm broad, apically acute to rounded, basally acute to attenuate, margins entire, round-toothed to lobed, thick, green, pubescent, ciliate; petioles 0-2 mm long; stipules linear to lanceolate, entire, 1.5-6 mm long, 0.8-1.1 mm broad at the base, persistent, ciliate, pubescent on both surfaces. Racemes with the lowermost 1 (2) nodes pistillate, the remaining 3-45 nodes staminate; bracts of the pistillate flowers lanceolate, ca, 1-2.5 mm long, ciliate, acute, entire or three-lobed, abaxially sparsely pubes- cent; bracts of the staminate flowers lanceolate, subcucullate, ca. 1-1.5 mm long, pubescent, entire. Staminate flowers: pedicels slender, 1.8-2 mm long, lower persistent part ca. 0.3-0.6 mm long; calyx- lobes 4(73%)-5(27%), broadly lanceolate to ovate, 1-1.5 mm long, 7. There are 8 specimens on 2 sheets of T. wrens in the Michaux Herbarium at Paris; these presumably represent the types of the varieties described in Michaux’s book, but the individual specimens are not labelled with the varietal epithets. 1967] Tragia — Miller and Webster 301 CN ET j Noy E Ug PL / er My nN) WZ die W Ny i v Y | c : Cy i P d ND p^ 7 AACS: con SN | EERE ~ S Fig. 29. Habit, Tragia urens (K. & L. Miller 776). 776 abaxially pubescent, entire, spreading at anthesis; stamens 2(91%) (3); filaments thickened and fleshy, 0.2-0.4 mm long, connate basally. Pistillate flowers: pedicels 0.6-1 mm long,. becoming 3.5-4 mm in fruit; calyx-lobes 6, linear, 1-1.5 mm long at anthesis, 1.5-1.8 mm long in fruit, acute, entire, ciliate, abaxially pubescent; styles con- nate 1/4-1/3 their length, stigmatic surfaces roughened but not papillate. Fruit 3-4 mm long, 7-8 mm broad; columella 1.9-2.6 mm long; seeds slightly elongated, 3-4 mm long, brownish with a tawny mottling when mature. 302 Rhodora [Vol. 69 DISTRIBUTION: Virginia south to Florida and west to east Texas, in dry sandy soil or woods and clearings (Map 10). REPRESENTATIVE SPECIMENS: — ALABAMA: MONTGOMERY CO., Montgomery, Sargent s.m. (16/V/1955) (sMU). FLORIDA: DADE CO., South Miami, Moldenke 5857 (NY). GEORGIA: RANDOLPH CO., 1 mi W of Coleman, Thorne & Muenscher 8154 (1A). LOUISIANA: RAPIDES PARISH, Alexandria, Hale s.n. (NY). MISSISSIPPI: JACKSON CO., Ocean Springs, Tracy 4728 (NY). NORTH CAROLINA: PENDER CO., ca. 15 mi NW of Wilmington on U. S. 421, Miller & Miller 776 (PUL). SOUTH CAROLINA: ALLENADE CO., 3.5 mi E of Sycamore, Radford 5347 (NCSC). TEXAS: HARDIN CO., 4 mi W of Silsbee, Thompson & Turner 103 (TEX). VIRGINIA: SOUTHAMPTON CO., 2 mi SW of Sebrell, Smith & Hodgdon 1069 (ARIZ, FLAS, GA, IA, ISC, KY, NCSC, NO, NY, PH, SMU, TEX, UARK). Linnaeus based T. urens on the treatment by Plukenet, citing Virginia as the type locality. Since there is no speci- men of Plukenet in the Sloane Herbarium (BM), the spe- cies must be typified by the illustration and description. The picture, although rather crude, is sufficiently charac- teristic — when taken in conjunction with the locality —to fix the application of the name. The specific epithet is not very appropriate, since T. wrens is the least painfully sting- ing of the U. S. spccies (hence Walter's substitute name) ; it seems possible that Plukenet may have somehow con- founded reports of the stinging qualities of T. urticifolia in providing his original description. As delimited here, the varieties proposed by Michaux and the T. lnearifolia of Elliott do not seem worthy of maintaining at any taxonomic level. Long linear entire leaves occur sporadically throughout the range of the spe- cies and certainly do not appear to show any correlations with geographic distribution or with reproductive differ- ences. The names of Rafinesque are as usual difficult to dispose of. The epithet and description of Allosandra ver- benifolia Raf. are somewhat suggestive of T. smallii, and it is possible that it could prove to be the earliest name for that plant. However, we have not been able to locate any 1967] Tragia — Miller and Webster 303 specimens of T. smallii collected by Baldwin or Ware, as mentioned by Rafinesque; the only Baldwin collection ex- amined (NY) proves to be T. urens. Until a type specimen is discovered, it seems quite unwarranted to take up a name as vaguely characterized as Allosandra verbenifolia. In many ways T. urens is the most isolated of the U. S. species. Although its pseudocolpate pollen is shared by T. smallii, together with the low stamen number, it is vegeta- tively unlike that species in producing some of its inflores- cences at the tip of leafy branches which give a ‘pinnate’ effect to the main stems. The inflorescence pattern ap- proaches that of neotropical species such as T. mexicana more closely than does that of the other U. S. taxa. Further- more, two collections were examined in which the plants were entirely pistillate; the inflorescences have 6-10 or more female flowers instead of only 1 or 2. Inflorescences with several pistillate flowers also occur in sects. Zuckertia and Bia, and it is thus of some interest that the inapertu- rate pollen grains in the latter group show a similarity to those of T. urens. It would appear that T. wrens is a phy- togeographically isolated taxon in the United States, with a completely different origin from the noseburn complex in Texas and adjacent states. DEPARTMENT OF BIOLOGY APPALACHIAN STATE TEACHERS COLLEGE, BOONE, NORTH CAROLINA (28607) DEPARTMENT OF BOTANY UNIVERSITY OF CALIFORNIA, DAVIS (95616) APPENDIX A TABULATION OF THE LINDHEIMER COLLECTIONS OF Tragia (based on examination of herbarium sheets and correlations by Blankinship (1907). Fascicle Fascicle Collector's Date no. no. Identity & herbaria I 176 1843 T. betonicifolia (MO) II 307 1844 T. brevispica (GH, NY, P, SMU) 304 Rhodora [Vol. 69 III 521 298 1846 T. ramosa (G, GH, MO, PH, SMU) III 522 299 1846 T. brevispica (GH, MO, TEX) T. urticifolia (PH) IV [?] ‘307’ 1847-48 T. brevispica (G, M, MO, NY, P) IV 521’ 1847-48 T. ramosa (GH) V 1154 139 1849 (Aug) T. brevispica (GH, MO, PH, SMU) V 1155 138 1849 (Jl) T. brevispica (GH, MO, PH, TEX) V 1156 74 1849 (J1) T. brevispica (GH) V 1156 74 1849 (Jl) T. ramosa (GH, MO, SMU, TEX) V 1156 128 ‘1847’ T. urticifolia (MO) LIST OF REFERENCES BAILLON, H. E. 1858. Etude générale du groupe des Euphorbiacées. 684 pp. Victor Masson, Paris. BENTHAM, GEORGE, & J. D. HooKER. 1880. Euphorbiaceae, in Genera Plantarum 3(1): 239-340, BLANKINSHIP, J. W. 1907. Plantae Lindheimerianae. Part III. Ann. Report Missouri Bot. Gard. 18: 123-223. CROIZAT, LEON. 1942. Peculiarities of the inflorescence in the Euphor- biaceae. Bot. Gaz. 103: 771-779. FERNALD, M. L. 1950. Gray's Manual of Botany, 8th ed. 1632 pp. American Book Co. New York. GLEASON, H. A. 1952. The new Britton and Brown illustrated flora, vol. 2. 655 pp. Lancaster Press. JOHNSTON, I. M. 1940. The floristic significance of shrubs common to North and South American deserts. Jour. Arnold Arb. 21: 356-363. JOHNSTON, M. C. 1962. The noseburn (Tragia, Euphorbiaceae) of western Texas. Rhodora 64: 137-142, JoNES, F. B., C. M. RowELL, & M. C. JOHNSTON. 1961. Flowering plants and ferns of the Texas coastal bend counties. 146 pp. Welder Wildlife Foundation. JUSSIEU, ADRIEN. 1824. De Euphorbiacearum generibus medicisque earumdem viribus tentamen. 118 pp. Didot, Paris. KNOLL, F. 1905. Die Brennhaare der Euphorbiaceen-Gattungen Dalechampia und Tragia. Sitzungsb. Math. Nat. Kl. Kais. Akad. Wiss. Wien 114: 29-48. 1967] Tragia — Miller and Webster 305 Lourteic, A., & C. A. O'DoNELL. 1941. Tragiae Argentinae. Lilloa 6: 347-380. McVauaH, RoGERs. 1961. Euphorbiaceae novae Novo-Galicianae. Brittonia 13: 145-205. METCALFE, C. R. & L. CHALK. 1950. Anatomy of the Dicotyledons. 1500 pp. in 2 vols. Clarendon Press, Oxford. MILLER, K. I. 1963. Preliminary studies on the genus Tragia (Eu- phorbiaceae). Proc. Indiana Acad. Sci. 72: 257. . 1964. A taxonomic study of the species of Tragia in the United States. 161 pp. Unpublished Ph.D. dissertation, Pur- due University. , & G. L. WEBSTER. 1962. Systematic position of Cnidos- colus and Jatropha. Brittonia 14: 174-180. MUELLER, JEAN. 1865. Euphorbiaceae. Linnaea 34: 1-224. MUELLER, JEAN. 1866. Euphorbiaceae [except Euphorbieae] in De Candolle, Alphonse. Prodromus systematis naturalis regni veg- etabilis. 15(2): 189-1286. 1874. Euphorbiaceae [part II, Acalypheae through Euphorbiaceae], in Martius, Flora Brasiliensis 11(2) : 293-752. Pax, F. 1890. Euphorbiaceae, in Engler & Prantl, Naturl. Pflanzenf. 3(5): 1-119. , and K. HorrMANN. 1919. Euphcrbiaceae — Acalypheae — Plukenetiinae, in Engler, A. Das Pflanzenreich IV. 147. IX, [Heft 68]: 32-101. PLUMIER, C. 1703. Novo plantarum Americanarum genera, 73 pp. Boudot, Paris. PUNT, WILLEM. 1962. Pollen morphology of the Euphorbiaceae with special reference to taxonomy. Wentia 7: 1-116. RITTERSHAUSEN, PAUL. 1892. Anatomisch-systematische Untersuchung von Blatt und Axe der Acalypheen. 123 pp. Inaugural-Disserta- tion, Munich. SHINNERS, L. H. 1958. Spring flora of the Dallas — Fort Worth area, Texas. 514 pp. Dallas. 4 1961. Tragia nepetaefolia var. leptophyl!a instead of var. ramosa (Euphorbiaceae). Southw. Nat. 6: 101. SMALL, J. K. 1913. Flora of the southeastern United States. 2nd ed. 1394 pp. New York. WopEHOUSE, R. P. 1935. Pollen grains. 574 pp. McGraw-Hill, New York. WooTON, E. O. 1906. Southwestern localities visited by Charles Wright. Bull. Torr. Bot. Club 33: 561-566. SEVENTEENTH REPORT OF THE COMMITTEE ON PLANT DISTRIBUTION The sixteenth report treated the Cyperaceae with the exception of the genus Carex. The present report deals with the Rosaceae with the exception of the genus Cratae- gus. It was felt by the committee that the New England Crataegi were still not sufficiently understood to make their inclusion meaningful in the present report. However an alphabetical list of Crataegi found in the Herbarium of the New England Botanical Club and in the Gray Herbarium is included together with the states in which they occur but no attempt has been made to arrange the species in distri- butional categories. In genus Rubus the work of A. R. Hodgdon and F. L. Steele has been followed (see ‘Rubus subgenus Eubatus in New England", Rhodora 68: 474-513 (1966) ) in which many of the species recognized by Fer- nald in Gray's Manual 8th ed. have been reduced to syn- onomy or to a hybrid category. This treatment seems to us to be more realistic than that of the Manual with its multiplicity of species. The data for these reports have been compiled from the material in the Herbarium of the New England Botanical Club, the Gray Herbarium and the Herbarium of the Pea- body Museum at Salem. Except as otherwise indicated, nomenclature follows that of Gray's Manual 8th ed. PRELIMINARY LISTS OF NEW ENGLAND PLANTS, XLI The sign + indicates that an herbarium specimen has been seen, the sign — that a reliable printed record has been found and the sign * is used for those plants which are not native in the New England area. ROSACEAE ie. N.H Vt Mass. R.I. Conn. Agrimonia gryposepala Wallr. + + + + + + A. parviflora Ait. + A. pubescens Wallr. + + ES + + A. rostellata Wallr. + + 306 1967] A. striata Michx. Alchemilla minor Huds. Amelanchier arborea (Michx. f.) Sb >>> > b> > > Pp Fern. . Bartramiana (Tausch) Roemer . canadensis (L.) Medic. . gaspensis (Wieg.) Fern. & Wieg. . humilis Wieg. . intermedia Spach . laevis Wieg. . nantucketensis Bickn. . neglecta Egglest. (A. Bartramiana X laevis) . sanguinea (Pursh) DC. . stolonifera Wieg. . Wiegandii Nielsen *Aruncus dioicus (Watt.) Fern. Chaenomeles lagenaria (Loisel.) Koidz. Crataegus anomala Sarg. . arnoldiana Sarg. . basilica Beadle . Beckwithae Sarg. . biltmoreana Beadle . Brainerdii Sarg. . Brainerdii var. asperifolia (Sarg.) Egglest. . Brainerdii var. cyclophylla (Sarg.) Palmer . Brainerdii var. Egglestoni (Sarg.) Robins. . Brainerdii var. scabrata (Sarg.) Egglest. . brumalis Ashe . Brunetiana Sarg. . Brunetiana var. Fernaldii (Sarg.) Palmer . chadsfordiana Sarg. . chrysocarpa Ashe . ehrysocarpa var. Bicknellii (Egglest.) Palmer . chrysocarpa var. phoenicea Palmer . crus-galli L. +++++++ +48 ++4+4 dude N.H. ear Committee on Plant Distribution Vt. + +++ ++ +++ Mass. + ++++ + +++ +++ +++ 4 + +++ ++ + R.I. ++ 307 Conn. + ++ +++ ++ 308 oonoo o a a a * * Qoooooo a * a0 aaaaaaaa o oooononn . crus-galli var. exigua (Sarg.) Egglest. . dilata Sarg. . Dodgei Ashe . Faxoni Sarg. Faxoni var. praetermissa (Sarg.) Palmer . flabellata (Spach) Kirchn. var. Grayana (Egglest.) Palmer . foetida Ashe . fretalis Sarg. Holmesiana Ashe Holmesiana var. villipes Ashe Idaea Sarg. . insolens Sarg. . intricata Lange . intricata var. straminea (Beadle) Palmer irrasa Sarg. var. Blanchardii (Sarg.) Egeglest. Jesupi Sarg. Jonesae Sarg. lemingtonensis Sarg. levis Sarg. . littoralis Sarg. . macrosperma Ashe . Macrosperma var. acutiloba (Savg.) Egglest. . macrosperma var. demissa (Sarg.) Egglest. . Macrosperma var. matura (Sarg.) Egglest. . macrosperma var. pentandra (Sarg.) Egglest. . media Sarg. membranacea Sarg. . monogyna Jacq. . Oakesiana Egglest. . Oxyacantha L. . pedicellata Sarg. . pedicellata var, albicans (Ashe) Palmer . Phaenopyrum (L. f.) Medic. . pilosa Sarg. Rhodora Me. +++ N.H. +++ +4 + + +++ + Vt. ++ H+ +4+++4+44+4+ | -- Mass. +++ ++ + | ++ [Vol. 69 R.I. + ope Conn. ++++ | + 1967] Committee on Plant Distribution C. Porteri Britt. var. caerulescens (Sarg.) Palmer . Pringlei Sarg. . Pringlei var. exclusa (Sarg.) Egglest. . pruinosa (Wendl.) K. Koch . pruinosa var. dissona (Sarg.) Egglest. . pruinosa var. latisepala (Sarg.) Egglest. . punetata Jacq. . punctata var. aurea Ait. Randiana Sarg. . rotundata Sarg. . rugosa Ashe . Shirleyensis Sarg. . Schizophylla Egglest. . spatiosa Sarg. Stonei Sarg. submollis Sarg. . Suborbiculata Sarg. . succulenta Link . suceulenta var. macracantha (Lodd.) Egglest. C. succulenta var. neofluvialis (Ashe) Palmer C. Websteri Sarg. Dalibarda repens L. *Duchesnea indica (Andr.) Focke *Exochorda grandiflora Hook. *Filipendula hexapetala Gilib. *F. rubra (Hill) Robins. *F. Ulmaria (L.) Maxim. *F. Ulmaria var. denudata (Hayne) Maxim. xFragaria Ananassa Duchesne F. vesca L. aa AQ Cy aaanaaaaaaaaa F. vesca var. americana Porter F. virginiana Duchesne F. virginiana var. terrae-novae (Rydb.) Fern. Geum aleppicum Jacq. var. strictum (Ait.) Fern. G. canadense Jacq. Me. N.H. Vt. + + + + — + + + + 4 Boc + + + + +o + + + + + + + + + Mass. + - ++ ++ +++ + ott X + b+ ++ -+ + + IE 309 Conn. ++ +p ttt 310 Rhodora [Vol. 69 Me. N.H. Vt Mass. R.I. Conn. G. canadense var. camporum (Rydb.) Fern. & Weath. -+ + + + G. laciniatum Murr. d + + + + H G. laciniatum var. trichocarpum Fern. + + + + + G. macrophyllum Willd. + + + G. Peckii Pursh + xG. pulchrum Fern. + + G. rivale L. -F + + + -+ + *G. urbanum L. + G. virginianum L. + + + *Kerria japonica (L.) DC. + + *Physocarpus opulifolius (L.) Moench + + + + + -+ Potentilla anserina L. + + + P. argentea L. + +- + + + + P. argentea var. pseudosalabra Wolf + + P. arguta Pursh + + + + + P. canadensis L. + + + ES — EE P. canadensis var. villosissima Fern. — *P. canescens Bess, + + P. Egedei Wormsk. var, groenlandica Polunin EE + + — EE . erecta (L.) Ràuschel — . fruticosa L. ++ + + + T P P P. gracilis Dougl. I P. gracilis var. pulcherrima (Lehm.) Fern. *P. intermedia L. 4- + + + *P. millegrana Engelm. P P ++ + + + . norvegica L. . norvegica var. labradorica (Lehm.) Fern. . palustris (L.) Scop. . palustris var. villosa (Pers.) Lehm. + . pectinata Raf. -F . recta L. E . reptans L. . Robbinsiana Oakes + . simplex Michx. +++ +++ 44 E +++ +++ 4 + . simplex var. calvescens Fern. + . tridentata Ait. + . verna L. FJ "U "V "U "ur * ++ +++ Prunus alleghaniensis Porter 1967] Committee on Plant Distribution 311 Me. Vt. Mass, R.I. g 3 2a ++ 4+44+4+4+4+4+44++ + ++ P. americana Marsh. *P. Avium L. + *P. cerasifera Ehrh. *P. Cerasus L. P. depressa Pursh T *P. domestica L. P. Gravesii Small *P. Insititia L. + *P. Mahaleb L. P. maritima Marsh. P. nigra Ait. P. pensylvanica L. f. *P. Persica (L.) Batsch P P P z oe +++ + . serotina Ehrh. . susquehanae Willd. . virginiana L. *P. spinosa L. Pyrus americana (Marsh.) DC. P. arbutifolia (L.) L. f. xP. arnoldiana (Rehd.) W. J. Bean *P. Aucuparia (L.) Gaertn. SPE: bsecata L *P. communis L. P. decora (Sarg.) Hyland P. decora var. groenlandica (Schneid.) Fern. P. floribunda Lindl. xP. magdaburgensis Schoch *P. Malus L. P. melanocarpa (Michx.) Willd. xP. mixta Fern. xP. pinnatifida Ehrh. *P. prunifolia Willd. xP. Shurangiaca Ilse xP. Soulardi Bailey *Rhodotypus scandens (Thunb.) Makino Rosa acicularis Lindl. TE + +++ +++ + +++ +++ + ++++++ ++ ++ ++ ++ ++++++++++++ +4+4+444 +++ ++ ++ + +4 + 4 ++ ee ++ ++ + ++ ++ iz ++ u R. acicularis var. Bourgeauiana Crepin + R. blanda Ait. + *R. canina L. KR. carolina L. L pU cT poorer dt ar iaceo af a LE 312 . carolina var. grandiflora (Baker) Rehd. . carolina var. villosa (Best) Rehd. . cinnamomea L. . Eglanteria L. . gallica L. . johannensis Fern. . micrantha Sm. . multiflora Thunb. . nitida Willd. . palustris Marsh. . rubrifolia Vill. . rugosa Thunb. . setigera Michx. . setigera var. tomentosa T. & G. *R. R. spinosissima L. virginiana Mill. Rubus allegheniensis Porter R. (R. pugnax Bailey, R. saltuensis Bailey) allegheniensis X hispidus (R. biformispinus Blanch., R. jactus Bailey, R. laevior (Bailey) Fern., R. permixtus Bailey) . allegheniensis X pensilvanicus (R. alumnus Bailey, R. paulus Bailey) . allegheniensis X setosus (R. ab- brevians Blanch., R. aculiferus Fern., R. frondisentis Blanch., R. glandicaulis Blanch., R. sceleratus Brainerd) . allegheniensis X vermontanus (R. ravus Bailey) allegheniensis var. Gravesii Fern. allegheniensis var. neoscoticus (Fern.) Bailey allecheniensis var. plausus Bailey . allegheniensis var. populifolius Fern. . arenicola Blanch. (R. Brainerdii Rydb., R. curtipes Bailey, Rhodora ++4+4+44++4++44 ++ d d- + + + + ++ + Vt. ++ +4++4++ 4 ++ Mass. ++4+4++ 4444 +++ + [Vol. 69 R.I. + + 444+ 444 Conn. +++++ 4444+ 444+ 4+ E 1967] * B *R. . Jaysmithii Bailey R. pauper Bailey, R. perpauper Bailey) . argutus Link (R. Blakei Bailey, R. jugosus Bailey, R. paludivagus Fern.) . bifrons Vest. . canadensis L. Chamaemorus L. . cuneifolius Pursh . elegantulus Blanch. (R. ami- calis Blanch., R. multiformis Blanch.) . elegantulus X vermontanus (R. multilicius Bailey) . Enslenii Tratt. (R. Baileyanus Britt.) . flagellaris Willd. (R. felix Bailey, R. scambens Bailey) . frondosus Bigel. (R. bellobatus Bailey, R. insulanus Bailey, R. multispinus Blanch., R. recurvans Blanch.) . hispidus L. (R. cubitans Blanch.) . hispidus X setosus (R. ad- jacens Fern., R. Blanchardianus Bailey, R. jacens Blanch., R. tholiformis Fern., R. trifrons Blanch., R. trifrons var. pudens (Bailey) Fern.) - hispidus X ? (R. spiculosus Fern.) . hispidus var. obovalis (Michx.) Fern. . idaeus L. . idaeus var. canadensis Richards . idaeus var. Egglestonii (Blanch.) Fern. . idaeus var. heterolasius Fern. . idaeus var. strigosus (Michx.) Maxim. illecebrosus Focke (R. tetricus Bailey) Me. E + + + + ES ++ 4d + ++ + N.H. + +++ ++ Committee on Plant Distribution Vt. t ++ + + ++ Mass. ++ 4 ++ ++ R.I. ++ ++ 313 Conn. + ++ ++ ++ ++ 314 Rhodora [Vol. 69 Me. N.H. Mass. R.I. Conn. *R. laciniatus Willd. + d xR. neglectus Peck (R. idaeus var. strigosus X occidentalis) R. occidentalis L. R. odoratus L. R. pensilvanicus Poir. (R. amnicola Blanch., R. avipes Bailey, R. bar- barus Bailey, R. facetus Bailey, R. floricomus Blanch.) . pensilvanicus X vermontanus (R. miscix Bailey) *R. phoenicolasius Maxim. R. pubescens Raf. R. pubescens var. pilosifolius A. F. Hill R. recurvicaulis Blanch. (R. arun- delanus Blanch., R. arundelanus var. Jeckylanus (Blanch.) Bailey, R. plicatifolius Blanch.) + + + R. recurvicaulis X setosus (R. arcuans Fern.) R. recurvicaulis X vermontanus (R. severus Brainerd) -+ R. semisetosus Blanch. (R. ascen- dens Blanch., R. hispidoides Bailey, R. ortivus Bigel., R. perinvisus Bailey) + + R. setosus Bigel. (R. Lawrencei Bailey) + + + + + R. setosus X vermontanus (R. Groutianus Blanch., R. Parlinii Bailey, R. univocus Bailey) + + + *R. triphyllus Thunb. R. vermontanus Blanch. (R. tardatus Blanch.) 4 c + + + +++ +g E de + ++ -4 +++ ++ e + + + + + + +++ + ++ ++ +. J- + ++ For the synonomy of species of Rubus which are recognized in Gray’s Manual, 8th ed. but are not treated as distinct species by Hodgdon and Steele, see Rhodora 68: 510-512 (1966). The following two taxa were unintentionally omitted from the latter list of Rejected Species. R. flavinanus Blanch. = hybrid involving allegheniensis R. floricomus Blanch. = pensilvanicus Sanguisorba canadensis L. -+ + + + + *S. minor Scop. + + + + 1967] Committee on Plant Distribution 815 Me. N.H. Vt. Mass. R.I. Conn. *S. officinalis L. + Sibbaldia procumbens L. *Sorbaria sorbifolia (L.) A. Br. + Spiraea alba DuRoi *S. japonica L. f. S. latifolia (Ait.) Borkh. S. latifolia var. septentrionalis + Fern. ES -4 + + + ++ ++ ++ *S. prunifolia Sieb. & Zucc. S. tomentosa L. Waldsteinia fragarioides (Michx.) Tratt. + + = + Te + + + + + + oc The geographical areas and the numbering of the cate- gories in this report are the same as in previous reports. The majority of the species treated are woody. The family contains some of the most perplexing taxa in our flora. I. GENERALLY DISTRIBUTED. — Amelanchier laevis, A. stolonifera, Fragaria vesca and its variety americana, F. virginiana, Potentilla norvegica, Prunus pensylvanica, P. virginiana, Pyrus floribunda, P. melanocarpa, Rosa carolina, R. palustris, R. virginiana, Rubus allegheniensis, R. canadensis, R. hispidus, R. hispidus var. obovalis, R. idaeus var. strigosus, R. pensilvanicus, R. recurvicaulis, R. setosus, R. vermontanus, Spiraea latifolia, S. tomentosa. Fragaria vesca var. americana has more stations in northern Maine than the species. Rubus hispidus and var. obovalis are not represented in Aroostook County but Hodgdon and Steele give the ranges of both taxa throughout New England at low altitudes. Our map shows no collection above 45°. The same is true of var. obovalis which does have one station north of 45^. There is need of more collecting. Ia. GENERAL EXCEPT MAINE COAST EAST OF KENNEBEC RIVER. — Rubus occidentalis. Ib. GENERAL EXCEPT WASHINGTON COUNTY COAST OF MAINE. — Agri- monia. gryposepala, Geum aleppicum var. strictum, Rubus idaeus var. canadensis. Ic. GENERAL EXCEPT CAPE COD. — Geum rivale, Potentilla fruticosa, P. palustris var. villosa, Rosa nitida, Rubus odoratus, R. pubescens. Potentilla fruticosa in general grows on sweet or neutral soils, but not exclusively. It seems to thrive except in extremely acid situ- ations. Rosa nitida is rare west of the Connecticut River. Ie. GENERAL EXCEPT CAPE COD AND WASHINGTON COUNTY COAST. — Agrimonia stricta, Geum laciniata, Potentilla arguta, Prunus nigra. 316 Rhodora [Vol. 69 Geum laciniatum, while widely distributed, is sporadic. Prunus nigra is apparently absent from southeastern Massachusetts and mainland Rhode Island and Connecticut except the Housatonic drain- age area. IIa. NORTHERN — NO STATIONS OR NOT MANY SOUTH OF 43?.— Amelanchier Bartramiana, X A. neglecta (Bartramiana X laevis), A. Wiegandii, Geum macrophyllum, Prunus depressa, Pyrus decora, Rubus elegantulus. IIb. — NORTHERN — NUMEROUS STATIONS SOUTH OF 43?. — Amelan- chier sanguinea, Dalibarda repens, Fragaria virginiana var. terrae- novae, Potentilla tridentata, Pyrus americana. Potentilla tridentata is a northern species found chiefly at higher elevations but locally in sterile acid areas, even at sea level. Pyrus americana is northern, becomes uncommon in southern New England and is apparently absent from southeastern Massachusetts, III. ARCTIC-ALPINE. — Geum Peckii, Potentilla norvegica var. lab- radorica, P. Robbinsiana, Pyrus decora var. groenlandica, Rubus Chamaemorus, R. pubescens var. pilosifolius, Sibbaldia procumbens. Rubus Chamaemorus occurs on the Washington County coast of eastern Maine and also in the White Mountains of New Hampshire. IV. SOUTHERN — GENERAL SOUTH OF 45°. — Amelanchier canaden- sis, Geum canadensis and var. camporum, Potentilla simplex var. calvescens, Rubus Enslenii, R. frondosus, R. Jaysmithii. IVa. SOUTHERN — GENERAL SOUTH OF 45°, BUT NOT ON THE MAINE COAST EAST OF THE KENNEBEC RIVER, — Prunus susquehannae, Rubus arenicola, R. flagellaris. IVb. SOUTHERN — GENERAL SOUTH OF 45°, BUT NOT IN WASHINGTON COUNTY. — Potentilla canadensis, P. simplex, Prunus serotina. IVd. SOUTHERN — GENERAL SOUTH OF 45°, BUT NEITHER ON CAPE COD NOR IN WASHINGTON COUNTY. — Amelanchier arborea, Geum laciniatum var. trichocarpum. V. CHIEFLY THE THREE SOUTHERN STATES. — Geum virginianum, Prunus americana, Rubus semisetosus. Although Rubus semisetosus belongs here, it also occurs sparingly on the coast of Maine and southern New Hampshire. Va. CHIEFLY THE THREE SOUTHERN STATES. BUT NOT IN WESTERN MASSACHUSETTS. — Pyrus arbutifolia, Rubus argutus. Pyrus arbuti- folia is a southern species, occurring in eastern Mass., R.I., and Conn., apparently only within fifty miles of the coast. Vb. CHIEFLY THE THREE SOUTHERN STATES, BUT NOT CAPE COD. — Agrimonia pubescens, A. rostellata, Sanguisorba canadensis. VI. SOUTHWESTERN NEW ENGLAND — CHIEFLY WEST OF THE CON- NECTICUT VALLEY. — Agrimonia parviflora, Rubus cuneifolius. 1967] Committee on Plant Distribution 317 VII. WESTERN NEW ENGLAND — CHIEFLY WEST OF THE CONNECTICUT VALLEY. — Waldsteinia fragarioides. Waldsteinia has two isolated stations in Maine, both in Kennebec County. VIII. COASTAL PLAIN. — Amelanchier nantucketensis. IX. CALCIPHILE. — Amelanchier gaspensis, A. humilis, Rosa blanda, R. johannensis. Amelanchier gaspensis has two stations in the Aroos- took River valley. A. hwmilis has been found at four stations in the Champlain valley. X. MARITIME. — Potentilla Egedei var. groenlandica, P. pectinata. XI. MARITIME WITH INLAND STATIONS. — Prunus maritima. XII. MISCELLANEOUS. — Amelanchier intermedia is northern. There are only six stations, three on the eastern Maine coast, the other three at Rutland and Bellows Falls, Vt. and Florida, Mass. Potentilla anserina is also a northern species. It occurs frequently in the river valleys of Aroostook County, Maine and on the shores of Lake Champlain. There are also Maine collections from Mattawamkeag on the Penobscot River and from Mt. Desert and Cape Elizabeth on the Maine coast. There are a few other stations in Vermont and a col- lection from Plainfield, N. H. on the Connecticut River. Prunus alleghaniensis is a southern species with only four stations, Lisbon in eastern Connecticut and Botsford, Bridgeport and Lyme in the western part of the state. Prunus Gravesii is known only from Groton and Noank in eastern Connecticut where it is apparently an endemic. Rosa acicularis, a western species, is represented by specimens from Randolph and Stratford Hollow, and Plainfield, N. H., Weybridge and Hartland, Vt., and Southbury, Connecticut. The var. Bourgeauiana has been collected at Matinicus, West Bath and Chebeague Island on the Maine coast, Randolph, N. H., Arlington, Ferrisburg, Man- chester, Middlebury and Weybridge, Vt., and on the Blue Hills Reser- vation, Mass., where it was probably introduced. Rosa carolina var. grandiflora occurs at Shelburne, N.H. and is occasional in the three southern states, The var. villosa might be classified as general but it is very sporadic with two stations in Me., one in N.H., one in western Vt., four in eastern Mass., and two in the Quinnipiac Valley, Conn. Rubus allegheniensis var. neoscoticus is represented by a specimen from Plymouth, Mass. and one from Errol, N. H., the determination of which is questionable. Rubus idaeus var. heterolasius is known only from the central Maine coast. The var. Egglestomii is very local, occurring in Windsor and Townsend, Vt. Rubus illicebrosus is represented from Portland, Me. and Huntington, Conn. Spiraea alba is known only from four stations in northwestern Vermont. 318 Rhodora [Vol. 69 XIIIa. INTRODUCED SPECIES — NORTHERN TENDENCIES, — Filipen- dula Ulmaria, Rosa spinosissima, Sorbaria sorbifolia. XIIIb. INTRODUCED SPECIES — SOUTHERN TENDENCIES. — Chaenome- les lagenaria, Duchesnea indica, Filipendula rubra, Fragaria Ana- nassa, Kerria japonica, Physocarpus opulifolius, Potentilla argentea, P. intermedia, P. millegrana, P. recta, P. verna, Prunus Avium, P. Cerasus, P. domestica, P. insititia, P. persica, Pyrus Acuparia, P. communis, P. Malus, Rosa canina, R. cinnamomea, R. Eglanteria, R. gallica, R. rugosa, Rubus idaeus, R. phoenicolasius, Spiraea japonica. XIIIc. INTRODUCED SPECIES — SPORADIC. — Alchemilla minor, Arun- cus dioicus, Exochorda grandiflora, Filipendula hexapetala, F. Ul- maria var. denudata, Gillenia trifoliata, Geum pulchrum, G. urbanum, Potentilla gracilis var. pulcherrima, P. repens, Prunus cerasifera, P. Mahaleb, P. spinosa, Pyrus baccata, P. pinnatifida, P. prunifolia, P. Soulardi, Rhodotypus scandens, Rosa multiflora, R. setigera and var. tomentosa, Rubus laciniatus, R. triphyllus, Spiraea prunifolia. XIIId. INTRODUCED SPECIES — LOCAL. — Sanguisorba minor, S. of- ficinalis. R. C. BEAN A. F. HILL R. J. EATON S. K. HARRIS A CORRECTION in the 16th Report of the Committee on Plant Distribution : In the Sixteenth Report (Rhodora 65:26), the occurrence of Eleocharis palustris (L.) R. & S. var. palustris in Rhode Island and Connecticut based on single specimens each was stated to be doubtful. The inclusion of plus (+) signs op- posite this species under “R. I.” and “Conn.” on p. 27 is therefore an error, and they should be deleted. Actually the two pertinent specimens involved have been re-examined and in our opinion should be referred unhesitatingly to E. palustris var. major Sonder. R. C. BEAN A. F. HILL R. J. EATON S. K. HARRIS PHYSALIS IN MEXICO, CENTRAL AMERICA AND THE WEST INDIES U. T. WATERFALL Continued from Vol. 69, No. 778 69. Physalis foetens Poiret, in Lamarck, Encyclopédie Méthodique. Botanique . . . Supplement 2:348. 1817. - Annual, 10-65 cm tall, spreading-hairy with varying mixtures of short hairs and longer, multicellular ones, many terminating in glands; leaf blades ovate to ovate-lanceolate, coarsely and irregularly toothed, sometimes saliently, sometimes sinuately so; principal blades 2-10 cm long and 15-55 mm wide, on petioles 1-5 cm long; flowering calyx hairy, often densely so, 4-8 mm long and 2.5-6 mm wide at base of lobes, upper 2-4 mm divided into usually lanceolate lobes, attenuate or subulate tipped; flowering pedicels 2-6 mm long; corolla immacu- late, or with slight, and not obvious spots, (8-) 10-15 mm long and 12-25 mm wide when fully expanded; anthers violet or bluish, 1.5-3 mm long, on slender, often blue, filaments 2-5 mm long; fruiting calyx 5-angled, basally cordate, hairy in varying amounts, (15-) 22-49 mm long and (13-) 18-40 mm wide, the upper 4-10 mm divided into attenuate or subulate-toothed lobes; fruiting pedicels 5-10 mm long, reflexed; berry nearly spheric, thin-walled, 10-15 mm in diameter. P. foetens is nearest P. subulata, differing in its immacu- late, or nearly immaculate, larger corollas, larger anthers, and its more abundant glandularity. SELECTED COLLECTIONS: MEXICO: DISTRITO FEDERAL: Sierra de Ajusco, 8500 ft, Oct. 2, 1895, Pringle 6219 ( BR, F, GH, MEXP, NY, UC, US, VT); DURANGO: Durango, 1896, Palmer 695 (US); GUANA- JUATO: near Dolores Hidalgo, Aug. 10, 1947, Kenoyer 1939 (GH); HIDALGO: near Tequixquiac, Aug. 30, 1903, Rose & Painter 6644 (NY, US); JALISCO: 5 miles s of Ojuelos on road to Lagos de Moreno, July 19, 1955, Weintraub & Roller 67 (MICH) ; MEXICO: cerca de Totolcingo, municipio de Acolman, July 18, 1965, Rzedowski 20217 (MEXP); OAXACA: Oaxaca, Sept. 25, 1930, Russell 220 (US): San Luis Tultit- lanapa, near Oaxaca, July 1908, Purpus 3582 (BM, F, GH, NY, UC, US) ; QUERETARO: silty flats with shrubs and cacti, Aug. 23, 1961, Waterfall 16547 (F, MICH, OKLA, US) ; SAN LUIS POTOSI: in arenosis circa Morales, 1867, Schaffner 701, (GH) ; SINALOA: Rosario, Apr. 4, 1910, Rose et al 14580 (US); VERACRUZ: Mt. Orizaba, Aug. 15, 1891, Seaton 372 (F, NY, US). 319 320 Rhodora [Vol. 69 Bourgeau 352 (BR, GH, P, US) Cacubaya, Vallée de Mex- ico, July, 1865-1866, is a collection that was in the Poiret Herbarium, accepted by him as P. foetens. 70. Physalis subulata Rydberg, Bull. Torr .Bot. Club 22: 306. 1895. Annual, 10-60 em tall, spreading-hairy with varying mixtures of short and longer, multicellular trichomes, few or many terminating in glands; leaf blades ovate to broadly ovate, sometimes basally oblique, margins irregularly and coarsely dentate, sometimes sinu- ately so; principal leaf blades 3-7 cm long and 2.5-6 cm wide on petioles 1.5-3 (-5) em long; flowering calyx hairy, 3-4.5 mm long, upper one-half to two-thirds divided into lanceolate, often subulate- tipped, lobes, on pedicels 1.5-3 (-5) mm long; corolla yellowish, macu- late, 6-7 mm long and 5-11 mm wide when fully expanded; anthers violet or bluish, (0.3-) 1-1.5 (-2) mm long on filaments 2-3 mm long; fruiting calyx hairy, 5-angled, basally cordate, 2-3 cm long and 2-2.5 cm wide, the upper 6-10 mm divided into narrow teeth, usually subu- late-tipped; fruiting peduncles 5-7 (-10) mm long, reflexed; berry nearly spheric 6-10 mm in diameter. 70a. var. subulata Leaves usually coarsely and irregularly dentate to sinuate-dentate. SPECIMENS EXAMINED. MEXICO: CHIHUAHUA: waste grounds, Guerrero, Sept. 8, 1887, Pringle 1344 (Type: GH; Isotypes: F, NY, US, VT); DURANGO: Tejamen, Aug. 21-27, 1906, Palmer 494 (F, GH, NY, US); Durango, Sept. 1896, Palmer 634 (F, GH, NY, UC, US). 70b. var. neomexicana (Rydberg) Waterfall, comb. nov., P. neo- mexicana Rydb., Mem. Torr. Bot. Club 4:325-326. 1895; P. foetens var. neomexicana (Rydb.) Waterfall, Rhodora 60: 168. 1958. This has the small maculate corollas of P. subulata, not the large, immaculate, or sometimes slightly maculate, corol- las of P. foetens, the closest relative. The var. neomexicana also has leaves less deeply, and more regularly toothed than either var. subulata, or P. foetens. COLLECTIONS EXAMINED. MEXICO: BAJA CALIFORNIA: Palm Valley, May 30, 1883, Orcutt s.n. (US). 71. Physalis angustiloba Waterfall, sp. nov. Planta herbacea, ad 1 meter alta; trichomatibus mollibus, articula- tis, partim glanduliferis, ad 1 mm longis; foliis ovatis, cuspidatis, inaequaliter magnodentatis vel integerrimis, principalibus 4-11 cm longis et 2.5-10 cm latis, petiolis 2-7.5 em longis; calycibus floriferis 3.5-6 mm longis et 3-5 mm latis ad basim loborum; calycis lobis 2-3.5 mm longis lanceolato-attenuatis vel angustioribus; pedicellis 2-5 mm longis; corollis luteis, maculatis, 9-15 mm longis et 13-15 mm latis; 1967] Physalis — Waterfall 321 antheris violaceis, 2-3 mm longis; calycibus fructiferis pentangulatis, 25-30 mm longis et 18-28 mm latis, vestitis; pedicellis fructiferis 3-5 (-15) mm longis; baccis 10-13 mm latis. TYPE: Rogers McVaugh 21141 (MICH) bushy annual 1 meter high, sandy soil in tropical deciduous forest, hills 2 miles n of La Cuesta, road to Talpa de Allende, JALISCO, Nov. 19, 1960, Isotype: (OKLA). COLLECTIONS SEEN. MEXICO: AGUASCALIENTES: oak forest near summit, Sierra del Laurel, 10 miles se of Calvillo, McVaugh 18439 (MICH, OKLA); GUERRERO: llano, San Antonio, Montes de Oca, Oct. 25, 1937, Hinton 11541 (GH, MICH, NY); JALISCO: McVaugh 21141 type, cited above. Each of the 3 collections cited above present differences which tempt one to assign each to a different species. Per- haps future collections will clarify the situation, 72. Physalis turbinatoides Waterfall, sp. nov. Planta herbacea; trichomatibus mollibus, articulatis, partim glandu- liferis; foliis ovatis, cuspidatis, inaequaliter et magne paucidentatis, principalibus 5-9 cm longis et 3-6 cm latis; petiolis 3-6 cm longis; calycibus floriferis vestitis, 2.5-3 mm longis et 2-2.8 mm latis ad basim loborum; calycis lobis anguste lanceolatis, 1.5-2 mm longis, pedicellis 3-5 mm longis; corollis immaculatis vel pallido-immaculatis, 5-6 mm longis et 5-7 mm latis; antheris luteis, 1-1.5 mm longis, filamentis filiformibus, 1-2 mm longis; calycibus fructiferis plus minusve vestitis praesertim ad venis, 24-32 mm longis et 15-20 mm latis, lobis attenua- tis, 6-10 mm longis; baccis 8-11 mm latis. TYPE: C. G. Pringle 7263 (vT) near Cuernavaca, 5000 ft, MORELOS, MEXICO Sept. 22, 1896; Isotype: (GH). 73. Physalis turbinata Medicus, in Acad. . . . Theod. - Palat. 4: 189- 190, t. 5, fig. 2. 1780; Alkekengi barbadense . . . Dillenius, Johannes Jacobus, Hortus Elthamensis, 10, t.9, f.9. 1732; Physalis barbadensis Jacquin, Nickolaus Joseph, Miscellanea Austriaca Sive Plantarum Selectarum, 1781; P. hirsuta var. barbadensis (Jacq.) Dunal in DC Prod. 13(1): 446. 1852. Annual, 15-100 em tall, stems spreading-hairy, the trichomes joint- ed, often gland-tipped; leaf blades ovate, often attenuate-tipped, margins entire to irregularly, often sinuate-dentate, 1-15 teeth on each side; principal leaf blades 5-11 cm long and 4-8 cm wide, ap- pressed, jointed hairs in varying abundance on both sides, sometimes restricted to the veins, sometimes some reduced to mere bases; petioles of principal leaves 2-7 cm long; flowering calyx (3-) 4-6 mm long, upper one-half to two-thirds divided into narrowly lanceolate lobes; flowering pedicels 4-6 mm long; corollas yellowish, maculate, hairy in the throat, 6-12 mm long and 7-15 mm wide; anthers bluish, 2-2.8 mm long on filaments 3-4 mm long; fruiting calyx 5-angled, cordate- 322 Rhodora [Vol. 69 based, pubescent, 3-4 cm long and 2-3 em wide, its lobes 5-10 mm long and usually somewhat attenuate; fruiting pedicels 5-15 mm long; berry nearly spheric, 9-13 mm in diameter, on a gynophore 1-2 mm long, the gynophore and lower inner surface of the fruiting calyx sometimes with gland-tipped hairs. This species is similar to P. pubescens, and tends to merge with it in its extreme manifestations, but the flowers and fruiting calyces are larger, and the latter have longer, and usually narrower calyx lobes, in this characteristic re- sembling P. cordata, which, however, has glabrous fruiting calyces, P. porrecta, and some phases of P. ignota. The occasional presence of gland-tipped hairs inside the fruiting calyx near its base is also reminiscent of P. ignota. Much material identified with the name P. turbinata in the past is here referred to P. cordata Miller. If the species- concepts of the two were identical, this course would be necessary due to the priority of the latter name, But the original concept of P. turbinata Medic, specifies a hairy plant, this actually being a proposed new name for Alke- kengi barbadense nanum, alliariaefolia Dillenius, Hort. Elth., the specific name being taken from part of the Dil- lenius description of the fruiting calyx, “ab initio angus- tiores, postia ampliores magis . . . angulosae, pentagone nempe...in mucronem longiorum productae . . . in quibus acini latent grandiusculi turbinati , . ." It is to such material, which has usually passed as large P. pubescens, that the name P. turbinata is here applied. SELECTED COLLECTIONS. MEXICO: coLiMA: Colima, Aug. 1897, Palmer 121 (MICH, UC, US) ; GUERRERO: second barranca east of Temisco, north of Río Balsas, Adama, Nov. 6, 1937, Mexia 8760 (F, GH, NY); JALISCO: wet banks near Guadalajara, Sept. 27, 1902, Prin- gle 1164315 (F, GH, US); MICHOACAN: pine forest on precipitous slopes 3 miles below lumber camp at Dos Aguas, Sept. 15, 1958, McVaugh 17865 (MICH); SINALOA: short tree forest, sand valley in foothills, Comedero, Jan. 27, 1940, Gentry 5391 (NY) ;YUCATAN: Gaumer 482 (MICH, UC, this number is P. viscosa var. yucatanensis at BM, F, GH, us). BRITISH HONDURAS: San Andres, Corozay, July 1933, Lundell 4932 (MICH). GUATEMALA: bank of river, Gulan, Jan. 17, 1905, Deam 387 (GH). HONDURAS: Amapala, Valle, Sept. 11, 1945, Rodriguez 3375 (F); between Juticalpa and airport, Mar. 14, 1949, Standley 18018 (F). PANAMA: near Madden Dam, Rio Chagres, 1967] Physalis — Waterfall 323 Canala Zone, July 30, 1935, Seibert 554 (GH, NY). COSTA RICA: Mar. 1900, Pittier 16082 (GH, NY, US); CUBA: San Jose, Santa Clara, June 1941, Howard 5170 (GH, NY). DOMINICAN REPUBLIC: bank of Maimon River, Piedra Blanca, La Vega, Jan. 18, 1946, Allard 14748 (NY). HAITI: Marmelada, du Nord, Dec. 18, 1925, Leonard 8110 (NY). PUERTO RICO: Rio Piedras, May 4, 1899, Heller 1277 (NY). ST. VINCENT: near Palmyra, Smith 1328a (GH). TORTOLA, V.I.: April 1913, Fishlock 5 (NY). 74. Physalis pubescens L., Species Plantarum 1:183, 1753; P. villosa Miller, Gard. Dict. ed 8, 1768; P. obscura Michx., var. glabra Michx. & var. viscido-pubescens Michx., Flora Boreali-Americana 1:149. 1803; P. hirsuta Dunal in D. C., Prod. 13(1) : 445. 1852; P. viscido-pubescens (Michx.) Dunal, Lc.; P. floridana Rydberg in Small, Flora of the Southeastern U. S. 983. 1903; P. pubescens var. minutifolia O. E. Schulz in Urban, Symbolae Antillanae . . . 6:145. 1909; P. pubescens var. glabra (Michx.) Waterfall, Rhodora 60:165. 1958 (See discus- sion under P. cordata). Annual, 8-90 cm tall, usually villous and sometimes viscid, varying to more or less glabrate; leaf blades ovate, often acuminate, bases sometimes inequilateral, margins usually irregularly several-toothed, sometimes entire or with 1-few teeth, surfaces varyingly soft-hairy with vestiture more abundant, and appressed, on the veins of the abaxial surface, sometimes nearly glabrous; principal blades usually 4-9 cm long and 2-4 cm wide on petioles 2-7 cm long; flowering calyx usually 4-10 mm long and 3-12 mm wide at base of lobes; upper 1-4 mm of calyx divided into ovate-deltoid to lanceolate lobes; flowering pedicels 3-6 mm long; corolla yellowish, dark-maculate, more or less matted-hairy in the throat below the maculations, 7-10 (-12) mm long and 10-15 mm wide; anthers bluish or violet, 1.5-3 mm long on fila- ments 2-3 mm long; fruiting calyx 5-angled, usually prominently so, usually soft-hairy, 18-30 mm long and 13-22 mm wide on pedicels 5-13 mm long; berry 10-18 mm in diameter, sessile or subsessile on the invaginated calyx-base. Physalis pubescens var. minutifolia O. E. Schulz is here considered as an unusually small extreme of the species, rather than as a distinct variety. It would be interesting to check further to see if there could be a geographic popula- tion in the Barbados similar to the type of the proposed variety. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: San José del Cabo, Jan. 20, 1928, Jones 24393 (F, GU) ; CHIAPAS: Cerro del Boqueron, Sept. 1913, Purpus 6993 (NY, UC) ; CHIHUAHUA: southwest- ern Chihuahua, 1885, Palmer 140 (GH, NY, US) ; JALISCO: Barranca of Guadalajara, Aug. 4, 1902, Pringle 8630 (BM, F, GH, NY, UC, US, VT); 324 Rhodora [Vol. 69 MICHOACAN: semi-desert scrub, 4 miles w of Apatzingan, Aug. 8, 1941, Leavenworth 1381 (F); NAYARIT: moist tropical forests, precipitous mountainsides, 12 miles w of Tepic, Sept. 8, 1960, McVaugh 18854 (MICH); Maria Madre Island, May 1897, Nelson 4255 (F, GH, US); NUEVO LEON: wocded hillsides, 19 miles w of Linares, McGregor et al 214 (KANU); OAXACA: Tomellin Canyon, Dec. 17, 1895, Pringle 7004 (GH, MEXP, VT); PUEBLA: Cerro de Coatepe, Aug. 1907, Purpus 2711 (UC); SINALOA: Culiacan, Oct. 20, 1904, Brandegee s.n. (UC); San Blas, Mar. 24, 1910, Rose et al 13382 (NY, US); thickets, Villa Union, Apr. 2, 1910, Rose et al 13885 (NY, US) ; SONORA: creek w of Pilares de Nacozari, Nov. 19, 1939, Drouet et al 3678 (F); TAMAULIPAS: Vic- toria, 1907, Palmer 53 (F, UC, US) ; TEPIC: ravine, Acaponeta, Apr. 9, 1910, Rose et al 14231 (GH, NY, US) ; VERACRUZ: Remudadero, March 1923, Purpus 9001 (F, GH, NY, UC, US); YUCATAN: Gaumer 23514 Kancabconot, Jan. 1917 (F, GH, Ny, US). BRITISH HONDURAS: river bank, Stann Creek District, Jan. 27, 1937, Gentle 1918 (GH, MICH, NY, US). COSTA RICA: La Lola, Limon, Oct. 21, 1953, Heiser 3654 (F); wet meadow, San Jose, Standley 41220 (Us). EL SALVA- DOR: San Salvador, Dec. 20, 1921, Standley 19552 (GH, NY, US). GUATEMALA: Quiriga, Izabal, May 15, 1922, Standley 24074 (GH, NY, US) ; Río Ixcan, Huehuetenango, July 23, 1942, Steyermark 49330 (F, US). HONDURAS: banana field, La Fragua, Atlantida, Dec. 7, 1927, Standley 52657 (F, US). NICARAGUA: El Recreo, Zelaya, Apr. 23, 1949, Standley 19110 (F). PANAMA: San Jose Island, Jan. 10, 1946, I. M. Johnston 1105 (GH, US) ; Gorgana Beach, Aug. 7, 1938, Woodson et al 1694 (GH, NY). BAHAMAS: New Providence, Apr. 7, 1904, Britton 123 (NY). CUBA: Sierra de Cabra, Pinar del Rio, Mar. 6, 1911, Britton et al 9811 (NY); San Pedro, Isle of Pines, Feb. 15, 1916, Britton et al 14334 (NY); Sierra Maestra prope Daiquiri, Ori- ente, Oct. 28, 1916, Ekman 8075 (NY). DOMINICAN REPUBLIC: Sierra de Ocoa, Ciudad Trujillo (Santo Domingo), Feb. 28, 1929, Ekman 11709 (NY—as var. minutifolia). HAITI: St. Michel de l'Atalaye, Nov. 20, 1925. (GH, NY). JAMAICA: New Market, Sept. 13, 1907, Britton 1602 (NY). PUERTO RICO: sand dunes, Catano, Feb. 14, 1914, Britton & Cowell 1552 (NY). TOBAGO: Charlotteville, Jan. 20, 1953, Hunnewell 19986 (GH). TORTOLA, V.I.: Fishlock 285 (GH). 75. Physalis leptophylla Robinson & Greenman, Proc. Amer. Acad. Arts & Sci. 29: 389-390. 1894. Annual, 0.5-1 meter tall; vestiture of soft, somewhat tangled, multi- cellular hairs of different lengths up to 1 or 1.3 mm long, some with elongate, terminal glands; leaf blades thin, ovate, acuminate, principal ones 4-8 cm long and 2.5 -6 cm wide, on petioles 2-8 cm long; margins of blades entire, or rarely with a few irregular, shallow teeth; sur- faces of blades glabrous, or nearly so, but margins more or less 1967] Physalis — Waterfall 325 ciliate, bases somewhat inequilateral; flowering calyx 2-3 mm long and 2-4 mm wide at base of the ovate-deltoid lobes fromed from the upper 1-1.5 mm of the calyx, pedicels 1-2 (-3) mm long; corolla light yellow or cream-colored, with brown or purplish maculations, glab- rous, or nearly so, within, 5-9 mm long and 6-10 mm wide; anthers violet to purplish, ovate to oblong, 0.8-1.5 mm long, on filaments 2-3 mm long; fruiting calyx thin, 5-angled, but seldom prominently so, 12-20 mm long and 10-15 mm wide, on pedicels 3-5 mm long; berry 4-8 mm long, sessile to subsessile on a calyx invaginated 1-1.5 mm. SELECTED COLLECTIONS. MEXICO: BAJA CALIFORNIA: near spring, north side of Cerro de la Giganta, Distritio del Sur, Nov. 28, 1947, Carter et al, 2073 (UC, US) ; CHIHUAHUA: canyon, Guasaremos, Rio Mayo, Aug. 25, 1936, Gentry 2442 (F, GH) ; COLIMA: steep hills, 14 miles wnw of Santiago in ravines, Nov. 7, 1960, MeVaugh 20782 (MICH, OKLA); DURANGO: shady canyon bottom, 10 miles w of Tama- zula, Dec. 18, 1939, Gentry 5270 (GH, NY); GUERRERO: Cutzamala, Coyuca, Sept. 26, 1936, Hinton 8493 (US); JALISCO: along stream in steep moist ravines, 9 miles n of road junction at w end of Bahia de Navidad, Dec. 12, 1959, McVaugh & Koelz 1730 (MICH, OKLA); Bar- ranca of Tepie, Oct. 11, 1893, Pringle 5455 (GH, US, VT); MEXICO: Bejucos, Temascaltepec, Nov. 13, 1933, Hinton 5204 (US); SINALOA: under shrubs, 64 miles s of Culiacan, Breedlove 1549 (US) ; SONORA: moist shady soil under trees or shrubs, Alamos, Oct. 28, 1939, Gentry 4771 (F, MICH, NY, OKLA, UC, US). GUATEMALA: brushy rocky slope between Zacapa and Chiquimula, Oct. 9, 1940, Standley 7381 (F). In its extremes, P. leptophylla tends to merge with P. pubescens. 76. P. hylophila Standley, Journ. Wash. Acad. Sci. 14: 243. 1924. Annual, 30-90 cm tall, stems soft-hairy with jointed whitish, more or less viscid trichomes, the longer ones 1-2 mm long, often tipped with pinkish or light brown glands, sometimes intermixed with shorter hairs which also may be glandular-capitate; leaf blades ovate, often acuminate, margins entire or with a few, irregularly sinuate teeth; young blades with varying amounts of hair similar to that of the stems but often appressed on the veins, more or less glabrate; principal blades 2.5-4 cm long and 2-2.5 cm wide, but all specimens examined have lost lower, possibly larger, leaves; petioles 1-2.5 em long; flowering calyx 3-5.5 mm long and 3-5 cm long; corolla yellowish, apparently immaculate or with spots not strongly contrast- ing, tube glabrous within; corolla 6-7 mm long and 7-8 mm wide when fully expanded; anthers yellowish, light greenish-yellow or bluish tinged, 1.1-1.3 mm long, on filaments 3-4 mm long; fruiting calyx 5-angled, but not strongly so, and with intermediate ribs, hairy, 14-16 mm long and 10-12 mm wide on pedicels 3-5 mm long. 326 Rhodora [Vol. 69 Physalis hylophila is similar to P. leptophylla in its small anthers, corolla tube essentially glabrous internally, and in having hairs often gland-tipped. They differ in P. lepto- phylla having contrastingly maculate corollas, smaller flow- ering calyces, lighter colored anthers, and a smaller, less obviously 5-angled fruiting calyx. COLLECTIONS EXAMINED. MEXICO: sonora: under large trees on lower slope of mountain w of Alamos, Dec. 12, 1939, Drouet & Richards 3971 (F). EL SALVADOR: wooded slope along lake, Laguna de Maquigue, La Union, Feb. 18, 1922, Standley 20971 (Type: US; Isotypes: GH, NY). PANAMA: EI Real, Darien, Oct. 7, 1938, Allen 956 (F, GH, NY, US). 77. Physalis vestita Waterfall, sp. nov. Planta herbacea, multicaulis e radice lignose elongata; caulibus 15-65 cm longis, lanatis, probabiliter adscendentibus vel decumbenti- bus, interdum ad basim fruticosis; foliis ovatis, dense canis lanosoves- titis, integerrimis vel inaequaliter paucidentatis vel undulatis, principalibus 25-40 mm longis et 15-28 mm latis, petiolis dense vestitis, 10-35 mm longis; calycibus floriferis dense vestitis, 3-4 mm longis et 4-5 mm latis; calycis lobis deltoideis, 1 mm longis; pedicellis floriferis 4-6 mm longis; corollis pallidoluteis, maculatis, tubis intus vestitis, 7-10 mm longis et 10-13 mm latis; antheris coeruleis vel violaceis, 3-4 mm longis; filamentis filiformibus, 2.5-4 mm longis; calycibus fructiferis pentangulatis, dense vestitis, 15-22 mm longis et 13-18 mm latis; pedicellis fructiferis 7-13 mm longis; baccis 8-10 mm latis. TYPE: J. N. Rose 13766 (US) vicinity of Mazatlan, Mar. 3, 1910, Sinaloa; Isotypes: (F, GH, NY). COLLECTIONS SEEN. MEXICO: siNALOA: Mazatlan, Nov. 20, 1926, Jones 22515 (F); cliff facing the ocean, Signal Hill, Mazatlan, Sept. 16, 1925, Mexia 32 (uc); Rose 13766, Type, cited above. 78. Physalis clarionensis Waterfall, sp nov. Planta herbacea; ramis 15-45 cm longis, glabris; foliis ovatis, acuminatis, inaequaliter crasse dentatis, principalibus 2-5 cm longis et 15-25 mm latis, petiolis 10-22 mm longis; calycibus floriferis glabris, 3.5-4 mm longis et 2-4 mm latis ad basim loborum; calycis lobis deltoideo-lanceolatis, 2 mm longis; pedicellis floriferis 4-5 mm longis; corollis pallido-luteis, obscure maculatis, jugulis glabris, 5-6 mm longis et 8-9 mm latis; antheris coeruleis vel violaceis, 1.5 mm longis; filamentis filiformibus, 2-3 mm longis; calycibus fructiferis pentangulatis, glabris, 23-27 mm longis et 16-18 mm latis; calycis lobis ovatis, mucronatis, 4-6 mm longis, pedicellis fructiferis 5-10 mm longis; baccis 10 mm latis. 1967] Physalis — Waterfall 327 TYPE: A. W. Anthony 411 (uc), Clarion Island, off the coast of Lower California, March-June, 1897; Isotypes: (F, GH, US). Physalis clarionensis resembles some phases of P. an- gulata, but has prominently 5-angled fruiting calyces in con- trast to the 10-ribbed ones of the latter species, and has maculate corollas, smaller than usual for P. angulata, as well as small anthers. Except for its relatively small corol- las and anthers and 5-angled fruiting calyces, it might seem referable to P. philadelphica (P. ixocarpa). The specific name refers to the Clarion Islands, the type locality. 79. Physalis minuta Griggs, Torreya 3: 138-139. 1903. Herbaceous, annual, stems 15-35 cm tall; vestiture of very short, appressed, antrorsely curled hairs; leaf blades ovate to lanceolate- ovate, principal ones 10-25 cm long and 6-12 mm wide on petioles 7-20 (-50) mm long; flowering calyx 2-3 mm long, divided one-fourth or one-fifth into attenuate lobes; corolla 4-b mm long, immaculated or with slightly contrasting spots; anthers 1.3-2 mm long, yellowish or bluish tinged; fruiting calyx 5-angled, 15-18 (-20) mm long and 12-15 mm wide on filiform pedicels 4-10 mm long; berry 7-8 mm in diameter, sessile, or nearly so, on the invaginated calyx-base, which is min- utely and sparsely capitate-hairy. Griggs, in describing this species, emphasized the alleged minute size of the corolla, basing the specific name on it. Speaking of them he said “. . . they would hardly be noticed except on careful search or by accident . . . corolla very small, about 2 mm in diameter when fully expanded . . .” On a sheet of the type collection (US) there is one opened corolla, irregularly flattened and pressed out on only one side, measuring 2.5 mm from its center to the margin of the one pressed lobe, therefore, 5 mm in diameter. On à duplicate sheet (GH) there are two pressed corollas, well- flattened; one is ca. 4 mm long and the other is ca. 5 mm long. Griggs must have been overly impressed by the minute undeveloped flower buds. In Physalis old plants reflowering under unfavorable conditions may produce small flower buds which either fall unopened, or, sometimes, produce small undeveloped flowers. This is another characteristic complicating the taxonomy of the genus. 328 Rhodora [Vol. 69 In any event, this species is not characterized by extreme- ly small flowers, although it is in the small-flowered series, so this emendation to the description is necessary to permit future identification of the species, and to prevent its being redescribed in the future on the basis of young, vigorous plants with well-developed flowers. It might also be noted that the type collection is of ma- terial from which the larger leaves have fallen. COLLECTIONS EXAMINED. MEXICO: coLiMA: along road be- tween El Ciruelo and Cuyutlan, Manzanillo, Mar. 11, 1943, Gilly et al 11 (MICH) ; GUERRERO: Acapulco, Oct. 1894 to March 1895, Palmer 304 (Type: US; Isotype: GH); Acapulco, Thiebaut 1101 (P). COSTA RICA: Braxilito Bay, July 2, 1932, Howell 10205 (GH, Us) ; Murcie- lago Bay, July 2, 1932, Howell 10211 (F); la base de Salises, July 1890, Pittier 2634 (BR, US); sand beach, 5 km east of Puntarenas, Aug. 24, 1958, Sauer, 2336 (wis). HONDURAS: vicinity of El Zomorano, Morazan, July 21, 1949, Standley 21520 (F). NICARA- GUA: shady places, Asseradores Islands, Jan. 15, 1903, Baker 147 (GH, MICH, NY, UC, US) ; sandy field, sea level, Corinto, Chinandege, July 19, 1947, Standley 11555 (F). 80. Physalis carnosa Standley & Steyermark, Field Mus. Publ. Bot. 23: 19. 1943. Herbaceous, much-branched, prostrate or decumbent, stems 15-20 cm long; vestiture of very short, antrorsely curled, appressed hairs; leaf blades succulent, ovate to rhombic-ovate, principal ones 15-25 mm long, on petioles 5-12 mm long, nearly glabrous, or with a few hairs similar to the stem-hairs; flowering calyx 2-2.5 mm long, upper one-third or one-fourth divided into narrowly triangular to ovate- triangular lobes; flowering pedicels ca. 2 mm long; corolla 3-4 mm long, immaculate (but possibly faded); anthers 0.8-1.2 mm long, bluish-green, on filiform filaments ca. 1.5 mm long; fruiting calyx 5-angled, 10-183 mm long and 7-9 mm wide on pedicels 4-5 mm long; berry ca. 6 mm in diameter. TYPE: Julian A. Steyermark 37766, sand dunes, alt. 1-2 meters, Ocos, San Marcos, GUATEMALA, Mar. 15, 1940. (F). DEPARTMENT OF BOTANY AND THE RESEARCH FOUNDATION OKLAHOMA STATE UNIVERSITY STILLWATER, OKLAHOMA 1967] Physalis — Waterfall 329 LITERATURE CITED No attempt is made in the following list to duplicate the many citations made under the species names and their synonyms. DUNAL, FELIX. 1852. Physalis in DeCandolle’s Prodromus Syste- maticus Naturalis Regni Vegetabilis. Paris: Crapalet. GREENMAN, J. M. 1900. Proc. Amer. Acad. Arts & Sci. 35: 311-312. HERMANN, PAUL, 1679-1686. Horti Academici lugduno-Batavia Cata- logus . . . Pl. 571. LANJouw, J. & F. A. STAFLEU. 1964, Index Herbariorum, Regnum Vegetabile 31(1), ed. 5. RYDBERG, P. A. 1896. The North American Species of Physalis and Related Genera. Mem. Torr. Bot. Club 4: 297-374, SLEUMER, H. 1950. Estudios sobre el género Dunalia. Lilloa 23: 117-142. STANDLEY, P. C. 1924. "Trees and Shrubs of Mexico. 1277-1288. STANDLEY, P. C. 1931. Eutheta. Field Mus. Publ. Bot. 8: 324. 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. PINE HILLS ELYMUS GEORGE L. CHURCH The wide variation from stout filaments to mere stubs in the glume structure of occasional populations of Elymus Hystrix L. strongly suggests that the origin may lie in the hybridization and introgression of the common glumeless forms with some other species of Elymus possessing well- developed glumes (Church 1967). In fact, experimental hybrids between E. Hystrix and E. canadensis L, var. cana- densis, although sterile, exhibit spikes with abundant and setaceous glumes, intermediate between those of the par- ents. Also noteworthy is the fact that the same strain of glumeless E. Hystrix has been crossed with E. Svensonii Church, an endemic Tennessee species with long, setaceous glumes, to produce a partly fertile F.. Elymus Svensonii in turn was crossed with E. virginicus L. var. glabriflorus (Vasey) Bush f. glabriflorus, a common form with rela- tively broad glumes in southeastern United States, to pro- duce a vigorous, intermediate and segregating F, (Church 1967). Hence it seemed reasonable to expect that in habi- tats where E. Hystrix and f. glabriflorus were sympatric, natural hybridization might give rise to partly fertile plants variously intermediate between the parents. An appropriate location in which to study hybridization of these two taxa was found on the limestone bluffs imme- diately adjacent to the Mississippi River drainage area near La Rue, Union County, Illinois. Fortunately, the area is maintained as a wildlife reservation by the University of Southern Illinois and through the kindness of Professor J. W. Voigt, the author was able to locate the station where collections of Elymus with setaceous glumes had been made. Here one may find E. Hystrix and abundant stands of E. virginicus that exhibit considerable variation in glume 330 1967] Elymus — Church 331 structure, and these populations constitute the major part of the present study.' A diagram of the Pine Hills area, based on a tracing of the geological survey map of the region (Alto Pass Quad. U.S.G.S.), is shown in figure 1. In the upper section, the bluff of Devonian, sandy limestone rises a sheer 500 feet in places on the west side, as indicated by the contour lines. Immediately adjacent is a low, springy marsh through which the Big Muddy River (left middle) cuts through to the Mississippi River (lower left). The heavy line at the lower right represents Illinois Route 3, from which a dirt road leads eastward across a ditch and old railroad bed to the foot of the bluff. At this point the path continues through mixed, dense woodland northward, skirting the bluff, before it turns westward (upper left) and rises from the moist, shady low area to the dry wooded crest. Here the forest dominants include the Post Oak (Quercus stellata Wang.), Black Hickory (Carya buckleyi Sarg.), and Red Cedar (Juniperus virginiana L. var. crebra Fern. and Grisc.) particularly on the edges of dry, hilltop prairies. The name of the locality, however, stems from the rare stands of the southern Shortleaf Pine (Pinus echinata Mill) just back from the bluff edges (Mohlenbrock and Voigt 1959, Voigt and Mohlenbrock 1964). The Elymus populations are confined largely to the edges of woodlands all along the path as is indicated in figure 1. From the author's observations during two collecting trips in 1960 and 1961, E. Hystrix is confined largely to the shady moist woodlands at the swamp level of the trail, with occa- sional stands at the beginning of the higher and drier wood- lands (upper left, fig. 1, “Government Rock", under one mile scale). Scattered along the lower trail but much more abundantly on the upper dry bluff area are stands of E. "These studies were greatly facilitated by generous support from the Section for Systematic Biology of the National Science Founda- tion — grants GB-180 and G-9046. For assistance in the breeding program, I wish to thank Roger W. Farrington, Michael S. Greenwood and Susan R. McGunigle. 332 Rhodora [Vol. 69 virginicus var. glabriflorus which is readily distinguished from other varieties by longer glumes, longer lemma awns and well-exserted spikes (Fernald 1933). It is distinguished also from E. canadensis L. var. canadensis, with which it could be confused, by the indurated, non-striated bases of the occasionally bowed glumes, the shorter, obtuse (rather than clearly bi-dentate) tipped paleas and the very erect spikes. One variant of var. glabriflorus has scabridulous glumes and lemmas: it is forma glabriflorus and will often be re- ferred to in the remainder of this paper as f. glabriflorus. The other variant of var. glabriflorus has densely hirsute glumes and lemmas: it is forma australis and will often be referred to as f. australis. Both taxa grow in close asso- ciation (figs. 1, 3, 4). Quite unique among all Elymus at Pine Hills are popula- tions with very setaceous glumes that appear to be natural hybrids of E. Hystrix and either f. glabriflorus or f. austra- lis (fig. 4). The glume structure has, in fact, been the basis of former identification of these stands as representatives of E. canadensis L. var. interruptus (Buckl.) Church, a very different taxon (Mohlenbrock and Voigt 1959, Church 1967). These putative hybrids lie in greatest concentration at the first part of the path along the dry bluff just above the moist basal section (fig. 1). The locality would appear to be an area where E. Hystrix and E. virginicus come into contact. Here the cut and scraped margins of woodland paths offer a disturbed and therefore favorable habitat for the spread of hybrids between these parents. THE NATURAL HYBRIDS A study of the variable populations with respect to spikelet structure, cytology of meiosis of the pollen-mother- cells and fertility, measured by percentage of mature seeds, has presented two basic types: — a, those with glabrous lemmas, represented by 2670* (fig. 4) and b, those with "Numbers refer to the author's collections in the Brown University Herbarium. 1967] Elymus — Church 333 hirsute lemmas 2662 (fig. 4). As the diagram in figure 1 indicates, the putative hybrids are closely associated with either f. glabriflorus or f. australis; 2652, 2613 respectively (fig. 4). Within group a, some populations have glumes that are heavily indurated and bowed at the base but taper to a setaceous awn and in these features resemble f. glabriflorus, 2595 (fig. 7). The longer (8 mm.) paleas, with truncate apices, however, suggest an E. Hystrix relationship. Based on a normal average of two seeds per spikelet, only 4% seed is formed. Much more frequent are forms (2668, 2670, figs. 2 and 4) with setaceous glumes, 0.2-0.5 mm. wide and pairs unequal in length, 12-15 or 18-30 mm. long. The reduction of occasional glumes to stubs and the partly spreading spikelets again suggest an E. Hystrix parentage. The ciliate margins of the paleas, on the other hand, are characteristic of f. glabriflorus. Examination of pollen-mother-cells (at least 50 in each stage) reveals chiasma frequencies of 1.73-1.92 (median 1.90), two or three bridges at first meiotic anaphase and extrusions in ca. 25% of the mature pollen. Seed set is ca. 50% (20-60%). In addition to the hirsute lemmas, the b group of puta- tive hybrids is characterized consistently by filiform glumes of various lengths and a pronounced, overall glaucous ap- pearance, which is frequently a characteristic of E. Hystria, (2669 — fig. 6). Cytological irregularities are comparable in nature and frequency to those found in group a, with the exception of the appearance of 3-4 accessory chromo- somes in the early stages of pollen-mother-cel] meiosis. Three specimens collected as living plants (2251, 2653, 2662) had seed sets of 0, 20, and 3596. It is significant that root-divided clones of these plants, grown in greenhouse pots and surrounded by closely grouped pots of E. Hystrix and f. australis, increased in seed set over three years as much as 80%. Evidence, then, for the possible origin of new forms by introgressive hybridization is rather convinc- ing. 334 Rhodora [Vol. 69 The following study is an attempt to demonstrate the ancestry of the hybrids by experimental crosses between the putative parents. EXPERIMENTAL HYBRIDS 1. E. Hystrix crossed with E. virginicus var. glabriflorus f. glabriflorus Sympatric crosses. A cross of 2669 (E. Hystrix) with pistillate 2652 (f. glabriflorus) produced a single interme- diate F, plant (1:3:220),° (fig. 5). The F, (17:17)* segre- gated to give 15 fertile plants resembling f. glabriflorus and 2 intermediates resembling the F,. One of the inter- mediates was sterile and the other 30% fertile. The latter had culms 60 cm. tall compared to parental culms of 100 cm., spikes 6 cm. long compared to 10-12 cm., glume width 1 mm. compared to the extremes of lacking or 1.5 mm, in the parents. The ratio of lemma body to awn length was 1:4 as in E. Hystrix compared to that of 1:3 in f. glabri- florus. The essentially reciprocal cross (5:5:43) (fig. 6) re- sulted in an F, characterized by more setaceous (0.3 mm. wide) glumes, unequal in length and with indurated bases. These plants resembled rather closely natural hybrids such as 2670 (fig. 4) but were sterile. Other crosses with the same strains were failures (0:110). Allopatric crosses. A collection of 2368 (E. Hystrix) from Rice County, Minnesota, consisted of vigorous plants with frequent vestigial glumes on the spikes. Material grown from seed in the greenhouse, however, was quite free of glumes and was employed successfully in crosses with f. glabriflorus from Pine Hills (figs. 9 and 10). "Three numbers in ( ) indicate: — surviving seedlings: seeds planted: florets emasculated. Occasional numbers in [ ] indicate original numbers of seeds germinated as opposed to number of sur- viving plants. "Two numbers in ( ) indicate: — plants obtained: F, seeds planted, unless otherwise specified, as in failure of original cross. 1967] Elymus — Church 335 In the case of crosses employing 2607 (f. glabriflorus) , one produced a single, sterile F, plant (1:1:64) but the other (fig. 9) (1:2:42) gave rise to a partly fertile, inter- mediate plant (glumes 0.4 mm. wide) that strongly re- sembled many of the wild, putative hybrids from Pine Hills (2670, fig. 4). The F, (8:60) consisted of 2 partly fertile plants like f. glabriflorus and 6 intermediates with filiform glumes. One of the latter plants produced seeds (20:224) from which 4 F, individuals arose; these have remained vegetative, The number of progeny obtained (10:51:121) was even greater in another cross in which 2603 (f. glabriflorus) was used as the pistillate parent (fig. 10), this collection being substantially identical to 2607. Again, the F; strongly resembled wild forms (2670, fig. 4) except for a slightly higher frequency of stubs among the setaceous glumes of the experimental hybrids. Three of this F, were fertile, 4 were sterile and the rest failed after the seedling stage. The survivors were ca. 15% fertile compared to 55% for the parents growing under the same greenhouse conditions. Studies of the pollen-mother-cells revealed an average chiasma frequency of 1.91 compared with 1.98 for the parents. Although one or two bridges occurred in 40% of the first meiotic anaphases, less than 5% of the pollen showed extruded chromatin in the cytoplasm. The F, (18[24] :43:274) consisted of 3 sterile plants re- sembling f. glabriflorus, one partly fertile E. Hystrix, 6 seedlings that failed and 8 partly fertile intermediates that looked very much like the F;. The reciprocal of the preceding cross produced an F, (4[5] :20:162) that consisted of one sterile and 3 interme- diates that were 25% fertile. The F. (14:20) of 10 plants have remained vegetative for 3 years. The other survivors were ca. 12% fertile and resembled E. Hystrix with fili- form glumes. 2. E. Hystrix crossed with E. virginicus var. glabriflorus f. australis 336 Rhodora [Vol. 69 Sympatric crosses. In attempts to reproduce the natural hybrids with hirsute lemmas (2662, fig. 4), 2669 (E. Hys- trix) was crossed with three morphologically identical strains 2613, 2647 and 2648 (all f. australis) with no suc- cess (0:110, reciprocal 0:77). Allopatric crosses. The seeds of a cross of 2647 (f. australis) from Pine Hills with 2368 (E. Hystrix) from Minnesota failed to germinate (0:4:60). A similar failure resulted from a cross of the same 2647 with 2642 (E. Hys- trix) from Michigan (0:5:56). In the cross of pistillate 2669 (E. Hystrix from Pine Hills) with 2615 (f. australis from Tennessee) 3 plants were obtained (3:3:30). One of this F, remained vegeta- tive, but the other two produced spikes that were glaucous like the pistillate parent but had hirsute lemmas inherited from the staminate parent, The long, setaceous glumes of irregular lengths gave these plants an appearance closely resembling some of the natural hybrids at Pine Hills 2662, fig. 4). After having remained sterile for 4 years, one of the spikes had a seed set of 18% which was appar- ently the result of selfing in this F,. Intermediate F, plants were also obtained when the same strain of f. australis from Tennessee was crossed with strains of E. Hystrix outside of Illinois. Pistillate 2615 (f. australis) crossed with 2643 (E. Hystrix from Michi- gan) resulted in 8 vigorous, intermediate F, plants (8:12: 214), only one of which was fertile (2%). The cross of pistillate 2369 (E. Hystrix from Wisconsin) with 2615 (f. australis) produced even more vigorous F, plants (3:5:18). These intermediates were sterile during the first year. In the second year, 19 spikes of this same F, were bagged and 24 seeds were harvested from 2578 florets. Ten F, plants were obtained (10:24). Although they were considerably less vigorous than the F,, they exhibited a wide range of combinations of glume structure and lemma sur- face. Some plants resembled E. Hystrix in having no glumes but the lemmas were very hirsute. Several re- sembled the natural hybrids with hirsute lemmas from Pine 1967] Elymus — Church 337 Hills mentioned above. All F, plants were sterile except one that was bagged and yielded 2 seeds in 480 florets. 3. E. Hystrix crossed with E. virginicus var. virginicus Since the Pine Hills population 2600 (var. virginicus, fig. 3) with wide, indurated glumes and included spikes did not appear to play a direct role in the origin of the hybrid forms under study, only a few attempts at crossing with sympatric E. Hystrix were made and all of these were failures. Among a series of unsuccessful allopatric crosses, the hybrid obtained by pollinating 2614 (var. virginicus) from Tennessee with 2642 (E. Hystrix) from Michigan yielded one fertile intermediate plant (1:1:72). Six surviving F: plants exhibited a variety of glume structures from seta- ceous to moderately wide and striated. The setaceous glume forms had a seed set of 10% compared with ca. 100% for the parents. One F, plant closely resembled E. Hystrix with occasional filiform glumes and was quite fertile. 4. Natural Hybrid E. Hystrix X E. virginicus f. glabri- florus crossed with f. glabriflorus A representative of the natural hybrid population 2668 is shown in figures 2 and 8. It has a spike 20 cm. long, spreading spikelets and setaceous (0.2-0.3 mm. wide) glumes of unequal lengths. Sympatric crosses employing 2659 (f. glabriflorus) pollen were made, but the seedlings obtained all died at an early stage (0[9]:31:131). One cross (fig. 8), however, was particularly successful (12:16: 63). This F, was intermediate between the parents and ca. 50% fertile. In comparison the parents were 10% (2668) and 90% (2659). The F, (12:42) segregated to give 4 fer- tile plants resembling 2668, 3 partly fertile plants re- sembling f. glabriflorus and 5 intermediate forms closely resembling the F, (fig. 8). These F, intermediates had an average seed set of 60% (with a range of 0-90%). The reciprocal cross with the same 2668 and 2663 (f. 338 Rhodora [Vol. 69 glabriflorus, similar to 2659) produced 2 intermediate F, plants (2:3:54) (fig. 2). One of these intermediates pro- duced a single seed that gave rise to a partly fertile F, which resembled f. glabriflorus except for a glaucous ap- pearance like that of the staminate 2668 (natural hybrid). The data seem to indicate strongly, then, that 2668 is a hybrid and the experimental crosses with f. glabriflorus represent backcrosses with one of its parents. No allo- patric crosses were made. 5. Natural Hybrid E. Hystrix X E. virginicus f. austra- lis crossed with E. Hystrix Sympatrie crosses. A cross of pistillate 2657 (Natural Hybrid from Pine Hills) and 2669 (E. Hystrix) was a failure (0:50). However, the reciprocal cross gave a single, intermediate plant (1:3:59). Whereas the parental strain of the Natural Hybrid had setaceous glumes (0.2 mm. wide) with frequent striations, the F, plant had mostly filament- ous, non-striated glumes and was strikingly fertile (40 seeds from 104 florets). The surviving F, of 38 plants has flourished in greenhouse pots for 3 years and has increased in fertility (as measured by seed set on 25 spikes) from 22-39%. The glumes of this F, are practically all fila- mentous from an indurated base or occasionally reduced to stubs and very rarely marked by a fine striation. This experimentally produced hybrid population had an even greater resemblance to strains of E. Hystrix with variable glume structure than did the parental strain 2657 (Na- tural Hybrid). Allopatrie erosses. Another strain 2653 (Natural Hy- brid) was pollinated with 2643 (E. Hystrix) from Michigan with even greater success (10:28:67) than in sympatric crosses. The 2643 parent had quite glabrous lemmas and no glumes. The F, in this experimental cross had glumes more reduced and setaceous than those of the Natural Hybrid 2653. In the F, (38:90), 22 of the plants remained vege- tative and the rest were intermediate in appearance like the F,. Only 3 seeds arose in the F.. 1967] Elymus — Church 339 6. Natural Hybrid E. Hystrix X E. virginicus f. austra- lis crossed with f. glabriflorus Allopatric cross. In a cross of pistillate 2694 (f. glabri- florus from North Carolina) with 2651 (Natural Hybrid from Pine Hills), an F, of 2 plants was obtained (2:3:116). One of these had glumes approaching those of f. glabriflorus in width, hirsute lemmas like the Natural Hybrid and was sterile. The other plant had hirsute lemmas but setaceous glumes and a seed set that increased from 10% to 35% in the same potted individual after 4 years. The F, of 10 plants (10:18) exhibited considerable variation in glume structure but very few seeds. A reciprocal cross (2:9:45) had glumes intermediate between those of the parents but was quite sterile. 7. Experimental intervarietal hybrids of E. virginicus Reciprocal crosses of 2600 (var. virginicus from Pine Hills) and sympatric 2663 (f. glabriflorus) failed (0:11:60 and (0:8:58). Reciprocal crosses of 2600 (var. virginicus) and allopatric 2615 (f. australis from Tennessee) did not even produce any seeds (0:64 and 0:90). As far as this limited data are concerned, these forms of E. virginicus appear genetically isolated from var. virginicus. The question to arise next is the degree of isolation be- tween f. glabriflorus and f. australis. After several failures, a cross of 2609 (f. glabriflorus) pistillate from Pine Hills and sympatric 2613 (f. australis) resulted in a single plant (1[8]:8:34). This F, individual had a seed set of 17% compared with parental yields of 53% pistillate and 49% staminate. Two F, survived (2:17). They had hirsute lemmas and were fertile. In the reciprocal cross with 2613 (f. australis) as the pistillate parent, 4 F, plants were ob- tained (4[10] :17:131). In contrast to the overall hirsute appearance of the spike in f. australis, these F, individuals were hirtellous only on the lemmas and were 20% fertile. The F, (0:18) was a failure. 340 Rhodora [Vol. 69 In a cross of 2648 (f. australis from Pine Hills) with allo- patrie 2688 (f. glabriflorus from Virginia) as the pistillate parent a rather high percentage of seed was obtained (10:18:27). These F, plants were hirsute like f. australis but were stunted and sterile. 8. Crosses of E. canadensis L. var. canadensis with sympatric E. virginicus f. australis Since E. canadensis var. canadensis is known to cross fairly readily with strains of E. virginicus in other areas (Church 1958, 1967), it was thought appropriate to test the degree of genetic isolation between representatives of these taxa at Pine Hills. A cross of 2611 (var. canadensis, fig. 3) with staminate 2597 (f. australis) produced 8 seed- lings that died at an early stage of development (0:8:82). The reciprocal cross failed. Although these data are limited, genetic barriers between these taxa appear strong. CONCLUSIONS AND SUMMARY Examination of Chart A presents a summary of the re- sults of the experimental crosses of Elymus species as they relate primarily to the origin of the Pine Hills, Illinois hybrid populations of E. Hystrix crossed with sympatric E. virginicus f. glabriflorus or f. australis. Chart B indi- cates the genetic compatibility of these parents and their hybrids from selected areas of their range in eastern United States. Natural Hybrid E. Hystrix X E. virginicus f. glabriflorus The data presented indicate clearly that the origin of forms of E. virginicus with setaceous glumes and glabrous lemmas lies in natural hybridization between E. Hystrix and E. virginicus f. glabriflorus. In sympatric parental populations (Chart A), barriers to hybridity are quite strong as is evident from the high percentage of failures or sterility in the F,. Since, however, partly fertile inter- mediates resembling natural hybrids appear in the F. 1967] Elymus — Church 341 from even a single F, plant, the origin of natural hybrids is demonstrated. Although similar barriers to crossing are encountered when allopatric E, Hystrix is employed, (Chart B) partly fertile forms similar to the natural hybrids have appeared in both the F, and F,. Again the parents of the wild hybrids are identified. In a series of the allopatric crosses of E. Hystrix with f. glabriflorus, it is noteworthy that in over half of the cases sterility was not encountered until the F, was fully developed. Probably related to these results is the fact that out of 10 inter-strain allopatric crosses of f. glabriflorus, only 5 gave offspring with seed set over 609%. The implied existence of genes that make for intra-specific incompati- bility might also explain the lack of uniform results in all allopatric crosses of E. Hystrix with f. glabriflorus. Of particular significance is the fact that the natural hybrids cross readily with sympatric f. glabriflorus (Chart A), and the offspring are still intermediate in appearance but with a much higher average of wider (0.5 mm. rather than 0.2 mm.) glumes and fewer reduced stubs than appear in the natural hybrid parent. The seed set of 25% would seem to be sufficient to maintain such back-crossed or intro- gressed (E. Hystrix into f. glabriflorus) populations under field conditions and thereby establish the basis for the evo- lution of a new taxon through further back-crossing and attendant increased fertility. Natural Hybrid E. Hystrix X E. virginicus f. australis It has been more difficult to produce the type b (hirsute lemmas) natural hybrids than in the case of type a (glab- rous lemmas). Although the b type hybrids are encountered frequently at Pine Hills, barriers to crossing between the sympatric E. Hystrix and f. australis populations are very strong. However, experimental partly fertile hybrids, closely resembling the natural hybrids were obtained in the F, in the case where the f. australis parent came from Tennessee (Charts A and B). 342 Rhodora [Vol. 69 Furthermore, the fact that partly fertile hybrids similar to those at Pine Hills appeared in the F, of experimental crosses in which neither parent came from Illinois, certainly is indicative of significant genetic compatibility between these taxa in broad segments of their ranges, Finally, the fact that the experimental crosses of the natural hybrids to either sympatric or allopatric E. Hystrix resulted in F, plants resembling the parental natural hy- brids but with increased fertility, is strong evidence of the ability of the natural hybrids to become the basis of new taxa under field conditions as in the case of the type a hybrids considered above. The varieties of E. virginicus Since the major effort of the program of experimental hybridization has focused on E. Hystrix, the data pertain- ing to the varieties of E. virginicus per se are rather limited. However, there would appear to be definite genetic barriers between f. glabriflorus and f. australis in all cases reported (Charts A and B), although in one instance these were not reached until the F,. In spite of the fact that f. australis appears as merely a robust, hirsute variant of E. glabriflorus var. glabriflorus, studies so far show strong genetic isolation from not only f. glabriflorus (Chart A) but also from the common var. virginicus (Chart B). The evolutionary status of E. Hystrix In the large number of inter-generic hybrids, natural or artificial, that have been reported in the Hordeae tribe of the Gramineae, all are sterile (Church 1967). In contrast, the many experimental hybrids involving E. Hystrix in these studies maintain at least partial fertility in the F.. It seems unjustifiable, therefore, to maintain this species in a separate genus (Hystrix patula Moench). As previously noted, many of the segregates of experi- mental crosses of E. Hystrix with E. virginicus f. glabri- florus are intermediate in glume structure between the parents. It should be noted further that these hybrids vary in spikelet structure: the majority have spikelets of several 1967] Elymus — Church 343 florets more or less adherent to the rachis but others have spreading spikelets of few florets. These latter segregates resemble E. Hystrix except for the abundant filiform glumes. Such “glumed forms" of E. Hystrix are encoun- tered in many parts of the range of the species (Church 1967). Samples of such a population from North Carolina (2698) have been observed to breed essentially true from seed with varying percentages of seed set in the progeny of the two generations studied. In general, the spikes have an appearance suggesting introgression from f. glabriflorus into E. Hystrix rather than the reverse pattern of gene flow that is indicated in the Pine Hills natural hybrids. Both of the putative parents lie well within the range of populations like 2698 in both North and South Carolina, and it is possible that hybrids intermediate between Æ. Hystrix and E. virginicus var. glabriflorus have arisen in this eastern part of their range as they have in Pine Hills. It is conceivable, on the other hand, that populations of E. Hystrix with glumes represent a primitive condition of the species, in which case the glumeless condition might be the result of subsequently established mutations occur- ring in populations ecologically isolated from other Elymus species. Possibly the glumeless condition is basically ho- mozygous recessive, since in a cross of E. virginicus f. glabriflorus with staminate E. Hystrix, only one of 20 F. plants was identical to E. Hystrix with glumes entirely missing in addition to being 50% fertile. The strikingly different overall appearance between EF. Hystrix and E. virginicus var. glabriflorus f. glabriflorus (or f. australis) masks their strong genetic affinity. How- ever, a parallel case within the genus is presented by the facility with which strains of E. virginicus var. virginicus and the very different appearing E. canadensis var. cana- densis will cross in some instances to produce F, popula- tions, some of which are slightly fertile (Brown and Pratt 1960, Church 1958, Pohl 1959). In relation to E. Hystrix attention is directed also to E. californicus Gould, an endemic of the San Francisco 344 Rhodora [Vol. 69 area, which was formerly placed in the genus Hystrix (Gould 1947). The somatic chromosome number of 56 (Stebbins, G. L. in Myers, E. M. 1947) would indicate that this species is octoploid in contrast to the somatic comple- ment of 28 for E. Hystrix. The author has succeeded in raising a few plants of E. californicus in the greenhouse at Brown University but in two growing seasons they have not reached maturity and, therefore, crossing experiments between these two species remain for future study. BROWN UNIVERSITY BOTANICAL LABORATORIES DIVISION OF BOTANICAL AND MEDICAL SCIENCES PROVIDENCE, RHODE ISLAND LITERATURE CITED BROWN, W. V. and G. A. Pratt. 1960. Hybridization and intro- gression in the grass genus Elymus. Amer. Jour. Bot. 47(8): 669-676, CHURCH, G. L. 1958. Artificial hybrids of Elymus virginicus with E. canadensis, interruptus, riparius and wiegandii, Amer. Jour. Bot. 45(5): 410-417. 1967. Taxonomic and Genetic Relationships of Eastern North American Species of Elymus with Setaceous Glumes. Rhodora 69: 121-162. FERNALD, M. L. 1933. "Types of some American species of Elymus. Rhodora 35: 187-198. GOULD, F. W. 1947. Nomenclatorial Changes in Elymus with a Key to the Californian Species. Madrofio 9: 120-128. MOHLENBROCK, R. H. and J. W. Voigt. 1959. A Flora of Southern Ilinois. Southern Illinois University Press. Pour, R. W. 1959. Morphology and cytology of some hybrids be- tween Hlymus canadensis and E. virginicus. Proc. Iowa Acad. Science 66: 157-159. STEBBINS, G. L. in Myers W. M. 1947. Cytology and Genetics of Forage Grasses. Bot. Rev. 13: 319. VorcT. J. W. and R. H. MOoHLENBROCK. 1964. Plant Communities of Southern Illinois. Southern Illinois University Press. Plate 1357. ELYMUS vir- vir @ x vir-~glab-glabO vir-glab-aust (o) can O vill ® Hystrix à HYBRIDS AGA Pine Hills Union Co Ill Fig. l. Topographic diagram of the Pine Hills area showing dis- tribution of parental and hybrid species of Elymus: see text for explanation. Fig. 2. left to right — E. virginicus f. glabriflorus 2663; experi- mental F;; natural hybrid of E. Hystrix X E. virginicus f. glabri- florus 2668. Plate 1358. 2652 2670 2662 261 E. canadensis var. canadensis 2611; E. vir- Fig. 3. left to right virginicus f. glabriflorus 2652; E. ginicus var. virginicus 2600; E. villosus 2685. Fig. 4. left to right — E. virginicus f. glabriflorus 2652; natural hybrid E. Hystrix X E. virgincus f. glabriflorus 2670; natural hybrid E. Hystrix X E. virginicus f. australis 2662; E. virginicus f. au- stralis 2613. Plate 1359. 7 P v AP "es — s 2669 M 2656 Fig. 5. experimental F, of E. virginicus f. glabriflorus 2652 (pistillate) X E. Hystrix 2669. Fig. 6. left to right — E. Hystrix 2669 (pistillate — one spikelet removed at each node for clarity); experimental F,; E. virginicus f. glabriflorus 2656 (staminate). Plate 1360. ' bi Li r3 ism a 2668 2668 x 2659 Fig. 7. natural hybrid E. virginicus f. glabriflorus X; E. Hystrix 2595. BUNS M g~ VA SO Fig. 8. left to right — natural hybrid E. Hystrix X E. virginicus f. glabriflorus 2668 (pistillate — some spikelets removed); experi- mental hybrid (probable backcross); E. virginicus f. glabriflorus 2659 (staminate). Plate 1361. 441 E ` F i fl |- | $ E E / t i | B i / WX 7 Fi h WE „= 2603 ff 2603 x 2368 Ñ 2368 Mi Fig. 9. left to right — E. 'Hystrix 2368 (pistillate, some spikelets removed); experimental F, hybrid; E. virginicus f. glabriflorus 2607 (staminate). Fig. 10. left to right — E. virginicus f. glabriflorus 2603 (pistil- late); experimental F, hybrid; E. Hystrix 2368 (staminate — some spikelets removed). 350 Rhodora [Vol. 69 SYMPATRIC CROSSES VIR. GLAB. VIR. AUST. Cross fails —— ......... Fy Sterile F] Seeds fail ono. Segregate — Sterile —_— partly fertile Fertile CATUUCVVWAEE Fertile p F4 - Vegetative zee Chart A 1967] Elymus — Church 351 ALLOPATRIC CROSSES H x VIR. AU st| H x VIR. GLAB Hit. l LLLI E TOII ILITI IIT Tenn. Tenn. Tenn. y r am bl alishi AAR | | VIR. GLAB. VIR. AUST. Va. Mich In. iil Tenn. mp Tenn. N.C. Tenn. He lil. "e HYSTRIX li. : m P ái N (um EL Chart B HYSTRIX — Elymus Hystrix VIR VIR — Elymus virginicus var. virginicus VIR GLAB — Elymus virginicus var. glabriflorus f. glabriflorus VIR AUST — Elymus virginicus var. glabriflorus f. australis H X VIR GLAB — Natural Hybrid between Elymus Hystrix and E. virginicus f. glabriflorus H X VIR AUST — Natural Hybrid between Elymus Hystrix and E. virginicus f. australis STUDIES IN THE EUPATORIEAE, (COMPOSITAE) IV R. M. KiNG HOFMEISTERIA Walpers in Walp. Repert. 6: 106. 1846. (Type species Helogyne fasciculata Benth.) Podophania Baill. Bull. Soc. Lin. Paris. 1: 268. 1880. (Type species Phania ? dissecta Hook. & Arn.) Woody sub-shrubs or perennial herbs. Leaves in 2/5 phyllotaxy (usually congested at flowering nodes, appearing whorled). Inflorescence monocephalic. Heads broadly campanulate, (100) 150-175 (250) flowered. Phyllaries numerous, multiseriate, densely imbricated, narrowly lance- olate. Receptacle convex, naked. Florets discoid, perfect, 5 lobed, white, pink, or lavender, glabrous, narrowly tubu- lar. Stamens 5, appendaged. Pollen tricolpate, surface nearly smooth. Pappus in a single series, usually consisting of setae and intervening squamae. Achenes narrowly pris- matic. 5 ribbed. ribs setose, sub-cuticular cells of ribs and lateral walls with many minute punctations, carpopodium somewhat flaring, with many rows of short or transversely elongate cells; basal foramen prominent with an even mar- gin. Chromosome number determined from one species as X = 18 (Kyhos, unpublished), X — 19 (Turner & Flyr 1966). The species of Hofmeisteria may be distinguished by the following key: 1. Anther appendages appressed, not visible at 10 X magnification Geen secnsseeavarrersaas ¢isoversonssanscnsssanasesusanseiiwbadeseastaaesesenesbanesatsswatssnecs4sisehensieistrie# 2 1. Anther appendages erect, visible at low magnification ............ 3 2. Pappus of 12-15 setose bristles of rather uniform length ........ NM 1. H. dissecta 2. Pappus of 5-6 long setae with intervening short, very lacerate 352 1967] Hofmeisteria — King 353 EEN SET e HIC HR UE 2. H. sinaloensis 3. Pappus consisting of 5-6 long setae with essentially no intervening memberssS a e aeaa araara eseina aeaa niea AEE ee ETS SEE ERPS ESERE 4 3. Pappus consisting of 3-5 long setae with distinct intervening groups of short setae or squamae ........ eee e 6 4. Leaves only lobed and dentate, not dissected .... 3. H. standleyi A, Leaves dissected .............csccccccccssssrscnrsserscccesssssessensnseneesseesossesssonees 5 5. Segments narrow with few teeth restricted distally ............ eRe MN NE EE R 4. H. schaffneri 5. Segments broad with many teeth ................ 5. H. urenifolia 6. Leaf lamina broad .................... eee 6. H. fasciculata 6. Leaf lamina narrow, filiform ..........:.. eee eem 7 7. Peduncles highly differentiated, arising from a cluster of DAVOS oes is ssa saen ladescvcarpoecevsodevonsaswanssamseces 7. H. crassifolia 7. Peduncles not highly differentiated, not arising from a cluster Ud l6. RII ARES Uc 8. H. filifolia 1. Hofmeisteria dissecta (Hook. & Arn.) R. M. King and H. Robinson. Phytologia 12: 466. 1966. Phania? dissecta Hook. & Arn. Bot. Beech. 433. 1841 Acapulco, Guerrero, Mexico. D. Sinclair sn (Holotype K!) Eupatorium dissectum (Hook. & Arn.) Benth. Bot. Sulph. 113. 1844. Not Eupatorium dissectum A. Gray, Proc. Am. Acad. 18: 100. 1883. Podophania Ghiesbreghtiana Baill. Bull. Soc. Lin. Paris 1: 268. 1880. Mexico, without precise locality, Ghiesbreght 315 in Oct. 1844, (Holotype P! isotype fragment F!) Podophania dissecta (Hook. & Arn.) B. L. Robinson Proc. Amer. Acad. 47: 192. 1911. Herb to 1 m. tall; dichotomously branched; leaves oppo- site, tripinnately dissected, blades triangular in outline, up to 7 cm long, 6-7 cm wide; peduncles up to 10 cm long; heads 6-7 mm high, 5-6 mm wide, 100-150 flowered; phyl- laries unequal, striate, ca 50; corolla tube white, ca 3.5 mm long; the anthers (including appendanges) ca .75 mm long, ca 100 » wide; appendages very small, cucullate; filaments ca 2 mm long; style branches greatly exserted at maturity ca. 2.5 mm long; style tube ca 2.5 mm long; pappus usually of 12-15 scabrate setae equaling the length of the corolla; achenes dark brown, ca 2 mm long; pollen ca 11 » in di- ameter; chromosome number not determined, [Vol. 69 Rhodora 354 Plate 1362. ie Plate 1362. Hofmeisteria dissecta, Holotype (K), Sinclair s.n. 1967] Hofmeisteria — King 355 SPECIMENS EXAMINED: MEXICO: CoLIMA: n. edge of Colima City, Langman 3212 (MEXU, PH); Manzanillo, Palmer 1371 in 1891 (GH 2 sheets, NY, US) ; GUERRERO: Pungarabato, Coyuca, Hinton 5453 (GH, K, NY, US) ; 20 miles s of Chilpancingo, Cronquist 9703 (MEXU, MICH, NY, TEX, US); Acapulco, ‘Howell 8506 (CAS); 30 miles n e of Acapulco, R. M. King 4176 (F, MICH, NY, TEX, UC, US); Temisco, Barranca del Consuelo, Mexia 8812 (CAS, F, G, NY, UC, us); Rincón de la Via, Kruse 73 (MEXU); El Calabaral, Langlassé 470 (G, GH, K, MICH, US); JALISCO: 8 miles s e of Pihuamo, McVaugh & Koelz 1511 (MICH); 9 miles w of Bahia de Navidad, McVaugh & Koelz 1723 (MICH) ; 2.5-4 miles n of La Cuesta, McVaugh 21218 (MICH) ; MEXICO: Pungarancho, Temascaltepec, Hinton 3127 (GH, K, US); MICHOACAN: Chila, 8 km n w of Aquila, Rzedowski 17943 (MiCH); La Salada, Nelson 6928 (GH, NY, US). Trojes, Coalcomán, Hinton 12292 (K, MICH, NY, P, UC, US) ; SINALOA: San Ignacio, Ortega 4972 (GH, MEXU, US). 2. Hofmeisteria sinaloensis H. S. Gentry, Brittonia 6: 329. 1948. Capadero, Sierra Tacuichamona, Sinaloa, Mexico. Feb. 18, 1940. H. S. Gentry 5615 (Holotype MICH ! isotypes DS! GH! NY! US!). Three to four dm tall, forming irregular spreading clumps; few branched, older wood glabrous, the younger pubescent; leaves alternate, pinnately dissected in 3 or rarely 5 lobed, thin leaflets; peduncles slender, 7-9 cm long, sparingly pubescent; heads 7-8 mm high, 6-7 mm wide ca 150 flowered. Phyllaries unequal, 5 nerved ca 60. corolla tube white, 3.5-4 mm long, the anthers ca 1.25 mm long (including appendages) ca 110, wide, appendages very small, cucullate; filaments 1.5-2 mm long; style branches greatly exserted at maturity, ca 3 mm long; style tube ca 3 mm long; pappus usually of 5-6 long scabrate setae equaling the length of the corolla, and 6-10 short inter- vening lacerate squamae usually less than .5 mm long; achenes ca 2 mm long, dark brown or black; pollen ca 16 p in diameter; chromosome number not determined. SPECIMENS EXAMINED: MEXICO: SINALOA: Capadero, Sierra Tachuichamona, Gentry 5581 (DS, F, GH, MICH, NY, US); 12 miles e of Esquinapa, along highway 15, Gentry, Barclay & Arguelles 19466 (US). 356 Rhodora [Vol. 69 Plate 1363. * Pd Sp T ` " * ¥ à * Type toiltc fien Hoffreisteria Sedge mbar Fee tres al Spe nov. Caindaro, Jtorrn Taouielusuxs ^ 194C Pob. sł Convo: olif^e in shado; Jnori- iolorast cuffruteseant i olupe on sückin brit lo! Plate 1363. Hofmeisteria sinaloensis, Holotype (MICH), Gentry 5615. - Porost, 1967] Hofmeisteria — King 357 3. Hofmeisteria standleyi (Blake) R. M, King and H. Rob- inson, Phytologia 12: 467. 1966. Fleischmannia standleyi Blake Contr. U.S. Nat. Herb. 22: 590. 1924. Mexico: Sonora, Sierra de los Alamos, March 19, 1910. J. N. Rose, P. C. Standley & P. G. Russell 13073 (Holotype U:! isotype NY!) Fleischmannia urenifolia (Hook. & Arn.) Benth. and Hook. var. Mexiae, B. L. Robinson, Contr. Gray Herb. 96: 18. 1931. Mexico: Jalisco, Hacienda del Ototal, east of San Sebastian, alt. 1500 meters, “herb growing in crevices of rock near water, steep wooded slope of narrow canon.” February 15, 1927. Ynes Mexia 1684a (Holotype GH! iso- types CAS! F! G! uc! US!). Perennial herb ascending from a procumbent stem 25-50 cm tall; few branched, rather densely pilose; leaves alter- nate; blades ovate, 2-4 cm long, 1.7-2.7 cm wide, acute or obtuse, unequally 4-6 toothed on each side or sometimes 3 lobed; peduncles up to 10.5 cm. long; heads 8-9 mm high, 8-11 mm wide, ca. 150 flowered; phyllaries unequal, 3 nerved, ca. 60; corolla tube white or tinged with purple 4.5-5 mm long; the anthers (including appendages) ca. 1.75 mm long ca 125 » wide, appendages obtuse, ca 220 , long; fila- ments ca. 2.5 mm long; style branches greatly exserted at maturity 1.5-2 mm long; style tube 3-3.5 mm long; pappus consisting of 5-7 scabrate setae ca 4.5 mm long and short intervening lacerate-margined squamae; achenes dark brown or black, 1.75-2.0 mm long; pollen ca 18 , in dia- meter ; chromosome number not determined. SPECIMENS EXAMINED: MEXICO: Jarısco: Hacienda del Ototal. e of San Sebastian, Mexia 1684 (CAS, F, MICH, NY) ; NAYARIT: 10 miles s e of Ahucatlan, McVaugh 15179 (MICH) ; SINALOA: Africa, Sierra Tacuichamona Gentry 5690 (DS, GH, MICH); San Ignacio, Ore- tega 288 (MEXU); Cerro de la Silla, San Ignacio, Montes & Salazar 288 (US). 4. Hofmeisteria schaffneri (A. Gray) R. M. King and H. Robinson, Phytologia 12: 467. 1966. Fleischmannia Schaffneri A. Gray, Proc. Am. Acad, 41: 101, 1881. San Luis Potosi “In montibus umbrosis humidis 358 Rhodora [Vol. 69 Plate 1364. he ni Fog Aha tha c id a "SW Plate 1364. Hofmeisteria standleyi, Holotype (vs), Rose, Stand- ley and Russell 10072. 1967] Plate 1365. Hofmeisteria schaffneri, Holytype (GH), Schaffner 349. Hofmeisteria — King Plate 1365. 359 Misco sites unosi aoi Den E HERBARIO BE m J. 6, SCHAFFNER Cre cst Neu Jy t x $ FLORA MEXICAN i 1 i^i rcnt Rome soa acc dote cie A Rem Vo a i x i yp m xt ste 360 Rhodora [Vol. 69 prope San Francisco." August 1876, J. G. Schaffner 349 (Holotype GH! isotype NY!). Ca 25-35 cm tall; stems few branched glandular pubes- cent; leaves chiefly alternate, the blades triangular in out- line ca 1 cm long and wide, dissected, lobes oblanceolate; peduncles up to 10 cm long; heads 9-10 mm high, 8-11 wide, ca 150 flowered; phyllaries unequal violaceous; striate, ca 50. corolla tube white or pink, 4.5-5 mm long; the anthers ca 1.5 mm long (including appendages) ca 125 , wide, appendages obtuse ca 125 , long; filaments 1.5-2 mm long; style branches greatly exserted at maturity 3-3.5 mm long; style tube ca 3 mm long; pappus consisting of 4-7 long maroon scabrate setae 4-4.5 mm long with a very short crown of intervening lacerate-margined squamae; achene, ca 1.5 mm long, dark brown; pollen 17-18 „ in diameter; chromosome number not determined. SPECIMENS EXAMINED: MEXICO: Jatisco: Real Alto, trail to Arroyo del Jaguay, Mexia 1752 (CAS, DS, F, G, GH, MICH, NY, UC, US). 7-8 miles n w of Los Volcanes, McVaugh 12205 (MEXU, MICH, US). Chiquilistlan, Jones 259 (Us). SAN Luis Potosi, M. Urbinox sn (MEXU). 5. Hofmeisteria urenifolia (Hook. & Arn.) Walp. in Walp. Rep. 6: 106. 1847. Phania? urenifolia Hook. & Arn. Bot. Beech. Voy. 297. 1840. Beechey, Mexico without precise locality (Holotype K!). Helogyne urenifolia Walp. in Walp. Rep. 4: 457. 1848, by typographical error. Phania urenaefolia Benth. & Hook, f. Gen. Pl. 2: 243. 1873. Fleischmannia urenaefolia Benth. & Hook. f. ex Hemsl. Biol. Cent. Am. Bot. 2: 91. 1881. Fleischmannia Langlassei B. L. Robinson, Proc. Am. Acad. 41: 273. 1905. Mexico: Michoacán, Arroyo de Barra- bas, E. Langlassé 27 (Holotype GH! isotypes G!, P!, US!) Fleischmannia urenifolia (Hook. & Arn.) Benth. & Hook. f. var. typica B. L. Robinson, Contr. Gray Herb. 96: 17. 1931. 1961] Hofmeisteria — King 361 Plate 1566. Plate 1366. Hofmeisteria urenifolia, Holotype (Kx), Beechey s.m. 362 Rhodora [Vol. 69 Perennial herb 25-30 cm tall; stems few branched, striate, glabrous; leaves alternate, the blades trifid or pinnately tri- sect, up to 4.5 cm long and wide; peduncles up to 5 cm long; heads 9-10 mm high, 8-11 mm wide, 150-175-flowered; phyllaries unequal, usually 3-nerved, ca 50; corolla tube white or pink, 4.5-5 mm long; the anthers (including ap- pendages) ca 1.25 mm long ca 125 » wide; appendages ob- tuse ca 125 y long; filaments 1.5-2 mm long; style branches greatly exserted at maturity, ca 3 mm long; style tube 2-2.5 mm long; pappus consisting of 5-7 maroon colored, scabrate setae 4-4.5 mm long and a low crown of very short lacerate- margined squamae; achenes dark brown, 1-1.2 mm long; pollen 15-16 » in diameter; chromosome number not deter- mined. ADDITIONAL SPECIMENS EXAMINED: MEXICO: CHIAPAS: Monserrate, Purpus 218 (GH, US); 10213 (NY); 10218 (UC). Du- RANGO: 3-15 km toward El Salto, MeVaugh 23609 (MICH) ; 50-54 road road miles w s w of El Salto, McVaugh 11564 (MEXU, MICH, US). GUERRERO: Tierras Blancas, Hinton 10143 (K, MICH); Chilacoyte, Hinton 14178 (F, NY, UC, US, W). JALISCO: Hacienda del Ototal, e of San Sebastian, Meria 1684 (CAS, DS, GH, UC); Barranca of Beltram, Pringle 5492 (GH); Saltillo, M. E. Jones June 1, 1892 (Pom). NAYARIT: 3 km n w of El Ocotillo, MeVaugh 23529 (MICH). SINALOA: 0.9 mi w of El Palmita, Breedlove 1719 (MICH). 6. Hofmeisteria fasciculata (Benth.) Walp. in Walp. Repert. 6:106. 1846. Helogyne fasciculata Benth. Bot. Voy. Sulph. 20: 14. 1844. Mexico: Baja California, Bay of Magdalena Hinds in 1841 (Holotype K!) Hofmeisteria fasciculata (Benth.) Walp. var. Xanti A. Gray, Brewer & Watson Bot. Calif. 1:299, 1876. Mexico: Baja California: Cape San Lucas. August 1859 — January 1860, L. J. Xantus 46 (Holotype GH!; isotypes K! US! P!) Hofmeisteria pubescens S. Watson, Proc. Amer. Acad. 24:54. 1889. Mexico: Baja California: Muleje Dr. Edward Palmer 422 on December 25, 1887. (Holotype GH!) Hof moeisteria fasciculata (Benth.) Walp. var. Grayi T. S. Brandeg. Zoe 5: 160. 1903. Mexico: Baja California: þe- 1967] Hofmeisteria — King 363 tween San Jose del Cabo and Cabo San Lucas, Purpus 208 (Holotype UC!; isotype US!) Hofmeisteria fasciculata (Benth.) Walp. var pubescens (Wats.) B. L. Robinson, Proc. Amer. Acad. 47: 192. 1911. Forming clumps to 6-7 dm. tall; usually few-branched, glabrous to densely pubescent; leaves opposite below, alter- nate above, rather fleshy, lobed to bipinnatisect, glabrous to densely pubescent, blades 1.5-4 cm long and wide, peduncles up to 20 em long, bearing scattered setaceous bracts. Heads 8-11 mm high, ca 1 cm wide, ca 150-flowered. Phyllaries unequal, striate, ca 100. Corolla tube white, pink or laven- der, 4-4.5 mm long, the anthers ca 1.75 mm long (including appendage) 120-130 „ wide, appendages obtuse ca 250 y long; filaments 1.5 - 2 mm long style branches greatly ex- serted at maturity ca 3-3.5 mm long; style tube 2.5-3 mm long; pappus of 2-3 long scabrate setae 4.5-5 mm long and 3-4 short squammae with lacerate margins .5-1.0 mm long; achenes dark brown or dull black 2-2.5 mm long ; pollen ca 18 &, in diameter; chromosome number determined as n = 18 (Kyhos unpublished) 19 (Turner & Flyr 1966). SPECIMENS EXAMINED: MEXICO: BAJA CALIFORNIA: Catalina Island, J. M. Johnston 4102 (CAS); Muertos Bay, Moran 3562 (Ds, us); Arroyo del Salto, Moran 7143 (CAS, DS, GH, TEX); 29 mi n of San Luis Gonzaga, Moran 8210 (DS, RSA); Puerto Escondido, Dawson 1103 (DS, F); ca 23 kms s of Loreto, Carter, Alexander & Kellogg 1998 (Ds, UC, US); 8 kms n of San Juanico, Gentry 4313 (DS); Malau- cita to La Cruz, Nelson & Goldman 7304 (US); between Rancho Santa Isabel and Santa Rosalillita, Carter & Ferris 3419 (DS, GH, MEXU, TEX, UC, US); San Francisquito Bay, J. M. Johnston 3574 (CAS); San Francisco Island, Moran 3729 (ps, UC); Collins, Kearney, & Kempton 192 (US); ca 18 kms of Loreto, Carter & Ferris 3874 (uc). San Lucas, Johansen 535 (CAS, DS); Arroyo Salado nr San Jose del Cabo, Purpus 408 (GH); San Jose del Cabo, Purpus 262 (K, US); Barrill, Harbison 41604 (RSA); Punta Frailes, Dawson 1143 (MICH); EI Coyote, Porter 113 (CAS, DS, G, MEXU, TEX, UC); south side of Punta Gasparino, Constance 3189 (DS, F, GH, MICH, UC, US); Aqua Verde, Rose 16577 (GH, NY, US); San Benito, T. S. Brandegee April 10, 1889 (UC); 16 mi s cf Puertocitos, J. R. Hastings & R. M. Turner 63-139 (Ds); % mi s of Puertocitos, Wiggins & Wiggins 15864 (DS, MICH, TEX); Santa Maria Bay, Rose 16265 (NY, US); Moran 3536 (DS); Espiritu Santo, Collins, Kearney & Kempton 140 (US); Rose 16867 Plate 1367. ws Y / V Plate 1367. Hofmeisteria fasciculata, Holotype (xK), Hinds 1841. in 1887. 1967] Hofmeisteria — King 365 (US); one mi n of Coyote Cove Bahia de la Concepcion, Wiggins 17411 (DS); Punta Santa Domingo, Wiggins & Wiggins 18270 (CAS, DS, MEXU, MICH); Bahia de la Concepcion, Wiggins & Wiggins 19000 (DS, MICH, RSA); Concepcion Bay, Rose 16680 (NY, US); Tortuga, I. M. Johnston 3600 (CAS) ; opposite Pond Island, Angel de la Guarda Island, I. M. Johnston 4233 (CAS, DS, GH, NY, UC, US); Las Animas Bay, I. M. Johnston 3493 (CAS, DS, GH, NY, UC, US); San Nicholas Bay, I. M. Johnston 3714 (CAS, DS, F, GH, NY, UC, US); Puerto Bal- landra, Carmen Island, I. M. Johnston 3814 (CAS, GH, UC, US); Bahia de los Angeles, Carter, Alexander & Kellogg 2540 (DS, UC, US); be- tween Bahia de los Angeles and San Borja, Cowan 2322 (TEX, US); 3.5 mi w of Bahia de los Angeles, Wiggins & Wiggins 16018 (Ds, MICH, TEX); 1.5 mi w of Bahia de los Angeles, Wiggins & Wiggins 14,792 (CAS, DS); Los Angeles Bay, I. M. Johnston 3456 (CAS, US); 4 mi s of Todos Santos, Thomas 7788 (US); 11 mi s of Todos Santos, Whitehead 874 (DS); 17 mi s of Todos Santos, Shreve 7227 (F, MICH) ; Mulege, Wiggins & Wiggins 18223 (CAS, DS, MEXU, MICH, TEX); La Paz, Wiggins 14569 (CAS, DS, GH); M. E. Jones 24080 (DS, F, GH, K, MICH, NY, PH, POM, RSA, SMU, TEX, UC); T. S. Brandegee 275 (F, NY, PH); Edward Palmer 137 in 1890 (GH, MEXU, NY, UC, US); Santa Cruz Island, Moran 3843 (ps, UC); Cerralvo Island, (DS, UC); Mono- serrate Island, Moran 3888 (DS, UC); Santa Catalina Island, Moran 3871 (DS, UC); Isla San Jose, Wiggins 17679 (US); Wiggins, Carter & Ernst 365 (ps, UC); Moran 3788 (DS, UC); Pichilinque Island, Rose 16526 (US); Ildefonso Island, Moran 4136 (Ds, UC); I. M. John- ston 3744 (CAS, DS, F, GH, NY, UC, US); San Marcos Island, Moran 3995 (DS, UC); Ferris 8654 (DS, MICH, NY, POM, RSA, US); Coronados Island, J. M. Johnston 3765 (CAS, GH, US); Carrisalitos, Magdalena Plain, Gentry 4447 (DS, GH); Magdalena Island, Moran 10808 (Ds, MEXU, MICH, RSA, UC); T. S. Brandegee January 1889 (Ds, UC); Orcutt 12 (US); Magdalena Bay, Mason 1897 (CAS, GH, US); T. Crocker August 9, 1932 (CAS, DS, GH). Isla Danzante, Wiggins 17553 (us); Carter & Sharsmith 4262 (UC); Wiggins 17550 (ps); Carmen Island, Dr. Edward Palmer 875 in 1890 (GH, NY, US); Rose 16618 (NY, US); Carter & Ferris 3717 (uc); Isla Partida, I. M. Johnston 322 (CAS, DS, GH, NY, UC, US); near Santa Rosalia, Shreve 7062 (F, GH, MICH); Santa Agueda, 5 m w of Santa Rosalia, Shreve 6539a (MICH); 10 mi n of Santa Rosalia, F. M. Reed 6265 (POM); 5 mi w of Santa Rosalia, Ferris 8716 (DS, US); Santa Rosalia, Edward Palmer 178 in 1889 (CAS, F, GH, NY, UC, US). SoNORA: Puerto de Lobos, Pringle 5960 (GH); Tepoca Bay, I. M. Johnston 3303 (CAS, GH, US). 7. Hofmeisteria crassifolia S, Wats. Proc. Am. Acad., 24: 53. 1889. “In crevices of rocks in high mountains about 966 Rhodora [Vol. 69 Guaymas, Sonora, Mexico" Edward Palmer 309 in 1887 (Holotype GH! isotypes DS! NY! US!) Suffrutescent, brittle plants 6-25 cm tall. Stems woody, glabrous, many-branched. Leaves alternate, linear, once or twice pinnatifid, triangular in outline, glabrous, succulent. Blade 1-1.5 cm long. Petioles 1.3-5 cm long, glabrous. Peduncles solitary, erect, glabrous with scattered setaceous bracts, 1-9 mm long, 5-18 cm tall. Heads broadly campanu- late to subhemispherical, 8 mm to 1 em high, ca 250 flow- ered. Phyllaries glabrous, ca 100, 1-3 nerved, imbricated, linear-lanceolate, apex attenuate to acuminate. Florets pinkish, tubular, 5-lobed 4-4.5 mm long. Stamens 5. An- thers (including appendages) ca 1.3 mm long ca 165 » wide, appendage ca 240 , long ca 145 u wide. Filaments ca 3 mm long ca 45 » wide. Style-branches 2, greatly exserted at maturity, ca 2 mm long ca 120 » wide. Pappus of 5 full length bristles nearly as long as the corolla often with 2 or 3 shorter supplementary bristles. Squamellae unequal, 10- 16, 1-4 between each bristle, 1-2.5 mm long. Achenes 4-5 angled, narrowly obprismatic, dull black, 1.5-2 mm long. Pollen tricolpate, spherical, nearly smooth, ca 13 y in dia- meter. Chromosome number not determined. SPECIMENS EXAMINED: MEXICO: Sonora: Guaymas, L. H. Bailey 265 (F); Palmer 165 in 1890 (CAS, GH, Ny, US); T. S. Brande- gee May 12, 1892 (r, PH); T. S. Brandegee sn in 1893 (Nv); Dr. Hisen April 1892 (Ds); San Pedro Nolasco Island, I. M. Johnston 3142 (CAS, GH, K, NY, US) ; San Pedro Bay, I. M. Johnston 4307 (Cas, US). 8. Hofmeisteria filifolia I. M. Johnston Proc. Cal. Acad. Sci. 12: 1185. 1924. "partially shaded rock crevices in Palm Canon, Angel de la Guarda Island," Baja California, Mexico. I. M. Johnston 3418 (Holotype CAS! isotypes DS! GH! NY! US!) Herbaceous perennial with numerous stems forming dense rounded clumps 1-6 dm tall; stems strictly branched above, striate, densely covered with glandular hairs; leaves alternate or sometimes opposite basally, lower leaves pal- mately five-parted into filiform lobes, these simple or three- 1967] Hofmeisteria — King 367 Plate 1368. Plate 1368. Hofmeisteria crassifolia, Holotype (GH), Palmer 309 368 Rhodora [Vol. 69 Leaves of Hofmeisteria. Figures 1-2, H. schaffneri. 1. McVaugh 12205 (MICH). 2. McVaugh 12205 (MICH). Figures 3-8, H. ureni- folia. 3. McVaugh 23609 (MiCH). 4. Hinton 14272 (F). 5. Langlassé 27 (GH). 6. Mexia 1684 (GH). T. McVaugh 11564 (MICH). 8. Mc- Vaugh 23529 (MICH). Figures 9-10, H. standleyi. 9. McVaugh 15179 (MICH). 10. Mexia 1684 (MICH). parted, 4-7 cm long, 2-6 cm wide; lobes 0.5-1 mm wide, 2-20 mm long; peduncles slender, 1.5-9 cm long, with scattered setaceous bracts 1-3 mm long; heads ca 8 mm high 4-5 mm wide, ca 160 flowered; phyllaries unequal, 3-nerved ca 50. corolla tube pale pink or white 3.5 - 4 mm long ; the anthers ca 1.75 mm long (including appendages) ca 165 , wide, appendages short, blunt ca 100 u long; filaments ca 1.5 mm long; style-branches slightly exserted at maturity 2-2.5 mm long; style-tube ca 2.5 mm long; pappus usually of 3 long scabrate setae 2.5 - 3 mm long and 8-10 short erosely dentate squamae 0.25 - 0.5 mm long; achenes ca 3 mm long, dark brown or black; pollen ca 18 „ in diameter; chromosome number not determined. SPECIMENS EXAMINED: MEXICO: Basa CALIFORNIA: Angel de la Guarda Island, Moran 10391 (CAs, Ds, MICH, RSA, UC, US); 369 Hofmeisteria — King 1967] Figure 11 A & a 3 e LJ ux Fig. 11. Distribution of Hofmeisteria. 370 Rhodora [Vol. 69 nafy # Plate 1369. Hofmeisteria filifolia, Holotype (cas), Johnston 3418. 1967] Hofmeisteria — King SB I. M. Johnston 3377 (CAS, GH, K, US) ; I. M. Johnston 3364 (CAS, US); Moran 10439 (DS, UC); Moran 7213 (CAS, DS, MEXU, MICH, NY, RSA, US, W). DEPARTMENT OF BOTANY SMITHSONIAN INSTITUTION WASHINGTON, D. C. 20560 LITERATURE CITED TURNER, B. L., and D. FLYR. 1966. Chromosome numbers in the Compositae. X. North American species. Amer. J. Bot. 53 (1) 24-33. FURTHER CHANGES AND ADDITIONS TO THE FLORA OF PORTO RICO AND THE VIRGIN ISLANDS BROTHER ALAIN H. LIOGIER After the publication of the nomenclatural changes and additions to the Flora of Porto Rico (cf. Rhodora 67: 315- 361. 1965), several more changes have come to my atten- tion, and by revising the literature and going through the herbaria, I have found more additions to the Flora of the area covered by Britton and Wilson. As in my previous paper, I shall give the page and after the valid name, 1 shall cite in parenthesis the name used by Britton and Wil- son in the Flora. This list is to be used as a supplement to the Flora, and reference must be made to the previous cita- tion by Britton and Wilson. Several errors have been made in my previous paper, and they are corrected here. Volume I of the Flora (Scientific Survey, vol. V). p. 18. Limnobium spongia Rich. To be added: Afiasco, Sintenis 5779. p. 32. Trichachne affinis Swallen, Rhodora 65: 365. 1963. To be added: Hess 428. p. 34. Digitaria diversifolia Swallen, Rhodora 65: 366. 1963. To be added: Juana Díaz. p. 51. Lasiacis maculata (Aubl) Urb. (Lasiacis sorghoidea H. & C.). p. 53. Echinochloa spectabilis (Nees) Link (Echinochloa poly- stachya Hitche.). v. 67. Chloris barbata Sw. (Chloris inflata Link; Chloris para- guayensis Steud.). p. 75. Eragrostis tenella (L.) Beauv. ex R. & S. (Eragrostis amabilis Wr. & Arn.). p. 77. Rhynchelytrum repens (Willd.) C. E. Hubb. (Tricholaena repens Nees). p. 108. Scleria melaleuca C. & S. (Scleria pterota C. Presl ex C. B. Clarke). p. 113. Calyptrogyne rivalis (Cook) León replaces Calyptrogyne occidentalis Maza (Calyptronoma rivalis Bailey). p. 117. Coecothrinax alta Bece. replaces Coccothrinax argentea of Britton & Wilson. p. 139. Tillandsia setacea L. (cited by Britton and Wilson as T. tenuifolia L.). 372 * Flora of Porto Rico — Liogier 373 Tillandsia tenuifolia L. To be added: Maricao, collected by Winters. This is the species that has been cited as Tillandsia pulchella Hook. (cf. Phytologia 8: 20. 1962). Commelina virginica L. (Commelina elegans HBK.). Curculigo capitulata (Lour.) O. Kuntze (Molineria hortensis Britt.). Atriplex littoralis (Jacq.) Fawc. & Rendle (Atriplex pentan- dra Standl.). Nymphaea jamesoniana Planch. (Castalia jamesoniana Britt. & Wils.). Lepidium pinnatisectum (O. E. Schulz) C. L. Hitche. To be added: Mayagiiez. Brassica hirta Moench. (Sinapis alba L.). Rorippa islandica var. hispida (Desv.) Butt. & Abbe (Radi- cula palustris Moench; Rorippa palustris Bess.). Nasturtium officinale R. Br. (Sisymbrium nasturtium-aquati- cum L.; Rorippa nasturtium-aquaticum Hayek). Crataeva tapia L. include as a synonym Crataeva gynandra L. Rubus coronarius Sweet. To be added: Cayey. Acacia macracantha H. & B. ex Willd. in L.; include as a synonym: Poponax macracanthoides Britt. & Rose, Sci. Surv. 6: 538. Acacia polyacantha Willd. in L. (Acacia suma Buch.-Harm ex Voigt). Acacia scleroxyla Tuss. To be added: Padre de las Casas (Holridge 78). Caesalpinia bonduc (L.) Roxb. (Caesalpinia crista or modern authors). Caesalpinia major (Medic.) Dandy & Exell (Caesalpinia bonduc of modern authors). Crotalaria mucronata Desv. (Crotalaria striata DC., not C. saltiana Andr.). Crotalaria lotifolia var. eggersii Senn. To be added: Vieques, St. Thomas. Crotalaria sagittalis var. fruticosa (Mill. Fawe. & Rendle. To be added: Bayamón. Crotalaria berteriana DC. Cultivated at Mayagüez Ex- perimental Station. Dalea carthaginensis (Jacq.) Macbr. (Parosela domingensis Millsp.; Dalea domingensis DC.). Tephrosia sessiflora (Poir.) Urb. (Cracca brevipes Kuntze). Tephrosia purpurea (L.) Pers. (Cracca purpurea L.). Lonchocarpus sericeus var. glabrescens Benth. (Lonchocarpus domingensis DC.). PPP? Rhodora [Vol. 69 . Geoffrea inermis W. Wright (Andira inermis HBK.). Lablab niger Medic. (Dolichos lablab L.). Oxalis rugeliana Urb. To be added: Coamo. Kallstroemia pubescens (G. Don) Dandy (Kallstroemia caribaea Rydb.). Galphimia glauca Cav. (Galphimia gracilis Benth.; Thryallis glauca Kuntze). Stigmaphyllon ledifolium (HBK.) Small (S. diversifolium A. Juss). Malpighia linearis Jacq. (Malpighia angustifolia L.). Zanthoxylum bifoliolatum Leon. To be added: Maricao. Guarea guidonia (L.) Sleumer (Guarea guara P. Wils.). Sapium jamaicense Sw. To be added: Luquillo Mts. Cedrela mexicana Roem. (C. odorata L.). Phyllanthus tenellus Roxb. To be added: Maricao. Euphorbia heterophylla L. (Poinsettia heterophylla K1. & G.). Euphorbia geniculata Ortega (Poinsettia geniculata Kl. & Garcke). Euphorbia cyathophora Murr. (Poinsettia cyathophora Kl. & Garcke). Chamaesyce hyssopifolia (L.) Small includes C. nutans (Lag.) Small, as a synonym. Chamaesyce adenoptera (Bertol.) Small ssp. pergamena (Small Burch (Chamaesyce monensis Millsp.). Chamaesyce turpinii (Boiss.) Millsp. (Chamaesyce anegaden- sis Millsp.; Chamaesyce portoricensis Millsp.). Comocladia dentata Jacq. To be added: Bayamón, Luquillo, Mayagüez. Comocladia cuneata Britt. To be added: Vega Baja (Britton, Brown & Chardon 6821). Conserve: Metopium toxiferum (L.) Krug & Urb. (Metopium linnaei Engl., in part). Ilex hypaneura Loes. To be added: El Verde, Luquillo Mts. (Alain 10846). Allophylus racemosus Sw. (Allophylus occidentalis Radlk.). Colubrina elliptica (Sw.) Briz. & Stern (Colubrina reclinata. Brongn.). Colubrina rufa var. antillana M. C. Johnst. To be added: Cambalache. Triumfetta rhomboidea Jacq. (Triumfetta bartramia L., nom. illegit.). Abutilon crispum Medic. (Gayoides crispum Small). Sida setifera Presl (Sida aggregata Presl). Sida alba L. is synonym to Sida spinosa L. Malachra alceifolia var. fasciata (Jacq.) Robyns (Malachra fasciata Jacq.). Flora of Porto Rico — Liogier 375 Urena lobata var. sinuata (L.) Hochr. (Urena sinuata L.). Pavonia spicata Cav. (Malache scagra Vogel; Pavonia scabra Ciferri). Hibiscus phoeniceus Jacq. (Hibiscus brasiliensis L., nom. nudum). Montezuma grandiflora (DC.) Urb. (Montezuma speciosis- sima, Sessé & Moc.). Gossypium barbadense L. (Gossypium peruvianum Cav.; Gossypium. brasiliense Macf.). Gossypium hirsutum L. var. punctatum (Schum.) Hutch. (Gossypium purpurascens Poir.). Doliocarpus major Gmel. in L. (Doliocarpus brevipedicellatus Garcke). Consolea moniliformis (L.) Britt. (Opuntia moniliformis L.). Consolea rubescens (Salm-Dyck ex DC.) Lam. (Opuntia rubescens Salm-Dyck). Pilosocereus nobilis (Haw.) Byles & Rowley (Cephalocereus nobilis Britt. & Rose). Leptocereus grantianus Britt. To be added: Culebra Isl. Daphnopsis americana ssp. caribaea (Griseb.) Nevl. (Daph- nopsis caribaea Griseb.) II. (Scientific Survey Vol. VI). Conostegia hotteana Urb. & Ekm. To be added: Luquillo Mts. Ossaea scalpta (Vent.) DC. (Ossaea domingensis Cogn.). Mitropsidium sintenisii (Kiaersk.) Burret (Calytropsidiuwm sintenisii Kiaersk.). Myrtus ? bellonis (Krub & Urb.) Burret (Eugenia ? bel- lonis Krug & Urb.). Zyzygium grande (Wight) Walp. To be added: Yauco. Ludwigia leptocarpa var. meyeriana (O. Kuntze) Alain. To be added: Manatí, Cabo Rojo, Bayamón, Humacao. Ardisia crenulata Vent. (Ardisia serrulata Mez, as to the portorican plant). Jacquinia arborea Vahl (Jacquinia basbasco Mez, in part). Manilkara albescens (Griseb.) Cron. To be added: Lajas. Jasminum fluminense Vell. (Jasminum azoricum sensu Britt. & Wils. in the Flora). . Jasminum multifiorum (Burm. f.) Andr. (Jasminum pubes- cens Willd.). Ipomoea spiralis House (cited as J. spirillus in Rhodora 67: 347, typographic error). Cordia obliqua Willd. To be added: Mameyes. Cordia wagnerorum Howard. To be added: Luquillo Mts. Cf. Journ. Arn. Arbor. 47: 137. 1966, p. 507. p. 507. p. 508. p. 512. Rhodora [Vol. 69 Stachys arvensis (L.) L. To be added: Cayey. Solanum nodiflorum Jacq. (Solanum nigrum sensu Britt. & Wils.; Solanum caribaeum Dunal). Solanum quitoense Lam. To be added: Toro Negro. Solanum woodburyi Howard. To be added: Luquillo Mts. Cf. Journ. Arn. Arbor. 47: 138. 1966. Buchnera longifolia HBK. (Buchnera elongata Sw., as to the Portorican plant). Utricularia gibba L. (Utricularia obtusa Sw.). Tabebuia heterophylla (DC.) Britt. includes Tabebuia pallida Miers, as synonym. Psychotria guadaloupensis (DC.) Howard (Psychotria gro- sourdyana Urb.). Psychotria nervosa Sw. (Psychotria undata Jacq.). Diodia apiculata (Willd.) R. & S. (Diodia rigida C. & S.). Coccinia grandis (L.) J. O. Voigt (Coccinia cordifolia Cogn.). Conyza chilensis Spreng. To be added: Yauco. Eclipta prostrata (L.) L. (Verbesina alba L.). . Polypodium attenuatum H. & B. To be added: Luquillo Mts. Polypodium triseriale Sw. To be added: Tortola Isl. Hypolepis nigrecens Hook. To be added: Jayuya. Asplenium rutaceum (Willd.) Mett. The presence of this species in Puerto Rico has been confirmed by Maxon. Thelypteris sprengelii (Kaulf.) Proctor (Dryopteris spren- gelii Kuntze, not Thelypteris balbisii Ching). Thelypteris rudis (Kuntze) Proctor. To be added: Jayuya. Thelypteris asplenioides (Sw.) Proctor. To be added: Jayuya. Thelypteris cordata (Fée) Proctor (Dryopteris cordata Urb.). The presence of the species in Puerto Rico has been confirmed: Utuado. Thelypteris resinifera (Desv.) Proctor (Dryopteris panamen- sis C. Chr.). To be added: Carolina. Hymenophyllum contortum v. d. Bosch (Hymenophyllum crispum HBK., sensu Maxon in Flora of Porto Rico). Hymenophyllum lanatum Fée. Replaces Hymenophyllum hir- sutum (L.) Sw.; this last species is not found in Puerto Rico. Hymenophyllum fucoides Sw. To be added: Mt. Guilarte. Hymenophyllum protrusum Hook, To be added: Luquillo Mts. Salvinia rotundifolia Willd. To be added: Mayagüez. Lycopodium verticillatum L.f. (Lycopodium setaceum Lam.). THE NEW YORK BOTANICAL GARDEN BRONX, NEW YORK NOTES FROM THE PRINGLE HERBARIUM, II. Since my previous article under this general title, infor- mation has continued to come to this Herbarium, indicating species which have not hitherto been reported in The Flora of Vermont, E. J. Dole, Editor, 1937. The list, only a part of which is given here, is longer than the earlier one; the other names are being left for subsequent mention. 1. PUCCINELLIA DISTANS (L.) Parl. var. ANGUSTIFOLIA (Blytt) Holmb. Burlington, Vermont, 21 June 1966, F. C. Seymour 24501 (VT) ; 7 Sept. 1966, same collector, 24,960 (VT). This alkali- loving plant has been found this year in two widely sepa- rated spots on the Campus of the University of Vermont. Near the Bailey Library it was growing abundantly, form- ing a sparse lawn in hard baked soil, introduced undoubt- edly in building operations. The other collection was from a roadside on Redstone campus, the women's part of the University campus. With much alkali soil in this State, this species might be expected. The strange fact is that it apparently has not been found before. 2. POA TRIVIALIS L. Wolcott, Vermont, Town Hill, meadow, 23 June 1966, Russell Kinerson, Jr. (VT). Newbury, 9 July 1966, F. C. Seymour 23,949. Another grass new to the State. So far north, it may be indigenous. See Gray's Manual, ed. 8, 1950, M. L. Fernald, p. 118, where it is said to be “indig. north w." 3. SCIRPUS POLYPHYLLUS Vahl. Newfane, Vermont, 17 Aug. 1911, L. A. Wheeler (NEBC). This is a southern species with its northernmost limit in *w. N.E." according to Gray's Manual, l.c. p. 274. Newfane is probably its northernmost known station. Found among unidentified specimens, this specimen was unrecognized probably because it had not previously been known in Ver- mont. It has been found in Buckland and Amherst, Massa- chusetts, both in or near the Connecticut River valley and is more frequent in Connecticut from the Connecticut River 377 378 Rhodora [Vol. 69 Valley westward. Newfane is only about 28 miles north of Buckland. 4. CAREX BUSHII Mack. Huntington-Richmond, Vermont, near the town line, in a pasture, 19 June 1964, Thomas G. Siccama and Alan N. Railsback (vT). In the course of ecological studies dealing with the flora of Camel’s Hump, these research students found the specimen so near the town line that it is uncer- tain in which of the two towns it was located. The other nearest known stations are Mount Desert, Me., and Con- necticut. Since the habitat is a pasture, it may have been introduced with live-stock. 5. ALNUS GLUTINOSA (L.) Gaertner, BLACK ALDER OF EUROPE. Essex Jct, Vermont, 20 Sept. 1965, William J. Gabriel (VT). Dr. Gabriel is Associate Geneticist in the U. S. Forest Service, Northeastern Forest Experiment Station, Burling- ton, Vt. He discovered a single large tree growing with the much smaller Alnus rugosa (Du Roi) Spengel. To find in that locality this species which is introduced from Europe is a bit surprising. In Pringle Herbarium is a specimen of A. glutinosa collected in cultivation in Burlington in 1901 by L. R. Jones. Conceivably the noncultivated tree could have come from the cultivated one in an almost adjacent city. 6. GEUM VERNUM (Raf.) T. & G. “Vermont” (vT). The only specimen in Pringle Herba- rium is mounted on a sheet which has the characteristic size, quality, etc. of the very early collections of Joseph Torrey, former President of the University of Vermont. The sheet bears no other indication of locality than merely “Vermont”. The specimen was identified, apparently by the collector, as Geum album. By an unknown hand, it was labeled Geum canadense Jacq. The very distinct long stipe of the receptacle and very small calyx mark it unmistakably as G. vernum. As the species is known in the State of New York, it is to be expected that further exploration will re- discover the now unknown locality where it grows in Ver- 1967] Pringle Herbarium — Seymour 379 mont. For the present this remains as the only known specimen from New England. 7. EUONYMUS EUROPAEUS L. Woodstock, Vermont, 16 Oct. 1966, Isabell R. Oktavec (Mrs. Frank) (vv). This species of Spindle-tree, escaping from cultivation in a number of places in New England, has made its debut in Vermont as indicated by the specimen cited. 8. OENOTHERA PILOSELLA Raf. Danville, Vermont, 14 July 1907, E. J. Dole (VT) ; Hart- land, 1 July 1911, Nancy Darling (vr). This plant with its showy flower is indigenous in the Midwest and is treated by L. H. Bailey in his Manual of Cultivated Plants, 1949, p. 738. Spreading from cultivation, it has become estab- lished in the East. The specimens cited although collected some years ago, were previously unrecognized and are now the only known specimens from Vermont, 9. ANTHRISCUS SYLVESTRIS (L.) Hoffm., CHERVIL. Braintree, Vermont, 25 June 1966 (VT). As a group was returning from the annual meeting of the Vermont Botani- cal and Bird Club, Mrs. Roberta G. Poland spotted from the moving car something unusual, In the party were Dr. & Mrs. Burdette Poland, Dr. & Mrs. Benjamin Shaub, Miss Marion L. Smith, Mrs. Mabel McIntosh and the writer of this article. None of us was able to identify in the field this accidental introduction of a white-flowered Umbellifer. The only other collections from New England which the writer has been able to find in herbaria are from Stony Brook Reservation and the region of Boston and Milton, both in Massachusetts. 10. CENTAUREA AUSTRIACA Willd. Dorset, Vermont, 6 & 22 Sept. 1966, A. H. Gilbert (VT). The collector reports a few plants growing near farm build- ings “‘persisting for several years" but not reproducing. This species is not reported in Gray's Manual, 8 ed. or in Britton & Brown's Illustrated Flora, 3 ed. Apparently it is a very rare adventive from Eurasia. Of the species de- scribed in the books mentioned, Centaurea austriaca most 380 Rhodora [Vol. 69 nearly resembles C. nigra L. with its entire to slightly lobed leaves and with its feather-like (pectinate) outer bracts of the involucre. In Centaurea austriaca, however, the body of the outer bracts is long attenuate or acuminate, not ovate as in C. nigra. FRANK C. SEYMOUR, PRINGLE HERBARIUM, UNIVERSITY OF VERMONT, BURLINGTON 05401 HALESIA CAROLINA L. IN KENTUCKY, INDIANA, AND OHIO: EDWARD W. CHESTER While studying the genus Halesia, new data has been found concerning the occurrence of H. carolina L. in three states where the presence of the species has been considered questionable. Most current manuals, e.g., Small (1933), Fernald (1950), Gleason (1952), and Gleason and Cron- quist (1963), do not include Kentucky within the range of this species. McFarland (1942) omits the species from his list of Kentucky plants and no specimens from naturally- occurring plants are present in the Kentucky State Her- barium (Dr. Edward Browne, personal communication, 1966). Gibson (1961) includes H. carolina in the state flora without specific locality. Braun (1943) reports the species from Harlan County and Garman (1913) notes its presence at “Star Lime Works, Lyon County." Braun (per- sonal communication, 1966) reports that the Harlan County population is probably now extinct due to destructive log- ging and mining in the area. Herbarium specimens have been examined from McCracken and Marshall Counties on the Tennessee River in western Kentucky and from Law- rence County on the Big Sandy River in eastern Kentucky. 'Contributions from the Botanical Laboratory, The University of Tennessee, N. Ser. 270. 1967] Halesia — Chester 381 The Lyon County record is considered authentic due to its proximity to documented stations even though voucher specimens are not available. Additional stations may be expected in counties adjacent to the Tennessee River and possibly in the extreme eastern section of the state. Halesia carolina L. reaches the northernmost limit of its natural range in southern Ohio where it is known from only two counties. The species is rare there and had previously been reported only from Scioto County (Schaffner, 1932; Jones, 1943; Braun, 1961) ; additional specimens have been observed from Pike County. Brendel (1858) first reported H. carolina from Indiana "Near Evansville on the Ohio River." Coulter (1899) also reported the species from this area but Deam (1940) specif- ically excluded the species from his Flora of Indiana since neither Coulter's nor Brendel’s report could be authen- ticated. Fernald (1950) credits the species to southern Indiana without specific locality, but other manuals (Small, Gleason, Gleason and Cronquist) exclude this state from its range. Specimens have been examined from two counties, Vanderburgh and Perry. It should be noted, however, that all specimens examined were collected before 1860, and due to expansion of Evansville and disturbance of surround- ing areas, it is quite possible that the species is now extinct in that area. Further exploration in southern Indiana, especially along the Ohio River, may possibly yield other stations. The following specimens were examined (herbarium ab- breviations follow Lanjouw and Stafleu, 1964; appreciation is extended to the curators of the respective herbaria for use of their material) : KENTUCKY: LAWRENCE COUNTY: Louisa, 1837, C. W. Short s.n. (TEX). MARSHALL COUNTY: 2 mi. s. of Tenn. River between Calvert City and Paducah, 22 Sept., 1945, W. H. Duncan 6186 (GA, MO). MCCRACKEN COUNTY: along the banks of the Tennessee River, 2 May, R. J. Fleetwood, s.m. (DPU). OHIO: PIKE COUNTY: Hay Hollow, Jackson Twp., 4 Oct., 1964, Floyd Bartley s.n. (08); e. of Omega, Scioto River, 16 Aug., 1965, C. S. Johnson 147 (NCU). SCIOTO COUNTY: Porter Twp., 13 April, 1929, Conrad Roth sm. (PH). INDIANA: 382 Rhodora [Vol. 69 VANDERBURGH COUNTY: near Evansville on Indiana Bank of Ohio River, 1848, C. W. Short s.n. (MO); near Evansville, 1850, C. W. Short s.n. (GH); banks of Ohio River near Evansville, 1858, C. W. Short s.m. (PH). UNIVERSITY OF TENNESSEE KNOXVILLE 37916 LITERATURE CITED BRAUN, E. Lucy, 1943. An annotated catalog of the spermatophytes of Kentucky. John S. Swift Co., Inc. Cincinnati. 161 p. , 1961. The woody plants of Ohio. State Univer- sity Press, Columbus, Ohio. 362 p. BRENDEL, FREDERICK, 1958. The trees and shrubs in Illinois. Trans. of the Ill. State Agr. Soe. 3: 588-604. COULTER, STANLEY, 1899. A catalogue of the flowering plants and the ferns and their allies indigenous to Indiana. 24th Annual Report of Dept. of Geology and Nat. Res. of Ind. 1073 p. Deam, C. C., 1940. Flora of Indiana. State of Indiana, Dept. of Cons. W. B. Buford Printing Co., Indianapols. 1236 p. FERNALD, M. L., 1950. Gray’s manual of botany, 8th ed. American Book Co., New York. 1632 p. GARMAN, H., 1913. The woody plants of Kentucky. Kentucky Agr. Exp. Station, Bull. 169. 62 p. GIBSON, DorotHy, 1961. Life forms of Kentucky flowering plants. Amer. Midl. Nat. 66: 1-60. GLEASON, H. A., 1952. The new Britton and Brown illustrated flora of the N.E.U.S. and adjacent Canada. Lancaster Press, Lan- caster, Pa. 3 vols. GLEASON, H. A., and A. CRONQUIST, 1963. Manual of Vascular Plants of N.E.U.S. and adjacent Canada. D. Van Nostrand Co., Inc., Princeton, New Jersey. 810 p. JoNES, C. H., 1943. Studies in Ohio floristies II. Rare plants in Ohio. Castanea 8: 81-108. LANJOUW, J., and F. A. STAFLEU, 1964. Index herbariorum, Part 1, the herbaria of the world. 5th ed. Utrecht, Netherlands. 251 p. MCFARLAND, F. T. 1942. A catalogue of the vascular plants of Kentucky. Castanea 7: 77-108. SCHAFFNER, JOHN, 1932. A revised catalog of Ohio vascular plants. Ohio Biol. Survey Bull. 25, 5(2) : 89-215. SMALL, J. K., 1933. Manual of the southeastern flora. Univ. of North Carolina Press, Chapel Hill, 1554. PETALOSTEMON FOLIOSUS IN ALABAMA In his revision of the genus Petalostemon (Leguminosae), Wemple (1965) suggested that P. foliosus Gray is perhaps now restricted in distribution to the Central Basin of Ten- nessee, although it was previously known to occur in north- ern Illinois as well. This assertion was based upon the observation that all herbarium specimens of the plant which he examined, collected in the past 50 years, were taken from limestone glades in Middle Tennessee. The purpose of this note is to report the discovery, August 23, 1966, of two populations of Petalostemon foliosus on limestone outcrops in northern Alabama, an area in which the plant was not previously known to occur. One of these is a small population, consisting of only a few plants, lo- cated in Franklin County, near Russellville, along county road 79, 0.4 mile south of the intersection with Alabama Route 24 (Baskin and Caudle 509). The other is an exten- sive population, consisting of hundreds of individuals, located in Morgan County, near McKendry, along Cedar Grove Church Road, 0.3 mile north of county road 55 (Bas- kin and Caudle 517). Documentary collections from both these locations have been deposited in the herbaria of Van- derbilt University and The New York Botanical Garden. The occurrence of Petalostemon foliosus in northern Ala- bama is not surprising in view of the environmental and floristic similarities of the two areas. Both are character- ized by shallow, rocky soil which is subject to wide fluctua- tions in moisture content, being extremely wet during the late winter and early spring and becoming very dry during the summer months. The ecological similarities of the areas have been described by Rollins (1963) in connection with his studies of evolution in Leavenworthia (Cruciferae). Although no comprehensive survey of the floristic affinities of the two areas has been made, several examples of species with restricted ranges are known to occur both in the glades 383 384 Rhodora [Vol. 69 of Central Tennessee and northern Alabama. Other species of legumes naturally distributed in both areas include As- tragalus tennesseensis Gray, Psoralea subacaulis 'T. & G., and Petalostemon gattingeri Heller. JERRY M. BASKIN AND CAROL CAUDLE VANDERBILT UNIVERSITY NASHVILLE, TENNESSEE LITERATURE CITED ROLLINS, REED C. 1963. The Evolution and Systematics of Leaven- worthia (Cruciferae). Contribution from The Gray Herbarium of Harvard Univ. No. CXCII. WEMPLE, D. K. 1965. Revision of the genus Petalostemon (Legumi- nosae). Unpublished Doctoral thesis, Iowa St. Univ. Library, Ames. Volume 69, No. 779 including pages 241-384, was issued Sept. 29, 1967 FARLOW REFERENCE LIBRARY 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 Paseo Editors ROLLA MILTON TRYON RADCLIFFE BARNES PIKE LORIN IVES NEVLING, JR. 3 Vol. 69 October-December, 1967 No. 780 CONTENTS: Annotated Checklist of the Dicotyledons of Tortola, Virgin Niands WU G. D'Atey a E E N E isa ana p RARE REDE ROS 985 Connecting Vowels in Epithets of Latin Origin. Berara: BoI -eennuridmieisn east i Deis cit 451 Another Tribulus Adventive in the New World. nean. M. Porter edet emt tetelecetidelesr ier ciere, 455 Book Review: The Flora of Eastern Himalaya. Rolla TEMOR. Lesser rit ten cintas sabe Crie s Sido Eon dee ERE 456 A New Station for Hamamelis virginiana L, in Minnesota. Albert G. Johnson and Margaret H. Smithberg .............. 457 Thlaspi tuberosum Nutt, a Taxonomic Synonym of Carda- mine douglassii Britt. Ronald L. Stuckey ........................ 460 (Continued on Inside Cover) The New England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. CONTENTS: — continued A New Talinum from the Cedar Glades of Middle Tennes- see, Stewart Ware wccccccsscccccssccecsscececscccesceecesecssecseecenseeeeeseeeee 466 A New Species of Machaeranthera Section Psilactis. R. S. Jackson and R. R. Johnsom ......... eee 476 Additions to the Vascular Flora of Oklahoma, II. R. John Taylor and Constance Taylor ............. ee 480 Saxifraga Aizoón in New Hampshire. John A. Churchill 483 The Saxifraga Aizoón Station Revisited. Frederic L. Steele ......... esee nnne ntn eee 484 Saxifraga cernua in New England. John A. Churchill and A. R. Hodgdon C-—-—-——A————»——— —— 485 Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 69 October-December 780 ANNOTATED CHECKLIST OF THE DICOTYLEDONS OF TORTOLA, VIRGIN ISLANDS W. G. D'ARCY SITUATION: Tortola (18° 23'— 18° 28' N latitude and 64° 33' — 64° 43’ W longitude) is 1114 miles long and 31% miles wide at its widest points, the axis of the island run- ning east north east and west south west. It is the largest of the British Virgin Islands, which are located at the northeast extremity of the Caribbean Sea. Surrounding Tortola itself are a number of small cays which shelter the island from the direct effects of ocean swells. Beef Island and Frenchman’s Cay, the nearest, are separated from Tor- tola by only a few yards of water, and have recently been connected by bridges. With the exception of Anegada and St. Croix, which are separated by distance and geology from the rest of the group, the Virgin Islands, British and Ameri- can, are separated by very short sea distances. TABLE I Distances Between Major Islands — British and U.S. Virgin Islands Statute Miles Kilometers Tortola - Virgin Gorda 5 fi Tortola - St. John 11% 214 Tortola - St. Thomas 9 14 St. John - St. Thomas 2 3 GEOLOGY : The geology of the island has been well described in other literature (Kemp & Meyerhoff 1926 & 1927, Martin- 385 1961 3sn2ny cc *q*5*M [Vol. 69 8321438 1d S391J4S1Q Alq S321I3SIq e391peu.squ] j9M iO 3S10q SROILVAH3SHO SAIIVLILNVDO-NON H3HIO ANV NOLLYIGOGA Ad Q3ONSQIAS SV CNOLLITNOO 3HÁLSIOW CNY TIVZNIYU AYD SNVWHON33J NY3( uod D Ava 3714dv 3417 Ava lOoBNBNvO Rhodora Ava $4338 SGNVTSTI NIOUWIA HSILIEUGH = VTIOLHOL 386 1967] Tortola — D’Arcy 387 Kaye 1959). It is essentially the West Indies Cretaceous formation which uplifted and metamorphosed in Eocene times with many resulting intrusions of igneous material such as quartz, granite and hornblende into the predominant limestone structure. The Virgin Islands present an ex- tremely local and varied geological pattern. Strata are usually tilted and folded, and cliffsides show a patchwork effect. Several different types of rock are to be found under- foot in the space of a few yards. No fossils have been found on Tortola, but their presence is to be expected as a number of references are made to fossils on the American Virgins. Fossils on the adjacent Coki Point, St. Thomas, are only visible when the rock is wet, and similar structures may yet be found on Tortola. PHYSIOGRAPHY: The mountain ridge forming the mass of the island drops steeply to the sea in most places, but is dissected by many sizable ghuts or ravines which fan out into large flat bottoms before meeting the sea. The ridge top consists of a broken and eroded peneplane of 800-1200 feet (240-360 m.) elevation with a higher ridge, Sage Mountain, reaching 1700 feet (450 m.) in the western third of the island. Martin-Kaye (1959) describes two other peneplanes, but they are difficult to see. Suggestion of a peneplane at 400-600 feet (120-180 m.) gives rise to a number of promontories extending seaward between the ghuts and bottoms. CLIMATE: A number of climatic factors strongly influence the pattern of vegetation. Seasons: Seasons are not strong- ly marked by temperature or length of day, but there is a significant seasonal change in vegetation independent of rainfall, e.g. Sabinea florida blooms in the early part of the year, Hibiscus sabdariffa, Eupatorium spp. and Melochia nodiflora are winter bloomers, Myrciaria floribunda blooms in the autumn and Acacia macracantha is usually deciduous in January or February. Rainfall variations often alter leaf drop or blooming for about a month forwards or back- wards, but the usual effect is to influence amount of flower- ing or set of fruit. Although a seasonal pattern of rainfall [Vol. 69 Rhodora 388 'spiooe1 3ueuryrede([ [e1inj[noLi2 y uro1g UJEL 'uAo[ poy 'uonejs jueuriedx;p IY} 3e UAI Jo sayouT, ‘PI ‘plover fo suvax ‘cg 'd ‘0961 uSnoiuj TOGT 40g Juewelddng 'sejejS POUN ey} jo ArewUING Orjeur[) 'e3eur[) Sojejg pojru[] JO snsuapD [eruuooe([ ‘Gp-9g 'ON Soje3g pejru[) eu; jo Audei3ojeur) wo1g 'uygof yg ‘Aeg zng *‘qloyualye g SIFQ, 'O ‘qd ‘UopsuryseM — 'A10jeAd9Sq() [VALN SIJ peju[] '9p6T ‘SMoeueydg ueotreury ƏL 0j Juewelddng YFM, pue 3esung 'estiung Jo se[qe[ 'qjuour jo Aep 4sTzZ, 889 $69 88°F 00°F 90'T 66 Y 6L'6 PTZ ESP $€9 SLL €6L vY'08 S'I8 €'68 8' I8 & 18 6 08 € 6L 6LL 6€ €9 09 09 T9 99 S9 L9 v9 09 66 66 66 $6 €6 $6 66 06 S6 06 SOqIl PLE GETE BOE ,88 «6I EOST ELST EOST BEZIT LOZI "2q 'a0oN 190 ydag ‘Ony fünf aung hoy ‘dy “LDN II adv L src 97'E [lesley O'LL L'9L uea 09 6€ Uurnururqy 06 88 Uurnulrxe w ,91n3e1eduie p OP ull (ET OTI ausi4eq Jo sanog “994A "WD spuv[ST wiry oY} Jo euro 1967] Tortola — D'Arcy 389 can be shown statistically, it is far from dependable and cannot be regarded as an important seasonal determinant. Farmers expect rain in October and plant accordingly but are often disappointed. Very heavy rains sometimes fall in May, but these too are irregular. From February to May, there is seldom much rain, and sometimes none at all. Winds are generally higher in the winter months, but the relation to seasonal vegetation patterns was not observed. Rainfall: Rainfall records are available only for the Experi- ment Station at Road Town, and these are assumed to indicate the general fluctuations at other points on the island. More important than seasonal fluctuations are periods of drought. Howard and Proctor (1958) assume that 4.0 inches of rain are necessary in each month for optimum plant growth in the West Indies. During the 63 year period between 1901 and 1964, 442 or 58% of months had less than four inches of rain, and 182 or 24% of months had less than two inches of rain; 64% of years had three or more months with less than two inches. Approximately one year in five has an extended drought which lowers the water table, turns wells brackish and which has a drastic effect on agriculture and vegetation in general, The effects of this drought pattern are much less pronounced in norm- ally wet districts; and in the very dry districts, it has imposed an almost xerophytic vegetation. Geographic rain- fall patterns must be considered in connection with wind and altitude. Wind: The easterly trade wind is almost continuous, and it is a rare day that it is absent altogether. Its drying effects are strikingly in the xerotic appearance of promontories at lower elevations. Fish Bay and Jean Hill, Duffs Bottom and Slaney Point are good examples where adjacent tracts with the same rainfall show very different patterns due mainly to the wind. Wind is prob- ably the major factor in the persistence of the Croton scrub on abandoned pastures, as similar areas with some shelter are largely covered with Eugenia biflora. Altitude: Sage Mountain, with its 1710 foot elevation is the principal land feature influencing precipitation. Reference should be made to the rainfall map of the island. Sage Mountain is 390 Rhodora [Vol. 69 frequently shrouded in cloud mists and provides the temper- ature drop to initiate precipitation. The trade winds tend to direct precipitation to the west and north sides of the ridge, and when winds are high, much of the rain falls in the sea between Tortola and St. Thomas. Higher moisture levels are needed to result in precipitation to the east of Sage Mountain, and this tends to follow the mountain ridge as far east as Hope. Some of this rain falls on lower levels, and because of the wind again, more of it falls on the north side, giving Brewers Bay and Cane Garden Bay districts a flourishing agriculture. East of Hope, and at sea level east of Brandywine Bay, rainfall is light, and there are frequent periods of several months without significant rain. Neither Virgin Gorda nor Beef Island has sufficient elevation to in- duce much rain and they have a chronic drought problem. Someone facetiously suggested the solution to Tortola’s water problem is to move Sage Mountain to Beef Island: it would undoubtedly have the desired result. Local residents frequently state that the island is “drying up"; this belief no doubt stems from memories of very high rainfall in the 1930’s, but reference to statistics from 1901 to the present indicate no trend in the amount of rainfall. HISTORY OF CULTIVATION: Artifacts found on Tortola indicate that in pre-Columbian times it was settled by Arawaks, an agrarian people. Dutch buccaneers took pos- session of the island in 1648, but it was not until 1700 that English planters from Anguilla made the first agricultural settlement, Slaves were brought in and almost the whole of the island was cleared for cotton, sugar, pasture and minor crops. Evidence of the extent of this cultivation is still to be found in half buried stone terracing in much of the present day mountain forest. With the end of slavery and the sugar era in the middle of the last century, much of this cleared land grew back into forest, although promontories and other areas of low rainfall achieved little more than a scrub vegetation and more promising lands remained de- voted to subsistence farming. Charcoal cutting and cattle raising have been traditional occupations on Tortola, but were accelerated in this century by government cattle and 1967] Tortola — D'Arcy 391 pasture programs. Planting of pasture grasses, especially the all-excluding pangola grass, has inhibited reforestation in many areas. On the south side near the top of the ridge, in spite of high rainfall, the strong winds have maintained open pasture and krumholz situations, whereas on the north, large areas of shrubs and some forest have reappeared. Early in this century, the government established a factory at the Experiment Station to encourage sugar, cotton and limes. Today, some sugar is cultivated for making local rum. In 1947, government cotton support was withdrawn and cultivation ceased, but Gossypiwm spp. are naturalized and wild trees are to be seen scattered through- out the island. Limes are no longer cultivated as a com- mercial crop, but plantation remnants in Cane Garden Bay provide a small household export to St. Thomas, and lime trees are a naturalized species. During the 1930's, tobacco was extensively cultivated in Free Bottom and other areas, but no plants are to be seen on the island today. Coconuts have been cultivated from time to time and are to be seen in many parts of the island. Over the centuries many other crops have been cultivated, and today bananas, yams, cas- sava, tanias, pidgeon peas, okra and peppers are standard crops exported on a household scale. The island is now a net importer of most of these crops. Seattered specimens of Tectona grandis, Theobroma cacao, Citrus sinensis and other species result from earlier agriculture attempts. Mangifera indica and Carica papaya are widely naturalized, the wild trees producing very inferior fruit. Judging from its present frequency, Annona muricata was probably culti- vated on an extensive scale for hog food. Introductions of Peltophorum inerme, Azadirachta indica and the wide- spread use of Samanea saman as a pasture tree reflect British colonial agricultural direction: while these species may occur on the American Virgins, they are not at all common there. Cultivation, forest clearing and the pre- vailing practice of burning off pastures have had a disastrous effect on the soil. Every sizable rain discolors Road Harbour with soil washed down from the hills, and the hillsides are again too dry for satisfactory growth in a 392 Rhodora [Vol. 69 few days. Sage Mountain forests, steep ghuts and rocky seaside cliffs fare much better as they have not had the same loss through soil disturbance, and therefore support a larger flora than the cultivated or pasture hillsides. On Camanoe, Peter Island and other cays are undisturbed areas which support floras comparable to areas of much higher rainfall on Tortola. Sage Mountain Forests: Once covering most of the central parts of the island, the forest is now reduced to a few dozen acres on the north and western parts of the ridge. Widely scattered standing boles of Manilkara balata left untouched by charcoal cutters attest to the former extent of forest in even recent times. Sage Mountain peak itself and Chalwell peak are now completely devoid of original vegetation. Probably very little of the present mountain forest is aboriginal, but most of it qualifies as sub-climax rain forest (Beard 1942, 1944, Stoffers 1956). Ficus trigonata, Tetragastris balsamifera, Manilkara balata, the luxuriant development of Clusia rosea trees, and the steep inarable condition of the summit argue for an abori- ginal condition, but Heliconia caribaea and Syzygium jam- bos suggest disturbance at some time. The presence of buried terracing at 1300 foot elevation on the north side is unarguable. The forest has been described (B.V.I. Report, 1964) as xerophytic rain forest, a formation characterized (Beard 1944) by peeling desiccated tree bark and small, much-cutinized leaves. These xeromorphic features are not conspicuous in the Tortola forests today, although Beard's description may have applied to areas which have since been cut off. The forest is no doubt modified to a considera- ble extent by the prolonged dry periods, and the unhappy term "xerophytic rain forest" might better be replaced by the appellation “aridulate rain forest." The almost complete absence of epiphytic orchids and the complete absence of gesneriads, and the fact that most visitors fail to penetrate the dense barrier of shrubs that surrounds the forest, and hence never see it at all, seem to be the main reasons for local avoidance of the term rain forest. Philodendrons, Aechmeas, Peperomias and Marcgravia rectiflora are fre- quent epiphytes. There are indications that the forest is ex- 1967] Tortola — D'Arcy 393 panding: Cyathea arborea, Heliconia caribaea, Cecropia peltata, Acnistus arboreus, Cupania americana and Syzy- gium jambos, characteristic species of Pioneer Rain Forest (Stoffers 1956), are found in fairly extensive areas at the periphery of the old formation. With the general decline of agriculture on the island, pastures are not being extended, and the competition of bottled cooking gas is fast reducing charcoal cutting. Sage Mountain peak and a substantial neighboring acreage have been planted to restore the forest, and if this does not re-establish the former flora, it will help provide humidity conditions favorable to natural extension of the present forest. Continued illegal charcoal cutting or establishment of proposed residential housing on the moun- tainsides will limit such expansion. FLORISTIC AFFILIATIONS: Although species are often not represented on all islands in the group, it is usual for a species to be found on more than one. The offshore Puerto Rico islands, Culebra, Icacos and Vieques, have essentially the same physical structure and flora as the Virgin Islands, which is a xerophytic extension of the Puerto Rico flora. Only perhaps 60 species of plants are found in the islands which are not found in Puerto Rico, and these include several endemics. A few plants clearly originate in the islands to the south and were not reported for Puerto Rico by Britton and Wilson (1923 & 1925). Morisonia americana is a striking example which occurs throughout the Virgin Islands and south to Venezuela. Chaetocalyx scandens was found on Tortola in 1965 and comes from the islands to the south. Some Tortola plants have widespread origins, e.g. Croton ovalifolius (South America and Greater An- tilles), Solanum elaeagnifolium (Mexico and S. U.S.), Diodia apiculata (Anegada and Hispaniola) and Stigma- phyllon diversifolium (Cuba and Hispaniola), but there are not many. Of course, many species have been brought to the island and naturalized through the agency of man, e.g. Pedilanthus tithymaloides (probably pre-Columbian), Kal- anchoà verticillata, Tamarindus indica, Haematoxylon campechianum, and Syzygium jambos, While floristic affili- ation is principally with Puerto Rico, the islands form a link 394 Rhodora [Vol. 69 in the ecological chain of windswept small islands stretching from the Bahamas to Tobago; many species are common throughout, and they reflect the xeric and saline habitats offered in this group of islands. One species of the archi- pelago conspicuously wanting in the wild state on Tortola is Casuarina equisetifolia which has not reproduced on any of the low seaside areas such as Sandy Point or Beef Island. Proximity is the principal reason for the greater affinity with Puerto Rico: Tortola is separated from Anguilla, the nearest of the Leeward Islands, by 92 miles (79 km.) of open sea, while between Tortola and Puerto Rico the farthest open distance (Culebra to Vieques) is only 9 miles (14 km.). ENDEMISM: While a number of endemics have been re- ported for other Virgin Islands, Tortola has only one, Calyptranthes kiaerskovii, based on a sterile collection by Eggers in 1879. Even for the Virgin Islands as a whole, Britton & Wilson (1923 & 1925) note only about 20 endemics, and some of them, e.g. Salvia thomasiana, are suspect. It is possible that further observations will turn up additional endemics on Tortola, but the number cannot be large, The presence of endemics on neighboring islands and the au- thor's own observations seem to indicate the possibility of endemics in a number of genera: Zanthoxylum, Malpighia, Brunfelsia, Galactia, Forestiera and Tephrosia. One species, Tillandsia lineatispica, which was reported as a St. John endemic on the basis of a single Eggers collection was found to be plentiful on Peter Island and Camanoe. Piptocoma antillana was found on Virgin Gorda and is plentiful on Peter Island, but is a rare plant on other Virgin Islands, the east end of Puerto Rico and Hispaniola, and its scant distribution begs the question of whether it is a relic or an expanding endemic. Small cays like Peter Island and Camanoe are well worth future attention as they have undergone less disturbance than the larger of the Virgin Islands, PREVIOUS COLLECTORS: Richard, Lowis-Claude, (1754- 1821). Collected on Tortola and St. Thomas in 1786-87, depositing in the Muséum National d’Histoire Naturelle, Paris. Reference to his collections was not made for this 1967] Tortola — D'Arcy 395 TABLE III Plants Growing Wild (indigenous, naturalized or escaping) Major Families Genera Species Leguminosae 47 73 Mimosaceae 12 TIT Caesalpiniaceae 7 I5 Fabaceae 28 41 Compositae 33 43 Euphorbiaceae 18 34 Rubiaceae 18 24 Malvaceae 12 24 Convolvulaceae ri 24 Myrtaceae 9 19 Boraginaceae 5 18 Solanaceae 7 16 Acanthaceae T 14 Amaranthaceae 8 12 Verbenaceae 7 11 Labiatae 7 10 Capparaceae 3 10 Total Major Families 188 329 Minor Families (79) 161 221 Total Wild Plants 349 550 Cultivated Plants LH Total Dicotyledons 661 checklist. L.-Cl. Richard was later professor of botany at the Ecole de Medicine in Paris. His son, Achille Richard, was author of Essai sur la Flore de Cuba, included in the work of Ramon de la Sagra on that country. Schomburgk, Sir Robert Herman (1804-65). Not to be confused with his younger brother Sir Otto Schomburgk who collected in the Guyanas, Sir Robert lived on Anegada in the 1830’s, and although he made no collections, his papers provided a use- ful introduction to further study. See Kemp (1926). Grise- bach, August Heinrich Rudolph (1814-79). His monu- mental Flora of the British West Indies makes no mention of Tortola. He worked with collections in the British Mu- 396 Rhodora [Vol. 69 seum, Kew Museum (sic.), the Academy of Stockholm and extensive other collections, and it is fair to assume he came across no Tortola items. Eggers, Baron Heinrich Franz Alexander (1844-1903). Resident of St. Croix and St. Thomas from 1869 to 1887, he visited Tortola over the Christmas season of 1887. While he took only 68 specimens, the localities recorded indicate he travelled widely during his brief stay. His Flora of St. Croix (1879) and other publications are important studies of the West Indies flora. Fishlock, Walter Charles (1875-1932). Born at Bathford, near Bath, England, he was a student gardener at Kew and then was employed in Hyde Park, London. On May 2, 1902, he arrived at the Experiment Station on Tortola, established the year before, and in about 1912 became Curator of the Station. In August 1919 after taking leave, he became Senior Curator in the Department of Agriculture, the Gold Coast. He collected extensively, sending specimens to Kew, New York Botanical Garden and other institutions. Older residents of the island credit him with introduction of many economic plants. His Handbook of the Virgin Islands con- tains little botanical information but is useful from a his- torical point of view. Most of his collections were published by Britton & Wilson (1923 & 1925). Britton, Nathanial Lord (1859-1934). Director of the New York Botanical Garden for many years, he was a towering leader in American botany. He was instigator of the now discarded American Code of nomenclature, and his publications follow this system. Between 1913 and 1923 he visited Tortola a number of times, sometimes accompanied by his wife, a distinguished bryologist, and by John Shafer, also of the New York Botanical Garden. This work was in preparation for his Botany of Porto Rico and the Virgin Islands, and while collecting was extensive, Tortola seems to have re- ceived less attention than any other major island in the study. Fishlock accompanied Britton on his Tortola ex- cursions, and many of the Britton and Wilson reports are based on Fishlock collections. Britton’s collections are de- posited in the Herbarium of the New York Botanical Garden. Velez, Ismael (1908—). Velez visited Tortola for 1967] Tortola — D'Arcy 397 about two weeks in 1950, collecting herbaceous species for his Lesser Antilles weed study (1957). His specimens are deposited with the Imperial College of Agriculture, Trinidad and with the Inter American University of Puerto Rico, his present address. In publishing his study Velez did not indicate on which of the Virgin Islands he collected a species. Little, Elbert L., Jr. (1907 —). As Dendrologist with the United States Department of Agriculture, Forest Service, he spent a number of years in Puerto Rico, visiting Tortola on July 19, 20 & 21, 1954. He collected 35 numbers of trees and other woody plants in preparation for his excel- lent Common Trees of Puerto Rico and the Virgin Islands (1964). His specimens are deposited with the United States National Museum, ; ORIGIN OF THE CHECKLIST: The author visited the island several times between 1959 and 1961 and was a resident for five years between 1961 and 1966. Only the final two or three years of residence was devoted to botanical work, and such work was necessarily secondary to other pursuits. A complete flora and vegetational study was envisaged, but this goal was interrupted early in 1966 when the author left the island for business reasons. All parts of the island were visited several times and over 800 numbers of vascular plants were collected. Virgin Gorda, St. John, St. Thomas and many of the smaller cays were visited several times with some collections taken. Tortola has been badly neglected in botanical literature, and it seems better, in spite of insufficient work, to provide a preliminary check- list of the plants growing there than to wait until an oppor- tunity for more study presents itself. The checklist includes 106 new reports by the author, as well as 11 by Eggers and 24 by Fishlock which had been overlooked in previous literature, or over 25% new reports in all. Specimens were sent to the Arnold Arboretum, the British Museum (Natural History), The Institute of Ja- maica, the University in Copenhagen and to the University of Florida and a complete set of specimens is retained by the author. Dr. R. A. Howard of the Arnold Arboretum, Mr. George Proctor of the Jamaica Institute and Dr. W. T. 398 Rhodora [Vol. 69 Stearn of the British Museum (N.H.) very kindly provided identification of plants they received; and Dr. A. Skovsted of the University in Copenhagen kindly supplied a list of plants collected by Eggers. Dr. Elbert L. Little, Jr., of the United States Forest Service, Bro, Alain Liogier of Man- hattan College, Dr. Ismael Velez of the Inter American University of Puerto Rico, Mr. J. E. Dandy of the British Museum (N.H.), M. G. Aymonin of the Museum National d'Histoire Naturelle, Paris, and Sir George Taylor, Kew, sent helpful correspondence, Gratitude is due to Mr. Malcolm Winter, Superintendent of Agriculture, for much advice and for permission to take specimens from the Experiment Station. Many house- holders were kind enough to supply plants from their grounds, and especial thanks must go to Mr. Rowan Roy, Mrs. A. A. Hurt and Mrs. Robert P. Nevin. Mr. Norman Kerr was most helpful with his thorough botanical back- ground and took part in several collecting trips. Dr. Richard A. Howard, Director, The Arnold Arboretum, was steadily encouraging in his prompt and helpful correspondence, and without his free assistance, this checklist would not have emerged. Dr. D. B. Ward and Mr. John Beckner, Univer- sity of Florida, patiently read this paper and offered many suggestions. CHECKLIST FORMAT: Families follow the order of Britton & Wilson. Genera are listed alphabetically and species alphabetically within genera. Nomenclature is taken mainly from Flora of Barbados, Bro. Alain Liogier's list and from Britton and Wilson, but in a few cases, reference to current monographs suggested other usage. Only names used by Britton and Wilson or widely used at the present time are indicated in synonomy. Unless otherwise indicated, cita- tions are for Tortola. The author's own collections are indi- cated by number only, and when no other citation is given, this is a new report for the species. A good description of the plant may usually be found in Britton and Wilson (1923 & 1925) or Little and Wadsworth (1964) when these authors are cited as finding a species on Tortola. Place names in the text are taken from the maps D.O.S. 346 1967] Tortola — D'Arcy 399 (Series E 837), First Edition 1959, but Eggers and Fish- lock collections are cited with place names as supplied. UNIVERSITY OF FLORIDA, GAINESVILLE 32601 BIBLIOGRAPHY ANON. British Virgin Islands, Reports for the Years 1961 and 1962. Her Majesty's Stationery Office, London 1964. BEARD, J. S. The Natural Vegetation of the Island of Tobago, British West Indies. Ecological Monographs, Vol. 14, No. 2. BEARD, J. S. Montane Vegetation in the Antilles. The Caribbean Forester, Vol. 3, No. 2, January 1942. BRITTON, N. L. and PERCY WiLsoN. Botany of Porto Rico and the Virgin Islands, Scientific Survey of Porto Rico and the Virgin Islands, Vols. V & VI: New York Academy of Sciences, 1923 & 1925. EGGERS, BARON H. F. A. The Flora of St. Croix and the Virgin Islands. Bulletin of the United States National Museum, No. 13, Washington, 1879. FisHLOCK, W. C. The Virgin Islands, B. W. L, A Handbook of General Information, London, 1912. GLEASON, H. A. and MEL. T. Cook. Plant Ecology of Porto Rico. Scientifie Survey of Porto Rico and the Virgin Islands, Vol. VII: Part 1, New York Academy of Sciences, 1927. Goopinc, E. G. B., A. R. LOVELESS, and G. R. PROCTOR. Flora of Barbados. Her Majesty’s Stationery Office, London, 1965. Howarp, R. A. & G. R. PROCTOR. The Vegetation on Bauxite Soils in Jamaica. Journal of the Arnold Arboretum, Vol. 38: 1-50, 151-169, Cambridge, Mass., 1958. KEMP, JAMES F. and HOWARD A. MEYERHOFF. Geology of the Virgin Islands, Culebra and Vieques, Scientific Survey of Porto Rico and the Virgin Islands, Vol. IV, New York Academy of Sciences, 1926 & 1927. LiTTLE, ELBERT L., JR. and FRANK H. WADSWORTH. Common Trees of Puerto Rico and the Virgin Islands. United States Department of Agriculture, Forest Service, Agriculture Handbook No. 249, Washington 1964. LioGIER, BRO. ALAIN. Nomenclatural Changes and Additions to Brit- ton and Wilson's *Flora of Porto Rico and the Virgin Islands." Rhodora Vol. 67: 315-361, 1965. MARTIN-KAYE, P. H. Reports on the Geology of the Leeward and British Islands. Castries, St. Lucia, 1959. MILLSPAUGH, CHARLES FREDRICK. Flora of the Island of St. Croix. Field Columbian Museum Publication 68, Chicago 1902. 400 Rhodora [Vol. 69 MOLDENKE, HAROLD N. A Resumé of the Verbenaceae, Avicenniaceae, Stilbaceae, Symphoremaceae and Eriocaulaceae of the World as to Valid Taxa, Geographic Distribution and Synonymy. Private Publication, Yonkers, N. Y. 1959. STOFFERS, A. L. The Vegetation of the Netherlands Antilles. Uit- gaven “Natuurwetenschappelijke Studiekring Voor Suriname en de Nederlandse Antillen," No. 15, Utrecht, 1956. VELEZ, ISMAEL. Herbaceous Angiosperms of the Lesser Antilles. Inter American University of Puerto Rico, 1957. CHECKLIST OF DICOTYLEDONS ON TORTOLA, VIRGIN ISLANDS Family 1. CASUARINACEAE Casuarina Adans. Casuarina equisetifolia J. R. & G. Forst. A few cultivated specimens; Government House, Experiment Station, Pasea Hall Estate. Family 2. PIPERACEAE Peperomia Ruiz & Pav. Peperomia glabella (Sw.) A. Dietr. Common at higher elevations. Tolerates a wide range of sunlight. Sometimes as an epiphyte with Epidendrum ciliare. 111, 112. Britt. & Wils. P. humilis (Vahl) A. Dietr. Rare on undisturbed rocks and cliffs. Seen at Huntum's Ghut (110 m.), and at west end of Brewers Bay (sea level). 466, 467, 468, 469, 470; Fishlock 243. Britt. & Wils. P. magnoliaefolia (Jacq.) A. Dietr. In shade, Sage Mt. forests. 87; Fishlock 369. Britt, & Wils. P. pellucida (L.) H. B. K. A common street weed in Road Town. 150, 152; Fishlock 230. Britt. & Wils. Piper L. Piper amalago L. Common in shady ghuts. 148, 169; Fishlock 149; Eggers 3189. Britt. & Wils. Pothomorphe Miq. Pothomorphe peltata (L.) Miq. Occasional at upper elevations. 465; Fishlock 338. Britt. & Wils. Family 3. ULMACEAE Celtis L. Celtis iguanaea (Jacq.) Sarg. Momisia iguanaea (Jacq.) Rose & Standley. Fishlock 360, Fishlock 361, and Fishlock 409 (Hunt- um's Ghut). Britt. & Wils. C. trinerva Lam. Britt. & Wils. Trema Lour. Trema micrantha (L.) Blume. Fishlock 219 (Lower Estate). Little & Wads., Britt. & Wils. 1967] Tortola — D’Arcy 401 Family 4. MORACEAE Artocarpus Forst. Artocarpus altilis (Parkinson) Fosberg. Artocarpus incisa L. f. Artocarpus communis Forst. Cultivated, chiefly in moist districts. “Breadfruit.” The seeded form, “Breadnut,” is rarely cultivated. Little & Wads. Cecropia L. Cecropia peltata L. Common on moist forest slopes out of the wind. “Trumpetwood.” Little & Wads. Ficus L. Ficus citrifolia Mill. Ficus laevigata Vahl. Common. A fine avenue occurs at Little Apple Bay. “Fig Tree." Fishlock 418 (Belle- view). Little & Wads., Britt. & Wils. F. trigonata L. Ficus crassinervia Desf. In Sage Mountain forests. 666. Little & Wads., Britt. & Wils. Family 5. URTICACEAE Fleurya Gaud. Fleurya aestuans (L.) Gaud. Very rare. Single plants seen at Call- woods, Cane Garden Bay, and at the Experiment Station. 212, Fishlock 244 (Lower Estate). Britt. & Wils. Pilea Lindl. Pilea herniarioides (Sw.) Lindl. Rare. Seen on Joe’s Hill, and on walls, Cane Garden Bay. May not be distinct from Pilea micro- phylla. 215, 216, 217. microphylla (L.) Liebm. Common cultivated. “Everlasting.” Fish- lock 375 (Slaney Point). P. nummulariaefolia (Sw.) Wedd. Very rare. Seen only at roadside, Experiment Station. 782, 783, 784. . sanctae-crucis Liebm. Common in mountain forests and coming down to 100 m. in Long Bush and Huntum's Ghuts. 100, Fish- lock 245 (Joe's Hill). Britt. & Wils. P. tenerrima Miq. Occasional patches seen in moist years. 293, 294, 295, 296. Britt & Wils. v — v Family 6. PROTEACEAE Grevillea R. Brown Grevillea robusta Cunn. The single fine specimen in the Road Town Experiment Station was surveyed at 49 feet tall on 25th April, 1965. Family 7. LORANTHACEAE Dendropemon Blume Dendropemon caribaeum Krug & Urban. Phthirusa caribaea (Krug & Urban) Engler. Common. This mistletoe seems particularly partial to Citrus. 831. Fishlock 397 (Old Plantation). Britt & Wils. 402 Rhodora [Vol. 69 Family 8. ARISTOLOCHACEAE Aristolochia L. Aristolochia littoralis Parodi. Aristolochia elegans Mast. Cultivated and escapes. “Dutchman’s Pipe." 355. A. trilobata L. Fishlock 433 (Long Bay). Family 9. POLYGONACEAE Antigonon Endl. Antigonon leptopus Hook. & Arn. Common in thickets, roadsides and towns; moist and dry districts. ‘Coral Vine," “Coralita.” 235. Coccoloba L. Coccoloba microstachya Willd. Coccolobis obtusifolia Jacq. 771, 772 (Beef Island); Fishlock 142 (woods). Little & Wads., Britt. & Wils. C. uvifera (L.) L. Coccolobis uvifera (L.) Jacq. Common along sea- coasts. Also plentiful near Meyers (300 m.). “Sea Grape." Little & Wads., Britt. & Wils. C. venosa L. Common in shady ghuts, mostly at lower elevations. "Chiggery Grape." 149. Little & Wads., Britt. & Wils. Family 10. CHENOPODIACEAE Atriplex L. Atriplex pentandra (Jacq.) Standley. Occasional along the sea shore. 105, 184. 2. Chenopodium L. Chenopodium ambrosioides L. Commonly cultivated as a medicinal, escapes, and naturalized at Cane Garden Bay. *Wormgrass." 329. Beta vulgaris L. (beetroot) is sometimes cultivated. Family 11. AMARANTHACEAE Achyranthes L. Achyranthes aspera L. Achyranthes indica (L.) Mill. Centrostachys indica (L.) Standley. Centrostachys aspera (L.) Standley. Perhaps the most annoying and troublesome weed on the island, covering abandoned cultivation very quickly. The form A. indica is perhaps more common on level ground in dry areas; and A. aspera occurs on level and sloping ground at higher elevations where there is more rainfall. They are distinct in appearance. “Man Catcher.” 673; Fishlock 209 (Pasea Hall Estate). Britt. & Wils. Alternanthera Forsk. Alternanthera ficoidea (L.) R. Br. ex R. & S. Achyranthes ficoidea (L.) Lam. Britt. & Wils. A. ficoidea var bettzickiana (Mich.) Baker. Achyranthes bettzickiana (Regel) Standley. Cultivated as an ornamental and occasionally escaping. A. paronychioides St. Hil. Achyranthes polygonoides (L). Lam. sensu Britt. & Wils. Alternanthera paronychoides St. Hil. Plentiful 1967] Tortola — D'Arcy 403 in sandy pastures and roadsides, disturbed areas. Perhaps more common in dry sections and near the sea. It is a very different plant from A. ficoidea var bettzickiana. Amaranthus L. Amaranthus caudatus L. Commonly cultivated and escaping in wet years. “Caterpillar.” 325. A. crassipes Schlechter. Fishlock 242 (Experiment Station). Britt. & Wils. A. dubius Mart. A troublesome weed in towns and gardens. Some- times cooked as greens. “Spinach.” Fishlock 134 (Experiment Station). Britt. & Wils. A. spinosus L. A troublesome streetside weed. 795; Fishlock 134 (Experiment Station). Britt. & Wils. Celosia L. Celosia argentea L. Celosia cristata L. Frequently cultivated as an ornamental. C. nitida Vahl. Fishlock 431 (Belmont); Eggers 3178 (in fruticetis maritimis). Britt. & Wils. Chamissoa H. B. K. Chamissoa altissima (Jacq.) H. B. K. Common in thickets along the sea coast. The tips of the spikes are quite pink. 240. Gomphrena L. Gomphrena decumbens Jacq. Gomphrena dispersa Standley. A com- mon weed. G. globosa L. Occasionally cultivated. “Bachelors Button.” Iresine P. Browne Iresine angustifolia Euph. Common in older secondary thickets, hill- sides. 581, 582, 583, 584, 585; Fishlock 61 (Great Mountain). Britt. & Wils. Philoxerus R. Brown Philoxerus vermicularis (L.) Beauv. Philoxerus vermicularis (L.) Nutt. A common halophyte, covering extensive areas at Sandy Point and on Beef Island. Britt. & Wils. Family 12. NYCTAGINACEAE Boerhaavia L. Boerhaavia diffusa L. Boerhaavia coccinea Mill. A common and trou- blesome weed, waste places and gardens. “Pigweed.” Fishlock 143 (waste places), Fishlock 144 (Experiment Station). Britt. & Wils. B. erecta L. Rare, but occurring in much the same way as B. diffusa. Fishlock 145 (Experiment Station). Britt. & Wils. Bougainvillea Comm. ex Juss. Bougainvillea spectabilis Willd. Seen growing wild near the sea at Fort Burt, probably rooting from refuse. Many forms and colors of this, B. peruviana H. & B. and B. glabra Choisy are grown as ornamentals. 404 Rhodora [Vol. 69 Commicarpus Standley Commicarpus scandens (L.) Standley. Seen near the sea at Coxheath and Little Apple Bay. 256 (Virgin Gorda), 231; Eggers 3173 (West End). Britt. & Wils. Mirabilis L. Mirabilis jalapa L. Cultivated and naturalized. Widespread in moist areas but not frequent. “Four O’Clock.” Fishlock 226 (Road Town). Britt. & Wils. Pisonia L. Pisonia fragrans Dumont-Cours. Torrubia fragrans (Dumont-Cours.) Standley. A common tree in forests and villages near sea level. “Water Mampo.” 284; Fishlock 402. Little & Wads. P. subcordata Sw. The common pasture tree of Tortola. Occurs at all elevations in forests and villages. *Mampo." Little & Wads., Britt. & Wils. Family 18. BATACEAE Batis L. Batis maritima L. A common halophyte on coastal flats. Plentiful at Slaney, Sandy Point and Beef Island. 219, 220, 221; Fishlock s.n. (saline soil). Britt. & Wils. Family 14. PHYTOLACCACEAE Petiveria L. Petiveria alliacea L. A common weed, often occurring with Rivina humilis and Achyranthes aspera. “Garlic Weed." 735, 736, 737, 738; Fishlock 133 (Experiment Station). Britt. & Wils. Phytolacea L. Phytolacca rivinoides Kunth & Bouche. A single immature plant growing in virgin forest at Arundel Estate was tentatively identi- fied as this species, 368. Rivina L. Rivina humilis L. Common in thickets and secondary growth, lower elevations. “Cat’s Blood." Fishlock 179 (Lower Estate). Britt. & Wils. Trichostigma A. Rich. Trichostigma octandrum (L.) H. Walter. Festooning in thickets in ghuts and secondary growth, lower elevations. Sometimes high climbing. Fishlock 161 (Road Town). Britt. & Wils. Family 15. AIZOACEAE Mollugo L. Mollugo verticillata L. 228 (Virgin Gorda). Britt. & Wils. Sesuvium L. Sesuvium portulacastrum L. A common halophyte on coastal flats. 726. Trianthema L. . Trianthema portulacastrum L. A common weed in the dry Long Look- East End section. Fishlock 487 (waste grounds). Britt, & Wils. 1967] Tortola — D’Arcy 405 Family 16. PORTULACACEAE Portulaca L. Portulaca oleracea L. A common weed in all areas. Often occurring as a halophyte forming part of the behind-the-beach cover. “Purselane.” Fishlock 193 (cultivated). Britt. & Wils. P. pilosa L. Waste places and forming large masses within the range of salt spray. “Rice Plant.” P. pilosa var. hortualis Bailey. Cultivated and escapes by vegetative means. “Jump-up-and-kiss-me.” P. quadrifida L. A weedy species chiefly of dry areas. The leaves are usually quite red. Fishlock 268 (Experiment Station). Britt. & Wils. P. rubricaulis H. B. K. Portulaca phaeosperma Urban. Widespread but rare along the south coast. 313. P. umbraticola H. B. K. Widely cultivated, but probably introduced in the last two or three years. 123. Talinum Adans. Talinum paniculatum (Jacq.) Gaertn. Rare. Seen only on rocks over the sea, Josia's Bay. 252, 253, 254; Fishlock 417 (Belleview). Britt. & Wils. T. triangulare (Jacq.) Willd. A low, yellow flowered form is common near the sea in western parts of Tortola. A taller (to 8 dm.), pink flowered form is cultivated and occasionally escapes. Fish- lock 233 (Experiment Station). Family 17. BASELLACEAE Basella L. Basella alba L. Basella rubra L. Ceylon spinach, grown for greens, often escapes but is not long persistent. Anredera Juss. Anredera leptostachys (Moq.) Steen. Boussingaultia leptostachys Moq. Festooning in thickets, moist districts up to about 300 m. 259; Fishlock 248 (Experiment Station). Britt. & Wils. Family 18. NYMPHACEAE Nymphaea L. Nymphaea ampla (Salisb.) DC. Castalia ampla Salisb. In the fresh- water pond below Windy Hill in Carrot Bay. This seems to be an intermediate between N. ampla and N. pulchella (DC.) Britt. as it has 60-70 stamens and 20-23 petals. 727, 728, 729. N. zanzibarensis Casp. Castalia zanzibarensis (Casp.) Britt. Rare- ly cultivated in lily ponds. Family 19. MENISPERMACEAE Cissampelos L. Cissampelos pareira L. Climbing in thickets and forests, moist dis- tricts. Sometimes employed as a medicinal. “Pistocle.” 489; Fishlock 306 (Lower Estate). Britt. & Wils. 406 Rhodora [Vol. 69 Family 20. ANNONACEAE Annona L. Annona glabra L. Common in salt marsh areas. “Pond Apple." Fish- lock 389 (Old Plantation). Little & Wads., Britt. & Wils. A. muricata L. Very common throughout the island, It is one of the dominant trees in the dry eastern district. Used medicinally, in confections and as pig food. “Soursop.” Little & Wads., Britt. & Wils. A. reticulata L. Occasionally cultivated along the north coast, and growing wild on the hillsides above Cane Garden Bay. Usually cooked before eaten. “Custard Apple." 331. Little & Wads. A. squamosa L. Common on open hillsides. It is one of the dominant trees in dry eastern districts. "Sugar Apple." Britt. & Wils. Family 21. LAURACEAE Cassytha L. Cassytha filiformis L. Only one small patch seen at Towers near Sandy Point, near the Eggers location. 346; Eggers 3174 (West End). Licaria Aubl. Licaria salicifolia (Sw.) Kosterm. Acrodiclidium salicifolium (Sw.) Griseb. Little & Wads. Nectandra Roland. Nectandra coriacea (Sw.) Griseb. Occasional in moist ghut hillsides. Seen in Johnson's and Purcell Ghuts. 204; Fishlock 5. Little & Wads. Ocotea Aubl. Ocotea floribunda (Sw.) Mez. In Sage Mountain forests. Fishlock 72 (Sage Mt.). Little & Wads., Britt. & Wils. O. leucoxylon (Sw.) Mez. Fairly common, Doty and Sage Mountain. 199, 724, 542, 705, 706, 756, 757; Fishlock 451 (Sage Mt.) ; Eggers $210, Little & Wads., Britt. & Wils. Persea Mill. Persea americana Mill. Persea persea (L.) Cockerell. Cultivated in moister districts, especially in mountain districts where it may escape. “Pear.” 703. Little & Wads. Family 22. PAPAVERACEAE Argemone L. Argemone mexicana L. An occasional weed and in roadsides. Fish- lock 10 (Experiment Station). Family 23. BRASSICACEAE Brassica L. Brassica integrifolia (West) O. E. Schultz. Rare. A clump of Brassica was seen growing wild at Hope one year: it may have been this species. Britt. & Wils. 1967] Tortola — D'Arcy 407 Cakile Mill. Cakile lanceolata (Willd.) O. E. Schultz. Rare along sandy sea coasts. Seen in numbers at Little Apple Bay. Fishlock 272 (Brandywine Bay, Fishlock 441 Long Bay). Britt. & Wils. Lepidium L. Lepidium virginicum L. Not common. Regularly seen along the Joe’s Hill-Soldier Hill road. Fishlock 223 (Road Town). Britt. & Wils. Brassica oleracea var. capitata (cabbage), Brassica caulorapa (kohl- rabi). Brassica chinense (Chinese cabbage), Brassica rapa (tur- nip) are all grown as market vegetables, Raphanus sativus (rad- ish) is sometimes grown for its large tuber-like roots which are boiled for stews. All are rarely spontaneous but not persistent. Family 24. CAPPARACEAE Capparis L. Capparis baducca L. Fishlock 412 (Huntum’s Ghut). Britt. & Wils. C. cynophallophora L. Common at lower elevations, especially by the sea. “Black Wattie." 796; Fishlock 96, Fishlock 476 (Harbours). Little & Wads., Britt. & Wils. C. flexuosa (L.) L. Common in thickets, lower elevations. Sometimes high climbing. “Bottle Wiss.” Fishlock 74 (Harbours). Britt. & Wils. C. indica (L.) Fawcett & Rendle. Common along the south coast east of Road Harbour but rare elsewhere. 88; Fishlock 97 (Harbours). Little & Wads., Britt. & Wils. C. portoricensis Urban. One tree seen on the Fort Hill and another at the top of the road leading down to Long Bay East. Cleome L. Cleome gynandra L. Gynandropsis gynandra (L.) Briq. Tall (to 1 m.) and very showy. Occasional in moist districts and on garbage „dumps. It covers over an acre of waste grounds at Little Apple Bay. “Massambee.” Used to prepare the West Indian vegetable dish “Calaloo.” 279. C. rutiosperma DC. Cleome ciliata Schum. & Thom. Occurring as a weed, usually prostrate or scandent. Fishlock 58, and Fishlock 159 (Experiment Station). Britt. Wils. C. speciosa H. B. K. Much like C. gynandra but less common. Seen by roadsides in Free Bottom. C. viscosa L. Cleome icosandra L. A common weed occurring almost everywhere on the island. “Wild Massambee." Fishlock 132 (Experiment Station). Britt. & Wils. Morisonia L. Morisonia americana L. Only one tree seen (Pleasant Valley), but reported to be common around Sea Cow Bay. “Wild Mesple." 154, 264. Britt. & Wils. 408 Rhodora [Vol. 69 Family 25. MORINGACEAE Moringa Adans. Moringa oleifera Lam. Moringa moringa (L.) Millspaugh. Moringa pterygosperma Gaertn. Occasional along roadsides. Cultivated and clearly naturalized. 811. Family 26. CRASSULACEAE Kalanchoé Adans. Kalanchoé brasiliensis Camb. Cultivated chiefly in dry eastern dis- tricts. 28, 30, 32. K. laxiflora Baker. Commonly cultivated. It is larger and showier than Kalanchoé pinnata and the tips of the petals have minute serrations, They may be varieties of the same species. 15, 16, 17, 18. K. pinnata Pers. Bryophyllum pinnatum (Lam.) Kurze. Common in all districts. “Love Plant.” 21, 22, 25, 26, 27, 29; Fishlock 359 (Huntum's Ghut). Britt. & Wils. K. somaliensis Hooker f. A widely cultivated trailer. Not seen to flower on Tortola. *Hen & Chickens." 607. K. verticillata Elliot. Occasionally cultivated and naturalized at East End. "Reptile." 19, 20. Family 27. ROSACEAE Chrysobalanus L. Chrysobalanus icaco L. Common at upper elevations; the most com- mon shrubby species around Doty. Harvested for pig food. "Coco Plum." Britt. & Wils. Species of Rosa (rose) are frequently cultivated around homes. Family 28A. LEGUMINOSAE — MIMOSACEAE Acacia Mill. Acacia farnesiana (L.) Willd. Vachellia farnesiana (L.) Wight & Arn. Little & Wads., Britt. & Wils. A. macracantha Humb. & Bonpl. ex Willd. Widespread and common, forming forests on sandy bottoms near the sea. “Casha.” 286. Britt. & Wils. A. muricata (L.) Willd. Senegalia muricata (L.) Britton & Rose. Seen on hillsides between Road Town and Fat Hog Bay. 9, 838; Fishlock 421 (Belleview). Britt. & Wils. A. riparia H. B. K. Senegalia westiana (DC.) Britton & Rose. A very troublesome weed species, often the first woody growth on abandoned hillside farming, and sometimes forming impenetrable thickets 20 m. across. “Catch-and-keep.” 241; Fishlock 183 (waste places). Britt. & Wils. A. tortuosa (L.) Willd. Fishlock 200 (Pasea Hall Estate). Little & Wads., Britt. & Wils. Adenanthera L. Adenanthera pavonina L. A few large trees along Long Bush ghut near sea level and a few along the Christophers-Doty road. May 1967] Tortola — D'Arcy 409 have been planted but some small seedlings were found in the second location. Little & Wils. Albizia Durazz. Albizia lebbeck (L.) Benth. Albizzia lebbeck (L.) Benth. A few tall trees (15 m.) at Carrot Bay. 71. Little & Wads. Calliandra Benth. Calliandra caracasana (Jacq.) Benth. Anneslia portoricensis (Jacq.) Britt. Occasionally throughout the island but common in dry eastern districts. 153, 448; Fishlock 581 (Edney's) ; Eggers 3185. Britt. & Wils. Desmanthus Willd. Desmanthus depressus Humb. & Bompl. Acuan depressum (H. & B.) Kuntze. Widespread. May not be distinct from Desmanthus virgatus. Britt & Wils. D. virgatus (L.) Willd. Acuan virgatum (L.) Medic. Throughout the island but common only in dry eastern districts, where it reaches a height of 2 m. 323; Fishlock 205. Britt. & Wils. Inga Mill. Inga fagifolia (L.) Willd. Inga lauwrina (Sw.) Willd. Common at middle and higher elevations, The pods are sometimes cooked as a vegetable. “Spanish Oak." 213. Little & Wads., Britt. & Wils. Leucaena Benth. Leucaena leucocephala (Lam.) DeWit. Leucaena glauca (L.) Benth. Common secondary growth on abandoned farming. “Wild Tama- rind." Fishlock 177 (Lower Estate). Little & Wads., Britt. & Wils. Mimosa L. Mimosa ceratonia L. Lomoplis ceratonia (L.) Raf. Common on moist hillsides. Large colonies were seen along the Road Town-Meyers road. Fishlock 267 (Huntum's Ghut). Britt. & Wils. M. pudica L. Seen only once in a pasture, Experiment Station. Pithecellobium Bentham in Mart. Pithecellobium unguis-cati (L.) Bentham in Mart. Common, especial- ly so at East End and at Flamingo Pond. The Tortola plants have red seed arils, although plants with white seed arils were seen in St. Thomas. Pongamia Vent. Pongamia pinnata (L.) Merr. One fine tree on Government House grounds. 437. Prosopis L. Prosopis juliflora (Sw.) DC. Neltuma juliflora (Sw.) Raf. Wide- spread and common, forming forests on sandy bottoms near the sea. Occurs much as Acacia macracantha but is more common in eastern parts of the island. “Casha.” Fishlock 392 (Old Plan- tation). Little & Wads., Britt. & Wils. 410 Rhodora [Vol. 69 Samanea Merrill Samanea saman (Willd.) Merrill. Pithecellobium saman (Jacq.) Benth. Planting of this tree as a pasture tree has had the strong encouragement of the Agriculture Department and it is becom- ing widespread. 12. Little & Wads. Family 28B. LEGUMINOSAE — CAESALPINIACEAE Bauhinia L. Bauhinia monandra Kurz. Caspareopsis monandra (Kurz) Britton & Rose. A common ornamental tree, occasionally spontaneous. “Napoleon’s Hat.” 234. Little & Wads. B. purpurea L. Only one tree seen; Treasure Isle Hotel grounds. “Poor Man’s Orchid.” 391. B. tomentosa L. Olvesia tomentosa (L.) Welw. A few shrubs, culti- vated and escaped. Pasea Hall Estate. 388. Caesalpinia L. Caesalpinia bonduc Roxb. Guilandina bonduc L. Children at East End were seen playing with large yellow seeds like those of this species but plants were not seen. C. crista L. Guilandina crista (L.) Small. Forming thickets along the south coast. Seeds were grey in all stages of development. Used medicinally. “Nickers.” “Ram Goat Bush." C. melanosperma (Eggers) Urban. Guilandina melanosperma Eggers. Reported only by Eggers. Endemic to Tortola, St. Croix and Mona. C. pulcherrima (L.) Sw. Poinciana pulcherrima L. Commonly culti- vated and seemingly naturalized in Baughers Bay. “Dwarf Poinciana.” Cassia L. Cassia alata L. Herpetica alta Raf. A few scattered plants and occasionally cultivated. Seen wild at Hannah and Long Look. "Sasparilla." 739; Fishlock 396 (Old Plantation). Little & Wads. . antillana (Britt. & Rose) Alain, Chamaefistula antillana Britton & Rose. Plentiful and showy at higher elevations and around Sage Mountain forest. Not seen below about 80 m. 147; Fish- lock 282 (Leonards). Britt. & Wils. C. bacillaris L. f. Chamaefistula bacillaris (L. f.) G. Don. Not seen. The Eggers specimen should be checked. Eggers 3188. C. bicapsularis L. Adipera bicapsularis (L.) Britton & Rose. Com- mon in all districts. Used medicinally. “Pissy bed." Fishlock 291 (Experiment Station). Britt, & Wils. C. fistula L. Two or three small specimens at the Experiment Station. C. glandulosa var. swartzii (Wickstr. MacBride. Chaemaecrista swartzii (Wickstr.) Britt. & Rose. Common, especially at higher elevations. “Scampesina.” 262; Fishlock 283 (Leonards). Britt. & Wils. e 1967] Tortola — D'Arcy 411 C. grandis L. f. One specimen on Pasea Hall Estate and another be- hind the Road Town Police Station. Although both trees flower well, they have very seldom produced seed. C. obtusifolia L. Emelista tora (L.) Britton & Rose. Common after dry spells. 297, 312; Fishlock 325 (Experiment Station). Britt. & Wils. C. occidentalis L. Ditremexa occidentalis (L.) Britton & Rose. A very common weed around towns. “Stinking Weed." 352, Fishlock 191 (Lower Estate). Britt. & Wils. Delonix Raf. Delonix regia (Bojer) Raf. Widely cultivated. Sometimes spontaneous but probably not persistent. “Flamboyant.” Little & Wads. Haematoxylon L. Haematoxylon campechianum L. Seen only in the Sea Cow Bay-Duff Bottom area, where it is common. “Dyewood Tree." 156; Fish- lock 36 (Sea Cow Bay). Britt. & Wils. Hymenaea L. Hymenaea courbaril L. Scattered specimens in moist districts. Chil- dren seem to enjoy the dry powdery pods. "Locust." Little & Wads., Britt. & Wils. Parkinsonia L. Parkinsonia aculeata L. Along south coast roadsides and plentiful at Towers. 812; Fishlock 396 (Old Plantation). Little & Wads., Britt. & Wils. Peltophorum Walp. Peltophorum inerme (Roxb.) Naves. Peltophorum ferrugineum (Dene.) Benth. Peltophorum dubium (Spreng.) Taub. Pelto- phorum pterocarpum (DC.) Baker. A number of fine trees in the Road Town area. 233. Tamarindus L. Tamarindus indica L. Common throughout the island, large trees occurring in moist bottoms. Because of its resistance to hurri- cane, it is a common surveying marker in land deeds. 230. Little & Wads., Britt. & Wils. Family 28C. LEGUMINOSAE — FABACEAE Abrus Adans. Abrus praecatorius L. Abrus abrus (L.) W. F. Wright. Common almost everywhere. Tortolians seem generally unaware of the poisonous nature of the seeds and they are common children’s playthings. “Jumbie Bead.” 298, 829; Fishlock 250 (Road Town). Britt. & Wils. Aeschynomene L. Aeschynomene americana L. Only one plant seen on a donkey path, Sage Mt. 541. Britt. & Wils. A. sensitiva Sw. Fishlock 126 (Experiment Station). Britt. & Wils. 412 Rhodora [Vol. 69 Alysicarpus Desv. Alysicarpus vaginalis (L.) DC. A common weed throughout the island. 357, 358, 359, 360, 361; Fishlock 265 (Slaney Point). Britt. & Wils. Andira Lam. Andira inermis (Wright) DC. Andira inermis H. B. K. A very common tree throughout the island. “Pictod.” Fishlock 111 (Experiment Station). Little & Wads., Britt. & Wils. Arachis L. Arachis hypogaea L. Spontaneous near candy shops, probably from seed dropped by children. Not known to be cultivated at the present time. Cajanus DC. Cajanus cajan (L.) Millsp. Cajan cajan (L.) Millsp. Several vari- eties are widely cultivated. "Pigeon Pea." 454. Britt. & Wils. Canavalia DC. Canavalia maritima (Aubl. Thou. Canavali maritima (Aubl.) Thou. Plentiful on sandy sea coasts. Fishlock 175 (Lower Estate) as C. lineata. Britt. & Wils. Centrosema (DC.) Benth. Centrosema virginianum Benth. Bradburya virginiana (L.) Kuntze. Common throughout the island, roadsides and thickets. Fishlock 160 (Lower Estate). Britt. & Wils. Chaetocalyx DC. Chaetocalyx scandens (L.) Urban var. pubescens (DC.) Rudd. One small colony seen by the roadside, Fish Bay. This is a stray from the Leeward Islands. 316. Clitoria L. Clitoria ternatea L. Widespread but not plentiful. 299, 722; Fishlock 195 (Experiment Station). Britt. & Wils. Cracca L. Cracca caribaea (Jacq.) Benth. Benthamantha caribaea (Jacq.) Kuntze. Only one plant seen beside road leading into Paraquita holding ground. 414; Fishlock 393 (Old Plantation). Britt. & Wils. Crotalaria L. Crotalaria incana L. Common by roadsides and as a weed in culti- vated lands. 452; Fishlock 195 (Experiment Station). Britt. & Wils. C. retusa L. Thickets and pastures, moist districts. 534; Fishlock 30 (Experiment Station). C. verrucosa L. A common weed, mainly at lower elevations. 404; Fishlock 308 (Purcells). Britt. & Wils. Dalbergia L. f. Dalbergia ecastophyllum (L.) Taub. Ecastophyllum ecastophyllum (L.) Britton. Plentiful in the salt pond at Cane Garden Bay. Also seen at Long Bay East. 10, 372. Britt. & Wils. 1967] Tortola — D’Arcy 413 Desmodium Desv. Desmodium axillare (Sw.) DC. Meibomia axillaris (Sw.) Kuntze. A large colony in a clearing, Sage Mt. forest. A procumbent vine to 4 m. long. 762, 763. D. canum (Gmel.) Schinz & Thell. Meibomia supina (Sw.) Britton. A common weed throughout the island. 7. D. glabrum (Mill) DC. Meibomia mollis (Vahl) Kuntze. Rare, seen on Kingstown Hill. 741; Eggers 3177. Britt. & Wils. D. procumbens (Mill) Hitche. Meibomia procumbens (Mill) Britt. In thickets and waste places. 405, 406, 407, 408. D. scorpiurus (Sw.) Desv. Meibomia scorpiurus (Sw.) Kuntze. Fish- lock 150 (Experiment Station). Britt. & Wils. D. tortuosum (Sw.) DC. Meibomia purpurea (Mill. Vahl. Desmo- dium purpureum (Mill.) Fawcett & Rendle. Common in pastures. 8. Britt. & Wils. D. triflorum (L.) DC. Sagotia triflora (L.) Duch. & Walp. A com- mon weed in most parts. 362, 363, 364, 365, 366. Dolichos L. Dolichos lablab L. Two forms of this species occur: a small (6 dm.) bushy plant with white or rarely blue flowers is cultivated and sparingly naturalized, *Bonavist Bean;" and a long (to 10 m.) vine closely resembling Pueraria lobata with large erect purple racemes which is rampant on trees and ruins in moist districts. Large masses seen in Long Bush Ghut and on Chalwell Estate. 157 (white form), 564, 565, 566 (purple form); Fishlock 167, (Experiment Station). Britt. & Wils. Erythrina L. Erythrina corallodendrum L. Rarely cultivated. 155. Little & Wads. E. sp. Two trees, Experiment Station, unarmed, with white flowers and seeds, the stigma not resembling the anthers. *White Im- mortelle." 725. E. variegata var. orientalis (L.) Merrill. Erythrina indica Lam. This tree, with vicious spines, red flowers and stigmas resembling the anthers is a favorite ornamental, growing rapidly wherever it can be protected from cattle. “Immortelle.” 724. Galactia P. Browne Galactia dubia DC. Infrequent along the south coast, perhaps more plentiful towards the west. 460. G. eggersii Urban. Endemic to Tortola, St. John and St. Thomas. Not seen. Eggers 3179. G. striata (Jacq.) Urban. Infrequent along the south coast. 257. Britt. & Wils. Gliricidia Kunth Gliricidia sepium (Jacq.) Kunth. Gliricidia sepium (Jacq.) Steud. Very common in the Road Harbour area. Much used for fence- posts, which sprout. May be naturalized. 159, Little & Wads. 414 Rhodora [Vol. 69 Indigofera L. Indigofera suffruticosa Mill. Common, roadsides and recently trimmed pastures. 575; Fishlock 142, and Fishlock 164 (Experiment Sta- tion). Britt. & Wils. Macroptilium (Bentham) Urban Macroptilium lathyroides (L.) Urban. Phaseolus lathyroides L. Widespread but rather rare. Seen in Baughers Bay and Sea Cow Bay. Fishlock 289 (Experiment Station). Britt. & Wils. Phaseolus L. Fhaseolus lunatus L. A small seeded white-flowered variety is culti- vated and sparingly naturalized. Pictetia DC. Pictetia aculeata (Vahl) Urban. Common in eastern parts of the island. “Thumb Tack.” 14, Fishlock 123 (Beef Island). Little & Wads. Piscidia L. Piscidia piscipula (L.) Sarg. Icthyomethia piscipula (L.) Hitch. Common near the sea, especially along the south coast. “Dog- wood." 4, 5; Fishlock 41 (Harbours). Britt. & Wils. Rhynchosia Lour. Rhynchosia minima (L.) DC. Dolicholus minimus (L.) Medic. Com- mon on fences and in thickets on hillsides. 425, 785; Fishlock 165 (Experiment Station). Britt. & Wils. R. reticulata (Sw.) DC. Dolicholus reticulatus (Sw.) Millsp. Fishlock 53 (Great Mountain), Fishlock 162 (Experiment Station). Britt. & Wils. Sabinea DC. Sabinea florida (Vahl) DC. Usually in ghutsides at higher elevations, although occasionally seen near sea level in moist districts. Perhaps the most spectacular species in the Tortola flora, and certainly worthy of cultivation. When it blooms in January or March the entire shrub is a mass of blue or mauve. *Wattapania." Fishlock 35 (Sage Mt.) Britt. & Wils. Sesbania R. Brown Sesbania grandifiora (L.) Pers. Agati grandiflora (L.) Desv. Both pink and white forms are rarely cultivated. Sophora L. Sophora tomentosa L. Britt. & Wils. Stylosanthes Sw. Stylosanthes hamata (L). Taubert. Forming large masses at Chapel Hil (East End), Pasea Hall Estate and Fort Burt. Generally near the sea. 196, 197, 198. Britt. & Wils. Tephrosia Pers. Tephrosia cinerea (L.) Pers. Occasional near the sea along the south coast. Seen at Welk Point and on Wickham's Cay. 455, 456, 457, 458 459; Fishlock 259 (Slaney Point), Fishlock 425 (West End). Britt. & Wils. 1967] Tortola — D'Arcy 415 T. aff. senna H. B. K. det. Proctor. Hillsides and pastures. Common. Proctor notes longer, coarser hairs on the pods than T. senna. 483, 484, 485, 486, 487. Teramnus P. Browne Teramnus labialis (L. f.) Spreng. Commonly festooning on trees and sometimes completely covering fences and shrubs. 637; Fishlock 153, and Fishlock 276, Experiment Station). Britt. & Wils. Vigna Savi Vigna antillana (Urban) Fawcett & Rendle. Although a conspicuous plant, it was not seen by the author on Tortola. Britt. & Wils. V. luteola (Jacq.) Benth. Vigna repens (L.) Kuntze. In occasional patches along the sea coasts. Seen at Purcell, Cane Garden Bay and in Road Town. 332; Fishlock s.n. Britt. & Wils. Phaseolus vulgaris L. (red bean) and Vigna unguiculata (L.) Walp. (black-eye pea) are occasionally grown for food and Lathyrus odoratus L. (sweet pea) is rarely grown for ornament from im- ported seed. Family 29. OXALIDACEAE Oxalis L. Oxalis corniculata L. Xanthoxalis corniculata (L.) Small. Common in cultivated fields, Doty-Sage Mountain area. 205, 206, 207. Britt. & Wils. Family 30. ERYTHROXYLACEAE Erythroxylum P. Browne Erythroxylum rotundifolium Lunan. Erythroxylon brevipes DC. Very plentiful in light woods, moist districts. When in flower, the woods are full of scent and bees. Fishlock 156 (Lower Estate). Species of Impatiens of the Family BALSAMINACEAE are rarely grown from imported seed under the name of “Lady’s Slipper.” Family 31. ZYGOPHYLLACEAE Guaiacum L. Guaiacum officinale L. At one time a plentiful wood, only planted trees are to be seen on Tortola at the present time. A number of wild trees are growing on Tamarind Point, Beef Island. “Lignum Vitae." 42. Kallstroemia Scop. Kallstroemia maxima (L.) T. & G. Fishlock 131 (Experiment Sta- tion). Britt. & Wils. K. pubescens (G. Don) Dandy. Kallstroemia caribaea Rydberg. A common and troublesome weed in all parts of the island. Fish- lock 131A (Experiment Station). Britt. & Wils. Tribulus L. Tribulus cistoides L. Seen only in three places: Chapel Hill, Little Apple Bay, and at Towers where it covers large areas. “Cockle- horn.” 696, 697, 698, 699; Fishlock 437 (West End); Eggers 3176. Britt. & Wils. 416 Rhodora [Vol. 69 Family 32. MALPIGHIACEAE Bunchosia Rich. ex Juss. Bunchosia glandulosa (Cav.) DC. Common at the edges of thickets and in light woods, moist districts. Fishlock 404 (Kingstown). Britt. & Wils. Byrsonima Rich. ex Juss. Byrsonima coriacea var. spicata (Cav.) Nied. Byrsonima spicata (Cav.) DC. Open pastures, north side of Sage Mountain. 197, 704; Fishlock 380 and Fishlock 450 (Sage Mt.). Little & Wads., Britt. & Wils. Heteropteris Kunth Heteropteris purpurea (L.) Kunth. Banisteria purpurea L. Festooned in thickets, Jean Hill and Fish Bay. 118, 809, 810. Britt. & Wils. Malpighia L. Malpighia glabra L. One bush seen at Pleasant Valley was destroyed by road building before a specimen could be taken. M. punicifolia L. A few planted shrubs, Baughers Bay and Kings- town Hill. “Barbados Cherry.” Stigmaphyllon A. Juss. Stigmaphyllon diversifolium A. Juss. Stigmaphyllon ledifolium (H. B. K.) Small. 793 (Baughers Bay). S. periplocifolium (Desf.) Juss. Stigmaphyllon lingulatum (Poir.) Small. Fishlock 196 (Experiment Station). Britt. & Wils. Species of Stigmaphyllon were seen commonly along roadsides at lower elevations and the sea coast, but the species were not dis- tinguished. Family 33. RUTACEAE Amyris L. Amyris elemifera L. Occasional along the south coast and in eastern sections. 806 (Camanoe), 463 (Peter Island). Citrus L. Citrus aurantifolia (L.) Swingle. Citrus aurantifolia (Christm.) Swingle. Widely cultivated and naturalized almost throughout the island. The Tortola fruits are small but the juice seems far more concentrated than that of larger fruits grown on other islands. “Lime.” Little & Wads. C. aurantium L. A few cultivated trees, Road Town. “Bitter Orange.” C. limon (L.) Burm. Citrus limonum Risso. Citrus limonia Osbeck. Only one or two trees at East End. The fruit was not examined. C. paradisi Macfayden. Citrus maxima (J. Burm.) Merrill. Citrus grandis (L.) Osbeck. A few cultivated trees, Road Town. “Grapefruit.” C. sinensis (L.) Osbeck. Occasionally cultivated, moist districts. Tor- tola oranges have a thick, rough or tubercled rind; and have a hint of lemon flavor. “Sweet Orange.” Little & Wads. 1967] Tortola — D'Arcy 417 Murraya Koenig ex L. Murraya exotica L. Chalcas exotica (L.) Millsp. Often cultivated for its fragrant flowers and long naturalized at the sides of ghuts. Fishlock 189 (Experiment Station). Britt. & Wils. Triphasia Lour. Triphasia trifolia (Burm. f.) P. Wilson. Common throughout the island, very numerous on hillsides above Road Town, and forming hedges at Little Apple Bay. “Sweet Lime." 158. Britt. & Wils. Zanthoxylum L. Zanthoxylum martinicense (Lam.) DC. Moist districts at all eleva- tions. “Yellow Prickle." Several trees of Zanthoxylum on Cane Garden Bay hillsides may not be this species. Fishlock 480 (Sage Mt.); Eggers 3227. Little & Wads., Britt. & Wils. Family 34. SIMAROUBACEAE Picramnia Sw, Picramnia pentandra Sw. Fishlock 113 (Huntum's Ghut), Fishlock 284 (Purcells). Little & Wads., Britt. & Wils. Suriana L. Suriana maritima L. On reefs and coastal sands, south side of Tortola. Seen on Road Reef, in Fish Bay, and in Hog Valley Bay. 104. Family 35. BURSERACEAE Bursera Jacq. Bursera simaruba (L.) Sarg. Lower elevations throughout the island, but only at sea level does the bark fully display its characteristic copper colour. A fine grove is to be seen at Cane Garden Bay. Regularly sprouting from fenceposts made of the branches. “Turpentine Tree." 52; Fishlock 394 (Old Plantation). Little & Wads., Britt. & Wils. Tetragastris Gaertn. Tetragastris balsamifera (Sw.) Kuntze. In ghuts, Sage Mountain forests. 665. Family 36. MELIACEAE Azadirachta A. Juss. Azadirachta indica A. Juss. A number of very fine specimens on the western edge of Road Town, some of them probably spontaneous. “Neme Tree." 6. Cedrela P. Browne Cedrela odorata L. Cedrela mexicana M. J. Roem. Widespread but not common, “Red Cedar.” Melia L. Melia azedarach L. Common throughout the island, but more so at lower elevations. When the tree is cut down it usually sprouts again. Used for tool handles. “Lilac.” 70; Fishlock 169 (copses). Little & Wads. 418 Rhodora [Vol. 69 Swietenia Jacq. Swietenia mahagoni (L.) Jacq. Swietenia mahagoni Jacq. Three or four thirty foot trees at the Experiment Station. “West Indian Mahogany." Little & Wads. Over the past few years, many acres of S. mahagoni, S. macrophylla King and a hybrid have been planted by the Agricultural Department in an attempt to restore something of the former rain forest on Sage Mountain peak. At time of writing, these were doing well and had reached a height of about 1 m. Family 37. EUPHORBIACEAE Acalypha L. Acalypha chamaedrifolia (Lam.) Muell.-Arg. Britt. & Wils. A. hispida Burm. f. Frequently cultivated. A. wilkesiana Muell.-Arg. Frequently cultivated. ‘“Copperleaf,” "Spotted Heart." 288. Adelia L. Adelia ricinella (L.) Britt. Ricinella ricinella (L.) Britt. Fishlock 14 (Clifton Hall), Fishlock 38 (Coastal hills). Britt. & Wils. Alchornea Sw. Alchornea latifolia Sw. Fishlock 95 (Sage Mt.). Britt. & Wils. Argythamnia Sw. Argythamnia candicans Sw. Plentiful in Belmont Bay; not seen else- where. 304, 305, 306. Britt. & Wils. A. fasciculata (Vahl) Muell.-Arg. Ditaxis fasciculata Vahl. Rare. Seen only along the main road at Pasea Hall Estate and at Flamingo Pond. 409, 668; Fishlock 395 (Old Plantation). Britt. & Wils. Breynia Forst. Breynia nivosa (W. G. Smith) Small. Although probably not spon- taneous, it may be seen long persistent on former house sites and in old ruins. Not common. 354. Chamaesyce S. F. Gray Chamaesyce articulata (Aubl) Brit. Chamaesyce vahlii (Willd.) Wilson. Euphorbia articulata Aubl. Rather rare, seen growing along roadsides at lower elevations. 342; Fishlock 263 (Slaney Point). Britt. & Wils. C. blodgettii (Englm.) Small. Euphorbia blodgettii Englm. The only record for this plant in the Lesser Antilles. Fishlock 222 (Lower Estate). Britt. & Wils. C. buxifolia (Lam.) Small. Euphorbia mesembrianthemifolia Jacq. Commonly growing on or immediately behind beaches, 62; Fish- lock 262 (Slaney Point). Britt. & Wils. C. hirta (L.) Millsp. Euphorbia hirta L. A troublesome weed every- where, but less common than C. hyssopifolia. 77; Fishlock 235 (Experiment Station). Britt. & Wils. C. hypericifolia (L.) Millsp. Euphorbia hypericifolia L. Fishlock 9 1967] Tortola — D'Arcy 419 and 144 (Experiment Station). Britt. & Wils. C. hyssopifolia (L.) Small. Euphorbia hyssopifolia L. A troublesome weed everywhere. 78. C. prostrata (Ait.) Small. Occurs with C. serpens, but not so plentiful. 79. C. serpens (H. B. K.) Small. Forming a mat on road verges, espe- cially common east of Road Harbour. Although perennial, it seems to disappear completely in dry seasons. 80. Codiaeum Juss. Codiaeum variegatum (L.) Blume. One of the most common orna- mentals. The many forms seen suggest this is a complex of species. “Croton.” Croton L. Croton astroites Dryand. Occurring with C. rigidus but less common. “Black Maran.” C. astroites and C. rigidus are reported to hybridize on other islands, but no evidence was seen of this here. 46; Fishlock 245 (Pasea Hall Estate). Britt. & Wils. C. betulinus Vahl. Very widespread but rather rare. 821; Fishlock 44 (Sea Cow Bay). Britt. & Wils. C. lobatus L. A common weed in Road Town streets and occasional in moist districts. 34; Fishlock 277 (Experiment Station). Britt. & Wils. C. ovalifolius Vahl. Common on dry hillsides, roadsides and in towns. The form seen commonly at Long Look has larger, softer and less pubescent leaves than forms seen in the Road Harbour area and on Beef Island. 797, Fishlock 125 (Experiment Station). Britt. & Wils. C. rigidus (Muell.-Arg.) Britton. One of the most troublesome and ubiquitous woody weeds on the island, one of the dominant species in rather stable association on abandoned hillside pastures. “Maran.” 45; Fishlock 39 (Sea Cow Bay). Britt. & Wils. Dalechampia L. Dalechampia scandens L. Fairly common, often in dry areas. Britt. & Wils. 803. Euphorbia L. Euphorbia heterophylla (L.) A. Rich. Poinsettia heterophylla (L.) Kl. & Garke. A widespread and troublesome weed, Fishlock 2 (Experiment Station), Fishlock 213 (Road Town). Britt. & Wils. E. lactea Haw. Rarely cultivated. Several fine abandoned trees on Peter Island. 672. E. milii Ch. des Moulins. Euphorbia splendens Hook. Occasionally cultivated. “Crown of Thorns.” 786. E. petiolaris Sims. Aklema petiolare (Sims) Millsp. Occurs mainly in western parts of the island. 83; Fishlock 42 (Sea Cow Bay). Little & Wads., Britt. & Wils. E. pulcherrima Willd. Poinsettia pulcherrima (Willd.) Graham. Very commonly planted at lower elevaticns. 420 Rhodora [Vol. 69 E. tirucalli L. One small tree at Doty and an avenue of large trees beside the sea at Towers. 209. Gymnanthes Sw. Gymnanthes lucida Sw. Eggers 3181. Hippomane L. Hippomane mancinella L. Rare, growing right at the edge of the sea. Because of the appealing appearance and smell of the fruit, this is the most dangerous poisonous plant growing in the West Indies. Britt. & Wils. Hura L. Hura crepitans L. Common in woods at lower elevations. “Sandbox Tree." Little & Wads. Jatropha L. Jatropha curcas L. Curcas curcas (L.) Britt. & Millsp. Rare. A large stand seen at Carrot Bay and a few plants in pastures at Free Bottom. “Physic Nut.” J. gossypifolia L. Adenoropiwm gossypifolium (L.) Pohl. A common weed, especially in disturbed areas near the sea. “Wild Physic Nut." 680; Fishlock 192, 428 (waste places). Britt. & Wils. J. integerrima Jacq. Jatropha hastata Jacq. Occasionally cultivated. 141. J. multifida L. Adenoropium multifidum (L.) Pohl. Increasingly cultivated but not seen to escape. Manihot L. Manihot utilissima Pohl. Manihot manihot (L.) Cockerell. Widely cultivated as a market crop, and occasionally spontaneous. “Cas- sava." Britt. & Wils. Margaritaria L. Margaritaria nobilis L. f. Phyllanthus nobilis (L. f.) Muell.-Arg. Fishlock 423 (Belleview); Eggers 3229. Little & Wads., Britt. & Wils. Pedilanthus Poiteau Pedilanthus tithymaloides (L.) Poit. Pedilanthus padifolius (L.) Poit. Pedilanthus angustifolius Poit. Four subspecies grow on Tortola: Pedilanthus tithymaloides ssp. angustifolius (Poit.) Dressler; ssp. padifolius (L.) Dressler; ssp. parasiticus (Kl. & Gke.) Dressler; ssp. tithymaloides. All four have been commonly seen both in cultivation and unattended, and it is difficult to say whether they are spontaneous. They are long persistent after cultivation, perhaps over thirty years. Phyllanthus L. Phyllanthus acidus (L.) Skeels. Cicca disticha L. Cicca acida (L.) Merr. Occasionally cultivated for fruit. Seen only in Road Town, Long Bush Ghut and Sea Cow Bay. Some of these trees were certainly spontaneous. “Gooseberry Tree." 535; Fishlock 350 (Road Town). Little & Wads., Britt. & Wils. 1967] Tortola — D'Arcy 421 P. amarus Schum. & Thonn. This and P. niruri are very much alike and were not distinguished during the author's stay. One or both of them were a common weed almost everywhere. 523, 524, 525, 526, 527. P. niruri L. Fishlock 147 (Experiment Station). Britt. & Wils. Ricinus L. Ricinus communis L. Common in waste places near the sea. 345; Fishlock 190 (Lower Estate). Britt. & Wils. Tragia L. Tragia volubilis L. Common throughout the island. Fishlock 196 (Experiment Station). Britt. & Wils. Unidentified species of Sapium were recorded by Eggers (3204), and by Little & Wads. Family 38. ANACARDIACEAE Anacardium L. Anacardium occidentale L. Cultivated in dry areas, especially eastern districts. “Cherry.” 145. Little & Wads., Britt. & Wils. Comocladia P. Br. Comocladia dodonaea (L.) Urban. In thickets and woods, south side of Tortola. This plant produces a strong “poison ivy" reaction. “Thumb Tack," “Poison Ash." Fishlock 103 (harbours). Britt. & Wils. Mangifera L. Mangifera indica L. Widely cultivated and naturalized in moist dis- tricts. “Mango.” Little & Wads., Britt. & Wils. Spondias L. Spondias mombin L. Scattered trees in moist districts. Seen at Sea Cow Bay and Belleview Ghut. “Ghut Plum," *Hog Plum." Fish- lock 414 and 443 (Experiment Station). Little & Wads., Britt. & Wils. S. purpurea L. Rarely cultivated, a few specimens in Baughers Bay. Eggers 3225 (in forest, Carrot Bay). *Jamaica Plum." S. sp. Rarely cultivated, a fine specimen in front of Treasure Isle Hotel and a few in Baughers Bay. “Paradise Plum.” Family 39. AQUIFOLIACEAE Tlex L. Ilex urbania Loes. Fishlock 78 (Sage Mountain). Britt. & Wils. Family 40. CELASTRACEAE Cassine L. Cassine xylocarpa Vent. Elaeodendron xylocarpum (Vent.) DC. Seen only near sea level between Road Town and Fish Bay. “Poison Tree." 168; Fishlock 255 and 475 (Slaney Point). Crossopetalum Crossopetalum rhacoma Crantz. Rhacoma crossopetalum L. Dry hillsides. 774 (Camanoe); Fishlock 266 and 427. Britt. & Wils. 422 Rhodora [Vol. 69 Schaefferia Jacq. Schaefferia frutescens Jacq. Dry forests and thickets at lower eleva- tions. Plentiful at Flamingo Pond. 679; Fishlock 257 (Slaney Point), Fishlock 391 (Old Plantation). Britt. & Wils. Family 41. sTAPHYLEACEAE Turpina Vent. Turpina paniculata Vent. Fishlock 453 (Sage Mountain). Little & Wads., Britt. & Wils. Family 42. SAPINDACEAE Cardiospermum L. Cardiospermum corindum L. Rare, seen at Flamingo Pond and at the top of Joe's Hill road. 96, 180. C. halicacabum L. Eggers 3180. Cupania L. Cupania americana L. A few trees in upper Huntum's Ghut and plentiful in Sage Mountain forests. 597. Melicoccus Browne Melicoccus bijugatus Jacq. Melicocca bijuga L. A most plentiful tree, pastures and woods at lower elevations. "Genip." Little & Wads. Paullinia L. Paullinia pinnata L. Wet pastures, mountain districts; mostly on the north side of Chalwell, and Sage Mountain. “Guard Wiss." 260, Fishlock 490 (s.1.). Britt. & Wils. P. plumieri Triana & Planchon. Reported by Brit. & Wils. on the basis of an incomplete collection by Shafer (1158). Serjania Plum. Serjania polyphylla (L.) Radlk. Common in moist districts, especially in hillside thickets. Fishlock 69, Fishlock 258 (Belleview). Family 43. SABIACEAE Meliosma Blume Meliosma herberti Rolfe. Britt. & Wils. Family 44. RHAMNACEAE Colubrina Rich. Colubrina arborescens (Mill.) Sarg. Colubrina Colubrina (Jacq.) Millsp. Common at lower elevations, especially behind beaches on the north side of the island. 61; Fishlock 98 (Harbours). Little & Wads., Britt. & Wils. C. reclinata (L’Hér.) Brongn. Years ago the bark of this tree was exported to other islands and to Europe, but it is rare now, and Tortola must import part of its requirements from Jost Van Dykes. “Mabi.” $07 (Camanoe), 191; Fishlock 251 (Duff Bot- tom), Fishlock 404a (Kingstown). Little & Wads. Britt. & Wils. 1961] Tortola — D'Arcy 423 Gouania Jacq. Gouania lupuloides (L.) Urban. Common in thickets, moist districts. 285; Fishlock 243 (Lower Estate). Britt. & Wils. Krugiodendron Urban Krugiodendron ferreum (Vahl) Urban. Occasional in thickets, mostly hillsides over the sea. “Ironwood.” 195. Little & Wads., Britt. & Wils. Reynosia Griseb. Reynosia uncinata Urban. Britt. & Wils. Ziziphus Torn. ex L. Ziziphus reticulata (Vahl) DC. Sarcomphalus reticulatus (Vahl) Urban. Eggers 3182. Little & Wads., Britt. & Wils. Z. rignonii Delp. Sarcomphalus domingensis (Spreng.) Kr. & Urban. Britt. & Wils. Family 45. VITACEAE Cissus L. Cissus caustica Tussac. Climbing on large trees in open pastures, north side of Sage Mountain, 669. C. sicyoides L. Common on fences, trees and in thickets. “Pudding Vine.” A variant with red petioles and calices, orange pistils and stamens, and pink or red fruits occurs on hillsides, Cane Garden Bay, Ballast Bay and Sophie Bay. 53, 107, 108; Fishlock 188 (Lower Estate). Britt. & Wils. C. trifoliata (L.) L. Common near the sea. In moist parts of West End and Belmont it covers all other vegetation at times, 91; Fishlock 407 (Sophie Bay). Britt. & Wils. Vitis L. Vitis vinifera L. Occasionally cultivated for ornament and food. Family 46. TILIACEAE Corchorus L. Corchorus aestuans L. A common weed in moist years. 93. Britt. & Wils. C. hirsutus L. A common weedy shrub in pastures and near the sea, chiefly in eastern districts. 190, 192; Fishlock 47 (Sea Cow Bay). Britt. & Wils. C. siliquosus L. A common weed. 278, 750. Britt. & Wils. Triumfetta L. Triumfetta bartramia L. Britt. & Wils. T. lappula L. Fishlock 301. Britt. & Wils. T. semitriloba Jacq. A troublesome weed throughout the island; es- pecially common in Jackass Ghut and Long Trench. All three Triumfettas were observed to intergrade on burr characters. 266, 267, 268, 269, 330, 367, 413, 644, 646, 647, 650; Fishlock 20 (Ex- periment Station). Britt. & Wils. 424 Rhodora [Vol. 69 Family 47. MALVACEAE Abelmoschus Medic. Abelmoschus esculentus (L.) Moench. Hibiscus esculentus L. Widely planted as a market vegetable. “Okra.” Abutilon Mill. Abutilon hirtum (Lam.) Sweet. Uncommon. Seen on Frenchmans Cay. 68. A. indicum (L.) Sweet. Seen mostly east of Road Town. Fishlock 220 (Pasea Hall Estate). Britt. & Wils. A. umbellatum (L.) Sweet. Very common weed in abandoned culti- vations and roadsides, lower elevations. 67, 411. Britt. & Wils. Althaea L. Althaea rosea Cav. Rarely cultivated. “Hollyhock.” Bastardia Kunth Bastardia viscosa (L.) Kunth. Plentiful, waste places. 415, 416, 417, 418, 419, Gossypium L, Gossypium arboreum L. Gossypium purpurascens Poir. Baughers Bay. Cultivation of cotton ceased when Government support was withdrawn in 1947, but it is persistent in many places, and seems to be naturalized. Other species may well be present. Hibiscus L. Hibiscus sabdariffa L. Rarely cultivated from local seed for orna- ment and for its red fruits which are used for jellies and drinks. A form with yellow flowers and fruits was seen in Road Town: it may be another species. “Sorrel.” H. vitifolius L. Common in Road Town, but not seen elsewhere. “Wild Cotton.” 764; Fishlock 181 (Road Town). Britt. & Wils. A number of species and hybrids of this genus are widely cultivated for ornament and sometimes escape. Malachra L. Malachra alceifolia Jacq. In occasional clumps, roadsides. 319; Fish- lock (Paraquita Bay). Britt. & Wils. M. capitata L. In occasional clumps, roadsides. 326. Malvastrum A. Gray Malvastrum americanum (L.) Torrey. Malvastrum spicatum (L). A. Gray. Common in eastern districts. Very plentiful in the bottom, Long Bay East. 412. M. corchorifolium (Desv.) Britt. Common in eastern districts. Very plentiful in Paraquita Bay. 393, 394, 395, 396, 397. M. coromandelianum (L.) Garke. Fishlock 24 (Experiment Station). Britt. & Wils. Malvaviscus Dill. ex Adans. Malvaviscus arboreus Cav. Occasionally cultivated. 387. Pavonia Cav. Pavonia spinifex (L.) Cav. Common in thickets and shaded roadsides. Fishlock 290 (Pasea Hall Estate). Britt. & Wils. 1967] Tortola — D'Arcy 425 Sida L. Sida acuminata DC. Fishlock 488 (hillside). Britt. & Wils. S. acuta Burm. f. Sida carpinifolia L. f. A very common, trouble- some weed almost everywhere. “Wireweed.” 300, 825, 826, 827, 828; Fishlock 27 (Experiment Station). S. ciliaris L. Widespread but not very common. Seen at Doty and a large patch at Chapel Hill (East End). 301, 302, 303. Britt. & Wils. S. eggersii E. G. Baker. This arboreal Sida is endemic to Tortola and Culebra. Not seen. Eggers 3183 (Coxheath 300'). Britt. & Wils. . glabra Mill. Britt. & Wils. . glomerata Cav. A common, troublesome weed. Fishlock 203 (Pasea Hall Estate). Britt. & Wils. S. humilis Cav. Widespread but not common. Usually in shady, well trimmed pastures and cultivation. Not troublesome. 587, 588, 589, 590, 591; Fishlock 1 (Experiment Station). S. salviaefolia Presl. Sida erecta Macf. Rare. Seen only once at Slaney Point, but not collected. Britt. & Wils. S. urens L. Widespread but rare. Seen on hillside at Doty and in Brewers Bay. 430. N N Thespesia Soland. Thespesia populnea (L.) Soland. Common along sea shores. “Heite- heite.” Urena L. Urena lobata L. Widespread but more common at higher elevations, growing in disturbed pastures, Chalwell and Doty. 382, 382, 384, 385, 386; Fishlock (Pasea Hall Estate). Britt. & Wils. Wissadula Medic. Wissadula amplissima (L.) R. E. Freis. 449; Fishlock 202 (Pasea Hall Estate). Britt. & Wils. Family 48. BOMBACACEAE Ceiba Mill. Ceiba pentandra (L.) Gaertn. Scattered throughout the island. Several quite large (12 m.) trees are growing in villages in dry eastern districts. “Silk Cotton Tree.” Little & Wads. Family 49. STERCULIACEAE Ayenia Loefl. Ayenia insularis Cristobal. Ayenia pusilla L. Britt. & Wils. Guazuma Adans. Guazuma ulmifolia Lam. Guazuma guazuma (L.) Cockerell. Rare, seen only in Huntum’s Ghut at about 100 m. elevation. 723; Fishlock 355 (Huntum's Ghut). Little & Wads., Britt. & Wils. Helicteres L. Helicteres jamaicensis Jacq. A few scattered bushes along the south coast between Road Town and Coxheath. 181, 671; Fishlock 252 (Slaney Point). Britt. & Wils. 426 Rhodora [Vol. 69 Melochia L. Melochia nodiflora Sw. A common and troublesome weed. 639; Fish- lock 15 (Experiment Station). M. pyramidata L. Moluchia pyramidata (L.) Britton. Common but most plentiful in moist districts. 640, 641, 642, 643; Fishlock 12 and 763 (Experiment Station). Britt. & Wils. M. tomentosa L. Moluchia tomentosa (L.) Britt. Common through- out the island, but most common on dry hillsides where it attains an ornamental appearance. 193; Fishlock 43 (Sea Cow Bay). Britt. & Wils. Theobroma L. Theobroma cacao L. Occasionally cultivated for its pods in moist districts. “Cocoa.” Waltheria L. Waltheria indica L. Waltheria americana L. Plentiful in Long Bay East and in Paraquita Bay. 392. Family 50. MARCGRAVIACEAE Marcgravia L. Marcgravia rectiflora Tr. & Pl. Plentiful in dense forests, Sage Mountain. Several specimens were also seen on rocks on the east side of Chalwell peak in full sun and wind. 438, 439, 440, 441, 442, 443, 444, 445, 446, 447; Fishlock 472 (Sage Mt.) ; Eggers 3239 (High Bush 1500'). Britt. & Wils. Family 51. GUTTIFERAE Calophyllum L. Calophyllum antillanum Britt. Calophyllum brasiliense Camb. A few trees seen near Pasea Hall Estate. “Wild Mammee." Clusea L. Clusia rosea Jacq. Scattered trees around the island and plentiful in mountain forests. Only in the mountains does it seem to adopt its renowned strangling epiphytic habit. “Pitch Apple," "Wild Mammee." Little & Wads., Britt, & Wils. Mammea L. Mammea americana L. Plentiful at upper elevations, rarely down to sea level. This is the author's favorite fruit of the island, but it is little esteemed locally and often left to rot on the ground. "Mammee Apple." Little & Wads., Britt. & Wils. Family 52. BIXACEAE Bixa L. Bixa orellana L. Only a few trees in Sea Cows Bay (not seen) and in Huntum's Ghut. One tree in Huntum's Ghut has been there over 30 years. “Roucou.” 161. 1967] Tortola — D’Arcy 427 Family 53. COCHLOSPERMACEAE Cochlospermum Kunth Cochlospermum vitifolium (Willd.) Spreng. Mazimiliania vitifolia (Willd. Krug & Urban. Until recently there was only one fine specimen in Little Apple Bay, but new introductions may make this an important ornamental. They are all of the double-flowered form. 149. Family 54. FLACOURTIACEAE Casearia Jacq. Casearia decandra Jacq. Seen only along the roadside between Joe's Hill and Meyers. 84; Fishlock 420 (Fort Hil). Little & Wads., Britt. & Wils. C. guianensis (Aubl. Urban. Common in thickets and along road- sides, upper elevations. ‘Wild Coffee." C. sylvestris Sw. Common in ghuts and along roadsides, moist dis- triets. “Wild Coffee.” 64; Fishlock 413 and 157 (Huntum’s Ghut); Eggers 3209 (High Bush 1000’). Little & Wads., Britt. & Wils. Samyda L. Samyda dodecandra Jacq. Not at all common, seen only in undis- turbed ghuts leading into Road Harbour. 1; Fishlock 246 (Lower Estate). Britt. & Wils. Family 55. TURNERACEAE Turnera L. Turnera ulmifolia L. Seen only on Long Bay Hill, where it was plentiful. 183. Family 56, PASSIFLORACEAE Passiflora L. Passiflora foetida L. Seldom to be seen in dry seasons, but in a wet year it is very plentiful on roadside fences, garbage dumps, and other waste places. «Love-in-the-Mist." 592; Fishlock 7 (Ex- periment Station). Britt. & Wils. P. laurifolia L. Seen in light woods, Sage Mountain above Carrot Bay. 702. P. multiflora L. Occurs rarely in moist districts. 599 (West End). Britt. & Wils. P. rubra L. In open places, north side of Sage Mountain. 186, 196. P. suberosa L. Passiflora pallida L. Common in thickets almost everywhere. 113, 253, 462 (Peter Island); Fishlock 152 (Experi- ment Station). Fishlock 264 (Slaney Point). Britt. & Wils. Family 57. CARICACEAE Carica L. Carica papaya L. Commonly planted for food and naturalized in moist districts. The wild forms are strikingly different in ap- pearance from those cultivated. They are tall and thin with 428 Rhodora [Vol. 69 branching and twining stems, and tiny, very sweet fruit. "Paw- paw." “Papai.” Little & Wads. Several species of Begonia are cultivated in houses and gardens for ornament. They are of the Family BEGONIACEAE. Family 58. CACTACEAE Hylocereus Britt. & Rose Hylocereus trigonus (Haw.) Safford. Common in the Long Trench district and also seen on Jean Hill at 50 m. Used to flavor rum in making the traditional “Miss Blyden" Christmas drink, and also for making tarts and pies. "Strawberry." Britt. & Wils. H. undatus (Haw.) Britt. & Rose. Seen only at Brittanic Hall, Road Town, where it appears as if naturalized. Mammillaria Haw. Mammillaria nivosa Link. Neomammillaria nivosa (Link.) Britt. & Rose. Seen only on cliffs overhanging the south coast road be- tween Road Town and West End. 776, 777. Britt. & Wils. Melocactus Link & Otto Melocactus intortus (Mill) Urb. Cactus intortus Mill. Common along seacoasts. Sometimes cooked for food. * Barrel Cactus." "Dildo." Britt. & Wils. Nopalea Salm-Dyck Nopalea cochenillifera (L.) Salm-Dyck. Commonly cultivated for ornament and medicinal use. Very long persistent after cultiva- tion, it is to be seen around old ruins and abandoned house sites giving the appearance of having naturalized. “French Prickly Pear." 834. Opuntia Mill. Opuntia antillana Britton & Rose. Two or three patches seen at East End. Seldom flowering. “Bull Suckers.” 109. Britt. & Wils. O. dillenii (Ker-Gawl.) Haw. Widespread and fairly plentiful. *Dildo," "Priekly Pear." O. repens Bello. One of the most noisome weeds of pastures. “Suckers.” Britt. & Wils. O. rubescens Salm-Dyck. Consolea rubescens (Salm-Dyck) Lemaire. Common along sea coasts. An almost spineless form occurs at Coxheath. “Dildo.” Little & Wads., Britt. & Wils. Pereskia Mill. Pereskia grandifolia Haw. Sometimes cultivated for ornament. Pilosocereus Lemaire Pilosocereus royenii (L.) Byles & Rowley. Cephalocereus royenii (L.) Britt. & Rose. Common near the sea. “Pipe Organ Cactus.” “Dildo.” Little & Wads., Britt. & Wils. Selenicereus Britt. & Rose Selenicereus grandiflorus (L.) Britt. & Rose. Rarely seen in gardens. Britt. & Wils. reported it as naturalized on Tortola. 1967] Tortola — D’Arcy 429 Family 59. MELASTOMATACEAE Clidemia D. Don Clidemia hirta (L.) D. Don. Very plentiful in thickets and becoming a pest in pastures, upper elevations. Fishlock 370 (Doty). Britt. & Wils. Miconia Ruiz & Pav. Miconia fothergilla Naud. Eggers 3198. M. laevigata (L.) DC. Very plentiful in thickets and roadsides, upper elevations. 85, 253; Fishlock 366 (Joe's Hill). Britt. & Wils. M. mirabilis (Aubl. L. O. Wms. Tamonea guianensis Aubl. Britt. & Wils. M. prasina (Sw.) DC. Eggers 3205 (Sage Mountain). Little & Wads., Britt. & Wils. M. thomasiana DC. Sage Mountain forests. Leaves growing in the wind are strikingly coriaceous while those in sheltered forest are membranacecus. 711,712; Fishlock 93 and 463 (Sage Mt.). Britt. & Wils. Tetrazygia L. Tetrazygia angustifolia (Sw.) DC. Plentiful on windswept south side of Sage Mountain. 101; Fishlock 32 (Sage Mt.); Eggers 3194. Britt. & Wils. T. elaeagnoides (Sw.) DC. Little & Wads., Britt. & Wils. Family 60. LYTHRACEAE Ammania L. Ammania latifolia L. Seen only by the main road at Hannah. 162, 163, 164; Fishlock 388 (Old Plantation). Britt. & Wils. Ginoria Jacq. Ginoria rohrii (Vahl) Koehne. Plentiful on Frenchman's Cay and around Little Apple Bay. Occasional plants at Fish Bay, Kings- town Bay and the Fort Hill, which bloom when leafless. Very showy when in flower, and often cut for its flowers. “Sugar Ant." 50, 182: Fishlock 399 (Old Plantation), Fishlock 400 (Baughers Bay). Britt. & Wils. Lagerstroemia L. Lagerstroemia indica L. Frequently grown for ornament and perhaps spontaneous. “Queen of Flowers." Family 61. PUNICACEAE Punica L. Punica granatum L. Both red and yellow forms are widely planted for ornament and escapes are sometimes seen. Fruit is very poor as the climate is too warm. Nevertheless, the remains of what seems to be a long abandoned plantation are to be seen on Jean Hill. 430 Rhodora [Vol. 69 Family 62. COMBRETACEAE Buchenavia Eichl. Buchenavia capitata (Vahl) Eichl. Mountain forests. 705. Little & Wads., Britt. & Wils. Bucida L. Bucida buceras L. Common in pastures and woods at lower eleva- tions, especially along the south coast west of Road Town. "Gregre." A valuable wood for boat building. Conocarpus L. Conocarpus erecta L. Common on reef formations and occasional along the sea coast. The leaves are green. “Buttonwood.” 75, 76; Fishlock 99 (Harbours). Little & Wads., Britt. & Wils. Laguncularia Gaertn. f. Laguncularia racemosa (L). Gaertn. f. Very common in marshy ground beside salt water. A favorite charcoal wood at lower elevations. “White Mangrove.” 247; Fishlock 217 (Pasea Hall Estate). Quisqualis L. Quisqualis indica L. Cultivated. A fine hedge is growing at Govern- ment House. “Rangoon Creeper.” Terminalia L. Terminalia catappa L. Very common by beaches and other coastal areas. "Almond." Fishlock 102 (Pockwood Pond). Little & Wads., Britt. & Wils. Family 63. MYRTACEAE Aulomyrcia Berg. Aulomyrcia citrifolia (Aubl.) Amsh. Myrcia citrifolia (Aubl.) Urban. Little & Wads., Britt. & Wils. Calyptranthes Sw. Calyptranthes kiaerskovii Krug & Urban. The lone Tortola endemic species to date. Based on a single sterile collection by Eggers; 3217, location not given. Eucalyptus L'Hérit. Eucalyptus camaldulensis Dehn. Eucalyptus rostrata Schlechtendal. Several large trees at Government House and a few scattered elsewhere, e.g., Hannah, Belleview. Both E. camaldulensis and E. tereticornis Smith may be represented at Government House. They were planted by the Agricultural Department many years ago. 2. Eugenia L. Eugenia axillaris (Sw.) Willd. Seen only behind the beach, Long Bay East. 390. E. biflora (L.) DC. Eugenia lancea Poir. One of the most common and troublesome weedy shrubs in pastures and hillsides. The wood is used for fish pot frames. “Birch Berry." 63, 102, 836, Britt. & Wils. 1967] Tortola — D'Arcy 431 E. cordata (Sw.) DC. Plentiful along Joe's Hill-Meyers Road. Flowers, mostly above the leaves, have an unusual but pleasant fragrance. 214; Fishlock 405 (Hog Valley Point). Britt. & Wils. E. monticola (Sw.) DC. Fishlock 120 (Belleview). Britt. & Wils. E. procera (Sw.) Poir. Fishlock 422 (Belleview), Fishlock 491 (ra- vine). Britt. & Wils. E. pseudopsidium Jacq. Seen at Doty and at Long Bay East near the sea. The shiny red or yellow fruits are very striking in appear- ance. “Wild Guava.” 273, 379, 623. Britt. & Wils. E. rhombea (Berg) Krug & Urban. Seen only in Belleview Ghut (150 m.). 208. E. sessiliflora Vahl. Fishlock 406 (Hog Valley Point). Britt. & Wils. E. sintensii Kiaersk. Britt. & Wils. Myrcia DC Myrcia splendens (Sw.) DC. Fishlock 454. Little & Wads., Britt. & Wils. Myrcianthes Berg. Myrcianthes fragrans (Sw.) McVaugh. Anamomis fragrans (Sw.) Griseb. Britt. & Wils. Myrciaria Berg Myrciaria floribunda (West) Berg. Eugenia floribunda West. Wide- spread. Although not planted, most trees receive some attention from farmers, as the fruits are esteemed for making the tradi- tional Christmas rum drink, “Guavaberry.” 194; Fishlock 119, 324 (Belleview); Eggers 3235 Britt. & Wils. Another yellow fruit, seemingly a Myrtaceae and also called 'Guava- berry," harvested in much the same way and at the same time as M. floribunda, is used for making tarts. Plants were not seen. Pimenta Lindl. Pimenta racemosa (Mill.) J. W. Moore. Amomis caryophyllata (Jacq.) Kr. & Urban. Occasional in ghuts at middle elevations, and very plentiful in Jackass Ghut. The fragrance is noticeable several yards away. “Bay-Rum-Tree.” 718; Fishlock 493 and 494. Little & Wads., Britt. & Wils. P. racemosa var grisea (Kiaersk.) Fosberg. Amomis grisea (Kiaersk.) Britton. Occasional in mountain forests. The bruised leaves emit a strong citronella odour. “Cinnamon Tree." Fish- lock 79 (Sage Mt.); Eggers 3196. Britt. & Wils. Psidium L. Psidium amplexicaule Pers. Plentiful in the Chalwell-Sage Mountain area. “Mountain Guava.” 146; Fishlock 372 (Doty); Eggers 3190. Britt. & Wils. P. guaiava L. Rarely cultivated. It is a troublesome weedy species around Doty and Chalwell Estate. “Guava.” Little & Wads., Britt. & Wils. 432 Rhodora [Vol. 69 Syzygium Gaertn. Syzygium jambos (L.) Alston. Jambosa jambos (L.) Millsp. Eugenia jambos L. Forming large copses in pastures and on the edges of the forest, north side of Sage Mountain. 761; Fishlock 88 (Sage Mountain). Little & Wads., Britt. & Wils. Family 64. LECYTHIDACEAE Couroupita Aubl. Couroupita guianensis Aubl. The fine tree at the Experiment Station is a continual object of curiosity. It was surveyed at 51 feet on 25th April, 1965. “Cannonball Tree." Family 65. RHIZOPHORACEAE Rhizophora L. Rhizophora mangle L. Coastal in salt swamps and on reefs. “Red Mangrove." 246; Fishlock 236 (Pasea Hall Estate). Little & Wads., Britt. & Wils. Family 66. ONAGRACEAE Ludwigia L. Ludwigia erecta (L) H. Hara. Jussiaea erecta L. Scattered clumps in marshy roadsides. 143, 248, 249, 250; Fishlock 16 and 289 (Turnbulls). L. octovalvis (Jacq.) Raven. Jussiaea angustifolia Lam, Scattered clumps in marshy roadsides. This and L. erecta are often found together. 830. Britt. & Wils. Family 67. ARALIACEAE Dendropanax Dene. & Pl. Dendropanax arboreum (L.) Dene. & Pl. A few scattered trees, mostly in the open, south side of Sage Mountain. 251; Fishlock 455 and 456 (Sage Mt.). Little & Wads., Britt. & Wils. Didymopanax Dene. & PI. Didymopanax morototoni (Aubl.) Dene. & Pl. Fairly plentiful, north side of Sage Mountain. 99. Little & Wads. Polyscias Forst. Polyscias filicifolia (Moore) Bailey. Commonly planted as a hedge or specimen shrub. “Angelica,” “Croton.” P. guilfoylei (Bull) Bailey. Commonly planted in moist districts, long persistent after cultivation. 567, 568, 569. Family 68. UMBELLIFERAE Anethum L. Anethum graveolens L. Widely cultivated for flavouring and medicinal purposes. Naturalized in moist districts, “Anise” (dill). Fish- lock 66 (Todmans). Britt. & Wils. Foeniculum Mill. Foeniculum vulgare Mill. Foeniculum foeniculum (L.) Karsten. Com- mon in upper districts, especially from Hope to Doty, growing in pastures and roadsides and one of the most plentiful plants in abandoned cultivation. “Wild Fennel.” 608, 609, 610. 1967] Tortola — D’Arcy 433 Pimpinella L. Pimpinella anisum L. Rarely cultivated from imported seed and locally persistent for a short time. (dill). Daucus carota L. (carrot) and Petroselinum crispum (Mill.) Mans- field (parsley) are cultivated as market vegetables. Family 69. MYRSINACEAE Ardisia Sw. Ardisia obovata Desv. Icacorea guadalupensis (Duch.) Britton. Com- mon in thickets and forests, north side of Chalwell, Doty, and Sage Mountain. 49, 203; Fishlock 340 (Joe's Hill); Eggers 3228. Little & Wads., Britt. & Wils. Myrsine L. Myrsine guianensis (Aubl.) Kuntze. Rapanea guianensis Aubl. Fish- lock 462 (Sage Mt.). Little & Wads., Britt. & Wils. Family 70. THEOPHRASTACEAE Jacquinia L. Jacquinia arborea Vahl. Jacquinia barbasco (Loefl.) Mez. From Paraquita Bay eastward, and common on Beef Island; rare else- where. 117. Britt. & Wils. J. berterii Spreng. Plentiful on Peter Island, Camanoe and perhaps Beef Island, but not seen cn Tortola itself. 488, 800. J. revoluta Jacq. Fishlock 75 (upper slopes of Sage Mt.). Britt. & Wils. Family 71. PLUMBAGINACEAE Plumbago L. Plumbago auriculata Lam. Plumbago capensis Thunb. A common ornamental. Because of dry conditions on the island, it is usually a tight formed, profusely flowering dwarf shrub. P. scandens L. Common by roadsides and in thickets, all but the driest areas. 533; Fishlock 249 (Experiment Station). Britt. & Wils. Family 72. SAPOTACEAE Bumelia Sw. Bumelia obovata (Lam.) A. DC. Near the sea and on dry promon- tories. Common in Baughers Bay. “Pintop.” 167, 586; Fishlock 446 (Baughers Bay); Eggers 3186. Britt. & Wils. Chrysophyllum L. Chrysophyllum argenteum Jacq. Common in thickets and forests, mountain districts. 211, 667; Fishlock 478 (Sage Mountain). Little & Wads., Britt. & Wils. C. cainito L. Only three fine trees near the Burial Ground, Road Town. “Star Apple." Dipholis A. DC Dipholis salicifolia (L.) A. DC. Fishlock 401 (Baughers Bay). Little & Wads., Britt. & Wils. 434 Rhodora [Vol. 69 Manilkara Adans Manilkara balata (Aubl. Dubard. Manilkara nitida (Sessé & Moq.) Dubard. Manilkara bidentata (A. DC.) Chev. Seen living only in dense forest, Sage Mountain. Many large boles dot the moun- tain hillsides, indicating the former extent of the Forest. 716; Fishlock 471 (Sage Mt.). Little & Wads., Britt. & Wils. M. zapota (L.) Royen. Sapota achras Mill. Hillsides, Sea Cow Bay and in Road Town. *Mesple." Mastichodendron (Engl) H. J. Lam. Mastichodendron foetidissimum (Jacq.) Lam. Sideroxylon foetidis- simwm Jacq. Little & Wads. Family 73. SYMPLOCACEAE Symplocos Jacq. Symplocos martinicensis Jacq. Thickets and pastures, wet mountain areas. 97, 714; Fishlock 453 (Sage Mt.). Little & Wads., Britt. & Wils. Family 74. OLEACEAE Forestiera Poir. Forestiera eggersiana Krug & Urban. Endemic to Virgin Islands, Culebra and Vieques. Rare, seen only at Pleasant Valley and Kingstown Hill. Both trees were staminate. 290. F. segregata (Jacq.) Krug & Urban. Eggers 3175 (West End in fruit). Britt. & Wils. Jasminum L. Jasminum fluminense Vell. Jasminum azoricum L. Cultivated and sometimes escaping. 389. J. officinale var grandiflorum (L.) L. H. Bailey. Jasminum grandi- florum L. Widely cultivated; not known to escape. J. pubescens (Retz.) Willd. Seen only on a wall at Britannic Hall, Road Town, where it was seemingly spontaneous, J. sambac (L.) Soland. Seen only at Harrigan's (Joe's Hill), where it was seemingly spontaneous. Family 75. APOCYNACEAE Allamanda L. Allamanda cathartica L. Frequently cultivated. *Yellow Allamanda." 343. Catharanthus G. Don Catharanthus roseus (L.) G. Don. Vinca rosea L. Cultivated and a common wildflower. White forms seem to stand more salt and red forms more drought. One of the few really *animal proof" ornamentals on the island. “Stinky Toes.” 596; Fishlock 307 (Norton Valley). Britt. & Wils. Nerium L. Nerium oleander L. One of the commonest and most easily grown ornamentals. Not spontaneous. “Oleander.” 224, 1961] Tortola — D'Arcy 435 Plumeria L. Plumeria alba L. Very common in coastal rocks and cliffs. “Sea Oleander.” “Frangipani.” Little & Wads., Britt. & Wils. P. rubra L. Occasionally cultivated. Both white and red forms are to be seen at Government House and Pasea Hall Estate. It is slow growing and has spectacular bloom, probably because of the dry climate. “Frangipani.” Little & Wads. Prestonia R. Brown Prestonia agglutinata (Jacq.) Woodson. Echites agglutinata Jacq. Rare in coastal thickets. 292 (Belmont Bay) ; Fishlock 25 (Slaney Point). Britt. & Wils. Rauwolfia L. Rauwolfia nitida Jacq. Rauwolfia tetraphylla L. sensu Britt. & Wils. Seen only in Cane Garden Bay. Fishlock 430 (Belmont). Little & Wads. “Bitter.” R. viridis R. & S. Rauwolfia lamarckii A. DC. A very common and troublesome weedy shrub throughout the island. “Bitter Ash.” Thevetia L. Thevetia peruviana (Pers.) K. Schum. Cerbera thevetia L. Seen only in one patch on the road from Soldier Hill to Cane Garden Bay (ca. 100 m.). “Yellow Oleander.” Urechites Muell.-Arg. Urechites lutea (L.) Britton. Scattered in coastal areas. Common in Duff Bottom. The sap is very poisonous, as the author learned to his discomfort. “Wild Allamanda.” 775 (Beef Island); Fish- lock 474 (Baughers Bay). Britt. & Wils. Family 76. ASCLEPIADACEAE Asclepias L Asclepias curassavica L. Occasional plants, moist districts. “Hittie McWanie.” 281, 282, 283; Fishlock 304 (Purcells). Britt. & Wils. Calotropis R. Brown Calotropis procera (Ait.) Ait. f. Conspicuous in pastures and waste places. “Cow Heel.” Cryptostegia R. Brown Cryptostegia grandiflora R. Br. Widely cultivated and sparingly naturalized. It is highly resistant to attacks of Cuscuta sp. “Purple Snake.” “Purple Allamanda.” Cynanchum L. Cynanchum decaisneanum (Schlechter) Alain. Metastelma decaisnea- num Schlechter. Coastal rocks and thickets. Seen at Fish Bay. 227. C. parviflorum Sw. Metastelma parviflorum R. Br. Coastal rocks and thickets, mostly in eastern districts. 318 (Beef Island). Britt. & Wils. 436 Rhodora [Vol. 69 Matelea Aubl. Matelea maritima (Jacq.) Woodson. Ibatia maritima (Jacq.) Dene. Isolated plants throughout the island. 69; Fishlock 465 (Fish Bay); Eggers 3240. Britt. & Wils. Stephanotis floribunda A. Brongn. and Hoya carnosa (L. f.) R. Br. are rarely cultivated. Family 77. CONVOLVULACEAE Cuscuta L. Cuscuta americana L. Britt. & Wils. C. globulosa Bentham. Fishlock 489 det. Y uncker. C. sp., “Yellow Love," is a common and serious pest throughout the island. (Distinctions between the above two species are slight and they are frequently confused. See Yuncker; Mem. Torr. Bot. Club vol. 18, pp. 217-222 (1931-32).) Evolvulus L. Evolvulus glaber Spreng. Britt. & Wils. E. sericeus Sw. Britt. & Wils. Ipomoea L. Ipomoea acuminata (Vahl) R. & S. Ipomoea cathartica Poir. Com- mon near the sea. Britt. & Wils. I. batatas (L.) L. Commonly cultivated for market. “Potato.” I. nil (L.) Roth. Seen only along the roadside between Welk Point and Fat Hog Bay. 315. I. pes-caprae (L.) Roth. Common on coastal sands. “Beach Morning Glory." Fishlock 294 (Pasea Hall Estate). Britt. & Wils. I. quamoclit L. Quamoclit quamoclit (L.) Britt. Rarely cultivated. I. repanda Jacq. Exogonium repandum (Jacq.) Choisy. Common at upper elevations in woods and thickets. 511; Fishlock 339 (Joe’s Hill). Britt. & Wils. I. steudelii Millsp. Exogonium arenarium Choisy. Seen at Welk Point and common on Camanoe. 765. I. tiliacea (Willd.) Choisy. Rare, seen at top of Joe's Hill and at Flamingo Pond. 496, 751. Britt. & Wils. I. triloba L. Common in eastern districts. "Pig-Vine." 317. I. tuba (Schlechter) G. Don. Calonyction tuba (Schlechter) Colla. A few large patches seen near the sea; Fish Bay, Towers and Carrot Bay. Jacquemontia Choisy Jacquemontia jamaicensis (Jacq.) Hallier f. Rare. Seen only beneath coconut trees on Wickhams Cay. 431. Britt. & Wils. J. nodiflora (Desv.) G. Don. Very common along the south coast. 320, 336; Fishlock 261 (Slaney Hill). Britt. & Wils. J. pentantha (Jacq.) G. Don. A common weed in all districts. A purple form grows on Jean Hill. 121; Fishlock 204 (Pasea Hall Estate). Britt. & Wils. 1967] Tortola — D'Arcy 437 Merremia Dennst. Merremia aegyptia (L.) Urban. Ipomoea aegyptia L. Common in moist districts, especially on the north side. 314; Fishlock 278 (Experiment Station). Britt. & Wils. M. dissecta (Jacq.) Hall. f. Ipomoea dissecta (Jacq.) Pursh. Not very common. Seen at Windy Hill and Belmont beach. 638; Fishlock 197 (Clifton Hall). Britt. & Wils. M. quinquefolia (L.) Hall. f. Ipomoea quinquefolia L. A common and troublesome weed in moist districts. Britt. & Wils. M. umbellata (L.) Hall. f. Ipomoea polyanthes R. & S. Common in cultivated areas and roadsides. “Yellow Morning Glory." 72, 740; Fishlock 308 (Purcells). Britt. & Wils. Porana Burm. f. Porana paniculata Roxb. Very long persistent after cultivation, but probably not spontaneous. 648. Stictocardia Hallier f. Stictocardia tiliaefolia (Desr.) Hall. f. Rivea campanulata | (L.) House. Fishlock 275 (Experiment Station). Britt. & Wils. Turbina L. Turbina corymbosa (L.) Raf. Plentiful in Ballast Bay. 122. Family 78. BORAGINACEAE Bourreria P. Browne Bourreria succulenta Jacq. Scattered trees throughout the island. In eastern districts, leaves are hispid, scabrous, but in the west they are glabrous and membranaceous; a result of rainfall dif- ferences. "Chinkwood." 237, 291; Fishlock 118 sl Little & Wads., Britt. & Wils. Cordia L. Cordia alliodora (R. & P.) Oken. Cerdana alliodora R. & P. Scatte ec high in ghuts. “Spanish Elm." 126, 218; Fishlock 106 (Lower Estate), Fishlock 121 (Purcells Mountain). Little & Wads., Britt. & Wils. C. collococca L. Cordia glabra L. Occasional in thickets and ghuts, mostly lower elevations. *Manjack." Fishlock 416 (Experiment Station). Britt. & Wils. C. nitida Vahl. Occasional at all elevations. *Red Manjack." 142, 179; Fishlock 410 (Huntums Ghut.). Little & Wads., Britt. & Wils. C. obliqua Willd. Plentiful in eastern districts; not seen elsewhere. Mucilage from the pink drupes is used to seal envelopes. ''Stick- ing Tree." 3. C. polycephala (Lam.) Johnst. Varronia corymbosa (L.) Desv. A troublesome woody weed in wet mountain districts, rarely seen down to sea level. “Black Sage." 124, 210, 730; Fishlock 114 (Belleview), Fishlock 216 (Experiment Station); Eggers 3234 as C. ulmifolia Juss. var ovata DC. Britt. & Wils. 438 Rhodora [Vol. 69 C. rickseckeri Millsp. Sebesten rickseckeri (Millsp.) Britt. Very plentiful along the south coast. "Dog Almond," *Black Manjack." 51. Little & Wads., Britt. & Wils. C. sebestena L. Sebesten sebestena (L.) Britton & Small. Very rarely cultivated. One spontaneous plant was found at Coxheath. 226; Fishlock 398 (Old Plantation). C. sulcata DC. Occasional at all elevations. "White Manjack." Fish- lock 447 (Experiment Station). Little & Wads., Britt. & Wils. Heliotropium L. Heliotropium amplexicaule Vahl. Cochranea anchusaefolia (Poir.) Giircke, Cultivated under the name of “Verbena.” H. angiospermum Murray. Schobera angiosperma (Murr.) Britt. A common and troublesome weed. “White Tip.” 554, 555, 556, 457, 558; Fishlock 139 (Experiment Station). Britt. & Wils. H. crispiflorum Urban. Large patches on Beef Island. Not seen else- where. 787. H. curassavicum L. Plentiful beside the sea. Britt. & Wils. H. indicum L. Tiaridium indicum (L.) Lehm. Common, especially near the coast. Fishlock 26 (Experiment Station). Britt. & Wils. Tournefortia L. Tournefortia bicolor Sw. Pastures, Doty and Sage Mountain. T. gnaphalodes (L.) R. Br. Mallotonia gnaphalodes (L.) Britt. Heliotropium gnaphalodes L. Rather rare. Coastal reefs and beaches. 481. T. hirsutissima L. Plentiful at upper elevations. 198; Fishlock 444 (ravines). Britt. & Wils. T. microphylla Bert. Common along roadsides and in thickets, south side of the island. Fishlock 297 (Pasea Hall Estate). Britt. & Wils. T. volubilis L. Eggers 3172 (West End). Family 79. VERBENACEAE Avicennia L. Avicennia germinans (L.) L Avicennia nitida Jacq. Occasional in coastal thickets. “Salt Pond.” Fishlock 218 (Pasea Hall Estate). Little & Wads., Britt. & Wils. Citharexylum L. Citharexylum fruticosum L. Common in thickets and hillsides at lower elevations. “Fiddlewood.” 553; Fishlock 198 (Clifton Hall). Little & Wads., Britt. & Wils. Clerodendrum L. Clerodendrum aculeatum (L.) Schlecht. Volkameria aculeata L. Coastal thickets, more common along the south coast. Occasionally high climbing. “Privet.” 41; Fishlock 184. Britt. & Wils. C. speciosissimum Paxt. A common ornamental in Road Town. C. thompsonae C. Balfouri. A common ornamental. “Danish Flag.” 1961] Tortola — D'Arcy 439 Lantana L. Lantana arida Britton. Moldenke 1959. L. camara L. A yellow and red flowered form is common in thickets and as a weed in most districts, *Yellow Sage;" a purple and buff flowered form with larger leaves and heavier construction occurs in Road Town. 42, 311; Fishlock 45 (Sea Cow Bay). Britt. & Wils. L. involucrata L. Common in coastal thickets. “Button Sage." 232; Fishlock 46 (Sea Cow Bay). L. reticulata Pers. Moldenke 1959. Petraea Houst. ex L. Petraea volubilis Jacq. A common ornamental. *Queen's Wreath." Priva Adans. Priva lappulacea (L.) Pers. A common weed. 548, 549, 550, 551, 552; Fishlock (Experiment Station). Britt. & Wils. Stachytarpheta Vahl Stachytarpheta jamaicensis (L.) Vahl. Valerianoides jamaicense (L.) Kuntze. Fishlock 158 (waste places). Britt. & Wils. Tectona L. f. Tectona grandis L. f. Tektona grandis L. (Tectena is conserved). A few widespread trees planted many years ago by the Agri- cultural Department. “Teak.” 236. Little & Wads. Verbena L. Verbena chamaedrifolia Juss. Cultivated. “Verbena.” Vitex L. Vitex agnus - castus L. A common ornamental, probably spontaneous. V. divaricata Sw. Little & Wads. Family 80. LABIATAE Coleus Lour. Coleus amboinicus Lour. Scattered in all parts of the island. Large patches were seen at Fort Burt, Huntums Ghut (110 m.) and Hope. Seldom flowering. “Wild Thyme." 606; Fishlock 108A. Britt. & Wils. C. blumei Benth. Commonly cultivated. Naturalized but rarely found, in ghuts. Seen in Belleview ghut (ca. 200 m.). "Joseph's Coat." Britt. & Wils. Hyptis Jacq. Hyptis capitata Jacq. Plentiful in wet mountain districts. 427; Fish- lock 333 (Joe's Hill) ; Eggers 3221. Britt. & Wils. H. pectinata (L.) Poit. Plentiful in moist seasons, roadsides and hill- side pastures. 398; Fishlock 305 (Purcells) ; Eggers 3223. Britt. & Wils. H. suaveolens (L.) Poit. Seen only in a large patch on the Fort Hill (30 m.). 328. 440 Rhodora [Vol. 69 Leonotis R. Brown Leonotis nepetaefolia (L.) R. Brown. A very common weed. Fishlock 122 (waste places). Britt. & Wils. Leonurus L. Leonurus sibiricus L. A common weed. “Rabbit Food." 239; Fish- lock 6 (Experiment Station). Britt. & Wils. Leucas R. Brown Leucas martinicensis (Jacq.) R. Br. Fishlock 473 (Baughers Bay). Britt. & Wils. Ocimum L. Ocimum basilicum L. Cultivated and perhaps spontaneous. “Basil.” O. micranthum Willd. Fairly common, especially along the south coast. “Wild Basil.” 450, 451, 452, 453; Fishlock 292 (Pasea Hall Estate). Britt. & Wils. Salvia L. Salvia occidentalis Sw. Plentiful in moist districts. Fishlock 52 and 288 (Great Mountain). Britt. & Wils. S. serotina L. Mountain districts, Seen at Cane Garden Bay. Many plants of this species could not be distinguished from the St. Thomas endemie S. thomasiana Urb. using Britt. & Wils. key and text. 333, 339, 340. Britt. & Wils. Thymus L. Thymus vulgaris L. Cultivated. “Thyme.” Family 81. SOLANACEAE Acnistus Schott. Acnistus arborescens (L.) Schlechtendal. Common in wet mountain districts. Seen in thickets at Meyers and Chalwell Estates. Brunfelsia L. Brunfelsia americana L. Cliffs and thickets near the coast, and on open hillsides above Road Town. Plants on Tortola exhibit con- siderable variation in aspect, leaf form, tube length, and berry size, and further investigation might show more than one species is represented. 344, 598; Fishlock 40 (Sea Cow Bay). B. undulata Sw. Widely cultivated for ornament. Capsicum L. Capsicum frutescens L. Plentiful at upper elevations, but seen down to sea level in wet seasons. “Jumbie Pepper," “Bird Pepper." 624; Fishlock 155 (Experiment Station). Britt. & Wils. Cestrum L. Cestrum diurnum L. Cultivated and naturalized near villages in moist districts. “Lady of the Day." 129, 347. C. laurifolium L'Her. A very plentiful shrub or tree at upper eleva- tions. "White Cinnamon." 128, 381, 617; Fishlock 55 (Great Mountain), Fishlock 449 (Sage Mt.). Britt. & Wils. C. nocturnum L. Cultivated and perhaps spontaneous. “Lady of the Night." 1961] Tortola — D'Arcy 441 Datura L. Datura innoxia Mill. Datura metel L. Common along the south coast. “Belladonna Bush." 94; Fishlock 215 (Experiment Station, as D. metel). Britt. & Wils. D. stramonium L. At West End jetty and in the ghut, Long Look. In wet seasons it extends its range, but in dry years it persists in only these two locations, both of them very dry. *Belladonna Bush." 95; Fishlock 492 (Experiment Station). Britt. & Wils. Physalis L. Physalis angulata L. Plentiful in all parts of the island. 238; Fish- lock 286 (Pasea Hall Estate). Britt. & Wils. P. pubescens L. Fishlock 5, 285, and 286 (Experiment Station). Britt. & Wils. P. turbinata Medic. Common in waste places near ghuts or dense thickets. 559, 560, 561, 562, 563; Fishlock 299 (Pasea Hall Es- tate). Britt. & Wils. Solanum L. Solanum elaeagnifolium Cav. Only one small patch east of the jetty, East End. 44. S. erianthum D. Don. Solanum verbascifolium L. Occasional plants, mostly in thickets, throughout the island. 54; Fishlock 415 (Pasea Hall). Little & Wads., Britt. & Wils. . ficifolium Ort. Solanum torvum Sw. Occasional plants throughout the island. Sometimes used medicinally. “Shoo Shoo Bush." 119: Fishlock 185 (Experiment Station). Britt. & Wils. S. lanceifolium Jacq. Rather rare. Seen at Sage Mountain, Brewers Bay (sea level), and in Road Town. The berries of the Sage Mt. plants were hispid, the others were glabrous. 86. S. melongena L. Widely cultivated, and occasionally spontaneous. *Melongae," *Egg Plant." . nodiflorum Jacq. Solanum migrum L. sensu Britt. & Wils. Occa- sional plants throughout the island. Some are believed to be poisonous, while others are eaten by children. The specific status of these plants needs further investigation. 543, 544, 545, 546, 547: Fishlock 8 (Experiment Station). S. persicaefolium Dunal. Common throughout the island but especially common in dry eastern districts. 33; Fishlock 168 (Lower Estate). Britt. & Wils. S. polygamum Vahl. Very common in coastal thickets. “Cakalaka Berry." Fishlock 70 (Belleview), Fishlock 101 (Harbours). Britt. & Wils. Nicotiana tabacum L. (tobacco) was extensively planted 25 years ago, but there is no sign of it now. Solanum tuberosum L. (Irish potato) is planted from time to time with indifferent success. Lycopersicum esculentum (L.) Mill. is cultivated as a market vegetable (tomato). N un 442 Rhodora [Vol. 69 Family 82. SCROPHULARIACEAE Bacopa Aubl. Bacopa monnieri (L.) Pennell. Bramia monnieri (L.) Drake. In freshwater ponds and marshy areas. Seen at Towers, Flamingo Pond, and Carrot Bay. 688, 689, 690, 691, 752, 753, 754, 755; Fishlock 311 (Purcells). Britt. & Wils. Capraria L. Capraria biflora L. Common almost everywhere. 327; Fishlock 211 (Experiment Station). Britt. & Wils. Russellia Jacq. Russellia equisetiformis Schlechtendal & Cham. Cultivated for orna- ment. Fishlock 212 (Road Town). Scoparia L. Scoparia dulcis L. Very plentiful in Long Bay East, uncommon else- where. 165, 166, 399, 400, 401, 402, 403; Fishlock 113 (Huntums Ghut), Fishlock 284 (Purcells). Britt. & Wils. Family 83. BIGNONIACEAE Crescentia L. Crescentia cujete L. At lower elevations, especially along the south coast, and occasionally seen in the mountains. “Calabash Tree." Fishlock 208 (Pasea Hall Estate). Little & Wads. Doxantha Miers Doxantha unguis-cati (L.) Rehder. Batocydia unguis (L.) Mart. Very plentiful in all moist districts. When it is in full bloom, the hillsides above Cane Garden Bay take on a glorious yellow color. “Cat’s Claw." 114, 116; Fishlock 37 (thickets). Britt. & Wils. Jacaranda Juss. Jacaranda mimosifolia D. Don. Jacaranda acutifolia H. & B. Rarely cultivated. The continuous salt spray in most parts of the island limits its expansion. Pandorea Spach Pandorea ricasoliana (Tanfani) Baill. Commonly cultivated. Phyrganocydia Mart. Phryganocydia corymbosa (Vent.) Bur. & Schum. Rarely cultivated, but large masses at Fonseca's Corner and Government House grounds attract attention. 432, 433, 434, 435, 436. Spathodea Beauv. Spathodea campanulata Beauv. Only three trees known, two in front of the Cottage Hospital, Road Town, and one in Huntum's Ghut. "African Tulip Tree." Tabebuia DC. Tabebuia pallida (Lindl) Miers. Tabebuia heterophylla (DC.) Brit- ton. One of the most plentiful tree species, occurring mostly at lower elevations. Several of the large leaved, large flowered 1967] Tortola — D'Arcy 443 form are on Government House grounds. “White Cedar." Fish- lock 334 (hillsides). Britt. & Wils. T. rufescens J. R. Johnston. Only known from Government House grounds and Kingstown Hill. A very large tree was accidentally destroyed at the Road Town Experiment Station about five years ago. 11, Tecoma Juss. Tecoma stans (L.) H. B. K. Very common along the south coast, but sometimes ascending to ca. 300 m. "Ginger Thomas." 261; Fish- lock 201 (Pasea Hall Estate). Little & Wads. Tecomaria Spach Tecomaria capensis (Thunb.) Spach. Commonly cultivated for orna- ment. Family 84. MARTYNIACEAE Martynia L. Martynia annua L. Seen only in the vicinity of the Road Town Experi- ment Station. *Wild Okra." 106; Fishlock 482 (waste grounds). Family 85. GESNERIACEAE Episcia Mart. Episcia fulgida Hook. Commonly cultivated and rarely escapes. *Strawberry Plant." Family 86. ACANTHACEAE Anthacanthus Nees. Anthacanthus spinosus (Jacq.) Nees. In coastal rocks; spiny and spineless forms are seen growing side by side. 808 (Camanoe), 837; Fishlock 105 (Harbours). Britt. & Wils. Asystasia Blume Asystasia gangetica (L.) T. Anders. Commonly cultivated and es- caped, perhaps naturalized. 307, 308, 309, 310. Blechum P. Browne Blechum pyramidatum (Lam.) Urban. Blechum blechum (L). Millsp. A common and troublesome weed. 135; Fishlock 353 (Experiment Station). Crossandra Salisb. Crossandra infundibuliformis (L.) Nees. Commonly cultivated. Dicliptera Juss. Dicliptera assurgens (L.) Juss. Diapedium assurgens (L.) Kuntze. On walls and waste places, Pasea Hall Estate. 742, 743, 744, 745, 746. Justicia L. Justicia carthaginensis Jacq. Fishlock 225 (Road Town). Britt. & Wils. J. periplocifolia Jacq. One plant seen, not collected, at Slaney Point. It is very plentiful on Peter Island. 464 (Peter Island). J. sessilis Jacq. Plentiful in moist districts. “Rock Balsam." 137, 138; Fishlock 362 (Huntum’s Ghut). Britt. & Wils. 444 Rhodora [Vol. 69 Pachystachys Nees Pachystachys coccinea (Aubl.) Nees. Cultivated for ornament, Gov- ernment House grounds. Pseuderanthemum Radlk. Pseuderanthemum atropurpureum Radlk. Commonly cultivated. Ruellia L. Ruellia coccinea (L.) Vahl. Fishlock 54 (Great Mountain). Britt. & Wils. R. tuberosa L. An extremely troublesome weed. "Many Roots." White forms occur with the blue forms at Kingstown Hill. 110, 242, 243, 244, 245; Fishlock 128 (Experiment Station). R. tweediana Griseb. Occasionally cultivated. Not seen to escape. Thunbergia Retz. Thunbergia alata Bojer. Fishlock 180 (Road Town). Britt. & Wils. T. erecta T. Anders. Meyenia erecta (T. Anders.) Benth. Cultivated for ornament. T. fragrans Roxb. Occasional plants, moist districts. Not seen in cultivation. 576, 577, 578, 579, 580; Fishlock 182 (Road Town). Britt. & Wils. Family 86. MYOPORACEAE Bontia L. Bontia daphnoides L. Seen only in a large stand in the salt marsh at The Towers. “Alling.” 131; Fishlock 100 (Harbours). Britt. & Wils. Family 87. PLANTAGINACEAE Plantago L. Plantago major L. Seen only in a small patch at Meyers. 127, 187, 188; Fishlock 63 (Great Mountain). Britt. & Wils. Family 88. RUBIACEAE Borreria G. F. W. Mey Borreria laevis (Lam.) Griseb. A troublesome weed almost every- where. Fishlock 57 (Experiment Station), Fishlock 280 (Joe's Hill). Britt. & Wils. B. ocimoides (Burm. f.) DC. Mountain pastures, not very plentiful. 428. Britt & Wils. Chiococca P. Browne Chiococca alba (L.) Hitchc. Occasional plants in ghuts and thickets throughout the island. Sometimes used to make a beverage. "Snakeroot." 65,66. Britt. & Wils. Chione DC. Chione venosa (Sw.) Urban. Britt. & Wils. Coffea L. Coffea arabica L. Several cultivated bushes in the Agricultural Sta- tion, Road Town, but not seen in fruit or flower. The same species is locally reported as naturalized in the Sage Mountain forest. Little & Wads. 1967] Tortola — D'Arcy 445 Diodia L. Diodia apiculata (Willd.) Schum. Seen only in large patches behind Trellis Bay, Beef Island. 789, 790, 798. Erithalis P. Browne Erithalis fruticosa L. Common along the sea coasts, especially in western parts of the island. “Black Torch." 258; Fishlock 429 (Belmont). Britt. & Wils. Ernodea Sw. Ernodea littoralis Sw. Seen only on Belmont Beach and in Trellis Bay, Beef Island. 593, 594, 595. Britt. & Wils. Exostema L. C. Rich. Exostema caribaeum (Jacq.) R. & S. Common along the seacoast. Conspicuous when in flower and fruit. Used medicinally. “Lorch,” 92: Fishlock 486 (Kingstown). Little & Wads., Britt. & Wils. Faramea Aubl. Faramea occidentalis (L.) A. Rich. Fairly common in Sage Mountain forests. 717; Fishlock 86 (Sage Mt.). Little & Wads., Britt. & Wils. Gardenia Ellis Gardenia jasminoides Ellis. Very commonly cultivated and long per- sistent, appearing spontaneous although probably not so. Rarely or never fruiting. Geophila D. Don Geophila repens (L.) J. M. Johnst. Geophila herbacea (Jacq.) Schum. Eggers 3236. Gonzalagunia Ruiz & Pav. Gonzalagunia spicata (Lam.) Duggena hirsuta (Jacq.) Britt. Plenti- ful in forests and pastures at upper elevations. Sometimes a shrub to 2 m., and rarely a high climbing vine. 140; Fishlock 327 (Joe’s Hill), Fishlock 786 (Great Mountain). Britt. & Wils. Guettarda L. Guettarda parviflora Vahl. Britt. & Wils. Little & Wils. G. scabra (L.) Vent. Guettarda scabra (L.) Lam. Britt. & Wils. Little & Wads. Ixora L. Ixora ferrea (Jacq.) Benth. Eggers 2195. Little & Wads., Britt. & Wils. Several species of Ixora are planted for ornament. Most common are: I. chinensis (salmon), I. coccinea (red), I. finlaysonia (white). Morinda L. Morinda citrifolia L. Common along sea coasts, especially in the Road Harbour and Sea Cow Bay areas. Used medicinally. “Painkiller Tree.” 794; Fishlock 488 (Road Town). Little & Wads., Britt. & Wils. Palicourea Aubl. Palicourea crocea (Sw.) R. & S. Palicourea riparia Benth. Common 446 Rhodora [Vol. 69 in wet mountain forests. Its showy red and orange flowers strong- ly recommend its use as an ornamental. *Yellow Cedar." 139, 369, 370, 371; Fishlock 94 and 363 (Sage Mt.); Eggers 3193 (High Bush, 1200’). Britt. & Wils. P. domingensis (Jacq.) DC. Rather rare in Sage Mountain forests. Its white flowers and wand-like aspect suggest suitability as an ornamental. 98; Fishlock 148 (Experiment Station). Britt. & Wils. Psychotria L. Psychotria brownei Spreng. Eggers 3233. Britt. & Wils. P. ligustrifolia (Northrop) Millsp. Plentiful between Joe's Hill and Meyers. 280; Eggers 3232. P. microdon (DC.) Urban. Psychotria pinularis Sesse & Mocino. Occasional in thickets throughout the island. 60, 820 (Camanoe) ; Fishlock 296 (Pasea Hall Estate). Britt. & Wils. P. undata Jacq. Britt. & Wils. Randia L. Randia aculeata L. Randia mitis L. Plentiful, especially on lower hill- sides and ghuts, sometimes troublesome as a weed. This is the traditional Christmas Tree on Tortola. “Fishing Rod.” 47, Fish- lock 210 (Pasea Hall Estate). Little & Wads., Britt. & Wils. Rondeletia L. Rondeletia pilosa Sw. Roadsides and thickets, moist districts. Plentiful at Hope and on Joe’s Hill. 350; Fishlock 434 (Zion Hill). Little & Wads., Britt, &. Wils. Spermacoce L. Spermacoce confusa Rendle. Spermacoce tenuior L. sensu Britt. & Wils. Fishlock 11 (Experiment Station). Britt. & Wils. Portlandia grandiflora L. and Mussaénda luteola Delile are rarely cultivated for ornament. Family 89. CAPRIFOLIACEAE Sambucus L. Sambucus simpsonii Rehder. Occasionally cultivated and long per- sistent, appearing as if naturalized. Fruits not seen. Family 90. CUCURBITACEAE Cayaponia Manso Cayaponia americana (Lam.) Cogn. Widespread in moist districts, forming large masses on Joe’s Hill and Brewers Bay. 123, 351; Fishlock 4 (Experiment Station). Britt. & Wils. C. racemosa (Mill. Cogn. Cayaponia racemosa (Sw.) Cogn. Britt. & Wils. Citrullus Forst. Citrullus lanatus (Thunb.) Mansfeld. Citrullus citrullus (L.) Karst. Occasionally cultivated. A fine crop comes to Tortola each year from Peter Island. *Watermelon." 1967] Tortola — D'Arcy 447 Cucumis L. Cucumis anguria L. Common, especially in dry districts. “Wild Cucumber.” Used for food. 324; Fishlock 154 (Experiment Station). Britt. & Wils. C. sativa L. Cultivated as a market vegetable. Lagenaria Ser. Lagenaria siceraria (Molina) Standl. Cucurbita lagenaria L. Widely cultivated, and escaping in wet seasons. “Sweet Gourd.” “Gourdie.” Fishlock 295 (Pasea Hall Estate). Britt. & Wils. Melothria L. Melothria guadalupensis (Spreng.) Cogn. Common in wet mountain districts, rare elsewhere. 334; Fishlock 21 (Experiment Station). Britt. & Wils. Momordica L. Momordica charantia L. A widespread and very troublesome vine. A decoction of the leaves is used medicinally as “Bitter Bark.” “Maiden Apple.” Fishlock 23 (Experiment Station). Britt. & Wils. Sechium P. Browne Sechium edule (Jacq.) Sw. Cultivated on the north side of the island. “Christophene.” Family 91. CAMPANULACEAE Hippobroma G. Don Hippobroma longiflora (L.) G. Don. Isotoma longiflora (L.) Presl. Common in pastures wet mountain areas, 82. Britt. & Wils. Family 92. GOODENIACEAE Scaevola L. Scaevola plumierii (L.) Vahl. Although all beaches were searched for this species, it was found only at Long Bay West. It is plentiful on the north side of Peter Island. 321; Fishlock 424 (West End). Britt. & Wils. Family 93. COMPOSITAE Acanthospermum Schrank Acanthospermum hispidum DC. A very plentiful and widespread weed, covering several acres in Free Bottom. 173, 174, 175; Fishlock 186 (Experiment Station). Britt. & Wils. Ageratum L. Ageratum conyzoides L. Common on hillsides, upper elevations. Dwarf forms, 4 cm. tall, are to be found in windswept pastures, Sage Mountain. 185, 681, 682, 683: Fishlock 281 (Leonards) ; Eggers 3222. Britt. & Wils. Ambrosia L. Ambrosia hispida Pursh. Seen in a large patch on Great Mountain (ca. 300 m.) 189. Britt. & Wils. 448 Rhodora [Vol. 69 Bidens L. Bidens cynapiifolia H. B. K. A common weed. 634, 635, 636; Fish- lock 29 (Experiment Station). Britt. & Wils. B. pilosa L. A common weed. 625, 626, 627. Borrichia Adans. Borrichia arborescens (L.) DC. Seen only at Trellis Bay, Beef Island. 773. Brachyrampus DC. Brachyrampus intybaceus (Jaeq.) DC. Fishlock 273 (Brandywine Bay). Britt. & Wils. Chaptalia Vent. Chaptalia nutans (L.) Polak. Fishlock 346 (Joe's Hill). Conyza Less. Conyza bonariensis (L.) Cron. Leptilon bonariense (L.) Small. Britt. & Wils. C. canadensis (L.) Cron. Erigeron canadensis L. Leptilon pusillum (Nutt.) Britt. Widespread. Plentiful in Sage Mountain pastures, and in abandoned cultivation, Josia's Bay. 200, 201, 202. Cosmos Cav. Cosmos caudatus H. B. K. Common in wet mountain pastures, and coming down to lower elevations in wet seasons. 348; Fishlock 48 (Great Mountain); Eggers 3226. Britt. & Wils. C. sulphureus Cav. Cultivated and sometimes spontaneous. Eclipta L. Eclipta prostrata (L.) L. Beside the Freshwater Pond, Towers. 674, 675, 676; Fishlock 439 (Zion Hill). Britt. & Wils. Elephantopus L. Elephantopus scaber L. Elephantopus mollis H. B. K. Common in pastures and roadsides, upper elevations. 420, 421, 422, 423, 424; Fishlock 309 (Purcells) ; Eggers 3224. Britt. & Wils. Emilia Cass. Emilia coccinea (Sims) Sweet. Common at upper elevations. 270, 271, 272. E. sonchifolia (L.) DC. Fishlock 238 and 356 (Experiment Station). Britt. & Wils. Erechtites Raf. Erechtites hieracifolia (L.) Raf. Fishlock 19 (Experiment Station). Britt. & Wils. Erigeron L. Erigeron cuneifolius DC. Britt. & Wils. Eupatorium L. Eupatorium corymbosum L. Osmia corymbosa (Aubl.) Britt, & Wils. Common at upper elevations. “Christmas Bush." 356, Britt. & Wils. 1967] Tortola — D’Arcy 449 E. odoratum L. Osmia odorata (L.) Schultz. Common in moist dis- tricts, at all elevations. “Christmas Bush." 410; Fishlock 319 (Experiment Station). Britt. & Wils. E. sinuata Lam. Osmia sinuata (Lam.) Britt. & Wils. Seen at sea level, Jackass Ghut. 731, 732, 733, 734. Lactuca L. Lactuca sativa L. Cultivated for market. “Lettuce.” Melanthera Rohr Melanthera nivea (L.) Small. Melanthera confusa Britt, Wet moun- tain districts. 684, 685, 686; Fishlock 440 (Zion Hill), Fishlock 484 (Jost Van Dyke). Britton used Fishlock 440 as the type for M. confusa. Britt. & Wils. Mikania Willd. Mikania cordifolia (L. f.) Willd. Common in thickets and old ruins, upper elevations. 426; Fishlock 112 (New Bush). Britt. & Wils. Neurolaena R. Br. Neurolaena lobata (L.) R. Br. Fishlock 87 (Sage Mountain). Britt. & Wils. Parthenium L. Parthenium hysterophorus L. À very common and extremely trouble- some weed. “Bitter Weed." 38, 39, 40; Fishlock 255. Britt. & Wils. Pectis L. Pectis linifolia L. Seen only by the main road, Pasea Hall Estate. 652, 653, 654, 655. Piptocoma Cass. Piptocoma antillana Urban. Piptocoma rufescens Cass. Common at sea level, north side of Peter Island, but not seen on Tortola. Probably of ornamental value. 373, 374 (Virgin Gorda), 375, 376, 377. Pluchea Cass. Pluchea carolinensis (Jacq.) D. Don. Pluchea odorata (L.) Cass. Forming large shrubby masses at Hannah, Paraquita Bay and Fat Hog Bay. 130; Fishlock 17 (Pasea Hall Estate). Britt. & Wils. P. purpurascens (Sw.) DC. Seen only in the Freshwater Pond, Towers. 677, 678, 700; Fishlock 438 (Zion Hill). Britt. & Wils. Pseudo-elephantopus Rohr Pseudo-elephantopus spicatus (B. Juss. ex Aubl.) Gleason. Pseudo- elephantopus spicatus (Juss.) Rohr. Common at upper elevations and coming down to sea level in moist ghuts. 133; Fishlock 337 (Joe's Hill); Eggers 3230. Britt. & Wils. Pterocaulon Ell. Pterocaulon virgatum (L.) DC. Fishlock 13 (Pasea Hall Estate). Britt. & Wils. 450 Rhodora [Vol. 69 Solidago (Vaill.) L. Solidago microglossa DC. Common at Meyers and widespread in mountain districts. Occasionally taken to sea level for cultiva- tion. 132, 628, 629, 630. Sonchus L. Sonchus oleraceus L. Fishlock 343 (Joe's Hill). Britt. & Wils. Synedrella Gaertn. Synedrella nodiflora (L.) Gaertn. Fishlock 22 (Experiment Station). Britt. & Wils. Tithonia Desf. Tithonia diversifolia (Helmsl.) A. Gray. In large patches at upper elevations; Great Mountain to Meyers. 134, 490. Tridax L. Tridax procumbens L. A common weed in the Road Town area. 170, 171, 172, 649, 650, 651. Verbesina L. Verbesina alata L. Tepion alatum (L.) Britton. Common although scattered, moist districts. Used medicinally. “Information (In- flammation?) Bush." 335, 337, 338. Vernonia L. Vernonia albicaulis Pers. Common at upper elevations. 799. V. cinerea (L.) Less. A very common and troublesome weed. 35, 36, 37; Fishlock 3, 28 (Experiment Station). Britt. & Wils. V. sericea L. C. Rich. Common at upper elevations. 275, 276, 277, 600, 601, 602. Wedelia Jacq. Wedelia calycina L. C. Rich. Seen only beside the airport, Beef Island. 813, 814, 815. W. parviflora L. C. Rich. Britt. & Wils. W. trilobata (L.) Hitche. Common in wet fields and ditches, lower elevations. 822, 823, 824; Fishlock 445 (Purcells). Britt. & Wils. Xanthium L. Xanthium strumarium L. Xanthium chinense Mill. Very plentiful at Long Look. 341; Fishlock 300 (Pasea Hall Estate). Britt. & Wils. CONNECTING VOWELS IN EPITHETS OF LATIN ORIGIN' BERNARD BOIVIN Latin compound words are commonly formed with the letter i as a connecting vowel. Certain other methods of forming compound words also occur in Latin, but are of rather limited application or usage. Article 73 note 2 of the International Code of Botanical Nomenclature calls for uniformity in the formation of names by stating that: “The use of a wrong connecting vowel or vowels (or the omission of a connecting vowel) in a name or an epithet is treated as an orthographic error". And the two examples given are opuntiaeflora, corrected to opuntiiflora, and napeaefolia corrected to napaeifolia. Some exceptions are given under Recommendation 73G. Before a vowel, a final vowel is normally elided: mult- angulus, not multi-angulus. The ablative ending may be retained as in atropurpureus and fuscovenetus. Forms re- vealing etymological distinctions may also be retained, such as caricaeformis from Carica versus cariciformis from Carex. Other examples of the intent of the Rules can be culled from various parts of the text of the Code. These are: acutiflorus, angustifolius, biflorum, brevifolia, brevipedunculata, eruciformis, fragiferum, gossipiifolia, grandiflora, hieraciifolium, latifolia, longisiliquum, menthifolius, multicaulis, multicolor, multi- florum, napifolius, 'neriifolia, rumicifolia, salicifolia, salviifolius, sorediiformis, tricolor, tricuspis, trinervis. Two epithets used in the Rules appear to be in need of correction. Fimbricalyx, Art. 42, should read: fimbriicalyx, compare: sorediiformis. Contortuplicatus, Rec. 21A, is not only contrary to Art. 73, it is also a very old word, the classical form of which is contortiplicatus. "Contribution No. 20, Faculty of Agriculture, Laval University, Quebec, Canada. Contribution No. 580, Plant Research Institute, Department of Agriculture, Ottawa, Canada. 451 452 Rhodora [Vol. 69 The full implication of Art. 73 may have escaped the attention of most botanists. A rapid sampling of the Kew Index, the Gray Index or the indices of various floras from a shelf at hand reveals that they all contain a fair number of epithets contrary to Art. 73. There is no intention on my part to carry out a full survey of all such names that need to be corrected under Art. 73; the task would be stu- pendous. However in skimming through M. L. Fernald, Gray's Manual of Botany, ed. 8, 1950, I have noticed that the following corrections are needed therein: NAME Acalypha ostryaefolia Agastache scrophulariaefolia Aristida longespica Aster dumosus var. subulaefolius Athyrium thelypterioides Beckmannia erucaeformis Betula caerulea-grandis Bidens heterodoxa var. monardaefolia Botrychium matricariaefolium Botrychium ternatum var. rutaefolium Bromus brizaeformis Carex livida var. rufinaeformis Cerastium longepedunculatum Comptonia peregrina var. asplenifolia Coreopsis delphinifolia Cyperus flavescens var. poaeformis Erectites hieracifolia Eupatorium urticaefolium Eupatorium verbenaefolium Gentiana Catesbaei var. nummulariaefolia Helianthus grosseserratus Helianthus trachelifolius Iva xanthifolia Kalmia polifolia Kosteletzkya virginica var. altheaefolia CORRECTION A. ostryifolia A. scrophulariifolia A. longispica var. subulifolius A. thelypteridoides B. eruciformis B. caeruleigrandis var. monardifolia B. matricariifolium var. rutifolium B. briziformis var. rufiniformis C. longipedunculatum var. aspleniif olia, C. delphiniifolia var. poiformis E. hieracüfolia E. urticifolium E. verbenifolium var. nummulariifolia H. grossiserratus H. trachelitfolius Iva xanthiifolia K. polüifolia var. altheifolia Latin Epithets — Boivin 453 Lindera melissaefolium Liparis lilifolia Lycopodium sabinaefolium Lyonia ligustrina var. capreaefolia Monarda fistulosa var. menthaefolia Monarda menthaefolia Myrica asplenifolia Osmunda cinnamomea f. cornucopiaef olia, Paspalum longepedunculatum Paspalum setaceum var. longepedunculatum Penstemon tubaeflorus Polygonum pensylvanicum var. rosaeflorum Potamogeton spathulaeformis Quercus falcata var. pagodaefolia Quercus macrocarpa f. olivaeformis Quercus pagodaefolia Ranunculus bulbosus var. valdepubens Ranunculus sicaeformis Salvia lanceaefolia Sambucus pubens f. rosaeflora Smilax Bona-Nox var. hederaefolia Sium cicutaefolium Tragia nepetaefolia Valerianella chenopodifolia Veronica hederaefolia Viola melissaefolia Zenobia cassinefolia L. melissifolium L. liliifolia L. sabinifolium var. capretfolia var. menthifolia M. menthifolia M. aspleniifola f. cornucopiifolia P. longipedunculatum var. longipedunculatum P. tubiflorus var. rosiflorum P. spathuliformis var. pagodifolia f. oliviformis Q. pagodifolia var. valdipubens R. siciformis S. lanceifolia f. rosiflora var. hederifolia S. cicutifolium T. nepetifolia V. chenopodiifolia V. hederifolia V. melissifolia Z. cassinif olia Corrections of connecting vowels cannot be done indis- criminately and I hope I have not made too many unjustified corrections. A few examples of special cases may help the reader recognize some of the unusual compounds that should be allowed to stand. Millefolium: this is an old classical word already in use by the third century A.D. and should be allowed to stand 454 Rhodora [Vol. 69 undisturbed. In compounds the forms mille- and milli- are equally acceptable. E.g., milli-formis, millepeda, millegrana. Similarly, with pedemontanus, Vincetoxicum, Cerefolium, graveolens, suaveolens, ete. Noveboracus: the latin form of York is Eboracus, hence the elision of a final vowel and the omission of a connecting vowel. Anethiodora: a better form would be anethodora, but the paragraph on the elision of a final vowel before an initial vowel is only a recommendation (73G), hence a cor- rection is not compulsory. Note also: asplenioides, brevi- aristatus. Antecedens: the preposition ante may either be used un- modified or it may be changed to anti. There is ample classical usage to justify both forms. Other prepositions and adverbs are also commonly used in compounds either unmodified or without connecting vowels; e.g.: semper- virens, interruptus, retroflexus, praematurus. Similarly with numerals: quinquefolius. Spica-venti: this and other similar epithets are termed "pseudo-compounds" in the Rules (73G), and are exempt from the provisions of Art. 73 note 2. They are also exempt from the provisions for the elimination of hyphens in com- pound words. Otherwise, hyphens should be eliminated in compounds. Muscaetoricum: appears to fall in the category of "pseudo-compounds" and should be allowed to stand un- changed. While reading Gray's Manual for the purpose of this survey, I paid attention only to connecting vowels in latin compounds. Yet I could not help noticing that a great many more orthographic errors should be eliminated. The most frequently needed correction is the lack of an additional 7 in certain genitives: Helianthus Maximiliani (for Mazi- milianii). Here is a sampling of certain other types of cor- rection also needed: Habenaria blephariglottis (blephari- diglottis); Panicum calliphyllum | (callophyllum); Acer Pseudo-Platanus (pseudoplatanus ); Franseria acanthicarpa 1967] Latin Epithets — Boivin 455 (acanthocarpa); Veronica Anagallis-aquatica f. anagalli- formis (anagallidiformis ). As it presently stands, Art. 73 is a rather mild attempt at spelling-standardization. Yet botanists seem pretty lackadaisical about its observance. Under the circumstances it hardly seems worth while to propose still further spelling- standardizations. Yet I would strongly support an Article 73 that would eliminate such variants as caeruleus, ceruleus and coeruleus; littoralis and litoralis; allegheniensis and alleghaniensis; pensylvanicus, pennsylvanicus and pensilva- nicus, and most of all, the multitude of alternate forms of kamtschaticus and alaschkensis. The latter are especially vexing. LOUIS-MARIE HERBARIUM, LAVAL UNIVERSITY, QUEBEC AND DEPARTMENT OF AGRICULTURE, OTTAWA, CANADA ANOTHER TRIBULUS ADVENTIVE IN THE NEW WORLD Tribulus L. (Zygophyllaceae) is an Old World genus of perhaps several dozen species, two of which are well-known adventives in the Americas. Tribulus cistoides L. is a com- mon weed in tropical Mexico and the Carribbean region, and T. terrestris L. a ubiquitous pest in the warm temperate areas of both North and South America. While examining specimens of this genus from South America, I discovered a third adventive, T. alatus Del. represented by the fol- lowing collections from western Peru: DEPT. PIURA: Pariñas Valley, Haught F-143 (F), 206 (GH); Talara, Johnston 3513 (F, GH); 1 km W Talara, Beetle 26201 (UC). DEPT. ICA: between Cocharcas and Quilque, Ferreyra 580 (US). Information on the labels of the Beetle and Johnston collections indicate that the plant was not common where found, but Haught F-143 states that the plant was “abundant after rains, especially north of Parifas Valley." This desert-dwelling annual from North Africa and the Middle East to India is easily recognized by its winged 456 Rhodora [Vol. 69 mericarps. Vegetatively, Tribulus alatus is similar to more pubescent examples of T. terrestris, which also is found in Peru. The three New World adventives may be separated readily by the following key: l. Perennial; flowers 2-4 cm in diameter; intrastaminal glands con- nate, forming a 5-lobed ring around ovary base ....... T. cistoides. 1. Annual; flowers 5-15 mm in diameter; intrastaminal glands free: 2. Flowers 5-10 mm in diameter; mericarps dorsally 2-4-spined —— RR T. terrestris. 2. Flowers 10-15 mm in diameter; mericarps winged on margin of COVSUM MM T. alatus. In addition to the adventives, a number of species of Tribulus have been described from the New World as indi- genes, Of these, T. alacranensis Millsp., from the Arrecife Alacrán, Yucatán, Mexico, and T. sericeus Anderss., from the Galápagos Islands, Ecuador, are synonyms of T. cis- toides. All others prove to be members of the closely related Kallstroemia Scop. DUNCAN M. PORTER DUDLEY HERBARIUM, STANFORD UNIVERSITY STANFORD, CALIFORNIA 94305 THE FLORA OF EASTERN HIMALAYA’ This handsome volume is a report of the botanical expe- ditions of the University of Tokyo to the eastern Himalayas in 1960 and 1963. It is a contribution to the flora of that area of high significance, not only because of the care with which it has been prepared but also because of its broad scope and biological orientation. The volume is based on some sixty thousand specimens representing about three "The Flora of Eastern Himalaya — Results of the Botanical Expe- ditions to Eastern Himalaya — Organized by the University of Tokyo 1960 and 1963. Compiled by Hiroshi Hara. i-X, pp. 1-744, pl. 1-40, figs. 1-68, route maps. University of Tokyo Press. 1966. ($32.00 from the University of Tokyo Press, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan.) 1967] Hamamelis — Johnson & Smithberg 457 thousand species of seed plants, ferns, bryophytes, lichens and fungi. The geographic records of the species are an important contribution but the catalogue of the collections is much more than that. The nomenclature and bibliog- raphy have received careful attention and the taxonomic work has been critical. Several new species are described and comments on old ones are frequent. There is a chapter on cytology which deals with sixty species and illustrates the somatic chromosomes of nine- teen of them. The chapter on phytogeography presents an excellent account of the vegetation and a detailed compari- son of the floristic relations between the eastern Himalaya and Japan. This latter study involves over two hundred and fifty Himalayan species and their Japanese counter- parts. A series of photographs, twenty-one of them in color, illustrate the principal vegetational features, selected species and floral details. This book will not only be widely used by botanists con- cerned with the Himalayan flora directly but will be of special interest to those interested in the eastern North American flora and its relationship to eastern Asia. Dr. Hara and his thirty-seven colleagues who contributed to the book are to be congratulated for this fine publication. ROLLA TRYON, GRAY HERBARIUM, HARVARD UNIVERSITY A NEW STATION FOR HAMAMELIS VIRGINIANA L. IN MINNESOTA. — Common witch-hazel (Hamamelis virginiana L.) is widely distributed in eastern North Amer- ica reaching the northwest limit of its range in Minnesota. Until recently it has been known only from Winona and Houston counties in the extreme southeast corner of the state. In 1960 Mr. Richard Brand, county agent of Todd County in central Minnesota requested verification of the identifi- cation of a witch-hazel plant found on a farm of Mr. Leon 458 Rhodora [Vol. 69 Robideau in Gray Eagle Township in that county (SW l4, of the NW 1⁄4, Sec. 24, T. 127N, R 32W). Information from Mr. Robideau, confirmed by Mr. Brand, indicated that this was a wild shrub growing in a wooded area never subjected to cultivation. Mr. Robideau further commented that this was the only known specimen on his farm and that no others were known to exist in the vicinity. This occurrence is a range extension of approximately 200 miles northwest of the nearest previously known station in Minne- sota at Quinn's Bluff, Winona County, and about 100 miles west of the nearest approach of witch-hazel in Polk County, Wisconsin. To obtain propagating material for the perpetuation of this unusual witch-hazel plant the site was visited on Octo- ber 19, 1963, Accompanied by Mr. Robideau we found the plant along a trail on his farm, Associated trees consisted in part of Ironwood (Ostrya virginiana (Mill.) K. Koch), Blue-Beech (Carpinus caroliniana Walt.), Quaking-Aspen (Populus tremuloides Michx.), Paper-Birch (Betula papyri- fera Marsh.), Red Oak (Quercus rubra L.), Bur-Oak (Q. macrocarpa Michx.). American Hazel (Corylus americana Walt.), Prickly-Ash (Zanthoxylum americanum Mill.), Wolfberry (Symphoricarpos occidentalis Hook.), and Lea- therwood (Direa palustris L.) were common associated shrubs. The witch-hazel plant was about 12 feet high, apparently quite old, as evidenced by the much thickened base and rotten stubs of former stems. The shrub was composed of several stems up to 4” in diameter and of a number of sprouts of moderate vigor. The plant was in flower at the time of observation but no fruit was present. The flowers of Hamamelis are perfect and self-incompatibility may be the factor involved in the failure to produce seed. Although fruit was absent, a number of fruit-like spiny galls was noted on the sterile fruit pedicels. These are pro- duced by the spiny witch-hazel gall aphid, Hamamelistes spinosus Shimer, an insect with a complex life cycle alter- nating between birch (in this case Betula papyrifera) and 1967] Hamamelis — Johnson & Smithberg 459 witch-hazel. The presence of this insect on an isolated plant suggests that until comparatively recent times there may have existed a more or less continuous population of witch- hazel between the Grey Eagle station and the present general northwest limit of the species along the St. Croix and Mississippi Rivers. Although no seed was obtained from this outlier speci- men of witch-hazel, the plant had been propagated in a modest way by Mr. Robideau through the removal of rooted basal sprouts. Two small plants thus obtained had been planted near the farm house on the property. One of these Mr. Robideau presented to the University of Minne- sota Landscape Arboretum. Cuttings obtained at the time of this visit rooted satisfactorily in a sand propagating bench but failed to become established when transferred to soil under greenhouse conditions. ALBERT G. JOHNSON AND MARGARET H. SMITHBERG DEPARTMENT OF HORTICULTURAL SCIENCE UNIVERSITY OF MINNESOTA, ST. PAUL THLASPI TUBEROSUM NUTT., A TAXONOMIC SYNONYM OF CARDAMINE DOUGLASSII BRITT. (CRUCIFERAE)! RONALD L. STUCKEY When Thomas Nuttall undertook his journey of scientific exploration into the “Old Northwest” (the Great Lakes Region) in the year 1810, he discovered several new species of plants, twenty of which he described as new in his now classic book, The Genera of North American Plants, and a Catalogue of the Species, to the Year 1817, published in 1818. In the course of preparing an account of Nuttall’s plant collections of this trip, I learned that his Thlaspi tuberosum, described from western Pennsylvania, has not been adequately understood by plant taxonomists. It would be an easy matter to make the proper determination of the plant and state the taxonomic position of this species if Nuttall’s original specimen were available for study today. As pointed out by Ewan ( 1952) a mystery surrounds the whereabouts and existence of certain of Nuttall’s plants believed to have been obtained on this trip. My own study of Nuttall’s writings and his plant specimens in the her- barium of the Academy of Natural Sciences of Philadelphia gives evidence that Nuttall may have lost many of his plants of this trip and then replaced some of these with plants obtained later on the same trip, or on another trip, or from another botanist (Stuckey, 1967). The specimen of Thlaspi tuberosum evidently was one of those which was not replaced. No specimen labeled as T. tuberosum has been located in the herbaria of the British Museum (London), the Royal Botanical Gardens (Kew), the City of Liverpool Museums, or the Academy of Natural Sciences, all of which have Nuttall's plants collected previous to the publication of his Genera. Nuttall's description of T. tuberosum in the Genera is quite brief and, as is shown below, even inaccurate at one i ‘Paper No. 731 from the Department of Botany, The Ohio State University, Columbus 43210. 460 1967] Thlaspi Tuberosum — Stuckey 461 point ; consequently later writers have not been able to deter- mine properly to what plant the description refers. De- Candolle (Syst. Nat. 2: 382. 1821 and Prodr. 1: 177. 1824) included T. tuberosum as a valid species, but placed it in a category of Species non satis notae because he did not have sufficient information to determine its relationship to other species. Eaton (Man. 3rd. ed. 486. 1822) wrote a descrip- tion of T. tuberosum nearly identical to that given by Nut- tall, and he continued this species and its description in all the succeeding editions of his Manual. Torrey and Gray (Fl. N. Am. 1: 114. 1838) also included T. tuberosum, but Gray did not mention or give a description of it in the first edition of his Manual (Gray, 1848), nor in the following four editions. Neither did Britton and Brown (1897, 1913) discuss this species. Payson (1926), who revised the genus Thlaspi for North America, stated that the genus was en- tirely lacking in northeastern America, and therefore he made no mention of Nuttall's T. tuberosum. Recent floristic works of northeastern North America, such as Fernald (1950), Gleason (1952), and Gleason and Cronquist (1963) also do not record this species. It therefore appears that the proper status of T. tuberosum has been lost sight of through the years and hàs never adequately: been deter- mined. In a situation where the original specimen is apparently not extant, it is often necessary to turn to other sources of supplementary data, if they are available, in order to gain a clearer understanding of the original description. These data may be found among the author's field notes, diaries, or letters. It is fortunate that Nuttall’s diary of the early portion of his 1810 trip was discovered and subsequently published by Graustein (1951). In this diary Nuttall wrote descriptions of plants he saw and studied. Whether he referred to these descriptions when he wrote the diagnoses in his Genera is not known. Detailed comparisons of these accounts in these two works might be made to determine their similarities and differences which in turn might lead to a«clearer understanding of certain species' diagnoses in 462 Rhodora [Vol. 69 Nuttall's Genera, Upon making several comparisons, I dis- covered that Nuttall’s diagnosis of T. tuberosum in the Genera, appears to be a condensed version of a much more detailed account of a plant he observed near Butler, about 27 miles north of Pittsburgh, and subsequently described in his diary (Graustein, 1951, p. 29). Pertinent data from these two sources are compared in the following chart. GENERA DIARY Name: “Thlaspi tuberosum” “a species of Cardamine?” Root: “tuberous and fibrous” “an irregular tho’ some what roundish tuber, (or rather bulb,) having several knobs or protuberances on its under sur- face" Stem: “pubescent; very short “roundish” and simple" Leaves: "rhomboid-ovate, "radical leaves are from reni- obsoletely toothed, form-cordate to obcordate, most- smooth, and sessile, ly entire tho' sometimes slightly radical ones upon dentate stem leaves, subcordate- long petioles" lanceolate, obtuse, distantly & bluntly toothed, the uppermost leaves sessile & somewhat cu- neate" Flowers: "rather large, like "rather large, . . . of a pale (petals) those of an Arabis, purplish tinge” rosaceous" Fruit: “Silicle suborbiculate, "Silique, erect, compressed. short and compressed" — obcordate" Height: “Not more than 40r 5 “about !4 foot high" inches” Flowering: “April and May” seen 26 April 1810 Habitat: Data not given “in moist shady woods generally nr. rivulets in a rich vegetable earth” Locality: “Western Near Butler, about 27 miles Pennsylvania” north of Pittsburgh Although the data from these two sources do not com- pletely agree, there is little question that the two descrip- tions refer to the same species. The notes on habitat, flowering time, and the diagosis of the plant’s morphology, except the description of the fruit, all point to its being 1967] Thlaspi Tuberosum — Stuckey 463 Cardamine douglassii Britt. Greene (1896), Schulz (1903), and Stuckey (1962), who have previously treated the tax- onomy and nomenclature of Cardamine douglassii or its synonyms, do not mention T. tuberosum, although Watson (A. Gray, Syn. Fl. N. Am. 1(1) : 156. 1895) had suggested that Nuttall’s T. tuberosum was probably the same as C. douglassii (C. rhomboidea (Pers.) DC. var. purpurea Tor- rey) because of its tuberous root, rose-colored flowers, and pubescence. Watson also stated that Nuttall described the pod “as orbicular.” The fruit of C. douglassii is long and somewhat tapered at the apex. To my knowledge there is no member of the Cruciferae of eastern North America which has an obcordate or suborbicular fruit with the vege- tative and flower characters of C. douglassii. This discrep- ancy in the original diagnosis has doubtless been the major reason why botanists have not been able to properly classify Nuttall’s species. Graustein (1951, p. 29, footnote 60) re- garded Nuttall's description to be in error with respect to fruit shape. Was this an error Nuttall made when looking at the plant? Or did Nuttall, in writing his notes, confuse this species with some other species? Examination of the Genera reveals that Nuttall did not give any description or mention any species under Cardamine or Arabis that would include the present-day understood C. douglassii. By com- paring the data in the Genera with the information in the diary, it is clear that Nuttall's note pertaining to Thlaspi arvense is the same species as the T. arvense of today. His reference to T. alliaceum is the Lepidium campestre, and his T. bursa pastoris is the Capsella bursa-pastoris. Thus, other possible species of Thlaspi or species in allied genera which might be considered to agree with Nuttall's descrip- tion of T. tuberosum (at least with respect to the fruit shape he gave) were otherwise included in his Genera. On the basis of publication date only, Nuttall's Thlaspi tuberosum (Gen. 2: 64. 1818) has priority over Britton’s Cardamine douglassii (Trans. N. Y. Acad. Sci. 9: 8. 1889). If T. tuberosum were transferred to Cardamine, it would create a later homonym because there already exists a 464 Rhodora [Vol. 69 Cardamine tuberosa described by DeCandolle (Syst. Nat. 2: 254. 1821) from Chile and illustrated by Delessert (Ic. Sel. Pl. 2: t. 29. 1823). A study of DeCandolle's diagnosis of C. tuberosa, the photograph of the type specimen on a microfiche card of the DeCandolle herbarium (Card No. 69), the drawing in Delessert, and the illustration in Schulz (19083, t. 10, f. 1) reveals that plants of C. tuberosa DC. are very similar to plants of C. douglassii Britt. with respect to size, tuberous base, basal leaves, pedicel position, and flower morphology. However, the cauline leaves are tri- foliate in the former and simple in the latter. Schulz (1903) separated C. douglassii Britt. (C. rhomboidea (Pers.) DC.) from C. tuberosa DC. primarily on the basis of the differ- ence in morphology of the cauline leaves. In addition, Schulz pointed out that the petals of C. tuberosa DC. are white. I conclude that C. tuberosa DC. and C. douglassii Britt. are indeed two different species based on different types, and therefore T. tuberosum Nutt. is properly placed in synonymy under C. douglassii Britt. My thanks are extended to Dr. Alfred E, Schuyler (The Academy of Natural Sciences) who searched for Nuttall's specimen of T. tuberosum at the British Museum, to Dr. Edward G. Voss (The University of Michigan) who has read the manuscript, and to Sir George Taylor (Royal Botanical Gardens, Kew) and Dr. E. F. Greenwood (City of Liverpool Museums) who have answered my letters of inquiry regarding the possibility of Nuttall's specimen of T. tuberosum being in their herbaria. DEPARTMENT OF BOTANY THE OHIO STATE UNIVERSITY, COLUMBUS 43210 REFERENCES BRITTON, NATHANIEL LORD, and ADDISON BROWN. 1897. An Illus- trated Flora of the Northern United States, Canada, and the British Possessions. Ist. ed. Vol. II. Charles Scribner's Sons, New York. 643 pp. 1913. An Illustrated Flora of the Northern United States, Canada, and the British Possessions. 2nd. ed. Vol. II. Charles Scribner’s Sons, New York. 735 pp. 1967] Thlaspi Tuberosum — Stuckey 465 EWAN, JosEPH. 1952. Nuttall’s diary of 1810 and some inquirendae. Rhodora 54: 234-236. FERNALD, MERRITT LYNDON, 1950. Gray’s Manual of Botany. 8th. ed. American Book Co., New York. 1xiv + 1632 pp. GLEASON, HENRY ALLAN. 1952. The New Britton and Brown Illus- strated Flora of the Northeastern United States and Adjacent Canada. Vol. II. [New York Bot. Gard., New York]. 655 pp. , and ARTHUR CRONQUIST. 1963. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. D. Van Nostrand Co., Inc., Princeton. 810 pp. GRAUSTEIN, JEANNETTE E. 1951. Nuttall’s travels into the Old Northwest. An unpublished 1810 diary. Chron. Bot. 14: 1-88. Gray, AsA. 1848. A Manual of the Botany of the Northern United States, from New England to Wisconsin and South to Ohio and Pennsylvania Inclusive, . . . , arranged according to the Natural System. James Munroe and Co., Boston and Cambridge. lxxii + 710 pp. GREENE, Epwarp L. 1896. Studies in the Cruciferae — I. 1. Carda- mine and Dentaria. Pittonia 3: 117-124. NUTTALL, THOMAS. 1818. The Genera of North American Plants, and a Catalogue of the Species, to the Year 1817. 2 Vols. Printed for the author by D. Heartt, Philadephia. 312 pp. and 254 pp. 4 index, eratum [sic], and additions. Payson, EpWIN BLAKE. 1926. Thlaspi, Oreocarya, and Erigeron. 1. The genus Thlaspi in North America. Univ. Wyo. Publ. Bot. 1: 145-163. SCHULZ O. E. 1903. Monographie der Gattung Cardamine. Bot. Jahrb. 32: 280-623. StucKEY, RONALD L. 1962. Characteristics and distribution of the spring cresses, Cardamine bulbosa and C. douglassii, in Michigan. Mich. Bot. 1: 27-34. 1967. The "lost" plants of Thomas Nuttall’s 1810 expedition into the Old Northwest. Mich. Bot. 6: 81-94. A NEW TALINUM (PORTULACACEAE) FROM THE CEDAR GLADES OF MIDDLE TENNESSEE STEWART WARE Talinum calcaricum Ware, sp. nov.— Herba perennis, glabra, 7.0-20.0 cm. alta; rhizoma tuberosum. Caulis in- ferior succulentus, abrupte decrescens ad longum sine foliis stipitem filo metallico similem. Folia linearia, ad 4.5 cm. longa, teretia, aggregata. Inflorescentia vulgo systema tri- cotomum cymarum scorpioidearum, raro cyma singula terminalis scorpioidea. Flores ephemeri, post meridiem aperti; sepala 2, ovata, 3-4 mm. longa, 2-3 mm. lata, ares- centia sed persistentia; petala 5, elliptica usque obovata, 8-10 mm. longa, 3-5 mm. lata, purpureo-rosea; stamina 25-45; stylus staminibus longior; stigma. breviter triloba- tum. Capsula pinguis, ovoidea usque obovoidea (raro ellip- soidalis), 4-6 mm. longa; semina matura 10-25 in capsula, 0.75 mm. lata, laevia, plumbea. Glabrous herb perennating by a tuberous rhizome in soil surface; roots fibrous ; stems one to several, lower succulent part to 8 cm. high, simple or branched, crowded with leaves, narrowing abruptly above to a leafless wiry stalk, to 15 cm. long. Leaves many, terete, fleshy, to 4.5 cm. long. Inflores- cense a system of scorpioid cymes, usually with three pri- mary branches, one or more of these sometimes secondarily branched, or occasionally a single terminal scorpioid cyme. Flowers regular, ephemeral, open from 1:00 p.m. to 6:00 p.m. CST (3:30 p.m. to 6:00 p.m. in Franklin Co., Ala- bama) ; sepals drying but persisting about fruit; petals 5, 8-10 mm. long; stamens 25-45; style longer than stamens; stigma with 3 short lobes, papillose ; capsule ovoid to obovoid, 4-6 mm. long, unilocular, dehiscing by three valves; pla- centation free central: mature seeds 10-25, smooth, gray, dull, 0.75 mm. long. Shallow soil at the edges of rock expo- sures in calcareous cedar glades of middle Tennessee and northern Alabama. Flowers in May through September. 466 1967] Talinum — Ware 467 TYPE: TENNESSEE. DAVIDSON CO: southeast of Nashville on U.S. 70S, across road from Mt. View School; cedar glade; in shallow soil on limestone outcrop. Aug. 21, 1966. Ware 215 (US). Isotypes at SMS, UT, VDB. OTHER SPECIMENS EXAMINED: TENNESSEE. DAVIDSON co: 8 mi. SE of Una, at jet. of Old Hickory Blvd. and Charlton Lane, across from Burnette's Chapel, Franklin, Freeman, and Sullivan 163 (VDB), Quarterman 1653 (VDB), Waits 059 (VDB), Woodruff, 16 July 1937 (UT) ; one mile N. of Burnette's Chapel on Laverge-Couchville Pike, Franklin, Freeman, and Sullivan 229 (VDB); at type locality, Ware 118 (VDB). Giles Co: S of Pulaski on limestone between Cedar Grove Church and quarry, Ware 117 (UT, VDB). MARSHALL CO: one mi. E. of jct. of Tenn. 99 and U.S. 431 on Tenn. 99, along roadside and in bare places in cedar glade, Frank- lin, Freeman, and Sullivan 147 (VDB). RUTHERFORD CO: N. of Mur- freesboro, limestone barren opposite Stones River Mil. Park, A. J. and Evelym Sharp 25905 (UT); 5.8 mi. E of Murfreesboro, in rocky dry creekbed, DeSelm 692 (UT); 1.9 m. NE of Rockvale on country road off Tenn. 99, occasional in openings, cedar glade near Snail Sheel Cave, Franklin, Freeman, and Sullivan 316 (VDB); just N. of Mur- freesboro, U.S. 70S, in roadside cedar glades, Ware 119 (VDB). WILSON co: S. of Gladesville, west end of Cedars of Lebanon State Park, glades along dirt road, Ware 21? (VDB). ALABAMA. FRANKLIN co: E. of Russellville, cedar glade at jet. of Alabama 24 and Franklin Co. 79, Ware 216 (UT, VDB). Six terete-leaved species of Talinum have previously been reported in North America east of the Mississippi River. Two of these, T. parviflorum Nutt. and T. calycinum Engelm. are widespread west of the Mississippi but have been found east of the river only in southern Illinois (Jones, 1950; Mohlenbrock, 1955). A third species, T. rugosper- mum Holz., ranges from northwestern Indiana through Illi- nois, Wisconsin and northeastern lowa to eastern Minnesota (Fernald, 1950; Gleason, 1952). The other three species are found only east of the Mississippi, and all occur in the southeastern United States. Talinum appalachianum Wolf is known only from two counties in central Alabama, grow- ing on granite outcrops along either side of the Coosa River (David Cotter, 1966, personal communication). Wolf (1939) reported T. mengesii Wolf to be confined to Ala- bama, but McVaugh (1943) expanded its range to include the granite outcrops of Georgia, and specimens examined 468 Rhodora [Vol. 69 in this study confirm its presence in Tennessee. The most widespread and best known of the eastern species of Tali- num is T. teretifolium Pursh, which ranges from Pennsyl- vania serpentine barrens to the Altamaha grit of Georgia, but is best known from the granite outcrops of the Pied- mont of the Carolinas and Georgia (Harper, 1926; Mc- Vaugh, 1943). In the early stages of an ecological study of Talinum in the cedar glades of middle Tennessee, the question arose of the proper identity of these cedar glade plants. All floristic and ecological workers in the cedar glades, from Gattinger (1901) through Weiss (1959), identified the Talinum there as T. teretifolium, except Svenson (1941), who reported it to be T. mengesii on the basis of a specimen identified by Wolf. While the cedar glade plants key out to T. teretifolium in Gray's Manual (Fernald, 1950), they key out to T. calyci- num in Gleason’s (1952) New Britton and Brown, but do not fit his description of that species. They do not fall safely within the description of either T. teretifolium or T. men- gesii in Small’s (1933) Manual, though they are more like the latter species. The cedar glade plants also key down to Talinum calycinum in Wilson (1932) and Von Poellnitz (1933), but they do not fit the description of that species given by either of these two authors. After herbarium specimens of cedar glade Talinum were examined and compared with specimens of T. teretifolium, T. mengesii, T. rugospermum, and T. calycinum from out- side the Central Basin, it became apparent that the delicate flowers and fruits of this genus are so damaged by pressing and drying that critical specific characteristics may be lost, especially those relating to flower color and size, fruit shape and size, Sepal persistence, and anther and style character- istics.. "Therefore, any meaningful study of the group would have to be based on living materials grown in the green- house, a situation first noted by Holzinger (1900). Using the taxonomic works cited above, comparisons were made between the cedar glade Talinum and the cháracteris- tics of the six species of terete-leaved Talinum known to 1967] Talinum — Ware 469 grow east of the Mississippi. A dichotomous key was pre- pared for the seven entities involved, and this key was used to identify living material from various locations, From greenhouse and field observations, a table was prepared comparing the cedar glade plants with the two common southeastern species, T. teretifoliwm and T. mengesü Talinum parviflorum and T. appalachianum have only 5 stamens, and can be eliminated from further considera- tion, for the cedar glade plants have many stamens per flower. Living material of five entities was studied: T. teretifolium (DeKalb Co., Georgia) ; T. mengesii (Marion and DeKalb Cos, Alabama); T. rugospermum (Tazewell Co., Illinois) ; T. calycinum (Saline Co., Arkansas) ; and cedar glade populations (Davidson, Rutherford, Marshall, and Giles Cos., Tennessee, and Franklin Co., Alabama). These population samples were grown in the greenhouse during the summers of 1965 and 1966, and compared with the sets of characters listed in the literature for each species, as well as with each other. Comparisons were also made with herbarium specimens from Vanderbilt University (VDB), the University of Tennessee (UT), and Southwest Missouri State College (SMS). On the basis of living material and herbarium specimens, the cedar glade plants appeared to be more like T. mengesii than any other species, so careful comparisons were made of qualitative and quantitative characteristics which differed in the two groups of populations. These studies revealed that the cedar glade populations constituted a new species, Talinum calearicum, named and described above. Most of the differences between T. calcaricum and T. mengesii are quantitative, and many of them overlap in field populations, but the general aspect differences hold up in greenhouse plants grown from seed. The qualitative differences are remarkably constant, and these plus the non- overlapping quantitative ones make separation of indi- viduals of the two species easy. A key to these and the other two southeastern terete-leaved species of Talinum is provided below. i 470 Rhodora [Vol. 69 Talinum calcaricum has fewer stamens, fewer seeds per capsule, darker colored petals, and is smaller throughout, except that it has considerably larger seeds than does T. mengesii. Talinum calcaricum has persistent sepals, a three- lobed stigma, an obovoid to ovoid capsule, and gray seeds, as compared with deciduous sepals, a subcapitate stigma, a subglobose capsule, and shiny black seeds in T. mengesii TABLE I. A comparison of T. calcaricum with the two common southeastern Talinums. T. teretifolium T. calcaricum T. mengesii Petal length 5-8 mm. 8-10 mm. 9-12 mm. Petal color pink purplish- pink pink Style equal to longer than longer than length stamens stamens stamens Stigma three three sub-capitate type short lobes short lobes Stamen number 15-20 25-45 50-90 Sepals deciduous persistent deciduous Seed size (see text) small large small Seed shiny dull shiny color black gray black Substrate Granite, limestone sandstone, serpentine, granite etc. Quantitative data were gathered from three areas: a cedar glade just north of Murfreesboro, Rutherford Co., Tennessee, on U.S. 70S (Ware 119); another cedar glade in 1967] Talinum — Ware 471 T. mengesii ---- (Northfork) 29 qu T. calcaricum $ 204 | i (Murfreesboro) —— a (Franklin Co, Ala.) ++--++. e: a è o 154 1-- Oo Y È = - o B post n cie po ! = EE pe- L—-—-i : - 1 04 = LEM LI Li T LI i Lg y LI LI V LI LI L] LI 1 5- 10- 15- 20- 25- 30- 35- 40- 45- 50- 55- 60- 65- 70- 75- 80- Number of seeds Fig. I. Comparison of two populations of T. calcaricum and a population of T. mengesii with respect to number of mature seeds per capsule. T. mengesii Erie ( Northfork) T. calcaricum 25 4 ( Murfreesboro) — v (Franklin Co., Ala.) -------- o 20- . z 2 Lu. [54d 5 e o (OA l L-4 E pe i = 5- r---2 ! = Lor [9] Tr TT 9 FE 5. 10- 15- 20- 25- 30- 35- 40- 45- 50- 55- 60- 65- 70- 75- 80- 85- 90- 95- Number of Stamens Fig. II. Comparison of two populations of T. calearicum and a population of T. mengesii with respect to number of stamens per flower. 472 Rhodora [Vol. 69 Franklin Co., Alabama, at the junction of Alabama 24 and Franklin Co. 79 (Ware 216); and a population of T. men- gesii from Marion Co., Alabama, on sandstone bluffs along Northfork Creek on U.S. 43 (Ware 123). Fifty flowers and fifty capsules were collected from each population at one pace intervals along line transects across the population. At Northfork Creek and in Franklin Co., Alabama, flowers and capsules were preserved in 3:1 acetic alcohol and stamens and seeds were counted later; at Murfreesboro, counts were made in the field from fresh flowers and cap- sules, Talinum calcaricum consistently had fewer stamens and seeds per capsule than T. mengesii (Figs. 1 & 2), and there was no overlap in stamen or seed numbers between the two species. To show a noticeable difference in size of seeds between T. mengesii and T. calcaricum, three replicates of 1500 seeds each were taken from the Northfork Creek (T. men- gesii) and Murfreesboro (T. calcaricum) populations. The average weights per 1500 seeds were 264 mg. for T. mengesii and 628 mg. for T. calcarieum. Thus, the seeds of the cedar glade species are on the average more than twice as heavy as those of T. mengesii. This visible size difference and the gray vs. black color of the seeds are perhaps the best dis- tinguishing characteristics on herbarium specimens. CROSSING RELATIONSHIPS Because the pistil is considerably longer than the stamens in both T. mengesii and T. calcaricum, capsules are not formed in the greenhouse unless pollination is manually effected ; thus, emasculation of the flowers was unnecessary in crossing experiments. Reciprocal crosses between the five populations of T. calcaricum gave 95% to 100% of the seed yield considered to be “normal” on the basis of field sampling. Crosses between the two populations of T. mengesii likewise produced yields comparable to those noted in the field. Crosses between T. calcaricum and T. mengesii gave quite different results. When Murfreesboro plants were polli- nated with T. mengesii pollen, not one of 18 pollinations 1961] Talinum — Ware 473 yielded mature seed. When Franklin Co., Alabama T. cal- caricum plants were pollinated with T. mengesii pollen, 4 of 9 pollinations yielded mature seed, but no capsule pro- duced as many as the minimum of 9 seeds/capsule found in field studies of that population (Table II). In reciprocal crosses, with pollen from Murfreesboro T. calcaricum plants placed on T. mengesii styles, 17 of 20 pollinations yielded mature seed, but only one produced the minimum number of 25 seeds/capsule found in field studies of T. mengesii, and only three capsules produced more than ten seeds. TABLE II. Results of crosses between T. calcaricum and two related species. Pistillate Staminate | Number Number parent parent of Number with polli- setting normal nations seed yield T. calcarieum X T. mengesti (Murfreesboro) (Northfork) 18 0 0 T. calearicum X T. mengesti (Franklin Co.) (Northfork) 8 4 0 T. mengesii X T. calcaricum (Northfork) (Murfreesboro) 20 IT 1 T. calcaricum X T. calycinum (Murfreesboro) (Arkansas) 9 4 0 After T. mengesii, T. calycinum is the species most like T. calcaricum. When Murfreesboro T. calcaricum plants were pollinated with pollen from Saline Co., Arkansas plants (Demaree 53659) which keyed out to T. calycinum, 5 of 8 pollinations produced mature seed, but no capsule produced the minimum number of 9 seeds/capsule expected for the Murfreesboro population. This demonstration of barriers to genetic exchange (one of 55 crosses produced normal seed yield) between the cedar glade Talinum and the two species most like it, T. mengesii and T. calycinum, strengthens the idea that T. 474 Rhodora [Vol. 69 calcaricum, recognized as a new species on morphological grounds, is a valid biological species. FIELD KEY TO THE SOUTHEASTERN TERETE-LEAVED TALINUMS PEMNICUDTBSMMMMM T. appalachianum. A. Stamens 15-many. B. Style not longer than the 15-20 stamens ........ T. teretifolium. B. Style noticeably longer than the 25-many stamens. C. Mature seeds black, shiny, more than 25 per capsule; stigma subcapitate; sepals deciduous; capsule subglobose; Stamens 45-90 ooccccccccccccccccccccccccecccsseceecssueeseeeeeeeeeees T. mengesii. C. Mature seeds gray, dull, 10-25 per capsule; stigma with three short lobes; sepals persistent; capsule ovoid to obo- void; stamens 25-45 .........sssseeee ee T. calearicum, ACKNOWLEDGMENTS Gratitude is expressed to fellow students at Vanderbilt University who aided in securing living materials for use in this study; to Miss Donna Marie Eggers, who aided in gathering of data; and to Professor Elsie Quarterman and other faculty members for help in preparing the manuscript. VANDERBILT UNIVERSITY NASHVILLE, TENNESSEE PRESENT ADDRESS COLLEGE OF WILLIAM AND MARY WILLIAMSBURG, VIRGINIA LITERATURE CITED FERNALD, M. L. 1950. Gray's Manual cf Botany. 8th ed. American Book Co. New York. 1632 p. GATTINGER, A. 1901. Flora of Tennessee and Philosophy of Botany. Gospel Advocate Press, Nashville, Tenn. 269 p. GLEASON, H. A. 1952. Illustrated Flora of the Northeastern U. S. and Adjacent Canada, Vol. II. New York Bot. Gard. 3 v. HARPER, R. M. 1926. The cedar glades of middle Tennessee. Ecol. 7: 48-54 HOLZINGER, J. M. 1900. The geographical distribution of the Tere- tifolium group of Talinum. Asa Gray Bull. 8: 36-39. JONES, G. N. 1950. Flora of Illinois. 2nd ed. U. of Notre Dame Press, Notre Dame, Indiana. 368 p. 1967] Talinum — Ware 475 McVaucąH, Rocers. 1943. The vegetation of the granitic flatrocks of the southeastern U. S. Ecol. Monog. 13: 119-166. MOHLENBROCK, R. H. 1955. Contributions to the flora of southern Illinois. Rhodora 57: 319-322. SMALL, J. K. 1933. Manual of the Southeastern Flora. The author, New York. 1554 p. SvENSON, H. K. 1941. Notes on the Tennessee flora. Jour. Tenn. Acad. Sci. 16: 1: 111-160. Von POELLNITZ, KARL. 1933. Zur Kenntnis der Gattung Talinum Adans. (Portulacaceae). Ber. Deut. Bot. Ges. 51: 112-127. Weiss, N. J. 1959. An Autecological Study of Three Mosses Im- portant in Xerarch Suecession in the Cedar Glades of Middle Tennessee. M. A. thesis, Vanderbilt Univ. 63 p. Witson, PERCY. 1932. Talinum, in Portulacaceae, N. Amer. Flora 21: 280-289. WoLr, W. 1939. The status of Talinum in Alabama. Amer. Midl. Nat. 22: 315-332. A NEW SPECIES OF MACHAERANTHERA SECTION PSILACTIS R. C. JACKSON AND R. R. JOHNSON! Machaeranthera arizonica Jackson & Johnson, sp. nov. Herba perennis 15-30 dm alta, glandularo-puberulentis (pilis glandularibus saepe trichomatibus longioribus eglan- dularibus intermixtis) ; foliis sessilibus, laminis lanceolatis, inaequaliter incise dentatis crispatisque; involucris 4-9-7.1 mm altis; phyllariis 35-62, lanceolatis, 4.2-4.7 mm longis; radiis 19-34, 6.8-10.8 mm longis, 1.7-2.2 mm latis; disci corollis 41-111, 3.4-5.2 mm longis; pappis florium discorum 33-51, 2.8-3.3 mm longis; pappus eorum radiorum nullis. Perennial herb from a strongly developed taproot, 15-30 dm tall, glandular-puberulent throughout with occasional intermixed non-glandular hairs; many branches from the somewhat woody base; leaves sessile, sometimes clasping at the base, alternate, crisped, irregularly incisely dentate, the lower ones more deeply so; heads on short peduncles with slightly reduced leaves; involucres 4.9-7.1 mm high; phyl- laries in 4-5 series, lanceolate with green tips, the scarious margins of the outer series white, the inner ones purple; rays 19-34, 6.8-10.9 mm long, 1.7-2.2 mm wide; disc corollas 47-111, 3.4-5.2 mm long ; pappus one series of 33-51 bristles, 2.8-3.3 mm long; achenes ca. 2 mm long. Chromosome num- ber n—5. TYPE. ARIZONA: PIMA COUNTY; Organ Pipe Cactus Na- tional Monument, low, rocky hillsides and sandy soil around Quitobaquito Springs, 31 March 1962, R. C. Jackson & R. R. Johnson 3043-1, KANU (Fig. 1) ; Isotype at ARIZ. Additional specimens in the University of Arizona Her- barium from the type locality: 5 March 1940, Benson 9934; 5 March 1940, Peebles 14560; 17 April 1952, Parker 7994; 28 April 1939, Nichols s.n.; 27 November 1939, Harbison 26176. The closest relatives of M. arizonica appear to be M. crispa (Brandegee) Turner & Horne and M. arida Turner & Horne. "This study was supported in part by NSF Grant GB-3071 476 1967] Machaeranthera — Jackson & Johnson ATT The second species is known from two localities in Baja California del Sur and the third occurs in the northwestern part of Sonora and the southern parts of California, Ari- zona, and Nevada. The springs at Quitobaquito, the type locality of M. ari- zonica, are fed from water forced through rock fissures to Plate 1370 Figure 1. Photograph of the holotype of Machaeranthera (Psi- lactis) arizonica Jackson & Johnson. 478 Rhodora [Vol. 69 the surface. The desert pupfish lives in these waters and is found also in the Sonoyta River, about one mile south, and in the Lower Colorado River, an indication that at some- time in the past the springs flowed to the river and the river to the Gulf of California. This could have provided a migra- tion route from the general locality of M. crispa, perhaps the ancestral or a closely related stock of M. arizonica, to Quitobaquito along more a mesic habitat. M. crispa and M. arida, however, appear to occupy drier habitats than M. arizonica at the present time. Machaeranthera arizonica is somewhat intermediate morphologically for characters of M. crispa and M. arida (Table 1) and may occupy the same position phylogenetical- ly since it is still biennial or perennial and occupies a some- what more mesic habitat than the annual M. arida. Artificial crosses between arida and arizonica yielded seed, but the F, hybrids have not been grown yet. Both species are self- sterile. Turner and Horne (1964) combined Psilactis with Ma- chaeranthera, proposing Psilactis as a series of the latter genus. This proposal was based on the break down of the diagnostic character of Psilactis, namely, the absence of pappus bristles on the ray achenes. One and possibly two species show variation for this character. In addition, a preliminary report on hybridization between Psilactis and Machaeranthera (Jackson, 1962) added weight to the pro- posed change. The presence or absence of a ray pappus as a diagnostic generic character does, indeed, seem a poor one for separating the two genera. The two taxa are genetically related as shown by the production of a hybrid which was, nonetheless, rather highly sterile. The degree of relation- ship and how this should weigh on the question of merging the taxa might be answered better when a detailed cytogene- tic analysis of inter- and intrataxon hybrids has been com- pleted. There are nevertheless some floral and leaf char- "Information supplied in *A Guide to the 40-Mile Scenic Loop Drive [of] Organ Pipe Cactus National Monument" by U.S. National Park Service. Machaeranthera — Jackson & Johnson 1967] "Dovuozi4D "PY pue DPN 'W 40g BVYVp ISuUBI PUB BSVAIBAB YF 4107 pasn əzəm sjuv[d gT 0} usej1no, 'Ajr[eoo[ 9edA3 OY} WOT pour?3qo SEM D2)402140 “PY JO peag '"?uozrtiy 'Aquno?) edoorrepy ‘JUALO enby 7e pejoe[[Oo paas SEM VPN "jy Jo 9o1inog IWIN aures ay} JE osnoquoa.ld sy} ur UAO.Ld Sjue[d Wort 9194 EXKL} IƏUYJO ay} IOF vje([ '(P96I[) 9UuioH 2p 1ou.in[, WOIJ ale PASUMI ` UL Sasuel IOJ CPA x (0°T-8'0) me aA D, 0% (TET) 9'T (uu) yp (o (8-9) mus (8'0I-89) 0*6 (26-99) GL (ww) Yysue, (q (0F-0£) p (78-61) r'6c (62-81) e peau dod ‘ou (v LIMO AVY (ese) p (g'8-8'6) og (&&v) 9% (ww) qidue[ sndded (p (08 `?) m (T9788) If (F&-Ic) 9'87 sepgsriq sndded ‘ou (ə (9-7) x (Z'S-7'8) rr (F'F-8°3) vg (ww) uq12uo[ eoo (q (001-6) x (PET) 108 (29-19) EZS peau Jod ‘ou (v LIMO] T Ist] ISULI X asuel UE o3uvl x 8.1932 €.18q2) Ddsto "W D291WOZi4D "I Dp. D "I x Ddsio "py pue VNUOZUD “PE (mmpi4o Da4oi]wnaom[on JI jo S1932€.1eq5 [Boy euros Jo suosrriedurioo |eorgo[oud.to]y ‘I e[qu;, 480 Rhodora [Vol. 69 acters that effectively set off Psilactis from Machaeranthera, either as a distinct genus or as a subgeneric taxon. On morphological grounds, Machaeranthera parviflora should be placed in series Psilactis also; it has the same chromosome number (n—5; Jackson 5241, KANU) as its other close relatives, M. arida, M. coulteri, M. crispa, and M. arizonica. UNIVERSITY OF KANSAS, LAWRENCE AND TEXAS WESTERN UNIVERSITY, EL PASO LITERATURE CITED TURNER, B. L. and D. HoRNE. 1964. Taxonomy of Machaeranthera sect. Psilactis (Compositae-Astereae). Brittonia 16:316-331. JACKSON, R. C. 1962. Intergeneric hybridization between Psilactis and Machaeranthera. Amer. Jour. Bot. 49:676. (Abstract). ADDITIONS TO THE VASCULAR FLORA OF OKLAHOMA — II It is a tendency for some botanists to think of the vascular flora of our area as rather thoroughly collected until for some reason they attempt to delineate the distribution of some species. Often when the specimens housed in our her- baria are examined, two conditions are encountered, (1) Although the species is often a rather common one, only a few specimens are available. (2) Again, sometimes there may be several folders of specimens, but all from a few loca- tions, The interest of more and more of our botanists today does not take them to the field, yet much general collecting is still to be done. I feel that the species discussed in this brief paper offer some support to these contentions. Al- though they are thought not to have been previously report- ed for the state, they occur in rather common habitats (old fields, rocky roadsides, sand dunes, etc.) and probably have been members of our flora for a long time. We wish to express our thanks to Betty Weisenhunt, who helped with the species of Schrankia. All specimens cited were collected by John and Constance Taylor unless otherwise indicated, and are presently located at the Bebb Herbarium, Norman. Ophioglossum crotalophoroides Walt. The genus Ophio- glossum has been known in our flora by the one species O. engelmannii Prantl for over 30 years. Now we have O. cro- 1967] Flora of Oklahoma — Taylor & Taylor 481 talophoroides as No. 3322, collected March 15, 1966 from a sandy old field just west of Durant city limits. This species differs in that it fruits about two to three weeks earlier in this area, is much smaller, is found mainly in sandy soils, the rhizome is almost spherical or bulb shaped, and in the older specimens, the old leaf bases remain to form a cylindri- cal black collar on their top. It has been reported from a number of counties in Texas: Newton ( Reverchon, 1903) ; Bastrop, Hardin, and Harris (Correll, 1956) ; and from Llano in the Edwards Plateau (Ridgway and Walne, 1965). One specimen of this species was located in the Bebb Her- barium from Faulkner Co., Arkansas by C. M. Moore, No. ILLT: Schrankia roemeriana (Scheele) Blankenship. In Texas this species of sensitive briar is known from the Red River south through the south central part of the state. Turner (1959) listed it as occurring in Fannin, Grayson, and Clay counties just south of the Red River, thus its presence in Oklahoma might have been expected. This species may be differentiated from S. uncinata by the lack of raised veins on the lower surface of the leaflets and the flattened seed pod. It usually occurs on soils derived from limestone or clay, while S. uncinata is usually found growing in sandy soil. Our specimen, No. 3355, is from along a county road from a limestone outcrop, collected 4 miles NE of Benning- ton in Bryan Co. on June 5, 1966. This species may occur over much of the southern half of the state where clay and limey soil are found and should be looked for in the Ar- buckle Mountains. Zanthoxylum hirsutum Buckley. This name is taken from Shinners (1958) who gives the following synonymy. Z. carolinianum var. fruticosum (Gray) S. Watson, Fagara fruticosa (Gray) Small, F. clava-herculis var. fruticosa (Gray) Sargent. The Texas distribution of Z. hirsutum is given as from the northcentral part of Texas south and west- ward to the Transpecos. We have this species from Okla- homa as No. 2095, rocky limestone hill, 11 miles W of Mari- etta, Love County ; No. 3427, limestone bluff above the Red 482 Rhodora [Vol. 69 River floodplain, 8 miles W of Waurika, Jefferson County ; No. 3427B, sand dune area 17 miles SE of Randlett, Cotton County ; and No. 2442, 1 mile S of Davidson, on sand dunes, Tillman County, which is our westernmost location. In our area it takes a low growth form with fruiting specimens 4-5 feet tall being common. Occasionally a height of 10 to 15 feet is attained, but these are usually branched at or near the ground. Zanthoxylum hirsutum differs from Z. clava- hereulis (under which it is sometimes placed as a variety) by smaller general growth form, leaves, and leaflets, and a smaller axillary inflorescence. Texas authors seem to dis- agree as to the exact status of this species, Vines (1960) lists Z. clava-herculis L. var. fruticosum Gray and gives essentially the same range as that listed by Shinners (1958) for Z. hirsutum. Gould (1962) lists Z. clava-herculis var. fruticosum (Gray) Wats. based on Fagara fruticosa (Gray) Small and gives a distribution including most of Texas except the high plains portion of the panhandle. He also lists a Z. hirsutum Buckl. from the coastal prairie and south Texas plains. Epilobium glandulosum Lehm. var. adenocaulon (Haussk) Fern. To the best of our knowledge, the genus Epilobium has been known in the Oklahoma flora by one species (E. coloratum) and one collection (E. L. Little, No. 3996) for almost 50 years (Hopkins, 1938). We now have E, glandu- losum var. adenocaulon as No. 2520 from a small spring in a branch of Tessequite Canyon, 6 miles S of Kenton, Cimar- ron Co., Oklahoma. This is a wide ranging species in North America, occurring from Alaska east to Newfoundland, south to West Virginia and westward to California. It was a mild surprise to find that similar material I had collected in the interior of Alaska in the summer of 1965 was con- sidered to belong to the same species. Rogers (1953) lists this species (as E. adenocaulon Haussk.) as common in the Colorado portion of the Black Mesa, which suggested the possibility of its occurring in Oklahoma. R. JoHN TAYLOR AND CONSTANCE TAYLOR SOUTHEASTERN STATE COLLEGE DURANT, OKLAHOMA 1967] Saxifraga Aizoón — Churchill 483 LITERATURE CITED CoRRELL, D. S. 1956. Ferns and Fern Allies of Texas. Renner, Texas: Texas Research Foundation. GouLp, F. W. 1962. Texas Plants — A checklist and ecological summary. Texas Agr. Exp. Sta. MP-585. Hopkins, M. 1938. Notes from the Herbarium of the University of Oklahoma. Rhodora 40:433. REVERCHON, J. 1903. The Fern Flora of Texas. Fern Bull. 11:33-38. Ripeway, J. and P. WALNE. 1965. Ophioglossum Crotalophoroides Walt. on Enchanted Rock, Llano County, Texas. Southwestern Naturalist 10(2) :143. Rocers, C. M. 1953. The vegetation cf the Mesa de Maya Region of Colorado, New Mexico, and Oklahoma. Lloydia 16(4) :257-290. SHINNERS, L. H. 1958. Spring flora of the Dallas-Ft. Worth area, Texas. Publ. by Lloyd H. Shinners, Dallas. TURNER, B. L. 1959. The Legumes of Texas. Austin: Univ. Texas Press. ViNES, R. A. 1960. Trees, Shrubs, and Woody Vines of the South- west. Austin: Univ. Tex. Press. SAXIFRAGA AIZOÓN IN NEW HAMPSHIRE* Saxifraga Aizoón Jacq. is an arctic-montane, amphi- Atlantic rock plant of Europe, Iceland, Greenland and North America'. In eastern North America the species, represented by var. neogaea Butters, ranges southward into cold localities of Nova Scotia, New Brunswick, Maine (Mt. Katahdin), Vermont, and northern New York’. The plant has not been reported previously from New Hampshire”. On 20 July 1939 I found a colony of Saxifraga Aizoóm, numbering about 80 flowering plants, in a deep chasm cut by glacial action into the flank of Mt. Washington, Coós County, New Hampshire. Most of the plants grew in a nearly horizontal fissure in a vertical rock wal] of impres- 'HULTEN, E. 1958. The amphi-Atlantic plants and their phytogeo- graphic connections. Kungl. Svensk. Vetensk. Handl. IV. 7: 1-340. "FERNALD, M. L. 1950. Gray's Manual of Botany, Ed. 8. American Book Co. New York. 1632 p. "PEASE, A. S. 1964. A Flora of Northern New Hampshire. New England Botanical Club, Inc. v + 278 p. *The author appreciated the help of John Beaman, Ph.D., Associate Professor and Curator of the Beal-Darlington Herbarium in the preparation of this article. 484 Rhodora [Vol. 69 sive height. The precipice faced northward, and thus was shaded most of the day. The fissure collected moisture from films of water seeping down the rock from above. The locality, at an altitude of about 3,500 ft, is often shrouded in cloud. Saxifraga Aizoón is generally a calciphile. Hence it is surprising that it should grow anywhere on the Presidential Range in New Hampshire where limestone is unknown‘. In an area as well botanized as Mt. Washington it is re- markable that the species could have remained this long undiscovered. I obtained five plants from crowded colonies, taking great care to prevent uprooting adjacent rosettes. The collection was donated recently to the Beal-Darlington Herbarium of Michigan State University, and one plant has been shared with the New England Botanical Club Herbarium. “BILLINGS, M. P. et al. 1946. Geology of the Mt. Washington Quad- rangle, New Hampshire. Bull. Geol. Soc. Amer. 57: 261-274, 1 pl. JOHN A. CHURCHILL LAFAYETTE CLINIC DETROIT, MICHIGAN THE SAXIFRAGA AIZOON STATION REVISITED The report by Dr. Churchill of Saxifraga Aizoön on Mt. Washington was so surprising that three of us decided to make an investigation. On July 29, 1967; A. R. Hodgdon, James Teeri, a graduate student at the University of New Hampshire, and I made a trip into Huntington Ravine. We were equipped with a good pair of field glasses but no rock- climbing gear. After some careful but fruitless searching we eventually located the Saxifraga well up on a cliff by means of field glasses. The plants were in full bloom and the colony appeared to be in the same condition as described by Churchill. After careful checking with a topographic map it appeared to us the elevation was close to 4500 ft. Naturally we were eager to examine the colony at close range and with some effort were able to reach a point about 1967] Saxifraga cernua — Churchill & Hodgdon 485 20 ft. below it. Even though two of us had some experience in rock-climbing, the last pitch was too much for us. It was steep and a rushing brook covered all the promising hand- holds. In a dry season, the ascent might have been easier, but it appeared to us that a rope would have been most desirable. We all felt considerable respect for Dr: Churchill's mountaineering ability. From a vantage point across from a narrow shelf covered with vegetation, it was possible to study the surrounding area with care through the glasses and we felt that it was possible that there were other interesting plants in ex- tremely hazardous or inaccessible sites in the gully. In any case we feel that Dr. Churchill’s discovery was the most outstanding bit of botanizing on Mt. Washington in the past 50 years. FREDERIC L. STEELE ST. MARY'S-IN-THE-MOUNTAINS LITTLETON, NEW HAMPSHIRE SAXIFRAGA CERNUA IN NEW ENGLAND* In the United States east of the Rocky Mountains, Saxi- fraga cernua L. has been found only in northern Minnesota.' Although it grows in the Gaspé Peninsula of Quebec, this chiefly northwestern and Eurasian arctic plant was hitherto unexpected in the well explored mountains of New England. The unusual circumstance attending the discovery of Saxi- fraga cernua on Mount Washington in Cóos County, New Hampshire, is the subject of this report. On July 20, 1939, one of the authors (Churchill) ex- plored a deep ravine cut into the flank of Mount Washington and, while exploring a cliff face there, found Saxifraga Aizoón. Near the Saxifraga Aizoón colony, Churchill also ‘FERNALD, M. L. 1950. Gray's Manual of Botany, Ed. 8. American Book Co. New York. 1632 P. "CHURCHILL, J. A. 1967. Saxifraga Aizoon in New Hampshire, Rhodora 69: 483-484. “The authors appreciated the help of John Beaman, Ph.D., Associate Professor and Curator of the Beal-Darlington Herbarium in the preparation of this article. 486 Rhodora [Vol. 69 explored a shelf situated high in a vertical rock trough down which a cataract poured, The shelf was covered with greenery, which was wet with spray. On the shelf grew plants, which the writer assumed were Saxifraga rivularis, a plant known from the Presidential Range. The story might easily have ended at this point, for the descent from the wet trough proved to be extremely treacherous. The other author ( Hodgdon) wanting to verify the Saxi- fraga Aizoön record, sought the plant on July 29, 1967 in company with F. L. Steele and James Teeri following direc- tions that Churchill had forwarded to Richard Eaton.* They decided they could not reach the shelf without excessive risk. However, they did scale a buttress at one edge of the trough reaching the level of the shelf. With glasses, Hodgdon could see Saxifrage plants on the shelf, and much to his surprise made out the features of Saxifraga cernua. When this news was related to Churchill at a meeting with Hodgdon at the Gray Herbarium, he had to admit that he had remembered seeing what he had thought was Saxi- fragia rivularis on the shelf, but that he feared he had made no collection of it. Fortunately, this fear proved to be unfounded and, when informed, Dr. John Beaman examined the collection, which he found was indeed Saxi- fraga cernua. The collection is deposited in the Beal-Darlington Her- barium of Michigan State University. JOHN A. CHURCHILL NATIONAL INSTITUTE OF NEUROLOGICAL DISEASES AND BLINDNESS BETHESDA, MARYLAND A. R. HODGDON UNIVERSITY OF NEW HAMPSHIRE DURHAM “STEELE, F. L. 1967. The Saxifraga Aizoön Station Revisited, Rhodora 69: 484-485. STATEMENT OF OWNERSHIP, MANAGEMENT AND CIRCULATION Date of Filing: October 1, 1967 Title of Publication: RHODORA Frequency of Issue: Quarterly Location of Known Office of Publication: Botanical Museum, Oxford Street, Cambridge, Massachusetts 02138 Location of the Headquarters of the Publishers: Botanical Museum, Oxford Street, Cambridge, Massachusetts 02138 Names and Addresses of Publisher, Editor and Business Manager: Publisher: New England Botanical Club, Inc., Botanical Museum, Oxford Street, Cambridge, Mass. 02138 Editor: Albion R. Hodgdon, Dept. of Botany, Univ. of New Hampshire, Durham, N. H. 03824 Business Manager: Herman R. Sweet, Botanical Museum, Oxford Street, Cambridge, Mass. 02138 Owner: New England Botanical Club, Inc., Botanical Museum, Oxford Street, Cambridge, Mass. 02138 No stockholders Known Bondholders, Mortgagees and other Security Holders: None Average No. Copies Single issue each issue during nearest to preceding 12 months filing date Total No. copies printed 775 775 Paid circulation Sales through dealers, etc. none none Mail subscriptions 719 750 Total paid circulation . 719 750 Free distribution 10 10 Total distribution 729 760 Office use, left-over, etc. 46 15 Total 775 775 I certify that the statements made by me above are correct and com- plete. Herman R. Sweet, Business Manager INDEX TO VOLUME 69 New scientific names and combinations are printed in bold face type Abelmoschus esculentus 424 Abrus abrus 411; praecatorius 411 Abutilon crispum 374; hirtum 424; indicum 424; umbellatum 424 Acacia farnesiana 408; macra- cantha 373, 387, 408; muricata 408; polyacantha 373; riparia 408; scleroxyla 373; suma 373; tortuosa 408 Acalypha chamaedrifolia 418; hispida 418; Acalypha ostryae- folia 452; ostryifolia 452; wilkesiana 418 Acanthaceae 395, 448 Acanthospermum hispidum 447 Acer Pseudo-Platanus 454; pseu- doplatanus 454 Achyranthes aspera 402, 404; bettzickiana 402; ficoidea 402; indica 402; polygonoides 402 Aciotis 172; fragilis 169 Acnistus arborescens 393, 440 Acrodiclidium salicifolium 406 Acuan depressum 409; virgatum 409 Adelia ricinella 418 Adenanthera pavonina 408 Adenoropium gossypifolium 420; multifidum 420 Adipera bicapsularis 410 Aechmea spp. 392 Aeschynoneme americana sensitiva 411 Agastache scrophulariaefolia 452; scrophulariifolia 452 Agati grandiflora 414 Ageratum conyzoides 447 Agrimonia gryposepala 306, 315; parviflora 306, 316; pubescens 306, 316; rostellata 306, 316; 411; striata 307, 315 (as stricta) Agropyron 122 Aklema petiolare 419 Alabama, Petalostemon foliosus in 383 Albizia Lebbeck 409 Albizzia Lebbeck 409 Alchemilla minor 307, 318 Alchornea latifolia 418 Alicabon 84 Alkekengi 82, 84; barbadense 321 Allamanda cathartica 434 Allium ampeloprasum 61, in Mississippi 61 Allophyllus occidentalis racemosus 374 Allosandra 296; lanceolata 296, 300; verbenifolia 300, 302, 303 Alnus glutinosa 378; rugosa 3778 Alsophila conjugata 65, 67; este- 374; lae 66-68; New taxa in the Genus 65; scabriuscula var. guatemalensis 66, 68, var. scabriuscula 68; Swartziana 67, 68; trichiata 65, 67; tryonorum 65-67 i Alternanthera ficoidea 402, var. bettzickiana 402, 403; parony- chioides 402 Althaea rosea 424 Alysicarpus vaginalis 412 Amaranthaceae 395, 402 Amaranthus caudatus 403; cras- sipes 403; dubius 403; spinosus 403 Ambrosia hispida 447 Amelanchier arborea 307, 816; Bartramiana 307, 316, X laevis 307, 316; canadensis S0%..316; gaspensis 307, 317; humilis 307, 317; intermedia 307, 317; laevis 307, 315; nantucketensis 307, A487 488 317; X neglecta 307, 316; san- guinea 307, 316; stolonifera 307, 315; Wiegandii 307, 316 Ammania latifolia 429 Amomis caryophyllata 431; gri- sea 431 Amyris elemifera 416 Anacardium occidentale 421 Anamomis fragrans 431 Andira inermis 374, 412 Anethum graveolens 432 Anisodontea 16 Anneslia portoricensis 409 Annona glabra 406; muricata 391, 406; reticulata 406; squa- mosa 406 Anredera leptostachys 405 Anthacanthus spinosus 443 Anthriscus sylvestris 379 Antigonon leptopus 402 Aquilegia canadensis 195; A spineless form of 195, f. ecalcarata 195, 196, f. flaviflora 195, f. Phippenii 195 Arabis 463; glabra 32; laevigata 32 Arachis hypogaea 412 Ardisia crenulata 375; obovata 433; serrulata 375 Argemone mexicana 406 Argythamnia candicans fasiculata 418 Arisaema 173; Stewardsonii 31 Aristida longespica 452; longi- spica 452 Aristolochia elegans 402; littoralis 402; trilobata 402 Artocarpus altilis 401; communis 401; incisa 401 Aruncus dioicus 307, 318 Asclepias curassavica 435 Aspens of the Columbia Plateau, Indications of Possible Mid- Cenozoic Hybridization in the 70 Aspidium coarctatum 5, lasiesthes 418; Rhodora [Vol. 69 7; pilosulum 7, 8 Asplenium rutaceum 376 Asprella Hystrix 135 Aster dumosus var. subulaefolius 452, var. subulifolius 452 Astragalus 48; canadensis 49, 50, var. carolinianus 49; carolini- ana 49, 50; glaber 48, 49; Michauxii 48, 49; multijugus 48, 49, 50; obcordatus 49; ten- nesseensis 384; villosus 49 (as villosa) Asystasia gangetica 443 Athenaea 83; cernua 99 Athyrium thelypteridoides 452; thelypterioides 452 Atriplex littoralis 373; pentandra 373, 402 Aulomyrcia citrifolia 430 Avicennia 60; germinans 438; nitida 438 Ayenia insularis 425; pusilla 425 Azadirachta indica 391, 417 Bacopa monnieri 442 Bahama Islands, Additions to the Flora of the Bimini Island Group 60 Banisteria purpurea 416 Barnes, Burton V., Indications of Possible Mid-Cenozoie Hybridi- zation in the Aspens of the Columbia Plateau 70 Basella alba 405; rubra 405 Baskin, Jerry M. and Carol Cau- dle, Penstemon foliosus in Ala- bama 383 Bastardia viscosa 424 Bates, David M., A Reconsidera- tion of Sidopsis Rydberg and Notes on Malvastrum A. Gray (Malvaceae) 9 Batis maritima 404 Batocydia unguis 442 Bauhinia monandra 410; pur- purea 410; tomentosa 410 1967] Bean, R. C, A. F. Hill, R. J. Eaton and S. K. Harris, A Cor- rection in the 16th Report 318; Seventeenth Report of the Com- mittee on Plant Distribution 306 Beckmannia arucaeformis eruciformis 452 Begonia spp. 428 Benthamantha caribaea 412 Beta vulgaris 402 Betula caerulea-grandis 452; caeruleigrandis 452; papyrifera 458 Bia 242, 252 Bidens cynapiifolia 448; Bidens hetercdoxa var. monardaefolia 452; monardifolia 452; pilosa 448 Bixa orellana 426 Blechum blechum 443; pyrami- datum 443 Boberella 84 Boerhaavia coccinea 403; diffusa 403; erecta 403 Boivan, Bernard, Connecting Vowels in Epithets of Latin Origin 451 Bontia daphnoides 444 Boraginaceae 395, 437 Borreria ocimoides 444; laevis 444 452; Botrychium matricariaefolium 452; matricariifolium 452; ternatum var. rufinaeformis 452, var. rufiniformis 452 Bougainvillea glabra 403; peru- viana 403; spectabilis 403 Bourreria succulenta 437 Boussingaultia leptostachys 405 Brachistus Pringlei 208 Brachyrampus intybaceus 448 Bradburya virginiana 412 Bramia monnieri 442 Brassica caulorapa 407; chinense 407; hirta 373; integrifolia 406; oleracea var. capitata 407; Index to Volume 69 489 rapa 407 Breynia nivosa 418 Britton, Donald M., Diploid Dry- opteris dilatata from Quebec 1 Bromus brizaeformis 452; brizi- formis 452; Dudleyi 31 Brunfelsia 394; americana 440; undulata 440 Bryophyllum pinnatum 408 Buchenavia capitata 430 Buchnera elongata 376; longifolia 376 Bucida buceras 430 Bumelia obovata 433 Bunchosia glandulosa 416 Bursera simaruba 417 Byrsonima coriacea var. spicata 416; spicata 416 Cacabus 83, 84, 100; hondurensis 84; mexicanus 82, 100, 101, 227 Cactus intortus 428 Caesalpinia bonduc 373, 410; crista 373, 410; major 373; melanosperma 410; pulcherrima 410 Caesalpiniaceae 395, 410 Cajan cajan 412 Cajanus cajan 412 Cakile lanceolata 407 Calliandra caracasana 409 Callitriche 173 Calonictyon tuba 436 Calophyllum antillanum 426; brasiliense 426 Calotropis procera 435 Calyptranthes kiaerskovii 394, 430 Calyptrogyne occidentalis 372; rivalis 372 Calyptronoma rivalis 372 Calyptropsidium sintensii 375 Canavali maritima 412 Canavalia maritima 412 Capparaceae 395, 407 Capparis baducca 407; cynophal- 490 lophora 407; flexuosa 407; indica 407; portoricensis 407 Capraria biflora 442 Capsella bursa-pastoris 463 Capsicum frutescens 440 Cardamine 463; Douglassii 463, 464, Thlaspi tuberosum a Taxo- nomic Synonym of 460; rhom- boidea var, purpurea 463; tu- berosa 464 Cardiospermum corindum 422; halicacabum 422 Carex 306, 451; Bushii 378; livida var. rufinaeformis 352, var. rufuniformis 352 Carica 451; papaya 391, 427 Carpinus caroliniana 458 Carterothamnus 45, 46; anomalo- chaeta 43-47 Carya buckleyi 331; pallida 54 Casearia decandra 427; guianen- sis 427; sylvestris 427 Caspareopsis monandra 410 Cassia alata 410; antillana 410; bacillaris 410; bicapsularis 410; fistula 410; glandulosa var. swartzii 410; grandis 411; ob- tusifolia 411; occidentalis 411 Cassine xylocarpa 421 Cassytha filiformis 406 Castalia ampla 405; jamesoniana 373; zanzibarensis 405 Casuarina equisetifolia 394, 400 Catharanthus roseus 434 Caudle, Carol and Jerry M. Bas- kin, Petalostemon foliosus in Alabama 383 Cayaponia americana 446; race- mosa 446 Cecropia peltata 393, 401 Cedrela mexicana 374, 417; odora- ta 374, 417 Ceiba pentandra 425 Celandine Poppy, The Persistence of the Double-Flowered Form of 179 Rhodora [Vol. 69 Celosia argentea 403; crista 403; nitida 403 Celtis iguanaea 400; trinerva 400 Centaurea austriaca 379, 380; nigra 380 Centradenia 173, 174 Central America, Physlis in Mexi- co, the West Indies and 82, 203, 319 Centrosema virginianum 412 Centrostachys aspera 402; indica 402 Cephalocereus nobilis 374; royenii 428 Cerastium longepedunculatum 452; longipedunculatum 452 Cerbera thevetia 435 Cerdana alliodora 437 Cestrum diurnum 440; laurifoli- um 440; nocturnum 440 Ceterach aspidioides 5 Chaenomeles lagenaria 307, 318 Chaetocalyx scandens 393, 412 Chalcas exotica 417 Chamaecrista swartzii 410 Chamaefistula antillana 410; ba- cillaris 410 Chamaesaracha physaloides 225 Chamaesyce adenoptera SSp. per- gamena 374; anegadensis 374; articulata 418; blodgettii 418; buxifolia 418; hirta 418; hy- pericifolia 418; hyssopifolia 374, 418, 419; monensis 374; nutans 374; portoricensis 374; prostrata 419; serpens 419; turpenii 374; vahlii 418 Chamissoa altissima 403 Chaptalia nutans 448 Chelidonium 180; grandiflorum 180; laciniatum 180; majus 32, 179, 180, 182, var. pleniflorum 181, var. plenum 181, 182 Chenopodium ambrosioides 402 Chester, Edward W., Halesia 1961] carolina L. in Kentucky, Indi- ana and Ohio 380 Chiococea alba 444 Chione venosa 444 Chloris barbata 372; inflata 372; paraguayensis 372 Chrysobalanus icaco 408 Chrysophyllum argenteum 433; cainito 433 Church, George L., Pine Hills Elymus 330; Taxonomic and Genetic Relationships of East- ern North American Species of Elymus with Setaceous Glumes 121 Churchill, John A., Saxifraga Aizoón in New Hampshire 483 Churchill, John A. and A. R. Hodgdon, Saxifraga cernua in New England 485 Cicca acida 420; disticha 420 Cissampelos pareira 405 Cissus caustica 423; 423; trifoliata 423 Cistaceae 184, 186 Cistus 186 Citharexylus fruticosum 438 Citrullus citrullus 446; lanatus 446 Citrus 401; aurantifolia 416; grandis 416; limon 416; limonia 416; limonum 416; maxima 416; paradisi 416; sinensis 391, 416 sicyoides Cleome ciliata 407; gynandra 407; icosandra 407; rutio- sperma 407; speciosa 407; viscosa 407 Clerodendron aculeatum 438; speciosissimum 438; thomp- sonae 438 Clidemia hirta 429 Clitoria ternatea 412 Closterium abruptum 57; Kuetz- ingii 57; libellula var. inter- Index to Volume 69 491 medium 57; setaceum 57; ulna 57 Clusia 173; rosea 392, 426 Coccinea cordifolia 376; grandis 376 Coccoloba microstachya 402; uvi- fera 402; venosa 402 Coccolobis obtusifolia 402; uvifera 402 Coccothrinax alta 372; argentea 372 Cochlospermum vitifolium 427 Cochranea anchusaefolia 438 Codiaeum variegatum 419 Coffea arabica 444 Coleus amboinicus 439; 439 Colubrina arborescens 422; colu- brina 422; elliptica 374; reclin- ata 374, 422 rufa var. antillana 374 Commelina elegans 373; virginica 373 Commicarpus scandens 404 Comocladia cuneata 374; dentata 374; dodonaea 421 Comolia 165 Compositae 35, 352, 395, 447 Comptonia peregrina var. aspleni- folia 452, var. aspleniifolia 452 Conocarpus erecta 430 Conostegia hotteana 375 Consolea moniliformis 375; rubes- cens 375, 428 Convolvulaceae 395, 436 Convolvulus sepium f. coloratus 33, f. sepium 33 Conyza bonariensis 448; canaden- sis 448; chilensis 376 Corchorus aestuans 423; hirsutus 423; siliquosus 423 Cordia alliodora 437; collococca 437; glabra 437; nitida 437; obliqua 375, 437; polycephala 437; rickseckeri 438; sebestena 438; sulcata 438; ulmifolia var. blumei 492 ovata 437; wagnerorum 375 Coreopsis delphinifolia 452; del- phiniifolia 452 Corylus americana 458 Cosmos caudatus 448; sulphureus 448 Couroupita guianensis 432 Cracca brevipes 373; caribaea 412; purpurea 373 Crataegus anomola 307; Arnoldi- ana 307; basilica 307; Beck- withiae 307; biltmoreana 307; Brainerdii 307, var. asperifolia 307, var. cyclophylla 307, var. Egglestoni 307, var. scabrata 307; brumalis 307; Brunetiana 307, var. Fernaldii 307; chads- fordiana 307; chrysocarpa 307, var. Bicknellii 307, var. phoeni- cea 307; crus-galli 307, var. exigua 308; dilata 308; Dodgei 308; Faxoni 308, var. praeter- missa 308; flabellata var. Gray- ana 308; foetida 308; fretalis 308; Holmesiana 308, var. villipes 308; Idaea 308; insolens 308; intricata 308, var. strami- nea 308; irrasa var. Blanch- ardii 308; Jesupi 308; Jonesae 308; lemingtonensis 308; levis 308; littoralis 308; macro- sperma 308, var. acutiloba 308, var. demissa 308, var. matura 308, var. pentandra 308; media 308; membranacea 308; mono- gyna 308; Oakesiana 308; Oxyacantha 308; pedicellata 308, var. albicans 308; Phaeno- pyrum 308; pilosa 308; Porteri var. caerulescens 309; Pringlei 309, var. exclusa 309; pruinosa 309, var. dissona 309, var. latisepala 309; punctata 309, var. aurea 309; Randiana 309; rotundata 309; rugosa 309; shirleyensis 309; schizophylla Rhodora [Vol. 69 309; spatiosa 309; Stonei 309; submollis 309; suborbiculata 309; succulenta 309, var. ma- cracantha 309, var. neofluvialis 309; Websteri 309 Crataeva gynandra 373; tapia 373 Crescentia cujete 442 Crocanthemum 184-186, 192; brasiliense 186; carolinianum 186; sect. spartioides 199 Crossandra infundibuliformis 443 Crossopetalum rhacoma 421 Crotalaria berteriana 373; incana 412; latifolia var, eggersii 378; mucronata 373; retusa 412; sagittalis var. fruticosa 373; saltiana 373; striata 373, ver- rucosa 412 Croton 389; astroites 419; betuli- nus 419; lobatus 419; ovali- folius 393, 419; rigidus 419 Cryptostegia grandiflora 435 Ctenomeria 242 Cucumis anguria 447; sativa 447 Cucurbita lagenaria 447 Cupania americana 393 Curcas curcas 420 Curculigo capitulata 373 Cuscuta americana 436; globulosa 436; sp. 435, 436 Cyathea arborea 393 Cynanchum decaisneanum 435; parviflorum 435 Cyperaceae 306 Cyperus flavescens var. poaeform- is 452, var. poiformis 452 Cypripedium acaule f. acaule 31, f. albiflorum 31 Dalbergia ecastophyllum 412 Dalea cathaginensis 353; domin- gensis 353 Dalechampia scandens 419 Dalibarda repens 309, 316 Daphnopsis americana SSp. cari- 1967] baea 375; caribaea 375 D'Arcy, W. G., Annotated Check- list of the Dicotyledons of Tor- tola, Virgin Islands 385 Datura innoxia 441; metal 441; stramonium 441 Daucus carota 433 Delonix regia 411 Dendropanax arboreum 432 Dendropemon caribaeum 401 Desmanthus depressus 409; vir- gatus 409 Desmids and Flagellates new to Massachusetts 55 Desmodium axillare 413; canum 413; glabrum 413; procumbens 413; purpureum 413; scorpiu- rus 413; tortuosum 413; tri- florum 413 Diapedium assurgens 443 Dicliptera assurgens 443 Dicoytledons of Tortola, Virgin Islands, Annotated Checklist of the 385 Didymopanax morototoni 432 Digitaria diversifolia 372; Ischae- mum 31; sanguinalis 31 Diodia apiculata 376, 393, 445; rigida 376 Dipholis salicifolia 433 Dirca palustris 458 Dissotis 165 Ditaxis fasiculata 418 Ditremexa occidentalis 411 Dolicholus minimus 414; reticu- latus 414 Dolichos lablab 374, 413 Doliocarpus brevipedicellatus 375; major 875 Doxantha unguis-cati 442 Dryopteris aspidioides 5; assimilis 2; assurgens 5; Bangii 5; bi- formata 5; boqueronensis 5; campyloptera 1; coarctata 5; consobrina 5; deflexa 5; deflex- um 5; densa 5; Desvauxii 7; Index to Volume 69 493 dilatata 1, 2, 3, diploid from Quebec 1, ssp. campyloptera 2, 3; diplazioides 6; dumetorum 5; ensiformis 6; falcata 6; fur- va 6; gigantea 6; glandulosa- lanosa 6; Herzogii 6; interme- dia 1, 2; Jamesonii 6; Jurgen- senii 6; laevigata 6; Leprieurii 6; leucothrix 6; Lindigii 5; Linkiana 6; longifolia 7; lugu- briformis 7; Macbridei 7; ma- crotis 7; membranacea 7; mollis 1; nemoralis 7; nervosa 7; nitens 7; panamemsis 376; Pa- voniana 7; permollis 7; peru- viana 7; phacelothrix 7; pilosula 8; pteroidea 8; Rolandii 8; Rosei 8; Rosenstockii 8; Rusbyi 8; simplicifrons 6; spinulosa 1, 2, var. americana 1, 3; Sprengelii 376; Standleyi 8; subandina 8; X triploidea 2; tristis 8; Warmingii 8 Duchesnea indica 309, 318 Duggena hirsuta 445 Eaton, R. J., R. C. Bean, A. F. Hill and S. K. Harris, A Cor- rection in the 16th Report 318; Seventeenth Report of the Committee on Plant Distribu- tion 306 Ecastophyllum ecastophyllum 412 Echinochloa crusgalli f. longiseta 31; polystachya 372; spectabilis 372 Echites agglutinata 435 Eclipta prostrata 376, 448 Elaeodendron xylocarpum 421 Eleocharis geniculata 60; palus- tris var. palustris 318, var. major 318 Elephantopus mollis 448; scaber 448 Eleusine indica 69 Elymus 121-123, 125, 137, 158; 494 californicus 343, 344; canaden- sis 126, 128, 132, 133, 160, var. brachystachys 132, 133, 138, 140, 141, var. canadensis 122- 125, 127-130, 132, 133, 138, 140, 7142-144, 146, 148-153, 158-160, 330, 332, 340, 343, 346, f. cana- densis 138, f. glaucifolius 138, var. interruptus 133, 134, 138- 141, 143, 148, 154-157, 160, 332; diversiglumis 122-128, 130, 132- 134, 137, 139, 140, 142, 144-147, 149-151, 153, 155, 156, 158-160, 330-351; hybrids 148-161, 332- 344; Hystrix 121, 122, 125, 128, 135-187, 144, 145, 147-150, 153, 154, 157-161, 330-351; interrup- tus 123, 128, 133; occidentalis 127; riparius 138, 146, 151, 159; Svensonii 134, 135, 137, 139, 142, 148, 153-156, 160, 330; villosus 138, 147, 157, 158, 160, 346; virginicus 122, 138, 157- 161, 330, 332, 339, 340, 342, var. glabriflorus 138, 148, 157, 158, 160, 332, 343, f. australis 138, 158, 382, 333, 335, 336, 338- 343, 346, 350, 351, f. glabri- florus 138, 143, 156-158, 330, 332-335, 337-343, 345-351, var. virginieus 122, 138, 143, 155, 156, 158, 160, 337, 339, 342, 343, 346, 350, 351, f. virginicus 157, 160; Wiegandii 122, 125, 128-130, 132, 137, 138, 142, 143, 151-153, 159, 160; with Setace- ous Glumes, Taxonomic and Genetic Relationships of East- ern Ncrth American Species of 121 Emelista tora 411 Emilia coccinea 448; sonchifolia 448 Epidendrum ciliare 400 Epilobium 482; adenocaulon 482; angustifolium f. albiflorum 33; coloratum 482; glandulosum Rhodora [Vol. 69 var. adenocaulon 482 Episcia fulgida 443 Equisetum arvense f. campestre 30; X litorale 30; variegatum var. Jesupi 31, var. variegatum 30, 31 Eragrostis amabilis 372; pecti- nacea 31; poaeoides 31; tenella 372 Erechtites hieracifolia 448, 452; hieraciifolia 452 Eremalche 15, 21 Erigeron canadensis 448; cunei- folius 448 Erithalis fruticosa 445 Ernodia littoralis 445 Erythrina corallodendron 413; indica 413; sp. 413; variegata var. orientalis 413 Erythroxylon brevipes 415 Erythroxylum rotundifolium 415 Eucalyptus camaldulensis 430; rostrata 430 Eugenia axillaris 430; ? bellonis 375; biflora 389, 430; cordata 431; floribunda 431; jambos 432; lancea 430; monticola 431; procera 4431; pseudopsidium 431; rhombea 431; sessiliflora 431; sintensii 431 Euglena acus var. rigida 55, 56 Euonymus europaeus 379 Eupatorieae (Compositae), Stud- ies in I-III 35; IV 352 Eupatorium corymbosum 448; dissectum 353; odoratum 449; sinuata 449; spp. 387; urticae- folium 452; urticifolium 452; verbenaefolium | 452; verbeni- folium 452 Euphorbia 249; articulata 418; blodgettii 418; cyathophora 374; geniculata 374; hetero- phylla 374, 419; hirta 418; hypericifolia 418; hyssopifolia 419; lactea 419; mesembrian- 1961] themifolia 418; milii 419; petio- laris 419; pulcherrima 419; splendens 419; tirucalli 420 Euphorbiaceae 241, 395, 418 Eutheta 84 Evolvulus glaber 436 Exochorda grandiflora 309, 318 Exogonium arenarium 436; re- pandum 436 Exostema caribaeum 445 Eyde, Richard H. and James A. Teeri, Floral Anatomy of Rhexia virginica (Melastoma- taceae) 163 436; sericeus Fabaceae 395, 411 Fagara clava-herculis var. fruti- cosa 481, 482; fruticosa 481, 482 Fagus grandifolia 51-54, f. mollis 53, var. caroliniana 53; South- eastern Limit of 51 Faramea occidentalis 445 Fern Notes V. New Combinations in Peruvian Species of Thelyp- teris 5 Ferns, An Annotated Bibliogra- phy cf Mexican (Review) 62 Ficus citrifolia 401; crassinervia 401; laevigata 401; trigonata 392, 401 Filipendula hexapetala 309, 318; rubra 32, 309, 318; Ulmaria 309, 318, var. denudata 309, 318 Flagellates and Desmids new to Massachusetts 55 Fleischmannia Langlassei 360; Schaffneri 357; Standleyi 357; urenaefolia 360; urenifolia var. mexiae 357, var. typica 360 Fleurya aestuans 401 Flora North America Launched— News Announcement 63 Foeniculum foeniculum 432; vul- gare 432 Index to Volume 69 495 Forestiera eggersiana 434; segre- gata 434 Fragaria X Ananassa 309, 318; vesca 309, 315, var. americana 309, 315; virginiana 309, 315, var. terrae-novae 309, 316 Franseria acanthicarpa acanthocarpa 455 454; Fumana 186 Galactia 394; dubia 413; eggersii 413; striata 413 Galphimia glauca 374; gracilis 374 Gardenia jasminoides 445 Gaylussacia baccata 59 Gayoides crispum 374 Gentiana Catesbaei var. num- mulariaefolia 452, var. num- mulariifolia 452 Geoffrea inermis 374 Geophila herbacea 445; repens 445 Geum album 378; aleppicum var. strictum 309, 315; canadense 309, 316, 378, var. camporum 310, 316; laciniatum 310, 315, 316, var. trichocarpum 310, 316; macrophyllum 310, 316; Peckii 310, 316; X pulchrum 310, 318; rivale 310, 315; ur- banum 310, 318; vernum 378; virginianum 310, 316 Gillenia trifoliata 318 Ginoria rohrii 429 Gliricidia sepium 413 Gomphrena decumbens 403; dis- persa 403; globosa 403 Gonzalagunia spicata 445 Gossypium arborescens 424; bar- badense 375; brasiliense 375; hirsutum var. punctatum 375; peruvianum 375; purpurascens 375, 424; spp. 391 Gouania lupuloides 423 496 Grammitis Linkiana 6 Grevillea robusta 401 Guaiacum officinale 415 Guarea guara 374; guidonia 374 Guazuma guazuma 425; ulmifolia 425 Guettarda parviflora 445; scabra 445 Guilandina bonduc 410; crista 410; melanosperma 410 Gymnanthes lucida 420 Gymnogramma diplazioides 6; polypodioides 6 Gynandropsis gynandra 407 Habenaria blepharidiglottis 454; blephariglottis 454 Haematoxylon campechianum 393, 411 Halesia carolina 380, 381; in Kentucky, Indiana and Ohio 380 Halimium 186, 192; sect. Sparti- oides 192 Hamamelis virginiana 457, A New Station in Minnesota for 457 Hamamelistes spinosus 458 Harris, Stuart K., Notes on the Flora of Coós County, New Hampshire II 29 Harris, S. K., R. C. Bean, A. F. Hill and R. J. Eaton, A Correc- tion in the 16th Report 318; Seventeenth Report of the Committee on Plant Distribu- tion 306 Harvill, A. M. Jr., Stipulicida in Gray’s Manual Range 59 Heeria 173 Helianthemum 184-186, 192; are- nicola 187, 190; argenteum 190; bicknellii 190; brasiliense 184, 185, 187, 189, 190, 192; cana- dense 188, 190, 192; carolinia- num 185, 187, 189, 190, 192, 193; Rhodora [Vol. 69 chihuahuense 189, 190; concolor 188, 190; corymbosum 187, 190; coulteri 190; dumosum 191; georgianum 191, 192; glomera- tum 191; greenei 188, 191, 193; nashii 191; nutans 188, 191- 193; Palynological Notes on American Species of 184 i patens 191, 193; pringlei 191; propinquum 191; rosmarinifoli- um 191; scoparium 191, 193; sect. Lechioides 193; sect. Spartioides 193 Helianthus grosseserratus 452; grossiserratus 452; Maximili- ani 454; Maximilianii 454; trachelifolius 452; trachelii- folius 452 Heliconia caribaea 392, 393 Helicteres jamaicensis 425 Helogyne fasiculata 352, 3062; urenifolia 360 Heliotropium amplexicaule 438; angiospermum 438; crispiflor- um 438; curassavicum 438; gnaphalodes 438; indicum 438 Hemidinium nasutum 56, 57 Herpetica alta 410 Herschellia 84 Heteromeris 184 Heteropteris purpurea 416 Hibiscus brasiliensis 375; esculen- tus 424; phoeniceus 375; sab- dariffa 387, 424; vitifolius 424 Hieracium paniculatum f. glan- dulosum 34 Hill, A. F, R. C. Bean, R. J. Eaton and S. K. Harris, A Cor- rection in the 16th Report 318; Seventeenth Report of the Committee on Plant Distribu- tion 306 Himalaya, The Flora of Eastern (Review) 456 Hippobroma longiflora 447 Hippomane mancinella 420 1961] Hodgdon, A. R. and John A. Churchill, Saxifraga cernua in New England 485 Hofmeisteria 46, 352, 368, 369; crassifolia 353, 365, 367, 369; dissecta 352-354, 369; fasicula- ta 353, 362, 364, 369, var. Grayi 362, var. pubescens 363, var. Xanti 362; filifolia 353, 366, 369, 370; gentryi 36; laphamioides 40, var. pau- ciseta 42; pleuriseta 35-37, var. laphamioides 40, var. pauciseta 42; pubescens 362; schaffneri 353, 357, 359, 368, 369; sinaloensis 353, 355, 356, 369; standleyi 353, 357, 358, 368, 369; tenuis 44; urenifolia 353, 360, 361, 368, 369; viscosa 37 Hordeae 122 Hordeum 122; jubatum 157 Howard, Richard A., The Per- sistence of the Double-Flowered Form of Celandine Poppy 179 Hoya carnosa 436 Hudsonia 186 Hura crepitans 420 Hylocereus trigonus 428; undatus 428 Hymenaea courbaril 411 Hymenophyllum contortum 376; crispum 376; fucoides 376; hir- sutum 376; lanatum 376; pro- trusum 376 Hypolepis nigrecens 376 Hyptis capitata 439; pectinata 439; suaveolens 439 Hystrix 344; patula 121-123, 135, 342, var. Bigeloviana 135 Ibatia maritima 436 Icacorea guadalupensis 433 Icthyomethia piscipula 414 Ilex hypaneura 374; urbania 421 Iliamna 21 Index to Volume 69 497 Impatiens spp. 415 Indiana, Halesia carolina in Ken- tucky, Ohio and 380 Indigofera suffruticosa 414 Inga fagifolia 409; laurina 409 Ipomoea acuminata 436; aegyptia 437; batatas 436; cathartica 436; dissecta 437; nil 486; pes- caprae 436; polyanthes 437; quamoclit 436; quinquefolia 437; repanda 436; spiralis 375; spirillus 375; steudelii 436; tiliacea 436; triloba 436; tuba 436 Iresine angustifolia 403 Isotoma longiflora 447 Iva xanthifolia 452; xanthiifolia 452 Ixora chinensis 445; coccinea 445; ferrea 445; finlaysonia 445 Jacaranda acutifolia 442; mimosi- folia 442 Jackson, R. C. and R. R. Johnson, A New Species of Machae- ranthera Section Psilactis 476 Jacquemontia jamaicensis 436; nodiflora 436; pentantha 436 Jacquinia arborea 375, 433; bar- basco 375, 433; berterii 433; revoluta 433 Jambosa jambos 432 Jasminum azoricum 375, 434; flu- minense 375, 434; grandiflorum 434; multiflorum 375; officinale var. grandiflorum 434; pubes- cens 375, 434; sambac 434 Jatropha curcas 420; gossypifolia 420; hastata 420; integerrima 420; multifida 420 Johnson, Albert G. and Margaret H. Smithberg, A New Station for Hamamelis virginiana L. in Minnesota 457 Johnson. R. R. and R. C. Jackson, A New Species of Machaeran- 498 thera Section Psilactis 476 Juniperus virginiana var. crebra 331 Jussiaea angustifolia 432; erecta 432 Justicia carthaginensis 443; peri- plocifolia 443; sessilis 443 Kalanchoé brasiliensis 408; laxi- flora 408; pinnata 408; somali- ensis 408; verticillata 393, 408 Kallstroemia 456; caribaea 374, 415; maxima 415; pubescens 374, 415 Kalmia angustifolia 59, polifolia 452; poliifolia 452 Kentucky, Halesia carolina in Indiana, Ohio and 380 Kerria japonica 310, 318 King, R. M, Studies in the Eupatorieae (Compositae) I-III 35; IV 352 Kosteletzkya virginica var. al- theaefolia 452, var. altheifolia 452 Krugiodendron ferreum 423 Labiatae 395, 439 Lablab niger 374 Lactuca sativa 449 Lagenaria siceraria 447 Lagerstroemia indica 429 Laguncularia racemosa 430 Lantana arida 439; camara 439; involucrata 439; reticulata 439 Lasiacis maculata 372; sorghoi- dea 372 Lathyrus odoratus 415 Leavenworthia 383 Lechea 186 Leguminosae 48, 395, 408 Leonotis nepetaefolia 440 Leontodon autumnalis var. pra- tensis 34 Leonurus sibiricus 440 Lepidium campestre 463; pinnati- Rhodora [Vol. 69 sectum 373; virginicum 407 Leptilon bonariense 448; pusillum 448 Leptobotrys 297; discolor 297, 300 Leptocereus grantianus 375 Leptocinclis acuta 56 Leptorachis 242 Leucaena glauca 409; leucocepha- la 409 Leucas martinicensis 440 Licaria salicifolia 406 Limnobium spongia 372 Lindera melissaefolium 453; me- lissifolium 453 Liogier, Brother Alain H., Fur- ther Changes and Additions to the Flora of Porto Rico and the Virgin Islands 372 Liparis lilifolia 453; liliifolia 453 Livingston, Robert B., A Spurless Form of Aquilegia canadensis L. 195 Lolium multiflorum var. diminu- tum 31 Lomopolis ceratonia 409 Lonchocarpus domingensis 373; sericeus var. glabrescens 373 Lotus corniculatus 33 Ludwigia erecta 432; leptocarpa var. meyeriana 375; octovalvis 432 Lychnis Flos-cuculi 32 Lycopersicum esculentum 441 Lycopedium sabinaefolium 453; sabinifolium — 453; setaceum 376; verticillatum 376 Lyonia ligustrina var. capreae- folia 453; var. capreifolia 453 Machaeranthera 478; arida 476, 478-480; arizonica 476-480; coulteri 480; crispa 476, 478- 480; parviflora 480; sect. Psilactis 478, 480 Macroptilium lathyroides 414 Malache scagra 375 1967] Index to Malachra alceifolia 424, var. fas- ciata 374; capitata 424; fascia- ta 375 Malacothamnus 14, 15, 21 Mallotonia gnaphalodes 438 Malparia 43; tenuis 43, 44 Malpighia 394; angustifolia 374; glabra 416; linearis 374; puni- cifolia 416 Malva 10, 14, 21; anomala 16; perpusilla 23; tricuspidata 11 Malvaceae 9, 10, 395, 424 Malveae 16, 21 Malvastrum 9, 11, 13-16, 18, 21- 23; amblyphyllum 19, 20; amer- icanum 11, 15, 18, 19, 22, 424; angustum 9-15, 23, 26; auran- tiacum 11, 19, 20; bicuspidatum 11, 20; coccineum 14, 15; cor- chorifolium 11, 12, 18, 424; coromandelianum 11-13, 15, 16, 18, 20, 22, 424; depressum 20; dimorphum 20; fremontii 14; grossulariaefolium 14; guarani- ticum 20; guatemalense 20; hispidum 14, 16, 21-23, 25, 26; interruptum 18, 19, 23; lacteum 15; munroanum 14; Notes on 9; Rugelii 18; scabrum 20, 21; scoparioides 20; scoparium 20; spicatum 18, 424; spiciflorum 18, 19; subtriflorum 15; tri- cuspidatum 11, 18; Wrightii 19 Malvaviscus arboreus 424 Malveopsis 15, 16; hispida 23 Mammea americana 426 Mammillaria nivosa 428 Mangifera indica 391, 421 Manihot manihot 420; utillisima 420 Manilkara albescens 375; balata 392, 434; bidentata 434; nitida 434; zapota 434 Maregravia rectiflora 392, 426 Margaranthus 84, 100, 224; sul- phureus 224 Volume 69 499 Margaritaria nobilis 420 Martynia annua 443 Massachusetts, Flagellates and Desmids new to 55 Mastichodendron .foetidissimum 434 Matelia maritima 436 Meibomia axillaris 413; mollis 413; procumbens 413; purpurea 413; scorpiurus 413; supina 413 Melastoma malabathricum 173, 176 Melastomataceae 163, 165, 174, 429 Melanthera confusa 449; 449 Melia azedarach 417 Melicocca bijuga 422 Melicoccus bijugatus 422 Meliosma herberti 422 Melocactus intortus 428 Melochia nodiflora 387, 426; pyra- midata 426; tomentosa 426 Melothria quadalupensis 447 Memecylon 174 Meniscium falcatum 6; giganteum 6; longifolium 7 Mentha arvensis f. lanata 33 Merremia aegyptia 437; dissecta 437; quinquefolia 437; umbella- ta 437 Metastelma decaisneanum 435; parviflorum 435 Metopium linnaei 374; toxiferum 374 Mexico, Physalis in Central America, the West Indies and 82, 203, 319 Meyenia erecta 444 Miconia 175; fothergilla 429; laevigata 429; mirabilis 429; prasina 429; thomasiana 429 Micrasterias pinnatifida 57, f. rhomboidea 56, 58; rotata 58 Mikania cordifolia 449 Miller, Kim I. and Grady L. nivea 500 Webster, A Preliminary Re- vision of Tragia (Euphorbia- ceae) in the United States 241 Mimosa ceratonia 409; pudica 409 Mimosaceae 395, 408 Minnesota, A New Station for Hamamelis virginiana in 457 Mirabilis jalapa 404; nyctaginea 32 Mississippi, Allium ampeloprasum in 61 Mitropsidium sintenisii 375 Molineria hortensis 373 Mollugo verticillata 404 Moluchia pyramidata 426; tomen- tosa 426 Momisia iguanaea 400 Momordica charantia 447 Monarda fistulosa var. menthae- folia 453, var. menthifolia 453; menthaefolia 453; menthifolia 453 Monochaetum 175 Montezuma grandiflora 375; spe- ciosissima 375 Morinda citrifolia 445 Moringa moringa 408; oleifera 408; pterygosperma 408 Morisonia americana 393, 407 Mouriri 174, 175; guianensis 173 Murraya exotica 417 Mussaënda luteola 446 Myrcia citrifolia 430; splendens 431 Myrcianthes fragrans 431 Myrciaria floribunda 387, 431 Myrica asplenifolia 453; asplenii- folia 453 Myrsine guianensis 433 Myrtaceae 395, 430, 431 Myrtus ? bellonis 375 Nasturtium officinale 373 Nectandra coriacea 406 Neltuma juliflora 409 Neomammillaria nivosa 428 Rhodora [Vol. 69 Nephrodium deflexa 5; Jamesonii 6; Leprieurii 6; macrotis 7; nemorale 7 Nerium oleander 434 Neurolaena lobata 449 New England, Saxifraga cernua in 485 New Hampshire, Notes on the Flora of Coós County II 29; Saxifraga Aizoón in 483 Nicotiana tabacum 441 Nopalea cochenillifera 428 Nototriche 21 Nymphaea ampla 405; jamesoni- ana 373; pulchella 405; zanzi- barensis 405 Ocimum basilicum 440; micran- thum 440 Ocotea floribunda 406; leucoxylon 406 Oenothera pilosella 379 Ohio, Halesia carolina in Ken- tucky, Indiana and 380 Oklahoma, Additions to the Vas- cular Flora of, II, 480 Olvesia tomentosa 410 Ophioglossum 480; crotalophoro- ides 480, 481; engelmannii 480 Opuntia antillana 428; dillenii 428; moniliformis 375; repens 428; rubescens 375, 428 Osmia corymbosa 448; odorata 449; sinuata 449 Osmunda cinnamomea f, cornu- copiaefolia 453, f. cornucopii- folia 453 Ossaea domingensis 375; scalpta 375 Ostrya virginiana 458 Oxalis corniculata 415; europea f. villicaulis 33; rugeliana 374 Pachystachys coccinea 444 Palicourea crocea 445; domingen- sis 446; riparia 445 1967] Panax trifolius 33 Pandorea ricasoliana 442 Panicum calliphyllus 454; callo- phyllum 454 Parosela domingensis 373 Parkinsonia aculeata 411 Parthenium hysterophorus 449 Paspalum longepedunculatum 453; longipedunculatum 453; millegrana 60; setaceum var. longepedunculatum 453, var. longipedunculatum 453 Passiflora foetida 427; laurifolia 427; multiflora 427; pallida 427; rubra 427; suberosa 427 Paullinia pinnata 422; plumieri 422 Pavonia scabra 375; spicata 375; spinifex 424 Pectis linifolia 449 Pedilanthus angustifolius 420; padifolius 420; tithymaloides 393, 420, ssp. angustifolius 420, ssp. padifolius 420, ssp. parasiticus 420, ssp. tithymal- oides 420 Pentaphiltrum 84 Peltophorum dubium 411; ferru- gineum 411; inerme 391, 411; pterocarpum 411 Penstemon tubaeflorum 453; tubi- florus 453 Peperomia glabella 400; humilis 400; magnoliaefolia 400; pellu- cida 400; spp. 392 Pereskia grandifolia 428 Perry, James D. and Robert L. Wilbur, Palynological Notes on American Species of Helianthe- mum (Cistaceae) 184 Persea americana 406; persea 406 Petalostemon foliosus 383; gattin- geri 384 Petiveria alliacea 404 Petraea volubilis 439 Petroselinum crispum 433 Index to Volume 69 501 Phaca 50 Phacus helikoides 55, 56; longi- cauda 55, 56; pyrum 55 Phania ? dissecta 352, 353; ureni- folia 360; urenaefolia 360 Phaseolus lathyroides 414; luna- tus 414; vulgaris 415 Phegopteris laevigata 6; branacea 7; mollis 7 Philodendron spp. 392 Philoxerus vermicularis 403 Phryganocydia corymbosa 442 Phthirusa caribaea 401 Phyllanthus acidus 420; amarus 421; niruri 421; nobilis 420; tenellus 374 Phymosia 21 Physalis 82-84, 100, 224, 327; acuminata 105, 106; acutifolia 83, 91, 223, 224; aequata 213; aggregata 86, 88, 108; Alkeken- gi 219; amica 85, 96; amphitrica 85, 86, 97; ampla 91, 219; angulata 90, 91, 216, 217, 219, 222, 327, 441, f. ramosissima 218, var. angulata 18, var. cap- sicifolia 218, var. lanceifolia 216, 219, var. ramosissima 218; angustior 86, 88, 89, 114; angus- tiloba 94, 320; angustiphysa 92, 228; arborescens 83, 85, 95; barbadensis 321; calidaria 85, 97; campanula 86 (as campula), 100; campanulata 208, 204; capsicifolia 218; cardiophylla 209; carnosa 95, 328; caudella 87, 89, 117, 119, var. caudella 118, 119, var. parva 118, 119; chenopodifolia 83, 90, 93, 107, 109, 110, 213, var. chenopodi- folia 109, f. immaculata 110, var. exigua 110, var. glandulosa 111, var. viridis 110; chimalteca 83, 86, 98; cinerascens 102; cinerea 86, 102; clarionensis 94, 326, 327; constricta 86, 99; cor- mem- 502 data 82, 94, 235, 236, 322, 323; Coztomatl 88; crassifolia 91, 206-209, 212, var. cardiophylla 209, var. crassifolia 207, 208, 210, f. crassifolia, f. muriculata 209, subvar. amplifolliata 209, 210, subvar. crassifolia 209, 210, var. infundibularis 207, 208, 211, var. versicolor 206, 207, 210, 212; curassavica 102, var. integrifolia 102, var. sinua- to-dentata 104; dentata 83, 224; dubia 217; Eggersii 224, 225; eximia 83, 100, 101, 227; Fendl- eri 206, var. cordifolia 206; filipendula 207, 208, 210; flava 207, 208, 211; floridana 323; foetens 93, 94, 319, 320, var. neomexicana 320; Galeottiana 113; Galeottii 113; genuicaulis 210; glabra 91, 113, 206, 207, 212; glutinosa 82, 83, 86, 100, 102, 227, var. eximia 101, var. glutinosa 101; gracilis 87, 111, 113; Greenei 206, 207; Green- manii 91, 226; Halicacabum 219; hastata 206; hastatula 90, 111; Hintonii 91, 226; hirsuta 323, var. barbadensis 321; hederaefolia 88, 204, 207, var. cordifolia 204, 206, var. hede- raefolia 205, var. puberula 116, 205; hylophila 94, 825, 326; ignota 93, 238, 322; in Mexico, Central America and the West Indies 82, 203, 319; ingrata 87, 107; ixocarpa 213, 327; jaliscen- sis 92, 231; laevigata 213, 214; Lagascae 91, 217, 220-222, var. glabrescens 221, 225, var. La- gascae 220, 221; lanceifolia 219; lanceolata var. laevigata 87, 117; latifolia 113; latiphysa 95, 238; leptophylla 94, 324- 326; lignescens 92, 231; Linki- ana 217; lobata 86, 225, 226; Rhodora [Vol. 69 longicaulis 87, 12; longifolia 119; longipedicellata 92, 230; luteoanthera 86, 113; maxima 82, 93, 232, 233; mayana 83, 95, 96; Mevaughii 89, 104; melanocystis 85, 98, var. cernua 85, 99, var. melanocystis 99; micrantha 220; microcarpa 91, 222; michoacanensis 90, 215; microphysa 88, 203, 204; Mimulus 90, 95, 211; minima 221, 222; minimaculata 89, 219; minuta 95, 327; mollis var. cinerascens 102, var. parviflora 102; Muelleri 88, 115; muricula- ta 209; neomexicana 320; nican- droides 98, 233, 234, var. attenuata 232, 235, var. nican- droides 234; obscura var. glabra 236, 323, var. viscido-pubescens 236, 323; orizabae 83, 89, 106, 111; Palmeri 205; parvianthera 90, 204; parviculea 220, 221; pedunculata 206, 207; Pennellii 89, 116, 117; pensylvanica 102, var. cinerascens 102; pentagona 238, 239; peruviana $83, 113, 114, var. latifolia 113; philadel- phica 87, 89-91, 113, 213, 218, 327, var. immaculata 215, var. minor 213, var. parviflora 215, var. philadelphica 214, f. phila- delphica 214, 215, f. pilosa 214; philippensis 87, 92, 117; por- phyrophysa 85, 96; porrecta 94, 237, 238, 322; Pringlei 91, 92, 227, var. curtiloba 228, var. Pringlei 228; puberula 107, 109, 110; pubescens 94, 221, 222, 236, 322, 323, 325, 441, var. glabra 236, 323, var. minuti- flora 323, 324; pumila var. sonorae 119; purpurea 91, 211; ramosissima 218; rigida 119; Rydbergii 88, 90, 212; Rzedow- ski 230; sabaena 225; saltillen- 1961] sis 102, 103; Sancti-josephi 86, 89, 114; Schiediana 111, 112; sonorensis 208, 210; sordida 89, 90, 115, 117; stapelioides 86, 105; subintegra 106, 107; subrepens 92, 229, 230; subulata 93, 319, 320, var. neomexicana 320, var. subulata 320; sul- phurea 91, 224, 225; tehuacan- ensis 87, 203; turbinata 92, 94, 236, 321, 322, 441; turbinatoides 94, 321; versicolor 208, 210, var. microphylla 208, 210; vestita 92, 236; villosa 323; virginiana 87 (as virginica), 89, 90, 118, 209, 213, f. macrophysa 209, var. longiseta 119, 120, var.nana 119, var. polyphylla 118, var. sonorae 119, var. subglabrata 213; viridoflava 92, 230; viscido- pubescens 323; viscosa 83, 86, 102, var. cinerascens 102, 104, var. sinuatodentata 103, var. spathulaefolia 103, var. yuca- tanensis 104, 322; volubilis 92, 229, 230; Wrightii 223 Physocarpus opulifolius 310, 318 Phytclacea rivinoides 404 Picramnia pentandra 417 Pictetia aculeata 414 Pilea herniarioides 401; micro- phylla 401; nummulariaefolia 401; sanctae-crucis 401; tener- rima 401 Pilosocereus nobilis 875; royenii 428 Pimenta racemosa 431; var. gri- sea 431 Pimpinella anisum 433 Pinus echinata 331 Piper amalago 400 Piptocoma antillana 394, rufescens 449 Piscidia piscipula 414 Pisonia fragrans 404; subcordata 404 449 ; Index to Volume 69 503 Pithecellobium saman 410; un- guis-cati 409 Plant Distribution, Seventeenth Report of the Committee on 306; A Correction in the 16th Report 318 Plantago major 444 Pleurocoronis 35, 43; gentryi 35- 37, 43, 240; laphamioides 36, 41-43, var. laphamioides 40, 42, 43, var. pauciseta 42, 43; leaves 36; pleuriseta 36, 38, 42, 43 Pleurotaenium nodosum 58; nodu- losum 58 Pluchea carolinensis 449; odorata 449; purpurascens 449 Plukenetia verrucosa 251, 252; volubilis 251, 252 Plumbago auriculata 433; capen- sis 433; scandens 433 Plumeria alba 435; rubra 435 Poa trivialis 377 Podophania 352; dissecta 353; Ghiesbreghtiana 353 Poinciana pulcherrima 410 Poinsettia cyathophora 374; geni- culata 374; heterophylla 374, 419; pulcherrima 419 Polyenemum majus 31 Polygonum pensylvanicum var. rosaeflorum 453, var. rosiflorum 453 Polypodium attenuatum 376; nervosum 7; nitens 7; Pavonia- num 7; pteroideum 8; triseriale 376; triste 8 Polyscias filicifolia 432; guilfoylei 432 Pongamia pinnata 409 Poponax macracanthoides 373 Populus 80; aurea 76, 77; bal- samifera 78; eotremuloides 78; grandidentata 70-78, 77-80, X tremuloides 72; heterophylla 78; lindgreni 78; sieboldii 80; tremula 78, 80, var. davidiana 504 80; tremuloides 70-79, 458; vancouveriana 77; voyana 78; washoensis 78, 80 Porana paniculata 437 Porter, Duncan M., Another Tri- bulus Adventive in the New World 455 Portlandia grandiflora 446 Porto Rico and the Virgin Is- lands, Further Changes and Additions to the Flora of 372 Portulaca oleracea 405; phaeo- sperma 405; pilosa 405, var. hortualis 405; quadrifida 405; rubricaulis 405; umbraticola 405 Portulacaceae 466 Potamogeton spathulaeformis 453; spathuliformis 453 Potentilla anserina 310, 317; argentea 310, 318, var. pseudo- salabra 310; arguta 310, 315; canadensis 310, 316, var. villo- sissima 310; canescens 310; Egedii var. groenlandica 310, 317; erecta 310; fruticosa 310, 315; gracilis 310, var. pulcher- rima 310, 318; intermedia 310, 318; millegrana 310, 318; nor- vegica 310, 315, var. labradori- ca 310, 316; palustris 310, var. villosa 310, 315; pectinata 310, 317; recta 310, 318; reptans 310, 318 (as repens) ; Robbinsi- ana 32, 310, 316; simplex 310, 316, var. calvescens 310, 316; tridentata 310, 316; verna 310, 318 Pothomorphe peltata 400 Prestonia agglutinata 435 Pringle Herbarium, Notes from the II 377 Priva lappulacea 439 Prosopis 257; juliflora 409 Prunus alleghaniensis 310, 317; americana 311, 316; Avium Rhodora [Vol. 69 311, 318; cerasifera 311, 318; Cerasus 311, 318; depressa 311, 316; domestica 311, 318; Gra- vesii 311, 317; Insititia 311, 318; Mahaleb 311, 318; mari- tima 311, 317; nigra 311, 315, 316; pensylvanica 311, 315; Persica 311, 318; serotina 311, 316; spinosa 311, 318; susque- hanae 311, 316; virginiana 311, 315 Pseuderanthemum atropurpureum 444 Pseudo-elephantopus spicatus 449 Psidium amplexicaule 431; guaia- va 431 Psilactis 478, 479 Psoralea multijuga 48-50, The Identity of 48; subacaulis 384 Psychotria brownei 446; grosour- dyana 376; guadaloupensis 376; ligustrifolia 446; microdon 446; nervosa 376; pinularis 446; un- data 376, 446 Pterocaulon virgatum 449 Puccinellia distans var. angusti- folia 377 Pullen, Thomas M., Allium ampel- oprasum in Mississippi 61 Punica granatum 429 Pyrus americana 311, 316; arbuti- folia 311, 316; X Arnoldiana 311; Aucuparia 32, 311, 318; baccata 311, 318; communis 311, 318; decora 311, 316, var. groenlandica 311, 316; floribun- da 311, 315; X magdaburgensis 311; Malus 811, 318; melano- carpa 311, 315; X mixta 311; X pinnatifida 311, 318; pruni- folia 311, 318; X Shurangiaca 311; X Soulardi 311, 318 Pyxidanthera barbulata 59 Quamoclit quamoclit 436 Quebec, Diploid Dryopteris dilata- 1961] Index to Volume 69 505 ta from 1 373; palustris 373 Quercus borealis 54; falcata var. Rosa acicularis 311, 317, var. pagodaefolia 453, var. pagodi- folia 453; macrocarpa 458, f. olivaeformis 453, f. oliviformis 453; pagodaefolia 453; pagodi- folia 453; rubra 458; stellata 331 Quincula 84; lepidota 225; lobata 225 Quisqualis indica 430 Radicula palustris 373 Randia aculeata 446; mitis 446 Ranunculus bulbosus var. valde- pubens 453, var. valdipubens 453; repens 32; sicaeformis 453; siciformis 453 Rapanea guianensis 433 Raphanus sativus 407 Rauwolfia lamarckii 435; nitida 435; tetraphylla 435; viridis 435 Reynosia uncinata 423 Rhacoma crossopetalum 421 Rhexia 163, 165, 168, 169, 172- 176; virginica 163-167, 170, 171, Floral Anatomy of 168, var. virginica 163 Rhinanthus Crista-galli 33 Rhizophora mangle 432 Rhodotypus scandens 311, 318 Rhynchelytrum repens 372 Rhynchosia minima 414; reticula- ta 414 Riba, Ramon, An Annotated Bibli- ography of Mexican Ferns (Re- view) 62; New Taxa in the Genus Alsophila 65 Ribes hirtellum var. saxosum 32 Ricinus communis 421 Rivea campanulata 437 Rivina humilis 404 Rondeletia pilosa 446 Rorippa islandica var. hispida 373; nasturtium-aquaticum Bourgeauiana 311, 317; blanda 311, 317; canina 311, 318; caro- lina 311, 315, var. grandiflora 312, 317, var. villosa 312, 317; cinnamomea 312, 318; Eglan- teria 312, 318; gallica 312, 318; johannensis 312, 317; micran- tha 312; multiflora 312, 318; nitida 312, 315; palustris 312, 315, rubrifolia 312; rugosa 312, 318; setigera 312, 318, var. to- mentosa 312, 318; spinosissima 312, 318; virginiana 312, 315 Rosaceae 306, 408 Rubiaceae 395, 444 Rubus abbrevians 312; aculiferus 312; adjacens 313; alleghanien- sis 312, 314, 315, var. Gravesii 312, var. neoscoticus 312, 317, var. plausus 312, var. populi- folius 312; alumnus 312; ami- calis 313; amnicola 314; arcu- ans 314; arenicola 312, 316; argutus 313, 316; arundelanus 314, var. Jeckylanus 314; as- cendens 314; avipes 314; Bai- leyanus 313; barbarus 314; bellobatus 313; biformispinus 312; bifrons 313; Blakei 313; Blanchardianus 313; Braineri- dii 312; canadensis 313, 315; Chamaemorus 313, 316; coro- narius 373; cubitans 313; cuneifolius 313, 316; curtipes 312; elegantulus 313, 316; En- slenii 313, 316; facetus 314; felix 313; flagellaris 313, 316; flavinanus 314; floricomus 314; frondisentis 312; frondosus 313, 316; glandicaulis 312; Grouti- anus 314; hispidioides 314; his- pidus 313, 315, var. obovalis 313, 315; hybrids 312-314; idaeus 313, 318, var. canadensis 506 313, 315, var. Egglestonii 313, 317, var. heterolasius 313, 317, var. strigosus 313, 315; illece- brosus 313, 317; insulanus 313; jacens 313; jactus 312; Jay- smithii 313, 316; jugosus 313; laciniatus 314, 318; Lawrencei 314; laevoir 312; miscix 314; multiformis 3813; multilicius 313; multispinus 313; X neglec- tus 314; occidentalis 314, 315; odoratus 314, 315; ortivus 314; paludivagus 313; Parlinii 314; paulus 312; pauper 313; pensil- vanicus 314, 315; perinvisus 314; permixtus 312; perpauper 313; plicatifolius 314; phoeni- colasius 314, 318; pubescens 314, 315, var. pilosifolius 314, 316; pugnax 312; ravus 33, 312; recurvans 313; recurvicau- lis 314, 315; saltuensis 312; scambens 313; sceleratus 312; semisetcsus 314, 316; setosus 314, 315; severus 314; spiculo- sus 313; tardatus 314; tavus 33; tetricus 313; tholiformis 313; trifrons 313, var. pudens 313; triphyllus 314, 318; uni- vocus 314; vermontanus 314, 315 Rudbeckia serotina f. homochroma 33, f. pleniflora 33 Ruellia coccinea 444; 444; tweediana 444 Russellia equisetiformis 442 tuberosa Sabinea florida 387, 414 Sagotia triflora 413 Salix fragilis 31; pentandra 31 Salvia lanceaefolia 453; lancei- folia 453; occidentalis 440; serotina 440; thomasiana 394, 440 Salvinia rotundifolia 376 Samanea saman 391, 410 Rhodora [Vol. 69 Sambucus pubens f. rosaeflora 4583, f. rosiflora 453; simpsonii 446 Samyda dodecandra 427 Sanguisorba canadensis 314, 316; minor 314, 318; officinalis 315, 318 Sapium jamaicense 374; spp. 421 Sapota achras 434 Saracha acutifolia 223; trica 97; stapeliaflora stapelioides 105 Sarcomphalus domingensis 423; reticulatus 423 Saxifraga Aizoón 483-486, in New Hampshire 483, The Station Revisited 484, var. neogaea 483; cernua 485, 486, in New Eng- land 485; rivularis 486 Scaevola plumierii 447 Schaefferia frutescens 422 Schobera angiosperma 438 Schrankia 480; roemeriana 481; uncinata 481 Schuyler, Alfred E., A New Status for an Eastern North American Scirpus 198 Scirpus, A New Status for an Eastern North American 198; atrovirens 198, 200, var. flacci- difolius 198, var. georgianus f. viviparus 31; flaccidifolius 198- 201; georgianus 198, 200; hat- torianus 198, 200; polyphillus 377 Scleria melaleuca 372; pterota 372 Scoparia dulcis 442 Sebesten rickseckeri 438; sebes- tena 438 Secale cereale 31 Sechium edule 447 Selenicereus grandiflorus 428 Senegalia muricata 408; westiana 408 Serjania polyphylla 422 Sesbania grandiflora 414 amphi- 105; 1961] Sesuvium portulacastrum 404 Seymour, Frank C., A Correction 69; Notes from the Pringle Herbarium II 377 Shetler, Stanwyn G., News An- nouncement — Flora North America Launched 63 Sibbaldia procumbens 315, 316 Sida 10, 12-14, 22; acuminata 425; acuta 425; aggregata 374; alba 374; carpinifolia 425; cili- aris 425; eggersii 425; erecta 425; glabra 425; glomerata 425; hispida 9-14, 23; humilis 425, rhombifolia 11; salviaefolia 425; setifera 374; spinosa 11, 374; urens 425 Sideroxylon foetidissimum 434 Sidopsis 9, 14, 16, 21, 22; A re- consideration of 9; hispida 23 Silene Cserei 32; Cucubalus 32; pumilio 32 Sinapis alba 373 Sisymbrium nasturtium-aquati- cum 373 Sitanion 122 Sium cicutaefolium 453; cicuti- folium 453 Smilax Bona-nox var. hederae- folia 453; var. hederifolia 453 Smithberg, Margaret H. and Albert G. Johnson, A New Sta- tion for Hamamelis virginiana L. in Minnesota 457 Solanaceae 395, 440 Solanum 82; caribaeum 376; eleagnifolium 398, 441; erian- thum 441; ficifolium 441; lan- ceifolium 441; melongena 441; nigrum 376, 441; nodiflorum 376, 441; persicaefolium 441; polygamum 441; quitoense 376; torvum 441; tuberosum 441; verbascifolium 441; Woodburyi 376 Solidago microglossa 450 Index to Volume 69 507 Sonchus oleraceus 450 Sonerila 165 Sophora tomentosa 414 Sorbaria sorbifolia 315, 318 Spathodea campanulata 442 Spermacoce confusa 446, tenuior 446 Sphaeralcea 13-16, 21; angusta 9, 13, 23 Spiraea alba 315, 317; japonica 315, 318; latifolia 315, var. septentrionalis 315; prunifolia 315, 318; tomentosa 315 Spondias mombin 421; purpurea 321; sp. 421 Stachys arvensis 376 Stachytarpheta jamaicensis 439 Staurastrum avicula var. subar- cuatum 57; biennianum var. ellipticum 57; controversum 57; eryptoserum 57; dejectum 57; Dickei 57; furcatum var. pisciforme 57; gracile 57; John- sonii var. depauperatum 57; paradoxum 57; pentacerum 57; polymorphum 57; setigerum 57 Steele, Frederic L., The Saxifraga Aizoön Station Revisited 484 Stephanotis floribunda 436 Stevia 46 Stictocardia tiliaefolia 437 Stigmaphyllon diversifolium 374, 393, 416; ledifolium 374, 416; lingulatum 416; periplocifolium 416; spp. 416 Stimson, William R., Additions to the Flora of the Bimini Island Group, Bahama Islands 60 Stipulicida, in Gray’s Manual Range 59; setacea 59 Stuckey, Ronald L., Thlaspi tu- berosum Nutt., a Taxonomic Synonym of Cardamine Doug- lassii Britt. (Cruciferae) 460 Stylosanthes hamata 414 Suriana maritima 417 508 Swietenia macrophylla 418; ma- hagoni 418 Symphoricarpus occidentalis 458 Symplocos martinicensis 434 Synedrella nodiflora 450 Syzygium jambos 392, 393, 432 Tabebuia heterophylla 376, 442; pallida 376, 442; rufescens 443 Taeniostema 184 Talinum 467, 468, 474, A New, from the Cedar Glades of Mid- dle Tennessee 466; appalachia- num 467, 469, 474; calcaricum 466, 469-474; calycinum 467- 469, 473; mengesii 467-474; paniculatum 405; parviflorum 467, 469; rugospermum 467-469; teretifolium 468-471, 474; tri- angulare 405 Tamarindus indica 393, 411 Tamonea guianensis 429 Tarasa 15, 21 Taylor, Constance and R. John Taylor, Additions to the Vascu- lar Flora of Oklahoma II 480 Taylor, R. John and Constance Taylor, Additions to the Vascu- lar Flora of Oklahoma II 480 Tecoma stans 443 Tecomaria capensis 443 Tectona grandis 391, 439 Teeri, James A. and Richard H. Eyde, Floral Anatomy of Rhexia virginica (Melastoma- taceae) 163 Tektona grandis 439 Tennessee, A New Talinum from the Cedar Glades of Middle 466 Tephrosa 394; aff. senna 415; cinerea 414; purpurea 373; senna 415; sessiliflora 373 Tepium alatum 450 Teramnus labialis 415 Terminalia catappa 430 Tetragastris balsamifera 392, 417 Rhodora [Vol. 69 Tetrazygia angustifolia 429; el- eagnoides 429 Thelypteris aspidioides 5; asple- nioides 376; assurgens 5; balbi- sii 376; Bangii 5; biformata 5; boqueronensis 5; coarctata 5; consobrina 5; cordata 376; de- flexa 5; densa 5; diplazioides 7; dumetorum 5; ensiformis 6; falcata 6; furva 6; gigantea 6; glandulosalanosa 6; Herzogii 6; Jamesonii 6; laevigata 6; Lep- rieurii 6; leucothrix 6; Lindigii 5; Linkiana 6; longifolia 7; lugubriformis 7; Macbridei 7; macrotis 7; membranacea 7; mollis 7; nemoralis 7; nervosa 7; New Combinations in Peru- vian Species of 5; nitens 7; Pavoniana 7; peruviana 7; pha- celothrix 7; pilosula 7; pteroi- dea 8; resinifera 376; Rolandii 8; Rosei 8; Rosenstockii 8; rudis 376; Rusbyi 8; sprengelii 376; Standleyi 8; subandina 8; tristis 8; Warmingii 8 Theobroma cacao 391, 426 Thespesia populnea 425 Thevetia peruviana 435 Thlaspi 461, 463; alliaceum 463; arvense 463; bursa pastoris 463; tuberosum 460-463, A Taxonomic Synonym of Carda- mine Douglassii 460 Thryallis glauca 374 Thunbergia alata 444; erecta 444; fragrans 444 Thymus vulgaris 440 Tiaridium indicum 438 Tibouchina 175 Tillandsia aloifolia 60; balbisiana 60; lineatispica 394; pulchella 3738; setacea 372; tenuifolia 372, 373 Tithonia diversifolia 450 Tium Michauxii 48 1967] Trachelomonas armata var. longi- spina 57; horrida 57; volvocina 57 Torrubia fragrans 404 Tournefortia bicolor 438; gnapha- lodes 438; hirsutissima 4388; microphylla 438; volubilis 438 Tragia 241, 242, 244-246, 249, 251-255, 257, 258, 297; affinis 255, 267; angustifolia 281; amblyodonta 242, 243, 245, 251, 254-257, 260, 268, 283, 286-290; bahiensis 257; bailloniana 241; betonicifolia 244, 251, 256, 259, 261, 270-274, 281, 283, 286, 292, 298, 299, 303; brevispica 242, 244, 251, 254-257, 260, 261, 266, 268, 273-278, 281, 283, 285, 286, 290, 303, 304; capensis 251; cordata 244, 245, 251, 255, 259, 261, 263, 265-267; discolor 300, [var.] euneata [f.] lanceolata 300, [f.] latifolia 300, [var.] linearis 300, [var.] subovalis 300; fallax 251; geraniifolia 251, 257; glanduligera 244, 246, 251, 254, 255, 257, 259, 261-264, 266, 278, 290; innocua 297, 300; in the United States, A Pre- liminary Revision of 241; laci- niata 241, 251, 256, 257, 259, 289, 292-294; leptophylla 244; linearifolia 300, 302; macro- carpa 266; mexicana 246, 303; Michauxii 266; nepetaefolia 453; nepetifolia 242-244, 251, 255-257, 260, 261, 263, 270, 277- 281, 283, 285, 286, 294, 453, [var.] amblyodonta 288, [var.] angustifolia 284, [var.] genuina 278, [var.] latifolia 283, var. leptophylla 284, 286, var. ra- mosa 283, 285-287, [var.] scutellarifolia 284, [var.] teu- criifolia 276, 277; nigricans 244, 245, 247, 249, 251, 256, 259, 289, Index to Volume 69 509 295; paxii 257; pinnata 257, 294; in the United States, A Preliminary Revision of 241; pungens 252; ramosa 242, 244, 248-251, 254, 256, 260, 261, 268, 273, 274, 277, 280-283, 285, 286, 290, 292, 304, [var.] latifolia 284, var.? leptophylla 283; saxi- cola 244, 251, 256, 260, 273, 290- 292, 299; scutelariifolia 285; sect. Eutragia 242, 243, 255, 260, 297; sect. Bia 254, 297, 303; sect. Leptobotrys 251, 254, 255, 259, 261, 285, 297; sect. Lepto- rachis 242; sect. Leucandra 242, 243, 255, 261, 285, 297; sect. Tragia 243, 251, 254, 255, 257, 259-261, 297; sect. Zuckertia 303; sellowiana 251; Smallii 249-252, 254, 255, 260, 261, 268, 273, 289, 297-299, 302, 303; stoltziana 251, 252; stylaris 249, 283, 285, 286, [var.] angusti- folia 283, [var.] latifolia 283, [var.] leptophylla 283; tenella 257; teucriifolia 242, 274, 2'76, 277, 290; tristis 251, 252; urens 245, 246, 249, 251, 252, 254, 255, 260, 261, 268, 289, 297, 299-303, [var.] innocua 300, [var.] lan- ceolata 300, [var.] linearis 300, [var.] subovalis 300; urticifolia 247, 248, 251, 254, 256, 259, 261, 263, 267-270, 274, 281, 285, 292, 302, 304, var. laciniata 292, var. texana 272, 274; volubilis 242, 243, 251, 260, 421 Trema micrantha 400 Trianthema portulacastrum 404 Tribulus 455, 456; alacranensis 456; alatus 455, 456; Another, Adventive in the New World 455; cistoides 415, 455, 456; sericeus 456; terrestris 455, 456 Trichachne affinis 372 Tricholaena repens 372 510 Trichostigma octandrum 404 Tridax procumbens 450 Triphasia trifolia 417 Triticum aestivum 31 Triumfetta bartramia 374, 423; lappula 423; rhomboidea 374; semitriloba 423 Tryon, Rolla, The Flora of East- ern Himalaya (Review) 456; Taxonomic Fern Notes V. New Combinations in Peruvian Species of Thelypteris 5 Tuberaria 186 Turbina corymbosa 437 Turnera ulmifolia 427 Turpina paniculata 422 Ulmus americana f. pendula 31 United States, A Preliminary Re- vision of Tragia in the 241 Urechtites lutea 435 Urena lobata 425, var. sinuata 375; sinuata 375 Urocarpidium 15, 21 Utricularia gibba 376; obtusa 376 Vaccinium crassifolium 59; elli- ottii 59; tenellum 59; vacillans 59 Vachellia farnesiana 408 Valerianella chenopodifolia 453; chenopodiifolia 453 Valerianoides jamaicense 439 Varronia corymbosa 437 Verbena chamaedrifolia 439 Verbenaceae 395, 438 Verbesina alata 450; alba 376 Vernonia albicaulis 450; cinerea 450; sericea 450 Veronica Anagallis-aquatica f. anagallidiformis 455, f. anagal- liformis 455; hederaefolia 453; hederifolia 453; peregrina var. peregrina 33 Vigna antillana 415; luteola 415; repens 415; unguiculata 415 Rhodora [Vol. 69 Vinca rosea 434 Viola papilionacea var. albiflora 33; melissaefolia 453; melissi- folia 453 Virgin Islands, Annotated Check- list of the Dicotyledons of Tor- tola 385; Further Changes and Additions to the Flora of Porto Rico and the 372 Vitex agnus-castus 439; divari- cata 439 Vitis vinifera 423 Volkameria aculeata 438 Waldsteinia fragarioides 315, 317 Waltheria americana 426; indica 426 Ward, Daniel B., Southeastern Limit of Fagus grandifolia 51 Ware, Stewart, A New Talinum (Portulacaceae) from the Cedar Glades of Middle Tennessee 466 Waterfall, U. T., Physalis in Mexico, Central America and the West Indies 82, 203, 319 Webber, Edgar E., Flagellates and Desmids New to Massachu- setts 55 Webster, Grady L. and Kim I Miller, A Preliminary Revision of Tragia (Euphorbiaceae) in the United States 241 Wedelia calycina 450; parviflora 450; trilobata 450 West Indies, Physalis in Mexico, Central America and the 82, 203, 319 Wilbur, Robert L., The Identity of Psoralea multijuga Ell. (Legu- minosae) 48 Wilbur, Robert L. and James D. Perry, Palynological Notes on American Species of Helianthe- mum (Cistaceae) 184 Wissadula amplissima 425 Withania 83; melanocystis 98 VIR 1967] Xanthidium antilopaeum var. hebridarum 58 Xanthium chinense 450; struma- rium 450 Xanthoxalis corniculata 415 Zanthoxylon americanum 458 Zanthoxylum 394; bifoliolatum 374; carolinianum var. fruti- cosum 481; clava-herculis 482, var. fruticosum 482; hirsutum Index to Volume 69 511 481, 482; martinicense 417; sp. 417 Zenobia cassinefolia 453; cassini- folia 453 Ziziphus reticulata 423; rignonii 423 Zuckertia 242 Zygophyllaceae 455 Zyzygium grande 375