Dodora 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 i RALPH CARLETON BEAN Associate Editors ROBERT CRICHTON FOSTER 3 ROLLA MILTON TRYON RADCLIFFE BARNES PIKE VOLUME 65 1963 Che Nem Lugland Botanical Club, Inc. Botanical Museum, Oxford St., Cambridge 38, Mass. Fy 7) ow R&eFLiEMCE LIBRARY, 73 APR 4 1963 Hodora 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 Associate Editors Vol. 65 January-March, 1963 No. 761 CONTENTS: | Plantae Austro-Americanae XI Richard Evans Schultes ........ 1 A Double-flowered form of Diapensia lapponica boa QU 9 8E. OO ERSTER e E 21 A New Station for Rhododendron maximum in Northern Vermont H. W. Vogelman and L. A. Charette ....................... 22 Sixteenth Report of the Committee on Plant Distribution .... 26 Scheuchzeria in Western Pennsylvania L. K. Henry ............. 34 Annotation of North American Polycarpaea Olga Lakela ..... 35 | Protandry in Two Species of Streptanthus (Cruciferae) Roed O Rolna o L asisnicmeMi M LL E 45 (Continued on Inside Cover) The Nem England Botanical Club, Ince. Botanical Museum, Oxford St., Cambridge 38, Mass. Lag tus LEE RESETS CONTENTS: — continued Studies in Linum: L. inbricatum and L. hudsonioides. C. Marvin Rogers ciceccccsscccssscesscesssecsssesssecesseseneeesseeseseeseeceeees 50 Notes on the Fern Genus Eriosorus Alice F. Tryon ............ 56 Contributions to an Illinois Flora I. The Genus Physostegia Robert H. Mohlenbrock ............ eene 58 The Intergradation of Senecio plattensis and Senecio pauperculus in Wisconsin T. M. Barkley .............. eee 65 The Geographic Location of Gorgonidium (Araceae) Dan H. Nicolson vircecccccccccccssccssccecccccccccccccececceveceeccceseecssesscceasscece 68 New Plants in Old Places Frank C. Seymour ................... 73 Tetroploid Passiflora incarnata in North Carolina Robert M. Lloyd .............. eese eene 79 A Recent Illinois Collection of Sanguisorba canadensis L. Glen S. Winterringer. Lees eee 80 The Flora of the Wolf Islands, New Brunswick, Part | Radcliffe B. Pike and Albion R. Hodgdon ............1:....... 82 Rbhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 65 January-March, 1963 No. 761 PLANTAE AUSTRO-AMERICANAE XI DE PLANTIS PRINCIPALITER EX COLOMBIAE ORIENTALIBUS PARTIBUS NOTULAE RICHARD EVANS SCHULTES The following miscellaneous notes on new or otherwise interesting plants, mainly from Colombia, are offered as a continuation of studies on the phytogeography of the north- western part of the Amazon. This research has been in part supported by a grant from the National Science Founda- tion, which also has provided a grant for the illustrations. The families are arranged in accordance with the Engler- Gilg system. CYCLANTHACEAE Asplundia ponderosa R. E. Schultes ex Harling in Acta Hortii Bergiani 18 (1958) 217, fig. 57, g-m. COLOMBIA: CoMisARÍA DEL AMAZONAS, Rio Apaporis, Raudal de Jirijirimo, March 1951, R. E. Schultes 12093; August 12, 1951, R. E. Schultes et I. Cabrera 13514; September 16, 1961, R. E. Schultes et I. Cabrera 14058. The Taiwano Indians living along the Rio Kananari (which enters the Apaporis immediately above Jirijirimo) reduce to ashes the leaves of this plant to add to the food of pregnant women. The Taiwano name of Asplundia ponder- osa is ka-ma-té-pa. With the original description of this species, Harling pub- lished analytical drawings of floral parts. The drawings published in the present article show the characteristic leaf and illustrate the habit of the plant. 1 bo Rhodora [Vol. 65 ASPLUNDIA ponderosa aN AE Schultes ex Harling Vo ` Figure 1. Asplundia ponderosa R. E. Schultes ex Harling. 1) Habit, X 1/4. 2) Leaf, X about 1/3. 3) Inflorescence, X about 1 1/2. Drawn by John Gumppenberg. 1943] Plantae Austro-Americanae XI — Schultes 3 Figure 2. Asplundia rhodea R. E. Schultes ex Harling. Inflorescence of Idrobo et Schultes 933. Photograph by R. E. Schultes. Plate 1274 4 Rhodora [Vol. 65 Asplundia rhodea R. E. Schultes ex Harling in Acta Hortii Bergiani 18 (1958) 214, fig. 57, d-f, t. 34. COLOMBIA: INTENDENCIA DEL META, Sierra de La Macarena, east- ern slope of Mt. Renjifo, alt. 600-1300 m., December 30, 1950 — Janu- ary 5, 1951, J. M. Idrobo et R. E. Schultes 933. When Asplundia rhodea was originally described, techni- Figure 3. Asplundia rhodea R. E. Schultes ex Harling. Roots of Idrobo et Schultes 933. Photograph by R. E. Schultes. Plate 1275 1963] Plantae Austro-Americanae XI — Schultes 5 cal drawings of floral parts and an illustration of an her- barium specimen were published. Two photographs of Idrobo et Schultes 933 are herewith offered as an aid in understanding the habit of the plant. Es "s P. V ES RE Figure 4. Dicranopygium omichlophilum R. E. Schultes ex Harling. Habit of the plant at the locality of Idrobo et Schultes 1184. Photo- graph by R. E. Schultes. Plate 1276 Dicranopygium omichlophilum R. E. Schultes ex Harling in Acta Hortii Bergiani 18 (1958) 294, fig. 76, a-c, t. 67. COLOMBIA: INTENDENCIA DEL META, Sierra da La Macarena, Mt. Renjifo, summit and environs, alt. 1300-1900 m., January 6-20, 1951, J. M. Idrobo et R. E. Schultes 1184. Dicranopygium omichlophilum is extremely abundant on 6 Rhodora [Vol. 65 moss-covered rocks in the cold streams at the top of the Sierra de La Macarena, where almost perpetual mist creates a dark and constantly wet environment. It covers the rocks in dense colonies together with a beautiful species of Spathi- phyllum. The flowers of both plants are excessively fra- grant. Figure 5. Dicranopygium omichlophilum R. E. Schultes ex Harling. Flower of Idrobo et Schultes 1184. Photograph by R. E. Schultes. Plate 1277 1963] Plantae Austro-Americanae XI — Schultes 7 Figure 6. Habitat of Dicranopygium omichlophilum R. E. Schultes ex Harling at the summit of La Macarena, Meta, Colombia. Photo- graph by R. E. Schultes. Plate 1278 The accompanying photographs illustrate the habit of Dicranopygium omichlophilum and two views of the habitat of the plant at the summit of La Macarena. BROMELIACEAE Pitcairnia mituensis L. B. Smith sp. nov. A Pitcairnia Maguirei L. B. Smith atque P. Wurdackii L. B. Smith, quibus maxime affinis, lamina foliorum subtus haud lepidota valde latiore differt. Florifera ultra 2 m. alta; foliis ad 1 m. et ultra longis; vaginis sub- orbicularibus, quam laminis paulo latioribus, intus castaneis; laminis 8 Rhodora [Vol. 65 linearibus, acuminatis, pungentibus, 7.5 cm. latis, planis, utrinque glabris, spinis atris curvatis 3 mm. longis laxe armatis; scapo erecto; scapi vaginis ignotis; inflorescentia pauciramosa, glabra; bracteis primariis parvis, ovatis, 25 mm. longis, quam basi sterili ramorum multo brevioribus; ramis adscendentibus, ad 6 dm. longis, robustis, Figure 7. Piteairnia mituensis L. B. Smith. Habit photograph taken at the type locality, near the summit of Cerro de Mitu, Vaupes, Co- lombia. Photograph by R. E. Schultes. Plate 1279 viridibus; bracteis florigeris late ellipticis, acutis, ad 20 mm. longis, pedicellos superantibus; floribus suberectis, flavo-viridibus, pedicellis cylindrieis, 5 mm, longis; sepalis lineari-lanceolatis, acutis, ad 45 mm, longis, ecarinatis; petalis 6 em. longis, nudis; staminibus inclusis; ovario ca. 1/2 infero; ovulis late alatis. 1963] Plantae Austro-Americanae XI — Schultes 9 COLOMBIA: COMISARÍA DEL VauPÉs, Río Vaupés, Cerro de Mita. alt. about 270 m. “Common on bare top of mountain, in clumps on rocks." August 20, 1960, Richard Evans Schultes 22711 (TYPE in Herb Gray). Pitcairnia mituensis is a member of the group of species characterized by broadly alate ovules and seeds and ranging through the Guayana Highland from British Guiana ic southeastern Colombia. MARANTACEAE Calathea acuminata Steyermark in Fieldiana, Bot. 28, no. 1 (1951) 161. COLOMBIA: COMISARÍA DEL VavrÉs, Río Negro, San Felipe. “Flowers white." December 12, 1947, R. E. Schultes et F. López 9332; “Flowers white.” R. E. Schultes et F. López 9337 A. The natives of the Río Guainía in Colombia and Venezuela boil the erushed leaves and stems of this (and possibly other) species of Calathea with the leaves of a species of Nectandra to prepare a poultice or dressing for infected ulcers of the legs and thighs. BALANOPHORACEAE Ombrophytum zamioides Weddell in Ann. Sci. Nat., ser. 3, 14 (1850) 184, t. 10. COLOMBIA: COMISARÍA DEL AMAZONAS, Río Loretoyacu, Isla Zan- cudo. “White saprophyte." R. E. Schultes et G. Black 8615. This curious saprophyte, identified by Dr. Joao Murça Pires of the Instituto Agronómico do Norte in Belem do Pará, has apparently not been reported from the Amazon Valley of Colombia or Brazil. The type was collected in Peru. MALPIGHIACEAE Banisteriopsis Caapi (Spruce ex Griseb.) Morton in Journ. Wash. Acad. Sci. 21 (1931) 485. PERU: DEPARTAMENTO DE SAN MARTÍN, PROVINCIA DE LAMAS, San José de Sisa-Nauta, alt. c. 550 m. “Flores rosadas. Trepador voluble (soga) cultivado. Shimba-ayahuasea.” July 26, 1958, O. Velarde Nunez 6577. (Herb. Morton Arb.; Econ. Herb. Oakes Ames). DEPARTAMENTO DE SAN MARTIN, PROVINCIA DE LAMAS, San José de Sisa-Nauta, alt. c. 600 m. “Trepador voluble. Cultivado. Su tronco se usa para preparar una bebida alucinógena. Ayahuasca legítima." July 29, 1958, O. Velarde Nuiiez 6586. (Herb. Morris Arb.). 10 Rhodora [Vol. 65 DEPARTAMENTO DE SAN MARTIN, PROVINCIA DE LAMAS, San José de Sisa-Nauta, alt. 600 m. “Flores rosadas. Trepador voluble. Cultivado. Su tronco se usa para preparar una bebida purgante y alucinógena. N.v. Cuchi-ayahuasca." July 29, 1959, O. Velarde Nuñez 6587. (Herb. Morris Arb.; Econ. Herb. Oakes Ames). It is rare that flowering material of Banisteriopsis Caapi is collected in the field. In fact, with the exception of several collections from vines brought to flower in botanical gardens or experiment stations (Ducke 25258, 25260; Murca Pires 19; Cuatrecasas et Dryander 14372), Banisteriopsis Caapi has been found in a flowering state only once since Spruce’s type — when Ducke collected topotypical material from the Rio Curicuriarí in Amazonian Brazil (Ducke 153). Velarde’s two excellent collections (6577, 6587), which are abundantly flowering, come from vines cultivated by natives on the eastern or Amazonian slope of the Peruvian Andes. An examination of the flowers discloses little if any significant variation from the general pattern of the flower of Banisteriopsis Caapi (see Cuatrecasas in Webbia 13 (1958) 506). Velarde 6586 is sterile, as are most collections of Banisteriopsis cultivated for medicinal or narcotic use, but may safely be referred to B. Caapi. Through the Velarde collections, we are able to cite several new vernacular names. Velarde 6586 is said to rep- resent ayahuasca legítima or “true ayahuasca"; Velarde 6577 is called shimba-ayahuasca; and Velarde 6587 has the name cuchi-ayahuasca. Banisteriopsis inebrians Morton in Journ. Wash. Acad. Sci. 21 (1981) 485. PERU: DEPARTAMENTO DE SAN MARTÍN, PROVINCIA DE LAMAS, San José de Sisa-Nauta, alt. 600 m. “Flores rosadas. Trepador voluble. Su tronco se utiliza como purgante y para preparar una bebida aluci- nógena. Purga-huasca." July 29, 1958, Octavio Velarde Nunez 6585 (Herb. Morris Arb.; Econ. Herb. Oakes Ames). This collection, in abundant (albeit young) flower, has the very firm-coriaceous leaves which seems to be the principal character separating Banisteriopsis inebrians from B. Caapi. Banisteriopsis inebrians, described from the Putumayo of 1963] Plantae Austro-Americanae XI — Schultes 11 Colombia, has apparently not hitherto been reported from Peru. Banisteriopsis Rusbyana (Ndzu.) Morton in Journ. Wash. Acad. Sci. 21 (1931) 487. PERU: DEPARTAMENTO DE SAN MARTÍN, PROVINCIA DE TARAPOTO, Tarapoto, alt. 600 m. “Trepador voluble. Cultivado. Su tronco se utiliza para preparar una bebida purgante y alucinógena. N.v. aya- hausca amarilla." August 2, 1958, O. Velarde Nuñez 6589 (Herb. Morris Arb.). This collection is sterile, but the membranaceous leaves match very closely what we have been accepting as repre- sentative of Banisteriopsis Rusbyana. Velarde 6589 is, in- deed, a very close match for Klug 1971 from the Putumayo of Colombia. The vernacular name ayahuasca amarilla, might possibly refer to the colour of the flowers. The type of Banisteriopsis Rusbyana was collected in Bolivia, and the species has hitherto been known only from Bolivia and Colombia. STERCULIACEAE A number of collections of Herrania have been studied subsequent to the publication of my synopsis of the genus in 1958 (Journ. Arn. Arb. 39 (1958) 216) and are herewith cited since some of them extend known ranges or are the source of interesting taxonomic details. Herrania albiflora Goudot in Ann. Sci. Nat. Paris, sér. 3, 2 (1844) 230, t. 5, figs. 1-10. VENEZUELA: ESTADO DE BARINAS, Fundo Paiva, Santa Barbara de Barinas, February 1953, L. Aristeguieta 1598. This represents the second Venezuelan collection of a species predominantly Colombian in distribution. Herrania Cuatrecasana García-Barriga in Caldasia 2 (1941) 57, t. x COLOMBIA: COMISARÍA DEL PUTUMAYO, Río San Miguel, Quebrada de Sipanae, alt. 400 m., December 12, 1940. J. Cuatrecasas 11012. This collection was inadvertently omitted from the synop- sis. Herrania kanukuensis R. E. Schultes in Caldasia 2 (1943) 11. BRAZIL: ESTADO DO AMAZONAS, Jart, Rio Branco. January 1913, J. G. Kuhlmann 3082. 12 Rhodora [Vol. €5 Herrania kanukuensis is known from Brazil through only two collections, both from the Rio Branco. The species is otherwise registered from British and Dutch Guiana. Herrania nitida (Poepp.) R. E. Schultes in Caldasia 2 (1943) 16, t. p. 17. BRAZIL: ESTADO po PARA, Belem. Cultivated in Museu Paraense, February 4, 1926, A. Ducke 21050. Rio Jary, December 16, 1912, E. Snethlage 12444. ESTADO DO AMAZONAS, Rio Solimoes, San Antonio do Ica, August 26, 1906, A. Ducke 7618; Rio Japurá, November 20, 1904, A. Ducke 14748; Sào Paulo de Olivenca, October 10, 1931, A. Ducke 23977; Rio Tocantíns, November 12, 1927; Rio Parintíns, January 20, 1936, A. Ducke 35406. Botanical Garden, Rio de Janeiro, (“from Ama- zonas"), March 31, 1922, Ducke 248. "Territorio do Acre, Seringal Orion, October 23, 1923, J. G. Kuhlmann 702. COLOMBIA: COMISARÍA DEL CAQUETÁ, Morelia, November 24, 1941, K. von Schneidern A 1366 bis. ECUADOR: PnoviNCIA NaPo-PasTAZa, Tiputini-Lagatococha, January-May, 1953, F. Fagerlind et G. Wibom 2283; 2348; Tena, October 18, 1939, E. Asplund 9428; December 30, 1939, E. Asplund 10198; Mera, December 7, 1955, E. Asplund 18720; Canelos, Novem- ber 15-23, 1958, G. Harling 3171; 3290; 3231. These newly cited collections, with the exception of Kuhl- mann 702 from the Brazilian Acre, do not alter our concept of the range of Herrania nitida. Herrania nitida (Poepp.) R. E. Schultes fma. sphenophylla R. E. Schultes in Bot. Mus. Leafl. Harvard Univ. 14 (1950) 131. PERU: DEPARTAMENTO DEL LORETO, Iquitos, November 20, 1940, E. Asplund 14780. This form has previously been cited twice from Amazoni- an Peru and once from Amazonian Brazil and Colombia. Herrania Mariae ( Mart.) Decaisne ex Goudot in Ann. Sci, Nat., sér. 5, 2 (1844) 233. Abroma Mariae Martius in Denkschr. Regensb. Bot. Gesell. 3 (1841) 297, tt. 6, 9. Theobroma Mariae (Mart.) Schumann in Martius Fl. Brasil. 12, pt. 5 (1886) 71, t. 15. Herrania atrorubens Huber in Bull. Soc. Genéve, ser. 2, 6 (1914) 187. BRAZIL: Estapo po PARA, Rio Trombetas, Castanhóes do Rio Cuminá-mirim, December 12, 1906, A. Ducke 7935. In my monograph of Herrania, I included H. atrorubens as a synonym of H. nitida (Poepp.) R. E. Schult. An opportu- nity for me to examine the type of Huber's concept has now arisen, and I find that it is referable to H. Mariae. Huber himself, in describing H. atrorubens, noted that it differed 1963] Plantae Austro-Americanae XI — Schultes 1 € Ó p entarg ' SAURAUIA "m Q^ pruinosa BE. Schultes PEN S TRU dus mean t Lem mensem p E Figure 8. Saurauia pruinosa R. E. Schultes. 1) Habit, X 1. 2) Nether surface of the leaf, X about 10 (single hairs X about 50). 3) Flower, X 3. 4) Sepals X 3 (single hair X about 30). Drawn by John Gumppenberg. 14 Rhodora [Vol. 65 from H. Mariae in being smaller and in having dark red flowers. Herrania pulcherrima Goudot var. pacifiea R. E. Schultes in Bot. Mus. Leafl. Harvard Univ. 14 (1950) 131, t. 28, low. fig. ECUADOR: PRoviNCIA ESMERALDAS, Río San Miguel, March 28- April 6, 1959, G. Harling 4569. Harling 4569 represents the second collection of this Pacifie coastal variety from Ecuador and the first from Esmeraldas. DILLENIACEAE Saurauia pruinosa FR. E. Schultes in Bot. Mus. Leafl. Harvard Univ. 16 (1953) 81. COLOMBIA: CoMisARÍA DEL PUTUMAYO, Valley of Sibundoy, Sibun- doy, alt. about 2225-2300 m. May 29, 1946. R. E. Schultes et M. Villar- real 7651. One of the most beautiful of the South American species of Saurauia because of the rose-pink hue of the leaf and pet- iole indumentum, this species has not hitherto been illus- trated. CARYOCARACEAE Caryocar gracile Wittmack Martius in Fl. Bras. 12, pt. 1 (1886) 350. COLOMBIA: AMAZONAS-VAUPÉS, Río Apaporis, Jinogojé (at mouth of Río Piraparaná) and vicinity, alt. about 700 feet, lat. 0°15’ S, long. 70*30' W. “Small tree, 25-35 feet tall. Flowers yellow. Fruit crushed for barbasco to poison fish. Makuna: gaw-gé. Puinave: ho.” June 8, 1952, R. E. Schultes et I. Cabrera 16668. Caryocar gracile, known from the Colombian Amazon basin through only several collections, is employed by the Indians of the Vaupés as a fish poison. The fruits are crushed and cast into still water. The oily seeds are some- times eaten as a food, and the bark is said by the Makunas to be employed like soap in washing. FLACOURTIACEAE Mayna muricida R. E. Schultes sp. nov. A Mayna amazonica folis lanceolatis (non ovatis), subtus gros- siuscule sparsissimeque (non molliter denseque) pilosis et supra glabris (non sparse pilosis) atque fructus aliis conspicue crispaturis (non integris) differt. COLOMBIA: CoMISARÍA DEL AMAZONAS, Trapecio amazónico, in- terior regions of trapecia between Amazon and Putumayo Rivers, alt. 1963] Plantae Austro-Americanae XI — Schultes 15 muricida RE. Schultes Figure 9. Mayna muricida R. E. Schultes. 1) Branch with leaves, 1. 2) Nether surface of the leaf, X 20. 3) Branch with fruit, X 1. 4) Cross section of fruit, X 1. 5) Seeds, X5. 5) Surface of fruit, x 20. Drawn by John Gumppenberg. 16 Rhodora [Vol. 65 slightly over 100 m. “‘Treelet. 20 ft. tall. Fruit white. Seeds used by Tikunas to kill rats. Tikuna name ~ ka-té-bo. In varial" October 1945, Richard Evans Schultes 6760 (TYPE in Herb. Gray). Mayna muricida, so named because the Tikuna Indians gather the seeds to scatter around their houses in the belief that they kill rodents, differs from M. amazonica especially in lacking a soft indumentum on the nether surface of the leaves and in having curiously crispate (marginally eroded and curled) wings on the fruit. Both species occur in the light “caatinga” or “varial”’ forest growing on sterile sand, not, as with most other species, in dense forests; M. ama- zonica, however, in many parts of its range (the western Amazon in Brazil, Colombia, Peru and Venezuela) is a typi- cal jungle tree. Mayna toxica R. E. Schultes sp. nov. A Mayna amazonica foliis subtus maxime mollissimis pilosis atque fructu chryseo-viride (non atroviride) et sine aliis (aliis ad costas reductis) principaliter differt. COLOMBIA: COMISARÍA DEL VAUPES, Río Vaupés, Tipiaca, (be- tween Mita and Javareté). “Bark poison for dogs, rats, man. Bush up to 12 ft. tall. Fruit light yellow-green. Desano name = bé-ra-man-^ o-hé-ree-gé.” May 14-24, 1953. Richard Evans Schultes et Isidoro Cabrera 19337 (TYPE in Econ. Herb. Oakes Ames). COMISARÍA DEL AMAZONAS, Rio Apaporis, Cano Peritomé, tributary below Rauda! Yayacopi, alt. about 750 ft. “Height 12 ft." February 18-20, 1952, R. E. Schultes et I. Cabrera 15519. COMISARÍA DEL VAUPÉS, Rio Apa- poris, Soratama, April 2, 1952, R. E. Schultes et I. Cabrera 16142. Mayna toxica is easily distinguished from related species by its golden yellow fruit which, instead of having conspicu- ous wings, is armed only with slightly thickened ribs. The specific name refers to the general belief amongst the natives that the bark (and, aecording to some, the seeds) are poisonous. It is stated that the rasped bark is employed to kill dogs and rodents. The active principle is not known. A test for alkaloids with modified Dragendorf reagent (see R. F. Raffauf in Econ. Bot. 16 (1962) 171) gave negative results. The fact that at least two species — Mayna muri- cida and M. toxica — are similarly employed for their toxic properties by Indians in far-separated parts of the Colom- bian Amazon suggests that an investigation into the chemi- cal eznstituents of this genus might be of interest. 1963] Plantae Austro-Americanae XI — Schultes iyi ad FeO RE. Schultes Ñ Figure 10. Mayna toxica R. E. Schultes. 1) Branch with leaves, X 1. 2) Fruit, X 1. 3) Enlargement of the surface of the fruit. 4) En- largement of the upper surface of the leaf. 5) Enlargement of the nether surface of the leaf. Drawn by E. W. Smith. 18 Rhodora [Vol. 65 MELASTOMACEAE The following notes on melastomaceous plants have been submitted by Dr. J. J. Wurdack of the Smithsonian Institu- tion, who writes: “Amone the collections thus far studied by me are represented a number of new records of Melastoma- taceae for Colombia. Dr. Schultes' Vaupés collections have especially emphasized the strong linkage with the flora of Dept. Loreto, Peru." Adelobotrys barbata Triana in Trans. Linn. Soc. 28 (1871) 68. COLOMBIA: CoMisARÍA DEL VAUPES, Río Piraparaná. R. E. Schultes et I. Cabrera, March 9, 1952, 15908; May, 1952, 17129; Sep- tember 19, 1952, 17574. Limited to the upper Río Negro drainage-area, this shrub- by species in a predominantly vining genus, is now well rep- resented from the Colombian Vaupés. Adelobotrys macrophylla Pilger in Verhandl. Bot. Ver. Branden- burg 47 (1905) 165. COLOMBIA: COMISARÍA DEL VAUPÉS, Río Piraparana, Cano Tee- meena, September 4, 1952, R. E. Schultes et I. Cabrera 17216; Río Pacoa, February 8, 1952, R. E. Schultes et I. Cabrera 15223. This species previously has been known only from the Department of Loreto in Peru. Branches of this bush are burned and the ashes are added to clay for the manufacture of pottery. Adelobotrys praetexta Pilger in Verhandl. Bot. Ver. Brandenburg 47 (1905) 167. COLOMBIA: CoMisARÍA DEL VAUPES, Río Piraparaná, September 1952, R. E. Schultes et I. Cabrera 17127; Río Apaporis, Río Popeyaca, February 25, 1952, R. E. Schultes et I. Cabrera 15661. Adelobotrys praetexta has hitherto been known only from Amazonian Bolivia and Peru. Bellucia umbellata Gleason in Bull. Torr. Bot. Club 58 (1931) 257. COLOMBIA: COMISARÍA DEL VAUPES, Río Piraparaná, March 9, 1952, R. E. Schultes et I. Cabrera 15916; Jinogojé, June 5, 1952, 16625; September 6, 1952, 17349. Río Apaporis, September 1952, R. E. Schultes et I. Cabrera 17094. Hitherto Bellucia umbellata has been reported only from Loreto, Peru. The Barasana Indian name of this bush is teé-na-mo. 1963] Plantae Austro-Americanae XI — Schultes 19 Graffenrieda candelabrum Macbride in Field Mus. Publ. Bot. 13, pt. 4 (1941) 321. COLOMBIA: COMISARÍA DEL AMAZONAS, Río Miritiparaná, March 2, 1952, R. E. Schultes et I. Cabrera 15728. Known hitherto only from the type collection from the Departmento de Loreto, Peru, this species is now registered from Amazonian Colombia. The bluish fruits are somewhat astringent and are chewed to “heal” bleeding gums. Tani- muka name = kweé-ma-mé. Yukuna name = ko-me-noo- ma-ré (‘“bird’s spice"). Leandra rhodopogon (DC.) Cogniaux in Martius Fl. Bras. 14, pt. 4 (1886) 109. COLOMBIA: COMISARÍA DEL VauPÉs, Río Apaporis, Soratama, June 20, 1951, R. E. Schultes et I. Cabrera 12727; February 25, 1952, 15695; August 17, 1952, 16951; January 1952, 19614. Río Piraparaná, September 1952, R. E. Schultes et I. Cabrera 17121. Río Vaupés, Mitu, L. Uribe Uribe 2920; J. Cuatrecasas 6803. Miconia astrotricha (DC.) Triana in Trans. Linn. Soc. 28 (1871) 113. COLOMBIA: COMISARÍA DEL AMAZONAS, Río Popeyaca, February 1952, R. E. Schultes et I. Cabrera 15603. The present collection, which has been compared with the type in Munich, represents apparently the first time the species has turned up since the original Martius material from Amazonian Brazil. Miconia filamentosa Gleason in Bull. Torr. Bot. Club 65 (19538) 579. COLOMBIA: COMISARÍA DEL VAUPÉS, Río Apaporis, Río Popeyacá, June 10, 1952, R. E. Schultes et I. Cabrera 16685; Jinogojé, March 1952, 19825. The previously known range of this species included only Amazonian Brazil and Peru. Miconia fissa Gleason in Bull. Torr. Bot. Club 59 (1932) 363. COLOMBIA: CoMisARÍA DEL AMAZONAS, Río Miritiparaná, May 8, 1952, R. E. Schultes et I. Cabrera 16406. Miconia fissa was previously known only from the type collection by Holt and Blake from the Río Maturaca, Ama- zonian Brazil. The Yukuna Indian name is hay-wa-ld, and the orange fruit is employed as a diuretic. Miconia marginata Triana in Trans. Linn. Soc. 28 (1871) 110. COLOMBIA: COMISARÍA DEL VaAUPÉS, Río Apaporis, September 1952, R. E. Schultes et I. Cabrera 17053. -Rio Piraparaná Septem- 20 Rhodora [Vol. 65 ber 10, 1952, R. E. Schultes et I. Cabrera 12367; Río Ricapuya, Sep- tember 27, 1952, R. E. Schultes et I. Cabrera 17624. Widespread in Amazonian Bolivia and Brazil and in Ven- ezuela and British Guiana, Miconia marginata is now regis- tered from Colombia. Ossaea araneifera Markgraf in Notizbl. 13 (1937) 462. COLOMBIA: COMISARÍA DEL VaurÉs, Río Apaporis, Jirijirímo, July 5, 1951, R. E. Schultes et I. Cabrera 12950. This species has been known only from the tvpe collection from Loreto, Peru. Ossaea cucullata Gleason in Bull. Torr. Bot. Club 58 (1931) 260. BRAZIL: ESTADO DO AMAZONAS, Muquentaua, Rio Teffé, R. L. Fróes 26299. COLOMBIA: COMISARÍA DEL VaurÉs, Río Apaporis, Soratama, July 17, 1951, R. E. Schultes et I. Cabrera 13096 ; August 4, 1951, 13349; January 1952, 19626. Río Kuduyarí, Cerro Yapoboda, April 1963, R. E. Schultes et I. Cabrera 20031. Previously registered only from Loreto, Peru, this species is now known from Amazonia Brazil and Colombia. MYRSINACEAE Conomorpha citrifolia Mez in Pflanzen. 4, Fam. 236 (1902) 256. COLOMBIA: COMISARÍA DEL VaurÉs, Río Piraparaná (tributary of Río Apaporis), Raudal Na-hoó-gaw-he, between lat. 0°15’ S, long. 70730'W and lat. 0*25'N, long. 70?30'W. “Flowers whitish yellow. Small tree along bank. Bark rasped into chicha to give it a peppery taste. Puinave: yoom-dd-ka. Barasana: gáw-hé-ké." September 1952, R. E, Schultes et I, Cabrera 17593. Conomorpha citrifolia, like the recently described C. litho- phyta R. E. Schult. of the Vaupés and C. magnoliifolia Mez in Dutch Guiana, is employed as a minor fish-poison. The Barasana Indians of the Rio Piraparaná crush the leaves and strew them on the surface of still or very slowly moving water. They also have an interesting use of the bark which is rasped into fermented chicha (made of Manihot esculenia Crantz or any of a number of edible fruits) as a spice to give the beverage a rather peppery taste. SOLANACEAE Solanum platyphyllum Humboldt e! Bonpland ex Dunal Sol. gen. aff. syn. (1816) 38. COLOMBIA: COMISARÍA DEL PuTUMAYO, road between El Pepino and Mocoa, alt. about 700 m. “Bush 2 feet. Fl. purplish white. Fr. 1963] Plantae Austro-Americanae XI — Schultes 21 ripens red with hairs. Lvs. with purplish cast underneath. Cult. Common name = wvilla.” July 28, 1960. R. E. Schultes 22556. This is the second collection of Solanum platyphyllum which has come to light from the Colombian Amazon. The difference in altitude between the other collection, Schultes 6642, which was from Leticia at 100 m. and Schultes 22556 at 700 m. is noteworthy. Both collections were taken from bushes cultivated for their edible berries. BOTANICAL MUSEUM, HARVARD UNIVERSITY A DOUBLE-FLOWERED FORM OF DIAPENSIA LAPPONICA. — Diapensia lapponica L. is a common plant in the alpine area of the Presidential Range of New Hampshire where it has been observed by many botanists. There is no record, how- ever, of any form occurring with more than 5 petals, the characteristic number. This June, while conducting a flower walk in the vicinity of Mt. Monroe in the Presidentials, my attention was called by one of the group to an unusual flower. It was a Diapensia with numerous white petals and no stamens, a state that might be designated as “double-flowered’’. Diapensia grows in dense tussocks, all the plants of which are in close associa- tion. This tussock contained a number of flowers of which most were double-flowered, but 2 or 3 were normal. A note was made of the location and a check will be made next year to see if the condition persists. A specimen of the plant has been deposited in the Herbarium of the University of New Hampshire. — FREDERIC L. STEELE, ST. MARY’S-IN-THE-MOUNTAINS, LITTLETON, N. H. A NEW STATION FOR RHODODENDRON MAXIMUM IN NORTHERN VERMONT H. W. VOGELMANN AND L. A. CHARETTE The most northerly known station in Vermont for Rhodo- dendron maximum L. has been found only 8 miles south of the Quebec border, not far from the village of Troy. The discovery was made in June 1957 by Dr. Charles G. Doll, Vermont State Geologist, while conducting field work in preparation for a new geological map of the state. Professor Doll guided the writers to the site on July 3, 1962 at which time a careful examination of the colony was made. Located in the north western section of Troy township, 14 mile south of Troy village, the rhododendron colony is on a level moss-covered site about 200 feet from the southeastern bank of the Missisquoi River. The oval-shaped colony, cover- ing an estimated l5 acre, is under an open stand of young trees of fir, yellow birch and red maple, with trunk diameters up to 10 inches. Surrounding this area is an older, denser second growth forest dominated by hemlock, fir, yellow birch, and red maple with an extensive cover of yew in the understory. The rhododendrons, toward the center of the colony, are dense and grow to heights of about 6 feet, while toward the edge of the colony they are lower and much sparser, finally becoming broken into small scattered clumps. Some of the latter grow under dense forest cover. No flowers or flower buds were seen and there was no indication that the plants had flowered the previous year. A careful search failed to disclose any seedlings in the area and it seems unlikely that the colony had flowered in recent years. A striking feature of the rhododendrons is the appearance of many drooping, browned leaves toward the ends of some of the branches, giving one the impression that the plants are lacking sufficient water. The mossy cover beneath the colony appeared to be unusually dry for early summer. Dr. Doll indicated that in the summer of 1957 he had noted the same leaf condition as well as the absence of flowers and 22 1963] Rhododendron — Vogelmann and Charette 23 fruits. Toward the periphery of the main stand, the stems of dead rhododendrons are conspicuous and these appear to have once linked some of the smaller scattered colonies with the main colony. A similar condition apparently occurs in the outlying colony in Lexington, Maine where many dead plants are reportedly found near the edge of the stand (4, 8). The general aspect of the Vermont colony may be a fur- ther indication of the widespread deterioration of rhododen- drons in northern and central New England (4). The dot map given by Iltis (7) shows the range of the species and clearly illustrates the disjunct distribution of colonies in the northern part of its range. Growing among the rhododendrons is an assemblage of northern shrubs and herbs. Nemopanthus mucronata and Viburnum cassinoides are common along with seedlings of Betula lutea which grow in the mossy cover with an abun- dance of Clintonia borealis and Coptis groenlandica. Occa- sional in the shrubby layer are Pyrus americana, Viburnum alnifolium and Vaccinium myrtilloides. Associated herbs include Osmunda cinnamomea, O. regalis, var. spectabilis, Dryopteris spinulosa, Cypripedium acaule, Cornus canaden- sis, Gaultheria hispidula, Maianthemum canadense, Trien- talis borealis, Trillium undulatum, Dalibarda repens, Lin- naea borealis var. americana and Carex intumescens. The Troy colony is 40 miles north of the uearest known rhododendron stations near Lanesboro and Peacham, Ver- mont, where several colonies grow along the borders of some of the ponds in the region. At a latitude of 44°53’ the new stand is almost as far north as the colony in Lexington, Maine, which is reported to be at about 45? north latitude. (8). The colony in Troy must be subjected to severe climatic stresses for the station is only 9 miles west of Newport, Ver- mont, which is noted for its cold winters, and in five of the last ten years has recorded temperatures of -30? F. or lower. In January 1957 the temperature dropped to -38° F. and a record low of -40° F. occurred in 1933. A summary of the known Vermont stations of Rhododen- dron maximum which are supported by herbarium speci- mens is as follows: ORLEANS COUNTY: Troy, C. G. Doll, H. W. Vogelmann and L. A. 4 Rhodora [Vol. 65 Ap | 7 t | | () f | | r S | N L | / ? f V c / 4 M N, EN (o H4 X / [oM v N < "n / N rv J \ on / v . [e] u ~ $ V E L OTTS, \ A M m L / 7M / Av / /^. PE M "V ~~, [77 | o ` ' S J ~ | ; =a — 4 "i —-. 5 / / / / / DISTRIBUTION OF p RHODODENDRON MAXIMUM Ls Vd. IN VERMONT 0M @ = locality based on herbarium specimens | O= locality based on literature citation ro o 5 a 20 4 n J miles Map. 1. Showing distribution of Rhododendron maximum colonies in Vermont. 1963] Rhododendron — Vogelmann and Charette 25 Charette (No. 2451) July 3, 1962 (VT., NEBC, HNH). WASHINGTON couNTY: Lanesboro, Miss M. P. Skinner (VT. NEBC); J. G. Jack, Aug. 16, 1901 (A). CALEDONIA COUNTY: Shores of Groton Pond, J. A. Bates, July 1897 (NEBC, HNH), L. R. Jones, July 4, 1898 (VT.); Shore of Long Pond, Ricker Station, C. G. Pringle, July 1879 (A); Groton, W. H. Blanchard, Oct. 7, 1911 (GH); Shore of Martin's Pond, Peacham, F. Blanchard, July 1881 (NHA); Peacham, Alice F. Stevens, July 26, 1892 (HNH), F. Blanchard, July 9, 1883 (A, HNH). BENNINGTON COUNTY: Pownal Center, D. S. Carpenter, June 28, 1922 (VT). A station at Niggerhead Pond in Washington County is given in the 1900 edition of the Vermont Flora (2) and it is also reported at Harvey's Pond in West Barnet (1, 3). Knowlton (8) records it at Ricker and Levi Ponds in the Groton area. These locations are plotted on Map 1 to show the distribu- tion of Rhodendron maximum colonies in Vermont. When these localities are coupled with those given in recent issues of Rhodora for New Hampshire and Maine (5, 6) a fairly accurate distribution of the species in New England is ob- tained. — UNIVERSITY OF VERMONT, BURLINGTON. LITERATURE CITED 1. ANoNYMOUS. Botanical Notes. Bulletin Vermont Botanical Club 2: 37-43. 1907. 2. BRAINERD, Ezra, L. R. JONES, and W. W. EGGLESTON. Flora of Vermont. Free Press Association, Burlington. 1900. 113 pages. 3. EGGLESTON, W. W., G. L. KIRK, and J. G. UNDERWOOD. Flora of Vermont. Bull. Vt. Agric. Exp. Station 187: 137-258. April 1915. 4. Hopapon, A. R. and RADCLIFFE PIKE. Recent Changes in Some Rhododendron Colonies in Maine and New Hampshire. Rhodora 62: 87-93. April 1960. Rhododendron maximum in Hopkinton and Har- risville, New Hampshire. Rhodora 63: 26-29, January 1961. 6. . An Ecological Interpretation of Rhododendron Colonies in Maine and New Hampshire. Rhodora 63: 61-70. March 1961. . ILTIS, HvGH. Studies in Virginia Plants II. Rhondodendron maxi- mum in the Virginia Coastal Plain and its Distribution in North America. Castanea 21:114-124. 1956. 8. KNowLTON, C. H. Rhododendron maximum in New England. Rhodora 52:215-218. September 1950. c! h -1 SIXTEENTH REPORT OF THE COMMITTEE ON PLANT DISTRIBUTION The fifteenth report included the genus Carex of the Cy- peraceae. The present report deals with the genera of the Cyperaceae other than Carex. The data for these reports have been compiled chiefly from the material found in the herbarium of the New England Botanical Club, in the Gray Herbarium and in the herbari- um of the Peabody Museum at Salem, Massachusetts. 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. CYPERACEAE Me. N.H. Vt. Mass. R.I. Conn. Bulbostylis capillaris (L.) C. B. Clarke + + + + + + B. capillaris var. crebra Fern. + + + + + + Cladium mariscoides (Muhl.) Torr. + + + + + + Cyperus brevifolius (Rottb.) Hassk. — C. dentatus Torr. + + + + + 4 C. diandrus Torr. + + + + + C. Engelmannii Steud. + + C. erythrorhyzos Muhl. + + + C. esculentus L. +t + + + + + C. ferruginescens Boeckl. + + C. filicinus Vahl t + + + 4 C. filiculmis Vahl + + + C. filiculmis var. macilentus Fern. + + + + + + *C, fuscus L. + C. Grayii Torr. + + + C. Houghtonii Torr. + + + + C. inflexus Muhl. + + + + cod C. odoratus L. + + + *C. ovularis (Michx.) Torr. + C. polystachyos Rottb. var. texensis (Torr.) Fern. + C. rivularis Kunth + t + + + c C. strigosus L. + + + + + + C. strigosus var. robustior Britt. + + X C. Weatherbianus Fern. (= Dulichium arundinaceum X -H Rhynchospora capitellata) 26 1963] Committee on Plant Distribution FA | Me. N.H. Vt. Mass. R.I. Conn. Dulichium arundinaceum (L.) Britt. T o GBMOGBRO vL Eleocharis acicularis (L.) R. & S. deU4M P EN Eom E. acicularis var. submersa (Hj. Nilss.) Svenson E EIS E. ambigens Fern. ay en E. calva Torr. eee ae eS + E. diandra C. Wright of 74 Ape Sees + E. elliptica Kunth + + + + t t E. Engelmanni Steud. + Tort E. equisetoides (Ell.) Torr. ur odia x E. fallax Weath. + E. halophila Fern. & Brack. Ld. - TEE E. intermedia (Muhl.) Schultes LEE. i err T E. melanocarpa Torr. te E. mierocarpa Torr. var. filiculmis Torr. 2 E. nitida Fern. ae E. obtusa (Willd.) Schultes 4o 4 a er E. obtusa var. ellipsoidalis Fern. d i x E. obtusa var. jejuna Fern. desc pe + E. obtusa var. Peasei Svenson Too E. olivacea Torr. + + + +t. + + E. ovata (Roth) R. & S. EL E oe “fe E. ovata var. Heuseri Uechtritz TUS E. palustris (L.) R. & S. LEM NS. + + E. palustris var. major Sonder Se ee ee a he on E. parvula (R. & S.) Link ilio dccem E. pauciflora (Lightf.) Link var. Fernaldii Svenson illi MON E. quadrangulata (Michx.) R. & S. var. crassior Fern. zl: + E. Robbinsii Oakes + ae + + + E. rostellata Torr. + ob = +t E. Smallii Britt. + + 4+ 4+ 4+ c E. tenuis (Willd.) Schultes "Iis + + + E. tricostata Torr. un ete" E. tuberculosa (Michx.) R. & S. v a nri dam Eriophorum angustifolium Honckeny + E. angustifolium var. majus Schultz + S E. gracile W. B. J. Koch Ulli, ME os E. spissum Fern. 4oxbexLBomcE X E. tenellum Nutt. 4 oco pM E. virginicum L. + sade ae E. viridi-carinatum (Engelm.) Fern. Tod BU + + + + + + Fimbristylis autumnalis (L.) R. & S. 28 Rhodora [Vol. 65 Me. N.H. Vt. Mass. R.I. Conn. F. autumnalis var. mucronulata (Michx.) Fern. + c Fuirena pumila Torr. t Hemicarpha micrantha (Vahl) Pax Psilocarya nitens (Vahl) Wood + P. scirpoides Torr, + 4 Rhynchospora alba (L.) Vahl + + + + + R. capillacea Torr. + + + + R. capitellata (Michx.) Vahl + + + + 4 R. fusca (L.) Ait. f. t + + + + R. inundata (Oakes) Fern. d + R. macrostachya Torr. + + 4 R. Torreyana Gray + + Scirpus acutus Muhl. S. americanus Pers. S. ancistrochaetus A. E. Schuyler S. atrocinctus Fern. S. atrovirens Willd. S. atrovirens var. georgianus (Harper) Fern. S. cespitosus L. var. callosus Bigel. S. cespitosus var. delicatulus Fern. S. Clintonii Gray S. cyperinus (L.) Kunth S. cyperinus var. pelius Fern. S. expansus Fern. S. fluviatilis (Torr.) Gray S. Hallii Gray . heterochaetus Chase . hudsonianus (Michx.) Fern. +o + c S. lineatus Michx. 1 S. Longii Fern. S. maritimus L. var. Fernaldii (Bickn.) Beetle S. Olneyi Gray S. paludosus Nels, var. atlanticus Fern. S. Peckii Britt. ++ ++ + E + +++ : + ++++4+4+44 — E + T + S. pedicellatus Fern. S. pedicellatus var, pullus Fern. S S ++++ S. polyphyllus Vahl S. Purshianus Fern. S. robustus Pursh S. rubricosus Fern. S. rubrotinetus Fern. S. Smithii Gray ++ $444 H+ al- 4 1963] Committee on Plant Distribution 29 Me. N.H. Vt. Mass. R.I. Conn. S. Steinmetzii Fern. 35 S. subterminalis Torr. ur decub I4 d S. Torreyi Olney "q- oie Ge GRE dei S. validus Vahl var. creber Fern. =P psig eae uncta ete Ul S. verecundus Fern. sai a ee ag Scleria pauciflora Muhl. var. caroliniana (Willd.) Wood EEr + + S. reticularis Michx. E e ER S. triglomerata Michx. + F S. verticillata Muhl. + The Cyperaceae, other than Carex, were treated by Prof. M. L. Fernald in RHODORA X: 135-144, 1908. As in the case of Carex the number of specimens available for study has become very much larger since then and there has aiso been a considerable change in nomenclature in the inter- vening years. At the conclusion of his report Prof. Fernaid included “Suggestions for Special Observation”. Most of the range extensions that he predicted have proved to be correct, but many of the species from the southern United States have not as yet been reported from the New England area. As in Carex the number of introduced species is insignifi- cant. In fact there are only two, Cyperus fuscus and C. ovu- laris. The latter is native farther south but in New England it is known only as a casual waif on made land in Boston and has not been collected recently. In Prof. Fernald's re- port he cites this single station for C. ovularis and none for C. fuscus. The geographical areas are in general the same as in the previous report. I. GENERALLY DISTRIBUTED. — Dulichium arundinaceum, Eleocharis acicularis, E. elliptica, E. obtusa, E. palustris var. major, Eriophorum spissum, E. tenellum, E. virginicum, Rhynchospora alba, R. capitellata, Scirpus acutus, S. atrocinctus, S. atrovirens var. georgianus, S. cyper- inus var. pelius, S. rubrotinctus and S. validus var. creber. Eleocharis obtusa is poorly represented in northern Maine. Scirpus acutus is infrequent except for the Champlain Basin and the western Maine Coast. There are no stations for it in Rhode Island. Scirpus atrocinctus has not yet been reported from southern Rhode Island or southern Connecticut and has only one station in southeastern Massa- 30 Rhodora [Vol. 65 chusetts. S. atrovirens var. georgianus has not yet been recorded from Aroostook County, Maine. Eriophorum spissum has not been reported from southern Connecticut and Rhode Island and is rare in southeast- ern Massachusetts, due to the absence of cold bogs in these areas. Ia. GENERAL, EXCEPT THE MAINE COAST EAST OF THE KENNEBEC RIVER. — None. Ib. GENERAL, EXCEPT THE WASHINGTON COUNTY COAST. — Cladium mariscoides. Ic. GENERAL, EXCEPT CAPE COD. — Eriophorum viridi-carinatum, Scir- pus atrovirens, S. pedicellatus and S. pedicellatus var, pullus. Scirpus atrovirens is infrequent. S. pedicellatus var. pullus has a distribution in New England similar to the species, but is much less frequent. Ila. NORTHERN — NONE OR NOT MANY STATIONS SOUTH OF 43°, — Eleocharis ovata and var. Heuseri, E. palustris, E. pauciflora var. Fernaldii, Eriophorum angustifolium, Scirpus cespitosus var. callosus and S. Clintonvi. The interpretation of the characters separating Eleocharis ovata and var. Heuseri is so difficult in many specimens that it might have been better to have treated the two together; E. ovata, as currently interpreted, appears to be local in southern New England with a single station in Massachusetts and two in central Connecticut. E. palustris occurs doubtfully at two stations in southern New England. Erio- phorum angustifolium and Seirpus Clintonii are confined to Maine only. IIb. NORTHERN — NUMEROUS STATIONS SOUTH OF 48°, — Scirpus hudsonianus. III. ARCTIC-ALPINE. — None. IV. SOUTHERN — GENERAL SOUTH OF 45^, — Bulbostylis capillaris var. crebra, Cyperus dentatus, Eleocharis Smallii, Rhynchospora fusca, Scirpus americanus, S. eyperinus and S. subterminalis. Scirpus americanus is more abundant near the coast. S. subtermi- "alis has been reported from only three stations in Vermont, all in upland areas. IVa. SOUTHERN — GENERAL SOUTH OF 45° BUT NOT ON MAINE COAST EAST OF THE KENNEBEC RIVER. — Cyperus esculentus, C. filiculmis var. macilentus, C. rivularis, Fimbristylis autumnalis, Scirpus Purshianus and S, Smithii. The only Vermont station for Fimbristylis autumnalis is near the Connecticut River in the extreme southeastern corner of the state. IVb. SOUTHERN — GENERAL SOUTH OF 45° BUT NOT IN WASHINGTON COUNTY. — Cyperus diandrus, C. strigosus, Eleocharis olivacea and E. Robbinsii. Eleocharis Robbinsii in New England is restricted to an area within a hundred miles of the coast. 1963] Committee on Plant Distribution 31 IVc. SOUTHERN — GENERAL SOUTH OF 45^ BUT NOT ON CAPE COD. — Eleocharis tenuis, Eriophorum gracile and Scirpus Torreyi. Eleocharis tenuis is frequent in Maine, and in New Hampshire occurs in the Merrimack River basin and eastward to the coast. It is occasional in eastern Massachusetts and has been collected at East Hartford and Wallingford, Connecticut. IVd. SOUTHERN — GENERAL SOUTH OF 45* BUT NEITHER ON CAPE COD NOR IN WASHINGTON COUNTY. — Bulbostylis capillaris, Cyperus inflex- us, Eleocharis calva, E. diandra, E. obtusa var. ellipsoidalis and var. jejuna and Scirpus expansus. Cyperus inflexus is infrequent throughout New England east of the Connecticut River. Eleocharis calva is also infrequent east of the Con- necticut River drainage area and is apparently absent from the coastal areas east of Connecticut. E. diandra is confined to the Kennebec, lower Merrimack, Thames and Connecticut River valleys. E. obtusa var. ellipsoidalis is frequent near the coast in eastern Massachusetts and southern Rhode Island. V. CHIEFLY THE THREE SOUTHERN STATES. — Hemicarpha micrantha. This species has a few stations in southeastern New Hampshire and adjacent southwestern Maine. Va. CHIEFLY THE THREE SOUTHERN STATES BUT NOT IN WESTERN MASSACHUSETTS. — Eleocharis Engelmanii, E. rostellata, Fimbristylis autumnalis var. mucronulata, Fuirena pumila, Rhynchospora macro- stachya and Scleria reticularis. Eleocharis Engelmanni is very local but with a marked concentra- tion in the Boston region. There is one outlying station at Falmouth, Cumberland County, Maine. Eleocharis rostellata has been reported in Massachusetts in the east only; there is a single station in Maine at Phippsburg, Sagadahoc County. In New England it is strictly mari- time. Fuirena pumila is local except on central Cape Cod; it has not been reported from Connecticut. Rhynchospora macrostachya is restricted to southeastern Massachusetts, Rhode Island and Connecti- cut with two outlying stations at Falmouth, Maine and at Leverett, Franklin County, Massachusetts. Scleria reticularis is restricted to within twenty miles of tidewater except for an outlying station in the Connecticut River valley near Springfield, Massachusetts. vb. CHIEFLY THE THREE SOUTHERN STATES BUT NOT ON CAPE COD. — Scirpus verecundus and Scleria triglomerata. Scirpus verecundus occurs at one station in southwestern Vermont. Scleria triglomerata is absent east of the Sudbury River valley and from all the southeastern counties in Massachusetts except Nantucket. Vc. CHIEFLY THE THREE SOUTHERN STATES BUT NEITHER CAPE COD NOR WESTERN MASSACHUSETTS. — Cyperus erythrorhyzos, C. filiculmis and Eleocharis tuberculosa. Cyperus erythrorhyzos is found chiefly in the lower Merrimack, 32 Rhodora [Vol. 65 upper Nashua, Sudbury, Charles, Taunton, Blackstone and lower Connecticut River valleys. Eleocharis tuberculosa has two New Hamp- shire stations at Madison and Tamworth, both in Carroll County. VI. SOUTHWESTERN NEW ENGLAND CHIEFLY, — Scirpus polyphyllus occurs at two Massachusetts stations in the Connecticut and Deerfield River valleys and at numerous riparian stations in Connecticut from the Connecticut River valley westward. VII. WESTERN NEW ENGLAND — CHIEFLY WEST OF THE CONNECTICUT VALLEY. — Scirpus Peckii. VIIa, COASTAL PLAIN PROPER. — Cyperus polystachyus var. texensis, Eleocharis tricostata and Rhynchospora Torreyana. Cyperus polystachyus var. texrensis occurs locally in specialized habitats on Cape Cod in southeastern Barnstable County and on Nan- tucket. Eleocharis tricostata is very local with one station on Nan- tucket and two in southern Rhode Island. Rhynchospora Torreyana is also local on Cape Cod in central Barnstable County, on Nantucket and in southern Rhode Island. VIIb. COASTAL PLAIN EXTENSIONS. — Eleocharis melanocarpa and Rhynchospora inundata. Eleocharis melanocarpa is vestricted to southeastern Massachusetts and Rhode Island. Rhynchospora inundata has three stations in Plymouth County, Massachusetts, one in Barnstable County on Cape Cod and one in Rhode Island. IXa. CALCIPHILE. — Rhynchospora capillacea and Scirpus cespitosus var. delicatulus. Rhynchospora capillacea is local with only 8 stations. Scirpus ces- pitosus var. delicatulus has a very restricted range; in New England it is confined to extreme northern Maine. IXb. CALCICOLOUS CHIEFLY WEST OF THE CONNECTICUT RIVER IN THE SOUTH; IF IN THE EAST, MOSTLY NORTH OF 45°. — Eleocharis inter- media. Xa. MARITIME IN VICINITY OF COAST. — Cyperus filicinus, C. Grayii, C. odoratus, Eleocharis halophila, E. parvula, Scirpus maritimus var. Fernaldii, S. Olneyi, S. paludosus var. atlanticus and S. robustus. Cyperus filicinus has not been collected in Maine east of the Kenne- bec River. C. odoratus does not appear north of Cape Ann, Essex County, Massachusetts. Scirpus Olneyi has not been found north of Cape Cod except for a station at Kents Island, Newburyport. Eleo- charis parvula, Scirpus maritimus var. Fernaldii and S. paludosus var. atlanticus are common all along the New England Coast. S. robustus is confined to the three southern states except for one station at Falmouth, Maine. Xb. MARITIME WITH INLAND STATIONS. -— None. XI, ESTUARINE. — None. XH. MISCELLANEOUS. — Cyperus brevifolius, C. Engelmannii, C. 1963] Committee on Plant Distribution 33 ferruginescens, C. Houghtonii, C. strigosus var. robustior, X C. Weatherbianus, Eleocharis acicularis var. submersa, E. ambigens, E. equisetoides, X E. fallax, E. microcarpa var. filiculmis, E. nitida, E. obtusa var. Peasei, E. quadrangulata var. crassior, Eriophorum an- gustifolium var. majus, Psilocarya nitens, P. scirpoides, Scirpus an- cistrochaetus, S. fluviatilis, S. Hallii, S. heterochaetus, S. lineatus, S. Longii, S. rubricosus, S. Steinmetzii, Scleria pauciflora var. caroliniana and S. verticillata. Cyperus brevifolius is a southern species which has been discovered at Greenwich, Connecticut and reported in Rhodora 51: 119-120, 1949. C. Engelmannii is found at Brattleboro, Vermont and at six stations in eastern Massachusetts. C. ferruginescens has one station in extreme southern Vermont and three in Connecticut, all along the Connecticut River. C. Houghtonii in New England occurs sparingly south of 45° but is not reported from Rhode Island and Connecticut. C. strigosus var. robustior, a variety of dubious validity, is seldom collected (six stations only!) ; as yet it has not been reported north of Boston. xU. Weatherbianus was described by Prof. Fernald (Rhodora 20: 189-191, 1918) as an intergeneric cross between Cyperus dentatus Torr. and Rhynchospora capitellata (Michx.) Vahl from a single tussock found by C. A. Weatherby at Dennis on Cape Cod in 1915. Mr. Marcel Ray- mond of the Montreal Botanical Garden has recently studied the type and is convinced that it should be treated as a chance hybrid between Dulichium arundinaceum (L.) Britt. and Rhynchospora capitellata (Michx.) Vahl. (See Rhodora 64: 349, 350, 1962). Eleocharis acicu- laris var. submersa seems to be indistinguishable from f. longicaulis. There are stations in Maine and Massachusetts. E. ambigens reaches its northeastern limit of range at Cuttyhunk, Massachusetts. The only other station in our area is at Block Island, Rhode Island where it was collected by Prof. Fernald in 1913. E. equisetoides is local at Welles- ley, Massachusetts, Cumberland and West Greenwich, Rhode Island and Thompson, Connecticut. X E. fallax has been collected at Yar- mouth, Massachusetts. E. microcarpa war. filiculmis is a southern variety which has one station in New England at Voluntown, Con- necticut. E. mitida is represented by specimens from Colebrook and Columbia in Coós County, New Hampshire. E. obtusa var. Peasei has been collected at Fryeburg, Maine and Ossipee, New Hampshire. E. quadrangulata var. crassior has been found at Wellesley, Massachu- setts and at Guilford and W. Guilford, Connecticut. Eriophorum an- gustifolium var. majus is, like the species, of northern distribution but is very local. It is represented by specimens from Matinicus and Oak- land, Maine and is cited by Prof. Fernald in Gray’s Manual, 8th ed., from Massachusetts. Psilocarya nitens occurs at Plymouth, Massachu- setts, the only New England station yet discovered. P. scirpoides has been found at Brewster and Harwich on Cape Cod, Springfield, Massa- 34 Rhodora [Vol]. 65 chusetts and East Providence, Rhode Island. Scirpus ancistrochaetus, a recently described species, has been reported from near Bellows Falls, Vermont (See Rhodora 64: 43-49, 1962). S. fluviatilis is local in river valleys near the coast south of 45°, but is frequent along the shores of Lake Champlain. S. Hallii has been found in New England only at Winter Pond, Winchester in eastern Massachusetts. S. hetero- chaetus is local in Massachusetts and Rhode Island but there are numerous stations on the shores of Lake Champlain. It has also been reported in the Connecticut River valley. S. lineatus has one station in central Maine, one in south central New Hampshire, two in the Champlain Valley in western Vermont and five in the Housatonic River valley in western Massachusetts and Connecticut. S. Longii, in New England a plant of fresh water river valleys, is local in Middlesex and Suffolk Counties in eastern Massachusetts; one station has been discovered at South Windsor in central Connecticut. S. rubricosus has five stations in eastern Massachusetts, three of them on Cape Cod. S. Steinmetzii is represented by specimens from Passadumkeag, Penob- scot County, Maine only. Scleria pauciflora var. caroliniana occurs at Uxbridge, Massachusetts and Columbia, Hartford and Waterford, Connecticut. It is cited from southwestern New Hampshire by Prof. Fernald in Gray's Manual 8th ed. S. verticillata has one station in northwestern Connecticut at Salisbury. R. C. BEAN A. F. HILL R. J. EATON SCHEUCHZERIA IN WESTERN PENNSYLVANIA. — In Sep- tember, 1961, while exploring a large sphagnum-cranberry bog, new to us, in Erie County, Pennsylvania, W. E. Buker and the writer found an old fruiting specimen of Scheuch- zeria palustris L. var. americana. Fern. We visited this bog again in early June, 1962, and saw another plant in fruit. Although we are familiar with many bogs here, this is our first and only record of this plant from the western part of the State. In our herbarium, we have specimens from Car- bon, Monroe, and Sullivan counties, all in eastern Pennsyl- vania. This species, of which ours in America is a variety, is widely distributed over the cooler regions of the Northern Hemisphere, growing in cold sphagnum bogs and on peaty shores. — L. K. HENRY, CURATOR OF PLANTS, CARNEGIE MUSEUM, PITTSBURGH, PA. ANNOTATION OF NORTH AMERICAN POLYCARPAEA' OLGA LAKELA Polycarpaea nebulosa spec. nov. Annua herbacea erecta. Radix tenuis ad 25 em. longa. Caulis 6-18 cm. altus corymbose ramosus pilo- sus vel glabrescens ad nodos incrassatus. Folia radicalia 8-12 mm. longa petiolata laminis oblongo-ovatis vel suborbicularibus; caulina 1.0-2.5 em. longa linearia carnosula glabra subrevoluta mucronata; axillaria fasciculata similia breviora. Stipulae scariosae plus minusve bifidae. Inflorescentia dichotomo-cymosa decomposita. Bracteae flor- ales scariosae. Flores 2.8-3.1 mm. longi perigynici. Hypanthium breve crateriforme. Sepala glabra scariosa nitida ovata vel lanceolata acuta. Petala ovata erosa vel integra sub anthesi incarnata cum staminum basibus connata. Filamenta subulata; antherae albae. Pistillum stipi- tatum; ovula plerumque 7; stigmata 3 subsessilia. Capsula trivalvis; semina reniforma pallide brunnea obscure transverse striata. Erect annual herb. Taproot slender to 25 cm. long. Stem 6-18 cm. high, corymbosely branched, pilose or glabrate, becoming enlarged at nodes. Radical leaves 8-12 mm. long, petioled, blades oblong-ovate or suborbicular; cauline leaves 1-2.5 cm. long, linear, fleshy, glabrous, somewhat revolute, nerve excurrent. Fascicular leaves similar, short- er. Stipules scarious, more or less cleft. Inflorescence a compound dichasial cyme. Floral bracts scarious. Flowers 2.8-3.1 mm. long, perigynous. Hypanthium short, crater-like. Sepals glabrous, scarious, lustrous, ovate, or lanceolate, acute. Petals ovate, erose or entire, pink in anthesis, connate with stamen bases. Filaments subulate; anthers white. Pistil stiped; ovules usually 7; stigmas 3, nearly sessile. Cap- sule 3-valved; seeds reniform, pale brown, with obscure tranverse striae. (Name derived from L. nebula, mist for cloud-like luster of the flowers). TYPE: FLORIDA, Temple Terrace, adjoining Tampa. South-facing slope of sandhill west of 56th St., east of Overlook Drive. O. Lakela 25565, 7 Nov. 1962. Pl. 1280 & 1281. HOLOTYPE: (USF). ISOTYPES: GH, SMU, US, F). ANNOTATION OF LIVING PLANT Subsequent to the discovery of Polycarpaea in North America, the Florida plant has been studied in the green- house and in the field in large numbers. Ripe seeds found in capsules of a mature plant, coll. 24893A, 26 Jan. 1962, "Contribution No. 2. Botanical Laboratories, University of South Flor da. 35 Rhodora 36 Plate 1280. Polycarpaea nebulosa. [Vol. 65 Hillsborough County Sandhill east of üverlosk Drive, northwest of inter- section ef Sóth St. with Temple Terrace Sishwar. l lants chewing in Marginal area of un- broken creund, with few daks, kupeterins, Liatris è Heliantherum, Petuis pink in life, fading to C oranges Atira neariy sessile: aces Wasally To or less, Teopile Terrace, ü, Lakela 2884 + oD Nav. deo à t QT Photograph of Holotype. 1963] Polycarpoea — Lakela 37 Plate 1281. Polycarpaea nebulosa. Enlarged portion of the inflores- cence of Holotype. were promptly planted in a greenhouse without artificial heat. Germination was evidenced on the 3rd of March. The spatulate cotyledons less than 1 mm. long and half as wide were more than tripled in size during the rosette develop- ment. They were fleshy, yellowish green and lustrous. The petioled radical leaves, 2-3(4) in number, 8-10 mm. long, with ovate, oblanceolate, or suborbicular blades, appeared singly. The blade appearing first became the largest. Fleshy in texture, and variegated with green and colorless tissue in pinnate fashion, they remained functional with tardily with- ering cotyledons through the growth of the lowest inter- nodes. Flowering in April and May, the habit and floral structures of the plants compared well with those grown in a natural environment. Meantime, the known sites of Polycarpaea were under close surveillance for signs of new growth. Summer rains were necessary to activate the heated sands. The first seed- lings were located on the 7th of June. Prior study facilitated Fig. 1 38 Rhodora [Vol. 65 the identification of the tiny seedlings appearing in masses in the openings of vegetation on grassy terraces and on sand- hills. The young plants of other species, e.g., Diodia and Siphonychia, were far beyond the seedling stages, thus elim- inating confusion in the identification of Polycarpaea. Pro- gressive germination rendered it possible to obtain various stages of growth in a single colony, as well as seeds with promise of germination (Fig. 1). Seedlings with complete rosettes, some already with withering leaves (no. 25136) were collected on the 23rd of June (PI. 1282). During anthesis the stems become denuded of foliage, excepting the fascicular leaves that may remain on branches below the inflorescences (Pl. 1283). The cauline leaves are best seen before flowering (Pl. 1283). The fleshy tissues of a living leaf appear striped or mottled with green and white. This pattern of variegation as seen under the microscope is effected by the distribution of chlorophyll-containing tissues over slightly elevated veins and the colorless parenchyma within interstices of the reticulum. The colorless cells reveal an abundance of white crystalline inclusions. Stomata occur on both surfaces. The midnerve, visible only on the lower surface, becomes an excurrent tip. It is soft and pliable, hardly a “bristle” in the true sense of the term. On drying, the leaf becomes longitudinally ridged or wrinkled with slightly revolute margins. FLORAL STRUCTURES Specialization in the genus is manifested by reduction of chlorophyll in stipules, floral bracts, and sepals. Structurally these organs have become translucent and scarious with reduced vascular supply. In stipules the recognizable mid- vein is produced to a filament; in floral bracts it is less prom- inent; in sepals the vascular trace entering into the some- what incrassate, usually colored basal area, becomes obsoles- cent toward the apex. Isolated spiral elements may be seen along the median line with thicker-walled parenchymatous Figure 1. Polycarpaea nebulosa, A, Seed, 0.5 mm. long. B, Seedling above the ground, 6 mm. long. C, Cotyledons and branch root, 12 mm. long. ith rosettes. ings w Seedl Polycarpaea nebulosa. Plate 1282 40 Rhodora [Vol. 65 Plate 1283. Polycarpaea nebulosa. Appearance of plant A, prean- thesis, B. anthesis. 1963] Polycarpoea — Lakela A1 cells. This remnant of the trace under a micro-projector casts a shadow which is more accented in Ceylon and Soutn American plants than in those of Florida and Lucknow. Often the sepals along this line appear folded or condupli- cate. Allowing average variations, the small flowers, 2-3.5 mm. long, examined in representative specimens, show marked uniformity in structural pattern. Notwithstanding, differ- ences are discernible. The Florida plant is similar to but not identical with South American and Old World plants passing under the epithet of Polycarpaea corymbosa. Despite the limited availability of materials from the various regions of this amphigaean genus, it was deemed expedient to record diagrammatically characters of possible diagnostic value. Boiled, dissected flowers were mounted in water under a coverslip. Sepals, ovaries with ovules, or capsules with seeds, and tips of leaves, if available, were traced to the same scale under a micro-projector. The style measurements were estimated across the field of the micro- scope. The stigmas were included for added length to facili- tate arriving at an approximate figure. (Fig. 2). DISCUSSION According to Martius (1872), styles in the genus vary greatly. In a large number of Florida plants studied, they have been found to be consistently short. The same cannot be reported for the Lucknow plant on the basis of a single available specimen with a few cymes in flower and without mature capsules. In habit the plant is lax, denuded of foliage at base, with spreading, linear, cauline leaves. Exchange specimens from the National Botanic Gardens, Lucknow, received at Tampa, 31 Oct. 1962, consist of a num- ber of plants in preanthesis and seedlings; their habit is similar to the cited specimen. The radical leaves in seedlings compare well with those of the Florida seedlings. However, the lowest cauline leaves are somewhat wider. The long styles and lanceolate, attenuate sepals are in common with Ceylon and South American plants. Vegetatively they are 42 Rhodora [Vol. 65 O OQ afe Fig. 2. Taxonomic characters of Polycarpeae. A. Sepals essentially ovate-lanceolate or lanceolate, acute or attenuate, glabrous or rarely ciliolate at base; style with stigmas 0.12 mm. long; ovules 11; excur- rent vein-tip less than 1 mm. long. Coll. Omar Shankar 12982, 20-9- 1963] Polycarpoea — Lakela 43 unlike. The cited Ceylon collection consists of ten plants in flower of which nine are strict, and mostly with persistent radical leaves. The cauline leaves at the lowest nodes are broadly linear or oblanceolate, 2-3.5 mm. wide, in sharp contrast to the narrower and shorter upper cauline leaves. Further study of the Lucknow and Ceylon material is recom- mended, especially since fruiting plants have not been seen. In the South American taxon, P. corymbosa, var. brasilien- sis, the cauline leaves throughout are narrowly linear or subulate, of firm texture, with strongly revolute margin and terminal seta 2-3 mm. long. The stipules are deeply cleft with prolonged filaments. The mature seed is dark brown, smooth, and shiny. In conclusion, P. nebulosa is distinguished by compact habit, reddish nodes, internodes and undersurfaces of leaves. The cauline leaves throughout, are narrowly linear. The ovate acute sepals are dorsally rounded. The petals are pink in contrast to the white and sordid yellow petals recorded for P. corymbosa. Mature seeds, comparable only with those of a Paraguayan plant, are of even light buff when fully cutinized, and the groove on the dorsal side disappears on soaking. A preliminary study of microsporogenesis in P. nebulosa indicates that the chromosome number is probably n = 3. Further investigation is necessary before this number can be confirmed (unpubl. Long, R. W., Associate Professor, University of South Florida.) DISTRIBUTION The present known distribution of Polycarpaea in Florida 56, Lucknow, India. (UsrF). B. Sepals lanceolate, attenuate, often ciliolate below with median shadow line; style with stigmas 0.24 mm. long; ovules 11-13; excurrent vein tip 1 mm. long or less. Coil. George R. Cooley, 14-1-57. Dambulla, Ceylon. (UsF). C. Sepals ovate, acute; style with stigmas 0.13 mm. long; seeds usually 7 (8), c. mature seed actual size 0.5 mm. 1.; excurrent vein tip 1 mm. long or less. Coll. Lakela 24779, 18-10-61. (USF). D. Sepals lanceolate, attenuate, pilose, with median shadow line; style with stigmas 0.26 mm. long; d. mature seed actual size 0.5 mm. long; ovules 6. Coll. E. Hassler 9808, 1907-08, Paraguay, S. A. (vs). 44 Rhodora [Vol. 65 is restricted to northeast Tampa and the western border of adjoining Temple Terrace. It occurs sporadically in an area approximating 12 square miles. It flowers from August to November. The question of Polycarpaea occuring in other parts of North America was aroused by a record from Mexico, listed in Index Kewensis. In reply to an inquiry, Dr. R. Llamas, Director, Instituto de Biologia, Ciudad Universitaria, in a communique referred to the “note of Hemsley (Biologia Centrali Americana Vol. I, p. TT) where he says: . . . The following is probably not a true Polycarpaea (P. cuspidata Ehrenb. Real del Monte).” Further, Dr. Llamas notes that the climate of Real del Monte, “is... a cold one: 2781 metres above sea level, with mean temperature in summer of 14.5 degrees Centigrade, which drops to from 2 to 10 degrees below zero Centigrade. For these reasons we are inclined to believe in the nonexistence of this species or, at least it has not been collected up to the present time." Pat- ently, the climate is unfavorable for survival of a tropical plant. The genus embraces a global distribution in the warm parts of both hemispheres. On the northern limit of the range approximating the same parallel, North America fills in the gap between the Canary Islands, China and Lucknow. ACKNOWLEDGEMENTS The author wishes to extend thanks to Dr. Lloyd H. Shinners for the Latin description and Miss Martha Gillon for painstaking aid with illustrations and the manuscript. THE UNIVERSITY OF SOUTH FLORIDA, TAMPA, FLORIDA. REFERENCES LAKELA, O. 1962. Occurence of species of Polycarpaea Lam, (Cary- phyllaceae) in North America. Rhodora 64:179-182. Martius, K. F. P. VON ET AL. 1872. Flora Brasiliensis, 14 (2): 253. PROTANDRY IN TWO SPECIES OF STREPTANTHUS (CRUCIFERAE) REED C. ROLLINS The widespread presence of self-incompatibility as an out- breeding mechanism in the Cruciferae has been emphasized particularly by Bateman (1955a). Of the eleven tribes re- ported upon, only the Streptantheae, in which but a single species of Caulanthus had been investigated, showed no self- incompatibility. During the last year, we have had two species of Streptanthus, S. cutleri Cory and S. carinatus Wright, in greenhouse cultures. These were grown from seed collected in western Texas. Isolation tests on both species show they are not only self-incompatible but that the flowers are protandrous as well The Streptantheae may now be added to the list of tribes showing self-incompatibili- ty. In his major paper, Bateman (l.c., p. 63) made the state- ment that there is no protandry, no protogyny and no dioecy in the Cruciferae. A later short note (Bateman, 1955b) acknowledged the presence of dioecy in Lepidium sisym- broides Hook. as an exception to the rule. Now, we bring to the attention of botanists an exception to the generally accepted idea that there is no protandry in the Cruciferae. At the same time, one of the most striking cases of zygo- morphy in this family, where actinomorphy characterizes the flowers of nearly all species, is emphasized. Both Streptanthus cutleri and S. carinatus have greatly elongated racemose inflorescences on which it is usual for a single flower to mature at a time. There are successively maturing flowers from base to apex of the inflorescence. These are usually spaced about 2 cm. apart at the time of full anthesis. In each flower, the anthers mature before the stigma and, as shown in figs. 1-4, the stamens project slight- ly. At the same time, they occupy the area at the center of the flower and effectively cover the stigma which remains unexpanded and situated toward the base of the flower until 45 46 Rhodora [Vol. 65 well after the pollen has been shed. At the stage of anther maturity, the outer floral parts tightly close the flower en- trance. Following anther maturation and the shedding of pollen, the ovary elongates, the stigma expands and the floral parts move to an open stance. The stigma becomes much en- larged and is brought by the elongating ovary and style to a central position below the entrance of the flower. The prim- ary flow of nectar from the nectaries located at the base of the filaments and near the point of insertion of the petals coincides with stigma maturation rather than anther matur- ity. This is significant not only because it correlates with the opportunity of flower entry by insects and the receptivity of the stigma to pollen but because it actually avoids the time of anther maturation as well. The question as to what it is about the flowers in the closed condition, where nectar is not available, that attracts appropriate insects for pollen pick- up, is a pertinent one. We do not have a definite answer but we did notice that flower odor seemed to be at its peak dur- ing the period of anther maturity and pollen shedding. Odor, together with flower shape and perhaps color, may fulfill the requirements for insect attraction. On the other hand, it may be that the insects pollinating these flowers do not dis- tinguish between the open and closed condition. If so, they would move from flower to flower indiscriminately, some- times being rewarded by available nectar and sometimes not. In any case, the timing of pollen pick-up by the insect is effectively separated from pollen deposit not only by protan- dry but by several correlated reinforcing mechanisms in addition. The situation is approximately the same in Strep- tanthus cutleri and S. carinatus, although the flower form in these two species is radically different because of the differ- ences in petal development. The petal color is also different. The petals of S. cutleri are light to reddish purple while those of S. carinatus are brownish white with the veins prominently purple. In both species the calyx is a deep blackish purple. Undoubtedly different insects pollinate these two species of Streptanthus. However, we have not observed insect visitation to the flowers of either species in 1963] Protandry in Streptanthus — Rollins 4T its native habitat. Both species occur wholly in extreme western Texas. Two levels of zygomorphy are shown by Streptanthus carinatus and S. cutleri. In the former, the stamen whorls Figs. 1-4. Flowers of Streptanthus, X 2. Fig. 1 & 2, face and side view respectively of S. cutleri. Fig. 3 & 4, face and side view respectively of S. carinatus. Drawings by Dorothy H. Marsh. are strongly zygomorphic but the other floral parts show little or no zygomorphy. The situation is readily seen by reference to fig. 9 and 4. The filaments of the paired sta- 48 Rhodora [Vol. 65 mens, above and below, are of different lengths and the usual radial symmetry of the androecium present in most crucifer flowers is thereby considerably changed. A further change is seen in the tendency of the two single stamens to become associated with the lower paired stamens into a grouping of four with the anthers at about the same level of exsertion from the flower. Recurved petals further empha- size the prominence of the anther position. The androecium of S. cutleri is quite similar to that of S. carinatus, but the upper stamen pair is less exserted and the tendency of the single stamens “to platform" with the lower paired stamens is perhaps more strongly developed. A strong zygomorphy is present in the corollas of Strep- tanthus cutleri. Here, the blades of the two lower petals have virtually been lost and the two upper bladed petals have become strictly upright. There is some variability be- tween different plants as to the amount of blade present on the lower petals. This ranges from no blade at all, as shown in fig. 1 and 2, to a distinct asymmetrical blade of consider- ably smaller size than that present on the upper petals. The calyx also shows a noticeable but not highly developed zygo- morphy in this species. The lower sepal is more pouched toward its base than any of the other three. This is shown in fig. 2. In summary, it may be stated that Streptanthus cutleri and S. carinatus are protandrous and self-incompatible. Thcse phenomena promote out-crossing between different plants of a given population. Protandry and self-incompati- bility reinforce each other and thereby probably produce a nearly foolproof system of out-crossing. Several facets of flower behavior strengthen the separation of pollen pick-up and pollen deposit in a particular flower provided by the maturation sequence characteristic of protandry. These are (1) flower closure, preventing insect penetration and conse- quent self-pollination during anther dehiscence, (2) flower opening during the period of stigma receptivity, (3) nectar flow timed- to coincide with stigma maturation and flower opening. — GRAY HERBARIUM, HARVARD UNIVERSITY. 1963] Protandry in Streptanthus — Rollins 49 LITERATURE CITED BATEMAN, A. J. 1955a. Self-incompatibility Systems in Angio- sperms III. Cruciferae. Heredity 9: 53-68. 1955 b. Note on Dioecy in the Cruciferae. Heredity 9: 415. VIOLA PALUSTRIS L. IN ARIZONA. — This past spring there were sent to me specimens of a species of violet previously unreported for Arizona. Kearney and Peebles (1960) listed nine species for the state which are named below. Viola palustris L. June 3, 1962. Collector V. W. Rhoton. Lake Number One (U.S. Forest Service name) on Heber- Payson Highway, about 25 miles from Heber, about 14. mile from edge of Mogollon rim. Elevation about 7200’. Coconino County, Arizona. Abundant in wet soil in a strip about 25’ x 500’ around the southwestern edge of the lake. Specimens of this collection are deposited in the herbaria of Arizona State College, Arizona State University and of Mr. V. W. Rhoton of Lakeside, Arizona. This is the white-flowered form which is apparently more common in the southern Rocky Mountains than the typical blue-flowered plant. The nearest locations to the above that I am aware of are in the high mountain passes north of Durango, Colorado. Only two species of violet are commonly found in the mountains of Arizona. These are V. canadensis L. and V. nephrophy!la Greene, which are occasional in pine and doug- las fir forests in northern, northeastern, and southeastern Arizona. Viola Rafinesquii Greene, V. pedatifida G. Don, V. aurea Kellogg, and V. nuttallii Pursh occur very rarely in coniferous forests in northern and northeastern Arizona. Viola charlestonensis Baker & Clausen has been reported from a single location. Viola umbraticola H. B. K. is a form of V. nephrophylla that I have been unable to separate and therefore consider synonymous. The number of species in the state remains therefore at nine. — NORMAN H. RUSSELL, DEPARTMENT OF BOTANY, ARIZONA STATE UNIVERSITY, TEMPE, ARIZONA. STUDIES IN LINUM: L. IMBRICATUM AND L. HUDSONIOIDES: C. MARVIN ROGERS On the basis of two or three seemingly minor features Planchon (London Jour. Bot. 7: 185-186. 1848) distin- guished two closely related species of flax of Texas, Linum multicaule Hook. (—L. imbricatum [Raf.] Shinners) and the new L. hudsonioides Planchon. Later Asa Gray (Pl. Lindheim., Boston Jour. Nat. Hist. 6: 155-156. 1850) some- what hesitantly maintained both, but subsequently (PI. Wright., Smithson. Contr. Knowl. 3: 27. 1852) decided that the two were not separable. Since that time they have been treated as a single species. This species has been readily distinguished from other flaxes with united styles by the numerous, small, closely appressed, imbricated leaves and the sepals with broadly scarious, nonglandular margins. An examination of the rather plentiful herbarium material reveals two clear-cut populations which, though very much alike in habit, are easily separated by a number of charac- teristics and it is evident that L. hudsonioides should proper- ly be recognized. The following table lists the more obvious TABLE 1. A comparison of L. imbricatum and L. hudsonioides.. L. imbricatum L. hudsonioides Pedicels and upper part of the Pedicels and upper part of the stem hirsute stem hirsutulous Upper leaves ciliate-margined Upper leaves not ciliate-margined Lower, longer pedicels (2-) 4-6 Lower, longer pedicels (3-) 8-13 (-11) mm averaging about 5 (-15) mm averaging about 10 mm long? mm long Outer sepals, like the inner, Outer sepals ^ntire or nearly so prominently toothed Cartilaginous portion of false Cartilaginous portion of false sep- septum conspicuously broad- tum uniformly narrow or ab- ened below sent throughout 'Contribution No. 94 from the Department of Biology, Wayne State University. "Dimensions, when based on 25 or more measurements, are shown in the following way: (2-) 4-6 (-11) mm long indicates a total range of 2-11 mm, with approximately 60% of the individuals being 4-6 mm long. 50 1963] Hyaline portion of false septum appressed pilose along the in- ner margin, otherwise essen- tially glabrous Seeds biconvex Unopened anthers 0.6-1.2 mm averaging less than 0.9 mm Linum — Rogers 51 Hyaline portion of false septum conspicuously tomentose near the summit Seeds three angled Unopened anthers 1-1.6 mm av- eraging more than 1.2 mm long long of these features, some of which are shown in the accom- panying illustrations (Figs. 1-9). Pedicel length and other more subtle differences in leafi- ness of the upper part of the stem, divergence of the branch- es and stature enable one to recognize perhaps nine of ten individuals without resorting to the other characters listed. Though probably not the most significant difference, the type of pubescence is perhaps the most easily observed and once the two taxa have been compared for this feature, specimens may be readily sorted on the basis of pubescence alone, since there appears to be virtually complete correlation between this and the other features. There is very little evidence of hybridization between the two species (Letterman, July 1880, Denison, Texas (MO) and perhaps Tharp, Apr. 22, 1931, Bastrop, Texas (TEX) can be interpreted as being intermediate) and, as shown in Fig. 10, they occupy nearly distinct geographical ranges. The species are clearly related to L. rigidum, which they resemble rather closely in the annual habit, shape and mode of dehiscence of the fruit, development of complete false septa, union of styles and pollen morphology. Linum imbricatum (Raf.) Shinners, Field & Lab. 25: 32. 1957. Fig. 1-4. Nezera (or Linum) imbricata Raf. New Fl. and Bot. N. Amer. 4: 66. (1836) 1838. Linum multicaule Hook. in T. & G. Fl. N. Amer. 1:678. 1840. Cathartolinum multicaule (Hook.) Small, N. Amer. Fl. 25: 84. 1907. Low annual herb (6-) 11-20 (-27) cm tall, from a slender tap root, branched at the base, sometimes divaricately so in plants along the Gulf shore; stems spreading-ascending or erect, terete or nearly so at the base, strongly striate above, glabrous below, conspicuously short-hirsute with stout-based hairs above; leaves narrow, opposite near the base, alternate. above, the larger (5-) 6-8 (-9) mm long, (0.5-) 0.7-0.9 (-1.2) mm wide, strongly imbricate, the lower spreading- ii! i x J.-L Sn Gn Es SSF EP ? |} p—s -——— i Ft z lo x " S EL imbricatum € @L.hudsonioides Figs. 1-8, X 6; Fig. 9, X 3/5. Figs. 1-4. Linum imbricatum. Fig. 1. Fruit. Fig. 2, Carpel. Fig. 3. Cross-section of seed. Fig. 4. Seed. Figs. 5-9. Linum hudsonioides. Fig. 5. Fruit. Fig. 6. Carpel. Fig. 7. Cross-section of seed. Fig. 8. Seed. Fig. 9. Habit Fig. 10. Distribution of L. imbricatum and L. hudsonioides. Not shown are two colleetions of L. hudsonioides from Saline and Sedgwick Counties, Kansas. 1963] Linum — Rogers 58 ascending, the cauline rather strongly appressed, with conspicuous cartilaginous-based midribs, short-awned, the upper ciliate-margined; stipular glands none; inflorescence few-flowered, the lower, longer pedicels (2-) 4-6 (-11) mm long, the upper mostly hidden by the sub- tending leaves; sepals 5, persistent, usually with a few stiff hairs along the midrib, ovate, (4.2-) 4.6-5.5 (-6.1) mm long, with broad, commonly purplish, scarious margins, prominently toothed above the middle, the median herbaceous portion extended into an awn; petals 5, fugacious, yellow, with or without a darker base, obovate, 6.5-8 mm long, pilose at the base inside; stamen tube short; staminodia none; stamens 5; filaments about 5 mm long; anthers elliptie, 0.6-1.2 mm long before dehiscence; fruit pale, completely 10-celled, broadly ovate, 2.6-3 mm high, 2.9-3.3 mm in diameter, splitting freely at the false septa into 5 parts; true septa cartilaginous; false septa appressed pilose along the inner margin, otherwise essentially glabrous, mostly hyaline, but with a cartilaginous portion along the ovary wall, this cartilaginous strip conspicuously broader toward the base of the carpel; styles 5, 2-4.3 mm long, united to within 0.3-0.8 mm of the stigmas; stigmas capitate; seeds 10, flattened, ovate, 2-2.6 mm long, 1.1-1.5 mm wide, reddish brown. In predominantly sandy soil in the Blackland and Coastal Prairies throughout east central Texas. Type: San Felipe (Austin County), Texas, Drummond 37, Coll. 3. Though the specimen, received from Torrey, from which Rafinesque drew his description of Nezera imbricata, is probably not extant, there is little question but that it was one of the several plants distributed under this number. A sheet from the Torrey herbarium (NY) was almost certainly erroneously labelled “II, 47”, accounting for some confusion as to the correct collection number. The original citation of Hooker's L. multicaule (in T. & G., Fl. of N. Amer. 1: 205. 1838, as L. selaginoides Lam.) was Collection 2, 47 (312), later (T. & G. op. cit.: 678. 1840) changed to Collection 2, 31, but the specimen in the Hooker herbarium (K), upon which the name was based is Collection 3, 37. Other similar specimens, here ecnsidered isotypes, have been seen at GH and NY. About 100 collections were examined of which the following are representative.’ TEXAS: ARANSAS: Whitehouse 18163 (SMU); Atascosa: Schulz 81 (US); BASTROP: Barkley et al. 7037 (OKLA, TEX) ;BEE: Shinners 9889 (SMU); BRAzOS: Parks, May 1, 1947 (OKLA, TEX) ;BROOKS: Cory 55231 (SMU, US); BURLESON: Fisher 31 (F); CALDWELL: McBoyce, 1931 (TEX) ; CALHOUN: Tharp, May 22, 1930 (TEX); COOKE: Shinners 3Specimens are cited alphabetically by state and then county and only one collection per county is given. For the eleven herbaria from which material was borrowed, the usual abbreviations are given. The assistance of the respective curators is gratefully acknowledged. 54 Rhodora [Vol. 65 12410 (SMU); DALLAS: Reverchon, May, 1875 (GH, NY); DE WITT: Riedel, Mar 15, 1942 (TEX); DiMMiT: Lundell 13603 (US); DUVAL: Croft 101 (NY); FAYETTE: Johnston & McCart 5167 (TEX); FORT BEND: Bray 106 (US); GoLIAD: Williams 41 (TEX); GONZALES: Tharp, May 21, 1936 (MICH); GRAYSON: Letterman, Jul, 1880 (M0) ; Harris: Tharp 2293 (TEX); Hays: Baker 35 (TEX); HIDALGO: Runyon 1703 (us); HUNT: Shinners 12245 (SMU); JIM WELLS: Drushel 6504 (M0); KLEBERG: Reed 54 (NY); LA SALLE: Reverchon, Apr 29, 1905 (OKL); LEE: Cory 55761 (SMU); MATAGORDA: Palmer 9674 (MO, US); MCLEN- NAN: Smith 426 (TEX); MILAM: Fisher 3151 (US); NAVARRO: Joor, June 8, 1880 (Mo, Us); NUECES: Heller 1389 (F, MICH, MO, NY, OKL, SMU, US); SAN Patricio: Williges 380 (SMU); TRaAvis: Rehm, May 1, 1935 (TEX p.p.); VAN ZANDT: Van Fleet 1217 (SMU); VICTORIA: Eggert, Apr 10, 1900 (MO); WALLER: Hall 64 (F p.p., GH, MO, NY, US); Wilson: Parks & Cory 7779 (TAES). Linum hudsonioides Planchon, London Jour. Bot. 7: 186. 1848. Fig. 5-9. Annual herb with a slender taproot, closely resembling L. imbri- catum in habit, (5-) 11-23 (-30) em tall; stems scabrous or hirsutulous on the angles above, otherwise glabrous, branching from the base, ascending to erect, nearly terete below, prominently striate above; leaves narrow, opposite near the base, alternate above, the larger (5-) 6-8.5 (-10) mm long, (0.5-) 0.7-0.9 (-1.1) mm wide, imbricated throughout, the lower spreading-ascending, the cauline closely ap- pressed, the lower sharp-pointed, the upper with a short slender terminal awn, all entire, the uppermost with a narrow scarious margin; stipular glands none; inflorescence few-flowered, the lower, longer fruiting pedicels (3-) 8-13 (-15) mm long, all usually con- spicuously exserted beyond the subtending leaves; sepals 5, persistent, the outer lanceolate-ovate, (4.5-) 5.5-6.2 (-7.2) mm long, with a very broad, scarious, entire or sparsely delicately toothed, or in age some- what lacerate, margin, the slender herbaceous median portion extend- ing into a conspicuous awn, the inner sepals similar, but more prom- inently toothed; petals 5, fugacious, obovate, yellow, with or without à darker basal portion, 8-12 mm long, pilose at the base inside; stamen tube short; staminodia none; stamens 5, filaments about 5 mm long; anthers elliptic, 1-1.6 mm long before dehiscence; fruit pale, completely 10-celled, broadly ovate, 2.7-3.5 mm long, 2.8-3.6 mm in diameter, splitting freely at the false septa into 5 parts; true septa cartilaginous; false septa tomentose near the summit, entirely hyaline, or with a uniformly very narrow outer cartilaginous portion; styles 5, 2.7-6.3 mm long, united to within 0.3-1.1 mm of the stigmas; stigmas capitate; seeds 10, ovate, 2-2.7 mm long, 1-1.2 mm wide, wedge-shaped in cross-section, reddish brown. In sandy or gravelly, granitic or occasionally calcareous soil, prin- cipally in the eastern Plains Country and the Edwards Plateau areas ~ 1963] Linum — Rogers 55 of Texas, with outlying stations in Trans-Pecos Texas and south- eastern. New Mexico and in the Wichita Mts. of Oklahoma. Two additional collections, perhaps questionable, have come from central Kansas. Type: Entre Bejar y el rio de la Trinidad, Berlandier 385, May, 1828 (K, not seen). Possible isotype, a poor specimen, at GH (dated Mar 20, 1828). About 100 collections were examined of which the following are representative. KANSAS: SALINE: Letterman, Aug 24, 1884 (Us); SEDGWICK: Sears, Jun 9, 1928 (OKL); NEW MEXICO: Eppy: Ripley & Barneby 2593 (NY); OKLAHOMA: COMANCHE: Orr 313 (SMU); GREER: Waterfall 7246 (OKL, OKLA); TEXAS: BASTROP: Thorp, Apr 22, 1931 (TEX); BELL: Wolff 2897 (TAES); BEXAR: Schulz 44 (US); BOSQUE: Van Fleet 52 (SMU); BREWSTER: Mueller 8056 (TEX); BROWN: Rev- erchon 1294 (MO); BURNET: Barkley & Rowell 62 (TEX) ; CALLAHAN: Letterman, Aug 10, 1882 (Mo, NY, US); COMAL: Pennell 5457 (NY); DALLAS: Reverchon 109 (US p.p) ; EASTLAND: Tharp, Apr, 1937 (NY); ERATH: Gough, May 16, 1921 (TEX); FRIO: Higdon 53-139 (TEX p.p-) ; GILLESPIE: Barkley et al. 47253 (TEX); HAMILTON: Reverchon, Apr, 1882 (GH); Hays: Stanfield, Apr 1, 1897 (NY); Hoop: Eggert, May 6, 1900 (mo); HOWARD: Parks & Cory 12520 (TAES) ; JACK: Shinners 12347 (SMU); JEFF Davis: Girard, Ft. Davis (GH); Jim Hoca: Tharp, Jun 15, 1928 (TEX); KERR: May 555 (SMU); LLANO: MceCart 5673 (sMU); LUBBOCK (or LAMB): Tharp 6309 (TEX); Mason: Shinners 26296 (SMU); MILLS: McWhorter, Apr 17 (TEX); MITCHELL: Pohl 5106 (SMU); NOLAN: Palmer 12469 (MO, us); STARR: Clover 1097 (NY); STEPHENS: Deaton, Jun 15, 1931 (GH, MICH, TEX); TAYLOR: Tolstead 7038 (MICH, NY, SMU, TEX) ; TOM GREEN: Tweedy 110 (US); Travis: Birge 999 (TEX); WALLER: Hall 64 (F p.p.) ; WEBB: Salazar et al. 10 (OKLA); WISE: Shinners 26105 (SMU). NOTES ON THE FERN GENUS ERIOSORUS ALICE F. TRYON Eriosorus Fée was taken up by Copeland (Genera Fili- cum, Chronica Botanica, Waltham, Mass. 1947) for the older, much used name, Gymnogramma. Several of the species are transfered with the explanation, “both Gymno- gramma and Neurogramma have the same type species as Gymnopteris and are thus synonyms”. Gymnogramma is a synonym of Gymnopteris but it must pass from botanical usage more precisely because it is superfluous (Internat. Code Bot. Nomencl., Montreal 1961, Art. 63). Gymnogram- ma Desvaux (Ges. Naturf. Freunde Berlin Mag. 5: 304. 1811) included Acrostichum rufum L. ( Gymnogramma rufum (L.) Desv.). However, Bernhardi had previously based Gymnopteris (Jour. Bot. Schrader 1: 297. 1799) on Acrostichum rufum L.; thus Desvaux's name is illegitimate because it included the type of Bernhardi's earlier name. Since Gymnogramma is superfluous its type is the same as that of Gymnopteris (Art. 7, note 4). Kuhn (Fests. 50 Jub. Reals Berl. (Chaetop.) 1882) pro- posed the name Psilogramme for these species and he was followed by Underwood (The genus Gymnogramma of the Synopsis Filicum, Bull. Torrey Club 29: 617-634. 1902) and Maxon (North American species of Psilogramme, Bull. Torrey Club 42: 79-86. 1915) neither of whom referred to the earlier name, Eriosorus. Copeland considers Eriosorus a “very natural genus of about 35 species, all tropical American, mostly Andean” but includes only a partial list of 14 species. The following new combinations and new species are added to these. These new names are published here for use in a treatment of the genus in Peru in which the species are illustrated. Eriosorus Orbignyanus (Kuhn) comb. nov. Gymnogramma Orbignyana Mett. ex Kuhn, Linnaea 36: 70. 1869. Eriosorus Lechleri (Kuhn) comb. nov. Gymnogramma Lechleri Mett. ex Kuhn, Linnaea 36: 71. 1869. Eriosorus rufescens (Fée) comb. nov. 56 1963] Eriosorus — Tryon 57 Gymnogramma, rufescens Fée, Gen. Fil. 181, to. 19C, f. 3. 1852. Eriosorus Stuebelii (Hieron.) comb. nov. Gymnogramma, Stuebelii Hieron. Hedwigia 48: 219, t. 9, f. 5. 1909. Eriosorus accrescens A. F. Tryon, sp. nov. Rhizoma ignotum, folia subscandentia (vel pendentia ?), laminae elongato-lanceolatae vel elongato-ovatae, bipinnato-pinnatifidae vel tripinnatae, apex indeterminatus gemma tomentosa fulva vel stra- minea, rhachis modice flexuosa, pinnae deltoideae subcoriaceae petio- lulatae, pinnulae deltoideae vel ovatae adaxialiter et abaxialiter fulvo- sericeae, nervis marginem attnigentes ad terminos flabellatis, sporae obscure fuscae. TYPUS: PERU, Dept. Cuzco, Prov. URUBAMBA, Puyupata — “Yun- capata", C. Vargas 2921 (US); PARATYPI: DEPT. Cuzco, Huad- quia, Bües 992 (US), Valle de Lares, Montana de Colca, Biies 1925 (US), Altura de Chaco, Biies 2135 (US). Petiole probably shorter than the lamina (specimens incomplete), atropurpureus, terete, plane or slightly channeled on the upper surface near the apex, slender, less than 1/2 the diameter of the rachis, pubes- cent with tan, patent trichomes. Lamina 15-30 cm. long, 8-15 cm. wide. Rachis castaneus becoming lighter colored toward the apex, terete, plane or channeled on the upper surface, tomentose, the trichomes tan, acuminate. Pinnae stalks ca. 1.0 cm. long. Ultimate segments bluntly lobed to crenulate, somewhat more densely pubescent on the lower surface along the veins, the trichomes tan, multiseriate with acumi- nate apex. Margin with a border of clear, linear cells. Spores triangu- lar-globose, the equatorial wing broad, sometimes lobed, the 3 angles slightly protruding, the proximal face with ridges or papillae adjacent to the commissural ridges, the distal face with 3 contiguous ridges forming a triangle. These specimens from Cuzco, although incomplete, are quite distinct from other species in having leaves with a large, tan, tomentose apical bud and stalked pinnae with soft, tan pubescence. They most closely resemble E. aureo- nitens, especially in the form of the bud, but in that species the leaves are more slender and densely covered with a bright, rust colored tomentum. Several other species also have sustained growth from the leaf apex but have smaller buds. The habit of the leaves is difficult to determine from the specimens but is either somewhat climbing or hanging, for the very slender petiole could not support the lamina. All of the specimens are from the Urubamba valley, north of Cuzco, from 3000-3350 m. GRAY HERBARIUM, HARVARD UNIVERSITY. CONTRIBUTIONS TO AN ILLINOIS FLORA I. THE GENUS PHYSOSTEGIA ROBERT H. MOHLENBROCK Linnaeus’ (1753) Dracocephalum included species now known to belong to more than one genus. Two groups of Illinois plants belong here. One usually has been called Dra- cocephalum, the other Physostegia. Nearly a score of years ago, there was considerable effort on the part of some bota- nists to consider Linnaeus’ Dracocephalum virginianum the type for the genus. This would have caused all Physostegias to be placed in Dracocephalum, and would have required the other Dracocephalums to be known by Adanson’s revived name Moldavica. Gleason (1952) has followed this plan. Fortunately, to avoid confusion, Dracocephalum, based on D. moldavica L., has been conserved under provisions of Article 14 of the International Code of Botanical Nomencla- ture (1961) so that Physostegia may still be retained for American species generally known by that name in the past. Jones (1945, 1950) and Jones, Fuller, et al. (1955) con- sistently have recognized only two species in Illinois. They list Physostegia speciosa and P. virginiana. They relegate P. angustifolia to synonymy under P. virginiana. This writ- er during this study has found sufficient evidence to justify the maintenance of Fernald's P. angustifolia as a distinct species. Neither P. intermedia nor P. parviflora is treated in Illinois by Jones (1945, 1950) and Jones, Fuller, et al. (1955), although Gleason (1952) attributes the former to Illinois and Fernald (1950) the latter. Both species were found to be present in Illinois in this study. Physostegia, because of its exserted stamens, of which the upper pair is shorter than the lower, generally is classified in tribe Stachydeae of the subfamily Stachydoideae. Tlie genus is phylogenetically close to the monotypic Synandra, but this latter genus possesses one nearly suppressed calyx lobe and anthers which are brought into contact due to in- curving of the filaments. The name Synandra refers to the “united anthers.” 1963] Physostegia — Mohlenbrock 59 Various authors have accorded different treatments to the genus. Fernald (1950) recognizes 7 species and 2 varieties over the same area for which Gleason (1952) lists merely 5 species. Further confusion has been brought about by the transfer of species to the genus Dracocephalum. The most valid characters to be used in distinguishing species are leaf size and shape, leaf margin, bracteal devel- opment, and calyx structure. Less important characters are corolla size and color and length and arrangement of the inflorescence. LEAF SIZE AND SHAPE. Although size and shape of leaves are generally unreliable diagnostic characters, they seem to stand up well in the separation of the species of Physostegia. Size is particularly important in the segregation of P. angus- tifolia, for it is the only species which always has all its leaves 1 em. broad or less. Shape is invaluable in distin- guishing P. parviflora, for it is the only species with the uppermost leaves rounded at the base. Neither leaf length nor leaf apex offers any reliable characters. LEAF MARGIN. Physostegia virginiana, P. speciosa, and P. angustifolia have coarsely serrate margins in which the teeth are frequently incurved-acerose. The teeth of P. parvi- flora are generally small and not acerose, while the margin of P. intermedia is repand or undulate. Some deviation from these general patterns exists, but for the most part the char- acter of the leaf margin is useful. BRACTEAL DEVELOPMENT. Perhaps the most conspicuous difference among the species of Physostegia is bracteal de- velopment. In P. speciosa and P. parviflora there is a grad- ual transition from the larger lower cauline leaves to the upper cauline and bracteal leaves; in the other species, there is an abrupt demarcation in size between the lower and upper leaves. The bracteal leaves may be reduced without reduction in size of teeth and thereby appear pectinate. In P. virginiana, the bracts may be entire. CALYX STRUCTURE. Several characters of the calyx may be useful in distinguishing the species. The calyx is always densely glandular in P. virginiana, eglandular in the others. 60 Rhodora [Vol. 65 It ranges in length from 3-6 mm. in P. intermedia, to 5-10 mm. in P. virginiana. Considerable overlapping of calyx lengths occurs, however. The relative lengths of the calyx teeth and tube are significant. The teeth are nearly one-haif as long as the tube in P. intermedia, one-third as long in P. virginiana and P. parviflora, and two-fifths as long in P. speciosa and P. angustifolia. COROLLA SIZE AND COLOR. The corolla is usually 2-3 cm. long in P. virginiana and P. angustifolia and only slightly smaller in P. speciosa; in P. parviflora and P. intermedia, the size range is 1.0-1.7 cm. long. Color of the corolla varies from all shades of purple to nearly white. INFLORESCENCE. The inflorescence ranges in length from 5-35 cm. ; it is shortest in P. intermedia. The spikes may be slender and interrupted (P. intermedia and P. angustifolia) or more robust and continuous in the other species; they may be stiffly erect or arched-ascending. ECOLOGY AND DISTRIBUTION OF THE SPECIES The species of Physostegia in Illinois occur in moist open areas. Collectors usually refer to these habitats as swales or low prairies. Physostegia virginiana seems to show a some- what wider tolerance in the moisture requirement. The five Illinois species of Physostegia exhibit three gen- eral distribution patterns. Physostegia virginiana and P. speciosa are eastern species generally ranging from New England to the Dakotas; P. virginiana extends south to the Gulf of Mexico, while P. speciosa reaches only Missouri, Tennessee, and the Carolinas. Physostegia angustifolia and P. intermedia are midwestern, occupying an area roughly bounded by Kentucky, Kansas, Mississippi, and Texas. Phy- sostegia parviflora is northwestern, ranging from Minne- sota and Indiana to the Pacific Ocean. In Illinois, Physostegia virginiana and P. speciosa are locally abundant throughout the state. Physostegia angusti- folia is fairly common, but thus far confined to the northern three-fourths of Illinois. Our most southern records are from St. Clair and Marion Counties. Physostegia parviflora 1963] Physostegia — Mohlenbrock 61 and P. intermedia are rare, the former known from Adams County, the latter from Lake and Lee Counties. The distributional data recorded in the paper are the results of study of all material in the following herbaria: University of Illinois, Illinois Natural History Survey, Illi- nois State Museum, Missouri Botanical Garden, Southern Illinois University. SYSTEMATIC TREATMENT Physostegia Benth. Lab. Gen. & Sp. 504. 1834. Dracocephalum L. Sp. Pl. 594. 1753, pro parte, non nom. conserv. Rather stiffly erect, branched or unbranched, perennials with glab- rous stems and leaves; leaves alternate, simple, serrate, dentate, undu- late, or rarely nearly entire, usually becoming reduced near the inflorescence; inflorescence terminal, spicate, simple or compound; flowers large, showy, purplish to whitish; bracts small, each bearing one flower; calyx campanulate or tubular, regular, pubescent, 10- nerved, slightly enlarging in fruit, the lobes deltoid; corolla tubular, bilabiate, with a dilated throat, the upper lip erect, emarginate to entire, the lower lip spreading, 3-lobed; stamens 4, the lower pair longer and ascending under the upper lip of the corolla; nutlets smooth. Key to the Illinois Species of Physostegia 1. Leaves more or less undulate ................... eene 1. P. intermedia 1. Leaves definitely serrate. 2. Leaves (at least the upper) broadly rounded at base, the teeth rarely more than 1 mm. long; corolla rarely longer than 1.5 cm. Maul dur tu E a E EER 2. P. parviflora 2. Leaves cuneate or subcuneate at base, the teeth regularly more than 1 mm. long; corolla 1.5-3.0 em. long. 3. Upper leaves abruptly reduced in size; spike appearing pedun- culate. 4. Broadest leaves never exceeding a width of 1 cm.; flowers Non RM 3. P. angustifolia 4. At least some of the leaves over 1 cm. broad; some of the flowers overlapping .......ccccecsecccessccessssecenseseeeesees 4. P. virginiana 3. Upper leaves gradually reduced in size; spike appearing sessile ——— ———— RR 5. P. speciosa 1. Physostegia intermedia (Nutt.) Engelm. & Gray, in Boston Journ. Nat. Hist. 5:257. 1845. Dracocephalum intermedium Nutt. in Trans. Am. Philos. Soc. 5:187. 1837. Slender upright perennial to 1 m. tall; leaves sessile, subcoriaceous, lanceolate to linear-lanceolate, acute at apex, cuneate at base, with low 62 Rhodora [Vol. 65 teeth or undulate along the margins, the lower leaves 1.0-1.2 cm. broad, the upper much reduced; inflorescence spicate, terminal and occasionally lateral, slender, interrupted, to nearly 30 cm, long; calyx campanulate, 3.5-5.0 mm. long, the lobes at least one-third as long as the tube; corolla 1.2-1.7 cm. long, purplish to rarely white. Range. — Kentucky to Kansas south to Texas and Alabama. Habitat. — Low prairies. This is our only species with an undulate leaf margin. The drastic reduction in size of the upper leaves recalls P. virginiana and P. an- gustifolia. Fernald (1950) records this species from Illinois, while Gleason (1952) states "reported from Ill." Jones, Fuller, et al. (1955) regard these references as errors for P. virginiana, but consider that the real P. intermedia does not occur in Illinois. A single station for this spe- cies in Illinois is the edge of Willow Slough in Adams County. 2. Physostegia parviflora Nutt. ex Benth. in DC. Prod. 12:454. 1825. Dracocephalum nuttallii Britt. Ill. Fl. ed. 2. 3:117. 1913. Erect perennial to a little less than 1 m. tall; leaves sessile, subcor- iaceous, broadly lanceolate, acute to acuminate at apex, the upper broadly rounded at base, serrate, the lower 1.5-2.0 em. broad, the upper gradually reduced; inflorescence spicate, terminal and occasionally lateral, rather thick, densely flowered, to 15 em. long; calyx campanu- late, 4-7 mm. long, the lobes about one-third as long as the tube; corolla 1.0-1.5 em. long, purplish to whitish. Range. — Minnesota to British Columbia south to Oregon and Illinois. Habitat. — Low prairies. The broadly rounded bases of the upper leaves are unique among the Illinois species of Physostegia. The corolla is the shortest in the genus. The gradual reduction in size of leaves from base to summit of stem is similar to the condition in P. speciosa. Fernald (1950) attributes this species to Illinois, although Gleason (1952), Jones (1945, 1950), and Jones, Fuller, et al. (1955) do not record it. Authentie specimens of P. parviflora in Illinois herbaria are from Lake and Lee counties. 3. Physostegia angustifolia Fern. in Rhodora 45:462. 1943. Rather robust erect perennial to nearly 1 m. tall; leaves sessile, subcoriaceous, narrowly lanceolate, acute to acuminate at apex, cune- ate at base, serrate, the broadest about 1 em. wide, the upper abruptly reduced in size; inflorescence spicate, slender, usually solitary, occa- sionally with lateral branches, remotely flowered, to about 30 cm, long; calyx tubular-campanulate, 4-8 mm. long, the teeth about two-fifths as long as the tube; corolla 2.5-3.0 cm. long, purplish to whitish. Range. — Tennessee to Illinois south to Texas and Mississippi. Habi- tat. — Low prairies. Fernald (1943) has discussed the naming of this species. 1963] Physostegia — Mohlenbrock 63 While the general aspect of the plant indicates its relationship with P. virginiana, the consistently narrow leaves and the very interrupted inflorescence are sufficient reasons for maintaining this as a distinct species. The distribution of this species is local throughout the state, except for the southern counties where it is apparently absent. Distribution. — Calhoun, Cass, Cook, Ford, Greene, Hancock, Hen- derson, Iroquois, LaSalle, Livingston, McLean, Madison, Marion, Mason, Morgan, Moultrie, Piatt, Pike, Shelby, Tazewell, Vermilion, Wabash, Will. 4. Physostegia virginiana (L.) Benth. Lab. Gen. & Sp. 504. 1834. Dracocephalum virginianum L. Sp. Pl. 594. 1753. Physostegia virginiana f. candida Benke, in Am. Midl. Nat. 16:423. 1935. Rather stout perennial to about 1 m. tall; leaves sessile, subcoriace- ous, lanceolate, acute at apex, cuneate at base, sharply serrate, the teeth upwardly curved, the lower leaves 1.2-2.2 em. wide, the upper abruptly reduced, usually entire; inflorescence spicate, usually branched, rather thick, continuous, to 20 cm. long; calyx short- tubular, glandular and puberulent, 5-10 mm. long, the teeth about one-third as long as the tube; corolla 2.2-3.0 cm. long, purple to white. Range. — Maine to Oklahoma south to Texas and Alabama. Habi- tat. — Low ground, particularly prairies. Opinion varies concerning the specific limitation of this species. Most botanists segregate P. speciosa as a separate species, although Fernald (1950) chooses to consider the latter as a variety. The chief diagnostic characters for P. virginiana are the greatly reduced and usually entire upper leaves and the large and glandular calyx. White-flowered forms have been called f. candida. This species has a general distribution throughout Illinois. Distribution. — Champaign, Christian, Cook, DeKalb, DuPage, Hancock, Henderson, Henry, Iroquois, Jackson, Jersey, Johnson, Kan- kakee, Lawrence, Lee, Livingston, Macon, Marion, Mason, Monroe, Pope, Randolph, Richland, St. Clair, Saline, Stark, Union, Vermilion. 5. Physostegia speciosa (Sweet) Sweet, Hort. Brit., ed. 2, 406. 1830. Dracocephalum speciosum Sweet, Brit. Fl. Gard. pl. 93. 1825. Physostegia formosior Lunell, in Bull. Leeds Herb. 2:7. 1908. Dracocephalum formosius (Lunell) Rydb. in Brittonia 1:95. 1931. Robust upright perennial to nearly 1 m. tall; leaves sessile, sub- coriaceous, lanceolate to oblanceolate, acute at apex, cuneate at base, sharply serrate, the lower leaves 2-3 cm. broad, the upper only gradu- ally reduced; inflorescence spicate, rather stout, terminal and usually lateral, continuous, 15-25 cm. long; calyx campanulate, 4.5-7.0 cm. 64 Rhodora [Vol. 65 long, the lobes about two-fifths as long as the tube, eglandular; corolla 1.5-3.0 em. long, purplish to rarely white. Although this species is similar to P. virginiana, it may be distin- guished by its eglandular calyx and its gradually reduced bracteal leaves. The flowers are generally slightly smaller in P. speciosa. Range. — Maine to North Dakota south to Nebraska and North Caro- lina. Habitat. — Low prairies. Distribution. — Adams, Boone, Brown, Champaign, Cook, DeWitt, Greene, Henry, Jackson, Jersey, JoDaviess, Kankakee, Lake, LaSalle, Lawrence, Lee, McHenry, Macon, Marshall, Mason, Ogle, Peoria, Piatt, Pike, Putnam, Rock Island, Sangamon, Tazewell, Union, Ver- milion, Wabash, Warren, Winnebago, Woodford. DEPARTMENT OF BOTANY, SOUTHERN ILLINOIS UNIVERSITY, CARBONDALE, ILLINOIS, LITERATURE CITED FERNALD, M. L. 1943. Virginia Botanizing Under Restrictions. Rho- dora 45: 445-480. FERNALD, M.L. 1950. Gray’s Manual of Botany. Ed. 8. American Book Company, New York. GLEASON, H. A. 1952. The New Britton and Brown Illustrated Flora. Vol. 3. The New York Botanical Garden, New York. INTERNATIONAL CODE OF BOTANICAL NOMENCLATURE. 1961. Utrecht, Netherlands. Jones, G. N. 1945. Flora of Illinois. Ed. 1. The University of Notre Dame Press, Notre Dame, Indiana. 1950. Flora of Illinois. Ed. 2. The University of Notre Dame Press, Notre Dame, Indiana. , G. D. FULLER, G. S. WINTERRINGER, H. E. AHLES, AND A. FLYNN. 1955. Vascular Plants of Illinois. The Illinois State Museum and The University of Illinois, Springfield and Urbana. LINNAEUS, C. 1753. Species Plantarum. THE INTEGRADATION OF SENECIO PLATTENSIS AND SENECIO PAUPERCULUS IN WISCONSIN' T. M. BARKLEY Dr. Hugh Iltis of the University of Wisconsin recently invited me to contribute Senecio to his Flora of Wisconsin project. Preparation of the treatment has focused attention on the intergradation of S. plattensis Nutt. with S. pauper- culus Michx. var. pauperculus, (hereafter referred to as S. pauperculus) and the related difficulty in delimiting the two taxa. The problem of species delimitation in Senecio is treated generally in my revision of S. aureus L. and allied species (in press). This paper, however, will serve to discuss a specific instance and its taxonomic implications. Senecio plattensis and S. pauperculus are fairly distinct taxa, each with its own range and ecological requirements. However, where the ranges and habitats overlap, the two taxa intergrade morphologically. The greatest extent of intergradation is in the upper Mississippi Valley from north- ern Missouri through eastern Iowa to Wisconsin and Min- nesota. Typical S. plattensis is distinguished by its persistent pubescence, its single-stemmed growth habit, its short erect caudex, its well developed pinnatifid lower cauline leaves, and its relatively compact inflorescence. It is fundamentally a plant of the prairies and plains of central North America. It also occurs in relict prairie areas in localities disjunct from the main distribution. Biologically typical S. pauperculus is distinguished by being glabrous or nearly so at maturity, by its frequent pro- duction of several loosely clustered stems, its short but often horizontal, branching caudex, its reduced cauline leaves and its generally loose inflorescence. It grows primarily in bor- eal woodlands, meadows, and open areas in woodland asso- ciations. TSupported in part by a grant from the Kansas State University Bureau of General Research. 65 66 Rhodora [Vol. 65 Each taxon is further distinguished by a series of more or less intangible features which collectively give it a charac- teristic gross aspect. In Wisconsin, typical S. plattensis is most abundant in the south western half of the state, i.e., in the areas primarily of gray-brown podzol soils. S. pauperculus occurs throughout the state, but it is most abundant and “most typical" in the northern half, in the areas of true podzol soils, (cf. soil map in U. S. Dept. Agr. Yearbook, 1938). Populations occur throughout Wisconsin which are more or less intermediate between S. plattensis and S. pauper- culus. These intermediate populations are most frequent in and near northern Washburn county in northwestern Wis- consin, and in Dane, Iowa, Green, and Rock counties at the southern edge of the state. The intermediates commonly resemble either S. plattensis or S. pauperculus rather closely, but they have conspicuous tendencies toward the other taxon. Two frequent "intermediates" in Wisconsin are: (a) otherwise typical S. pauperculus, but with light, persistent tomentum, and (b) fairly typical S. plattensis but with strongly reduced cauline leaves and a branching caudex. Plants which are exactly midway between the two taxa are rather uncommon. In North America the genus Senecio appears to be com- posed of numerous more or less discrete taxa which have broad but definite natural ranges. These intergrade mor- phologically with related taxa wherever their ranges and habitats overlap. It is one thing to recognize the biological patterns within the genus, but quite another to reflect these patterns in the taxonomic system. If there were no senecios other than S. plattensis and S. pauperculus, the taxonomic treatment could be simple; the two entities could be recog- nized as infraspecific taxa of one species. However, in other localities, both S. plattensis and S. pauperculus intergrade just as completely with other taxa as they do with each other in the upper Mississippi Valley. S. plattensis, for example, intergrades with both S. obovatus Muhl. ex Willd. and S. tomentosus Michx., while S. pauperculus intergrades with S. streptanthifolius Greene and S. smallii Britt. 1963] Senecio — Barkley 67 Combining S. plattensis and S. pauperculus into a single species could not stop there; it would be necessary to include S. obovatus, S. tomentosus, S. smallii, and S. streptanthifoli- us, plus the other taxa with which these in turn intergrade. Continued combining of all the entities which intergrade through S. streptanthif olius would be an almost limitless process, eventually taking much of what is included in Ryd- berg's sections Aurei, Lobati, and Tomentosi. The one result- ing species would contain so many and such diverse entities, and would encompass so much variation as to be ridiculous. Therefore, S. plattensis and S. pauperculus are recognized as separate species. The necessity for maintaining these entities as species becomes apparent only when all of the related senecios are considered. This is of small consolation to the floristic botanist working in the upper Mississippi Valley area, and one can sympathize with Davidson, who reduced S. plattensis to S. pauperculus in his flora of southeastern Iowa (1959). __ DEPARTMENT OF BOTANY, KANSAS STATE UNIVERSITY, MAN- HATTAN, KANSAS. LITERATURE CITED BARKLEY, T. M. (in press). A Revision of Senecio aureus Linn. and Allied Species. Trans. Kan. Acad. Davipson, R. A. 1959. The Vascular Flora of Southeastern Iowa. State Univ. Iowa, Stud. in Natural History XX (2) : 1-102. U. S. Dept. Agr. 1938. Yearbook of Agriculture, Soils and Men. Washington. THE GEOGRAPHIC LOCATION OF GORGONIDIUM (ARACEAE) DAN H. NICOLSON! The type-species of the monotypic genus Gorgonidium, G. mirabile Schott, until just recently was believed to have been collected by Gaudichaud in the Papuan area, but has now been found to have been collected by d’Orbigny in Bolivia. The type-material of G. mirabile (Schott, 1864) was cited as being from “Carari, Sualica (?)", the query reflecting Schott's uncertainty as to the locality. Schott credited the collection to Gaudichaud and cited the field description, "Arum à fleur violette." The type was stated to be at Leiden. At Leiden, the holotype was found, with the following information which Schott had not reported. The specimen is labelled as being from “Herb. Gaudichaud" (which does not mean that it was collected by Gaudichaud). A collection number, 484, appears on the label that bears the field de- scription. Engler (1920), the last monographer of the Araceae, re- ported, without any explanation, that this “Gaudichaud” collection was from “Monsungebiet. — Papuasische Pro- vinz: Insel Carari im Mare sualicum." There is a Cavili Island in the Sulu Sea, but that is between the Philippine Islands and Borneo, not in the Papuan area, and there is no evidence that Gaudichaud visited it. Members of several institutions (Leiden, Kew, British Museum) agreed that something was wrong with this locality. The solution of the type-locality problem came when the present author noted that an isotype of Spathantheum or- bignyanum Schott (1859) at Leiden had a field label saying "Arum alique verte de Cavari, Sicalica," in handwriting ^ "This problem was investigated while the author held an inermediate graduate fellowship, No. 21024, from the U.S. National Science Foundation. The author wishes to thank Drs. A. Lourteig (Paris), H. Sleumer (Leiden), R. Foster (Harvard) and W. Dress (Cornell) who contributed much of their personal time and enthusiasm to this problem. 68 1963] Gorgonidium — Nicolson 69 identical with that on the field label of the G. mirabile holo- type. The collector's name, d'Orbigny, which was the basis of the specific epithet, was not to be found on the specimen, although it, too, was noted as being from ''Herb. Gaudi- chaud." Dr. Alicia Lourteig, of the Paris Museum of Natural His- tory, kindly investigated the “Gaudichaud” and d'Orbigny herbaria at Paris. She reported that there is not, and never has been, an officially recognized Gaudichaud herbarium, and that the materials so labelled at Leiden probably only record that they were received from Gaudichaud. However, the d'Orbigny herbarium, never before investigated for its iso- types of Araceae, produced isotypes of both S. orbignyanum (holotype at Geneva) and G. mirabile (holotype at Leiden) and yielded the following complete information from the original field labels, written in the hand of d'Orbigny. G. mirabile — Cavari, Sicasica Prov., Bolivia — d'Orbigny 484 —Arum à fleur violette. S. orbignyanum — Cavari, Sicasica Prov., Bolivia — d'Orbigny 485 — Arum à tige verte. In the early days of the Rijksherbarium (Leiden) and in other herbaria, there was a good deal of label-copying and sometimes the original labels were discarded. For instance, the same handwriting on the field labels of the Leiden types of G. mirabile and S. orbignyanum may be seen on the holo- type of Xenophya branceaefolia Schott, which was collected by Zippel in New Guinea. However, in the case of the latter, the original field labels are present. The change from *Sicasica" to “Sualica” in handwriting can best be seen in two steps. First, changing it "ic" to “u” happens when the dot over the “i” is omitted and the top of the “c” is not rounded. Second, “sica” to “lica” is explicable in terms of the old-style internal “s” that looked more like an uncrossed “f”. This internal “s” can easily be misread as an “l” in handwriting. A copying clerk, unfamiliar with the place-names, might easily have read “Sualica” (viz. Gorgo- nidium) or “Sicalica” (viz. Spathantheum) where “Sicasi- ca” was originally written. 70 Rhodora [Vol. 65 It is probable that d'Orbigny gave some of his duplicates to Gaudichaud, who, in turn, gave them to Leiden. The original d'Orbigny field labels (Paris) do not bear the col- lector's name. In the absence of this, they were recorded at Leiden as being from *Herb. Gaudichaud". Both Scott and Engler, not knowing of the original duplicates at Paris, in- correctly interpreted this to mean that Gaudichaud had collected the holotype of G. mirabile. Schott (died 1865) had 3282 color and black and white drawings of Araceae prepared, for which he himself paid. This set of drawings (of which less than 200 have been pub- lished) is deposited at the Natural History Museum, Vienna. The unpublished folio drawing numbered “Icon. Herb. Palat. Vindob./Schott Aroideae Nr. 1990" was drawn by Nickelli from the Leiden holotype of Gorgonidium mirabile. Engler (1920, Pl. 2, fig. F-H) illustrates a pistillate flower, a stamen, and a staminate flower. Figure F is copied from the Schottian drawing. Figures G and H are highly stylized, but were undoubtedly based on the Schottian draw- ing, rather than on the holotype. The illustration published here includes fourteen figures which were copied by Miss Mitsu Nakayama from photo- graphs of Schott's Nr. 1990, which contains twenty-seven figures. The author wishes to thank Prof. Dr. K. H. Rechin- ger, Director of the Natural History Museum in Vienna for permission to publish these figures. The following description of Gorgonidium mirabile is syn- thesized from the original description (Schott, 1864), Eng- ler's description. (1920), and photographs (taken by the author) of the holotype at Leiden and the Schottian drawing Nr. 1990 at Vienna. Leaf unknown. Peduncle incomplete. Spathe purple, about 20 cm. long and 5 cm. wide, open almost to the base. Spadix purple, approxi- mately 14 cm. long; stipe 1.7 cm. long, pistillate portion 1.8 cm. long, and staminate portion 11.5 cm. long. Pistillate flowers more or less whorled, each surrounded by 6-8 irregularly arranged staminodia; ovary 4-locular, locules 1l-ovular; ovule orthotropous; style filiform; stigma 4-lobed. Staminate flowers contiguous with the pistillate, with 6-8 stamens; lower flowers with free stamens, often irregularly ar- 1963] Gorgonidium — Nicolson 7l Figs. 1-4b Gorgonidium mirabile Schott. 1. Inflorescence with one half of the spathe removed. 2. Details from lower (pistillate) portion of the inflorescence: 2a. Several pistillate flowers and their staminodia; 2b. Single pistillate flower and its staminodia : 2c. Longitudinal section of pistil; 2d. Longitudinal section of an ovule; 2e. Cross- 72 Rhodora [Vol. 65 ranged; upper flowers with stamens more or less connate by their filaments; stamens with long filaments, much longer than the spher- oidal anthers, these paired, stipitate below the free end of the con- nective, and dehiscent by a terminal pore. SUMMARY The type-specimen of the type-species of Gorgonidium Schott has hitherto been considered as having been collected by Gaudichaud in the Papuan area. In fact, however, it was collected by d’Orbigny in Bolivia. The error developed from the omission of d’Orbigny’s name on the field labels, the dis- tribution of duplicates by Gaudichaud, and a clerk’s miscopy- ing of ''Sicasica" (a province of Bolivia) as *Sualica", which was thought to be a sea in the Papuan area. The solution was found by a study of previously unrecognized isotypes of Spathantheum orbignyanum Schott and Gorgonidium mir- abile Schott in the Paris Museum of Natural History. — L. H. BAILEY HORTORIUM, CORNELL UNIVERSITY. LITERATURE CITED ENGLER, A. 1920. Das Pflanzenreich, IV, Fam. 23F : 52. ScHoTT, H. W. 1859. Spathantheum orbignyanum Schott in Bon- plandia 7:164-165. . 1864. Gorgonidium mirabile Schott in Ann. Mus. Bot. Lugd.-Bat. 1: 282. section of the ovary; 2f. Top view of stigma; 2g. Side view of stigma. 3. Details from central (staminate) portion of inflorescence. 3a. Several staminate flowers; 3b. Single stamen; 3c. Longitudinal section of immature anther; 3d. Longitudinal section of anther at anthesis. 4. Details from upper (staminate) portion of inflorescence: 4a. Several staminate flowers; 4b. Single staminate flower. NEW PLANTS IN OLD PLACES FRANK C. SEYMOUR Justification for this title lies in the fact that the plants listed in this article are mostly new, so far as our knowledge goes, to the localities in which they were collected. Old places refers to the fact that their localities are mostly areas which have been extensively botanized previously. Historically, Bristol County, Massachusetts, from which most of the records come is one of the earliest regions in America to be settled by white men. Botanically, while not one of the earliest areas to be explored, it has been the scene of collecting activity for nearly a century, at least. The earliest collectors in this county, so far as has been ascer- tained, are George Mackie who collected in 1872 in New Bedford, and S. M. Holman who collected in 1880 in Attle- boro. A few specimens from each of these are in the Her- barium of the New England Botanical Club. The earliest botanist who collected a considerable number of specimens was Carl Blomberg. His specimens (NEBC) are from Easton and range in date from 1890 to 1902. Much better known as a botanist is Eliphalet Williams Hervey, who appears to have begun collecting at about the date when Blomberg's collecting ceased. Although many of his specimens in the Club Herbarium are not dated, from those which are dated, we learn that he collected in New Bedford and vicinity during the years 1899 to 1916. His “Flora of New Bedford" (the 3rd edition, revised, published in 1911) seems to have marked the climax of his botanical investigations in Bristol County. Other botanists whose specimens are in the Herbarium of the Club include S. N. F. Sanford. His specimens are from various towns in this County during the years 1903 to 1929 and probably later. Many others, — outstandingly F. F. Forbes, F. W. Hunnewell and C. H. Knowlton, — have con- tributed extensively to our knowledge of this fruitful region. In spite of so much botanical activity, Dighton seems to 73 74 Rhodora [Vol. 65 have been almost entirely overlooked, perhaps because it is on the other side of Taunton River from New Bedford. Not many bridges cross that river. During the present writer's eight years of residence in Dighton, his botanical eye picked up a few plants which seem worthy of note. Of the 47 species listed below, 40 are from Bristol County, Massachusetts. 6 of the 40 have been reported before but are mentioned for reasons given below. It is surprising that of the 34 believed not to have been reported before, only 20 are introduced while as many as 14 are native. Thus only 59% were probably not there in Hervey’s day. Of the 20 introduced species not previously reported, 6 or 30% are introduced from elsewhere in America while 14 or 70% are introduced from Europe or Asia. The list below is divided into groups to represent these aspects. Usually the number of the collection is cited so that in the future the identifica- tion may be verified or, if necessary, corrected. Many of the specimens here mentioned are to be found in the herbarium of the New England Botanical Club and in the writer's private herbarium. Some are also in the her- barium of Southern Methodist University. The writer wishes to thank Dr. Reed C. Rollins, Director of the Gray Herbari- um, and Dr. Richard A. Howard, Director of the Arnold Arboretum, for the privilege of referring to those respective herbaria. I. BRISTOL COUNTY, MASSACHUSETTS, introduced species believed not to have been previously reported. 1. Picea Abies (L.) Karst. A few young trees surrounded by woods, removed from any building, appear to have been self- sown. Dighton, 1957, Seymour 17,072. Native of Europe. 2. Pinus resinosa Ait. In some cases, this tree has been planted, but in a number of other cases, a group of young trees are irregularly arranged and show no signs of human agency, as in the case of Seymour 16,507, Dighton, 1956. Native as near as northern Mass. 3. Eragrostis multicaulis Steud. Plentiful along a roadside in Dighton, 1961, Seymour 19,383. Not nearly so frequent in the writer's observation as the very similar E. pectinacea (Michx.) Nees. Native of e. Asia. 4. Calamagrostis epigejos (L.) Roth, var. georgica (K. Koch) 1963] New Plants — Seymour 75 Ledeb. Several plants of this grass, as yet rare in New England, were in a neglected field near buildings in Dighton, 1957, Seymour 17,752. Native of s. Russia. 5. Tradescantia virginiana L. Hardy in cultivation and persisting in more than one locality where it has escaped, as by a roadside, Dighton, 1961, Seymour 19,279a. Native as near as Conn. 6. Salix babylonica L. A single tree in an open field with no evi- dence of cultivation. Taunton, 1961, Seymour 19,146; 19,240, both from the same tree at different seasons. Native of Eurasia. 7. Polygonum cuspidatum Sieb. & Zucc. Large colonies in several spots in and near Taunton, 1961, Seymour 19,528. Native of e. Asia. 8. Polygonum sachalinense F. Schmidt. Only one colony known. Near site of burned building, Dighton, 1961, Seymour 19,527. Native of e. Asia. 9. Rosa gallica L. Several bushes in a vacant lot near woods, Dighton, 1961, Seymour 19,232. Native of Europe. 10. Lotus corniculatus L. First seen by the writer in Topsham, Vt., 1959. Abundant on a hillside and adjacent road, Seymour 18,180. New to Vt. The following year, it appeared in consid- erable quantity along a roadside in Dighton, 1960. Seymour 18,566. First collection in southeastern Mass. In 1961, collected along a roadside, Warren, R. I. Seymour 19,448. New to Bristol Co., R. I. Native of Europe. 11. Lathyrus latifolius L. Along a roadside, Dighton, 1961. Catherine M. Seymour 19,471a. Cultivated and escaping. Thor- oughly established. Native of Europe. 12. Erodium cicutarium (L.) L’Her. Many plants as weeds in a cultivated field, Dighton, 1961. Wyman E. Hawkes. Native of Europe. 13. Catalpa speciosa Warder. Frequently escaped from cultiva- tion. Small trees seen in various places. Collection from a small tree with several others near a large tree along a roadside, Taun- ton, 1961. Seymour 19,311. A group of small trees in a similar situation. but not near any large tree, Raynham, 1961, Seymour 19,330, Native as near as Tenn. 14. Catalpa bignonioides Walt. Apparently less frequent than C. speciosa. Along a roadside, Easton, 1961, Seymour 19,575. Native farther south. 15. Plantago lanceolata L., var. sphaerostachya Mert. & Koch. Roadside, Taunton, 1961, Seymour 19,242. This variety new to Bristol Co., Mass. Native of Europe. 16. Silphium perfoliatum L. A large colony growing luxuriantly in a vacant lot, in moist soil; observed for several years; Taunton, 1961, Seymour 19,530. Native as near as Ont. 76 Rhodora [Vol. 65 17. Helenium flexuosum Raf. (H. nudiflorum Nutt.). Many plants in a meadow along upper tidal part of Taunton River, Taunton, 1956, Seymour 16,991. Native as near as N. C. 18. Hieracium Pilosella L. A few plants in an open field far from human habitation, Dighton, 1957, Seymour 17,081. Observed in abundance on more than one lawn in Taunton, 1962. It is desirable to record with specimens the first date when plants like this, which are spreading, reach each locality so that the rate and direction of spread may be plotted. Native of Europe. 19. Hieracium pratense Tausch. A few plants, Dighton, 1961, Seymour 19,231. Glocester, R. L, 1961, with Richard J. Eaton, along roadside, (Previously reported only from Smithfield, R. I.) Seymour 19,225. Native of Europe- 20. Hieracium vulgatum Fries. Three plants on a lawn where it certainly did not occur the year before, Dighton, 1961, Seymour 19,473. Native of Europe. Il. BRISTOL COUNTY, MASSACHUSETTS, native species believed not to have been reported previously. 21. Botrychium matricariaefolium A. Br. A few plants in shade of young trees and shrubs which were about 1.5 m tall in a neglected field with Orobanche uniflora L. Seymour 19,741. 22. Zostera marina L., var. stenophylla Aschers. & Graebn. Washed up on shore, Horseneck Beach, Westport, Mass., Seymour 19,307. 23. Potamogeton gemmiparus Robbins. A few plants floating free in Barstow Pond, Taunton, Mass., Seymour 19,516. 24. Potamogeton alpinus Balbis, var. subellipticus (Fern.) Ogden. In rapid current, tidal part of Segreganset River, Dighton, 1957, 1961 (with P. epihydrus Raf., var. Nuttallii (C. & S.) Fern.). Seymour 17,182; 19,481. 25. Spartina X caespitosa A. A. Eaton. Several plants in salt marsh along Taunton River, Dighton, 1961, Seymour 19,384. 26. Zizania aquatica L., var. aquatica. Sizable colony on tidal shore of Segreganset River, Dighton, (near Potamogeton alpinus, var. subellipticus) 1957, Seymour 17,183. 27. Scirpus robustus Pursh. A few plants at edge of tidal pool along Taunton River, Dighton, with S. paludosus Nels., var. atlanticus Fern., 1959, Seymour 18,393. 28. Carex Deweyana Schwein. Occasional in damp woods, Digh- ton, 1956, Seymour 16,512. 29. Carex abscondita MacM. In rich, usually damp woods. Wide- spread in Dighton and vicinity. In 1916, when the writer was studying the flora of Martha’s Vineyard, Mass., it was surprising that although this species occurred in R. I. and Martha’s Vineyard, it was unknown in Bristol Co., Mass. Now it is found not only [1963 New Plants — Seymour 77 III. IV. to occur in the latter county, as was to be expected, but it is widespread and not uncommon. Dighton, 1957. Raynham, 1960. Rehoboth, 1961. Taunton, 1961. Bristol, R. L, 1961. Seymour 17,122; 18,569; 19,227; 19,251; 19,460; 19,493. 30. Carya ovata (Mill) K. Koch. Common in Dighton, 1955. From several collections, two are cited from the same tree at different seasons. Seymour 17,144; 17,679. 31. Betula nigra L. Several young trees along upper tidal part of Taunton River, Taunton, 1955. The trees are irregularly arranged on uncultivated land and appear to be self-sown. Seymour 16,459; 16,497. 32. Hepatica americana (DC.) Ker. Several plants in thin decidu- ous woods; known only in this one spot, Dighton, 1959. Guided to the spot by Elliott Walker. Seymour 18,492; 18,496. 33. Prunus americana Marsh. Occasional in Dighton, 1961. Along a hedgerow, Seymour 19,230. 34. Callitriche verna L. (C. palustris L.) Along Terry Brook, Assonet, Freetown, Mass., Seymour 19,107. BRISTOL COUNTY, MASSACHUSETTS, introduced species previously reported. 35. Thuja occidentalis L. Much cultivated in this county, it is escaping and becoming established. Two medium sized trees not near any building and surrounded by woods show no sign of having been planted. Berkley, Mass., 1961, Seymour 19,128. Collected also by Hervey in New Bedford. Native in western New England. BRISTOL COUNTY, MASSACHUSETTS, native species pre- viously reported. 36. Phragmites communis Trin., var. Berlandieri (Fourn.) Fern. Small colony of a dozen plants in swampy land recently cleared for gas pipe line, Dighton, with Lycopodium inundatum L., var. inundatum, 1961. Not previously observed in this spot although the writer lived within one quarter mile for eight years. He had collected in that spot the year before. Was it there all the time? Was it self-planted after the land was cleared? Was it growing there but stunted until the land was cleared? Any of these is possible; Seymour 19,529. Listed by Hervey from New Bedford, but no specimen seen. 37. Carex Emmonsii Dew. Common in Dighton and vicinity, 1956, Seymour 16,513; 17,078; 18,501. There is only one other record from Bristol Co.; North Easton, A. A. Eaton, 1903. 398. Trillium erectum. L. A number of plants in swampy woods of hemlock, Tsuga canadensis (L.) Carr., 1960. Reported earlier by Wilfred A. Hathaway but no specimen seen previously. Sey- mour 18,509, 78 V. VI. Rhodora [Vol. 65 39. Rhododendron canadense (L.) Torr. Border of swamp, one good-sized clump, Raynham, Mass., 1961, Seymour 19,180. Dis- covered by William N. Andrews, Sr., who directed the writer to the spot. Rare in this county; previously collected only in Attle- boro. 40. Physalis pubescens L., var. grisea Waterfall. One large plant about 1.5 m tall along a roadside, Dighton, 1959. Not found the following year. Seymour 18,488. In this county, previously col- lected only by Hervey in New Bedford. MISCELLANEOUS collections of introduced species from outside Bristol County, Mass. 41. Nymphaea tuberosa Paine. Large root picked up, floating near shore, Winter Pond, Winchester, Mass., 1962, with Richard J. Eaton. The main root bears several small fusiform tubers. Seymour 19,702. Not previously reported from Mass. Native as near as Vt. and southwestern Conn. 42. Erysimum inconspicuum (S. Wats.) MacM. A single plant in a vacant lot, formerly a poultry farm, Stoughton, 1961, Sey- mour 19,310. New to Mass. Native as near as Ont. MISCELLANEOUS collections of native species from outside Bristol Co., Mass. 43. Potamogeton nodosus Poir. A few plants with P. gramineus L. in Pine River, Ossipee, N. H., Seymour 19,840. New to N. H. 44. Trisetum melicoides (Michx.) Vasey, var. majus (Gray) Hitche. Only one plant seen; Lake Morley, Fairlee, Vt., 1959 Seymour 18,216. Reported from northern Vt. in Gray's Man., ed. 8. Rare. 45. Sphenopholis nitida (Biehler) Scribn. Scarce; in woods near crags near Asplenium montanum Willd., Glocester, Providence Co., R. I. with Richard J. Eaton. Seymour 19,205. Collected in Cum- berland, R. L, 1875, and 1878; not collected in this county since. 46. Rhynchospora capillacea Torr. Numerous plants in shallow water of shore of Robbins Cove, Berry Bay, Freedom, N. H., with Ranunculus reptans L., 1958, Seymour 17,887. New to N. H. 47. Viola pubescens Ait. Several plants in two different spots in one of the few remaining relatively unspoiled patches of rich woods in Brockton, Plymouth Co., Mass., 1962, Seymour 19,801. Not previously reported from Plymouth Co., Mass. STOUGHTON, MASS. 1963] Passiflora — Lloyd 79 TETRAPLOID PASSIFLORA INCARNATA IN NORTH CAROLINA. — The genus Passiflora L. in North Carolina is represented by two relatively distinct species. Passiflora lutea L., the yellow passion-flower, is infrequent and occurs in more mesic environments than the common P. incarnata L., the passion-flower or maypops (Blomquist and Oosting, 1959). The latter species has been seen on road banks or railroad banks in dry soils harboring little other vegetation and pro- viding rapid runoff of surface water after summer showers. These are often fully exposed to sunlight throughout the day. Specimens of P. incarnata were collected in July, 1962, one mile north of U. S. Highway 15-501 & 70 bypass on Duke Street, Durham Co., Durham, North Carolina (Lloyd 1001, Duke University Herbarium). Buds were opened and fixed in 1:3 acetic alcohol at 11:00 AM, and later squashed in acetocarmine. Microsporogenesis of sporocytes showed 18 bivalents at metaphase 1. The finding of tetraploid P. incarnata coupled with pre- vious counts of n — 9 by Storey (1950) and Bowden (1945) establishes the presence of polyploidy in this species. In view of the lack of quadrivalents in the plants I examined they may have been of allotetraploid origin. Diploid (Bald- win, 1949) and tetraploid (Bowden, 1945) strains of P. lutea also occur and both species should be investigated fur- ther to determine the origin and distribution of their includ- ed polyploid races, which may have some application to horticulture. Special thanks are due the staff of Duke University for the use of their facilities and Dr. Peter Raven for his com- ments in the preparation of the manuscript. —ROBERT M. LLOYD, DEPARTMENT OF BOTANY, DUKE UNIVERSITY, DURHAM, NORTH CAROLINA. ! LITERATURE CITED BALDWIN, J. T., JR. 1949. Polyploidy in Passiflora lutea L. Rhodora 51: 29. BLoMQuisT, H. L. and H. J. OosriNG. 1959. A Guide to the Spring 1lPresent address: Dept. of Botany, University of California, Los Angeles 24, California. 80 Rhodora Vol. 65 and Early Summer Flora of the Piedmont, North Carolina. 6th. revised ed. Published by the authors. Durham. BOWDEN, WRays, M. 1945. A list of chromosome numbers in higher plants, II. Menispermaceae to Verbenaceae. Am. Journ. Bot. 32: 191-201. STOREY, W. B. 1950. Chromosome numbers of some species of Passiflora occurring in Hawaii. Pac. Sci. 4: 37-42. A RECENT ILLINOIS COLLECTION OF SANGUISORBA CANA- DENSIS L. — The last definite date for a specimen of Ameri- can burnet, Sanguisorba canadensis L., in Illinois is fifty-five years ago. This material was collected by E. J. Hill, Sept. 25, 1907, moist prairie by R. I. Railroad, Joliet, Will County. Previous collections by H. L. Boltwood and A. B. Seymour are known from Ottawa, LaSalle County, dated 1880 and 1882. A statement in VASCULAR PLANTS OF ILLINOIS (Jones, Fuller et al, 1955) says: “Moist ground, very rare, or possibly extinct in Illinois.” An S. B. Mead specimen is cited in the same publication: “Cass Co.: Beardstown, Aug. 27 — ?". In view of the interval since the last collection of this species and in connection with the questionable date of the Cass County specimen, the following information seems pertinent. Illinois is perhaps near the southwestern limit of distri- bution of S. canadensis L., and a recent collection by Mr. R. T. Rexroat of Virginia, Illinois, will verify the continued occurrence of this species in the state. Collection data are as follows: Cass Co., Illinois, south of Beardstown, once swampy area (moist only when the specimen was collected), Sept. 15, 1962, R. T. Rexroat 8577, (ISM). Mr. Rexroat, an amateur botanist and collector, has dem- onstrated the value of such collectors and their contributions to a continued study of the flora of Illinois. Since 1953 Mr. Rexroat has collected eight species of vascular plants previ- ously not recorded for Illinois and six of these are from Cass County. The following list is furnished although many of them were reported in previous issues of RHODORA. Cass 1963] Passiflora — Lloyd 81 County: Fimbristylis vahlii (Lam.) Link, Scleria reticularis Michx., Lipocarpha maculata (Michx.) Torr., Cyperus grayioides Mohlenbrock, Anagallis arvense forma caerulea (Schreb.) Baumg., and Arctium tomentosum Mill. Morgan County: Eragrostis curvula (Schrad.) Nees, and from Schuyler County, Falcaria sioides (Wibel) Aschers. Mr. Rexroat’s work is an example of what careful and continued collecting can accomplish in a state which has been consid- ered well botanized during a century and a half. In spite of opinions of some professional botanists, who regard further collecting in Illinois unnecessary, the effort seems worth- while. — GLEN S. WINTERRINGER, ILLINOIS STATE MUSEUM, SPRINGFIELD. THE FLORA OF THE WOLF ISLANDS NEW BRUNSWICK* PART I RADCLIFFE B. PIKE AND ALBION R. HODGDON Dwarfed by the nearby Grand Manan Archipelago, the Wolf Islands have received scant attention since their dis- covery in 1603 by Champlain who gave them the picturesque name, “Les Isles des Perdreaux". The “Wolves”, as these islands are known, have been neglected for excellent reasons; they are both literally and figuratively overshadowed by Grand Manan which has long attracted visitors including botanists; they are so situated in the hazardous waters of the Bay of Fundy as to be un- attractive to the usual run of visitors, except fishermen and rare yachtsmen. The islands are not remote; North Head on Grand Manan is about 12 miles to the south, Campobello Island is 10 miles to the west and the mainland of New Brunswick not more than 8 miles away to the north. The most important factor in their isolation has probably been the nature of the Bay of Fundy itself, its extreme tidal currents and notorious meteorological conditions coupled with the rugged shore lines of the islands themselves. Of all the accessible islands that lie off the North Atlantic Coast, these are the last to have yielded to the botanist. It * This research is part of a project entitled “Floristie and Phyto- geographic Investigations of the Wolf Islands and other islands in the Bay of Fundy" which was supported by a grant from the Central University Research Fund of the Graduate School of the University of New Hampshire, and the Society of the Sigma Xi. Published with the approval of the Director of the New Hampshire Agricultural Experiment Station as Scientific Contribution No. 307. * Collections of all taxa except Lycopodium clavatum were made at least once and often for most of the individual islands, but in the case of common and well-defined species, the records are based in major part of field observations made in July 1962, Specimens are deposited in the University of New Hampshire Herbarium. 82 1963] Wolf Islands — Pike and Hodgdon 83 East WSF Gult GG Reek Fut Wel Fat pot 7 South We(f Fundy cz li mile Woff Islands a 5 ap Fig. 1. Map of Wolf Islands, Pennfeld Parish, Charlotte Co., New Brunswick, Canada. is doubtful indeed if there is anywhere a comparably neglected insular area so near a highly developed coast line. It is the usual fate of islands in inhabited regions to be ravaged by man. This is true of the Grand Manan group, of Machias Seal and other islands that we have visited in the Bay of Fundy. But the Wolves, to a very considerable extent, have been preserved in nearly their original state. Of the 5 islands (see map), Fatpot and Flat Wolf seem to have been totally undisturbed, also South Wolf in major part and a very considerable portion of East Wolf. On South Wolf the Canadian Government maintains a lighthouse, the 84 Rhodora [Vol. 65 keeper of which is the only year-round inhabitant. The area immediately around the lighthouse and a mere fringe of meadow on the western side of the island show the effects of man. Also there has been irregular summer occupancy of the Wolves by fishermen for a long time. The Hawkins family of Beaver Harbor, New Brunswick, the present owners of East Wolf, have been very helpful in giving us information about the human history of the islands and also have pointed out several interesting species of plants. The occupancy of East Wolf has been sporadic over the years depending on the supply and demand for herring. Cows, pigs and chickens have been kept on the island during the summer fishing season at times, but no sheep were intro- duced. Thus the destruction of native vegetation has been confined to a few acres in the vicinity of the camps on Paul's Cove on the western side of the island. As is customary on many islands, rabbits have been introduced repeatedly to provide emergency food. These creatures were exces- sively abundant in 1959 on East Wolf but had declined almost to the point of disappearance by 1962. A great fire set by visiting fishermen many years ago burned the eastern end of East Wolf. Large areas nearer the center of East Wolf also were burned off or cleared many years ago pro- viding at present a savannah-like aspect with scattered large trees mostly of Betula papyrifera var. cordifolia. The dominant forest trees, Picea glauca and Abies balsamea, seem to be very slow or quite ineffective in regenerating these cleared areas, many of which are now carpeted with almost unbroken mats of Cornus canadensis. Despite these alterations and undoubtedly the cutting of trees at times, the impression one forms after visiting any one of the Wolf Islands is of essentially wild and unspoiled nature — quite in contrast to the feeling evoked by a visit to the Grand Manan group. The virgin character of the coniferous forest is indicated by the deep carpet of lichens and mosses and perhaps by the abundance of three species of orchids; Habenaria obtusata, H. clavellata, and Listera cordata. The granitic base of the islands rises sheer and bold with 1963] Wolf Islands — Pike and Hodgdon 85 very few landing places and only one semi-protected anchor- age— at Paul's Cove on East Wolf. The islands are not high, the maximum elevation being about 100 feet, but there are imposing sea cliffs and yawning chasms on the seaward or eastern sides of all the islands, the best development of which occurs on South Wolf. Unlike Grand Manan and the principal other Bay of Fundy islands, the Wolves have no rocks of calcareous affinity nor basaltic origin and as might be expected no species of calcareous affinity occur there. On the smaller islands of the Wolves the soil seems to be entirely litter or humus down to bed rock, On South Wolf the land- bridge connecting the main island with the granite nubble where the lighthouse is placed, is a ridge with much gravel and water-worn cobbles indicating water deposition and a probable former beach line. Marine gravels and clays are revealed at the heads of small coves where the soil profiles are exposed by erosion. Meteorological conditions can be judged somewhat from the weather bureau records at Eastport, and from Coast Guard and lighthouse records from West Quoddy Head in Lubec, both stations being in Maine within 10-15 miles from the Wolves. The temperature of the water of the Bay of Fundy varies little between the high of summer and the low of winter due to the great turbulence and mixing caused by the tides which range from 27.5 to 13.5 ft, the average for the year being 19 ft. The highest surface temperatures are between 52? and 54? F. in summer and the lowest winter temperature of the surface water is between 34? and 32? F. The winter extremes only occur during unusually cold, win- dy weather and usually in the winter the water is little cold- er than in the summer. This small range of water surface temperatures has a strong influence on the air temperatures surrounding the islands tending to moderate the seasonal extremes. These same water temperatures strongly influence the atmospheric moisture producing “vapor” or “sea smoke" in winter and fog in spring and summer. Nearby Quoddy 86 Rhodora [Vol. 65 Head in Maine has recorded periods of well over 300 hours of continuous fog. Thus the climate of the Wolves is not only decidedly maritime, but somewhat boreal, and in some respects almost montane. This distinctive local climate should be kept in mind during discussions of the flora of these islands. While attending a botanical meeting in Montreal in August 1959, we discussed together the possibilities of doing some research on the Wolf Islands. The senior author had known of the Wolves since childhood and had frequently seen them on fogless days from Campobello. Neither of us had ever heard of any botanical work being carried on there, nor could we learn of any from the Canadian botanists with whom we talked. Therefore in September the senior author made two brief visits to the Wolves and brought back 62 taxa as representative and interesting samples of the flora. When we studied these it seemed appropriate to compare them with the “List of the Vascular Plants of Grand Manan” (Weatherby and Adams, Contr. Gray Herb. CLVIII, 1945.) Several taxa which were common on the Wolves were not listed at all from Grand Manan. Plans were made for further investigation of the Wolves in particular, but also of Grand Manan (Rhodora 64: 98-108, 1962). All subsequent visits have strengthened the early indications that there are significant differences between these two groups of islands. More detailed analyses of these island differences and other features of phytogeographic interest relating to the Wolves and other Fundian Islands will be covered in subsequent papers. Comparative studies of the floras of islands in North America have been grossly neglected though many fine studies of individual islands have been made. After four seasons of study involving a total of fifteen days spent on the Wolves extending over a period from June to October in different years, we are impressed with the fact that there are many potential problems for investiga- tion. The ecological features of islands so nearly virgin in character deserve attention. The small size of the islands 1963] Wolf Islands — Pike and Hodgdon 87 combined with their isolation have concentrated certain populations of unique character there, affording a natural laboratory for population studies in taxonomy. LIST OF PLANTS OF THE WOLF ISLANDS* East Fat Flat South Gull Wolf Pot Wolf Wolf Rock Equisetum arvense L. var. boreale (Bong.) Ledeb. x x E. sylvaticum L. var. pauciramosum Milde X X Lycopodium annotinum L. var. alpestre Hartm. x L. annotinum L, var. pungens (La Pylaie) Desv. x L. clavatum L. x L. obscurum L. var. dendroideum (Michx.) D. C. Eat. x x Botrychium simplex E. Hitchc. var. tenebrosum (A. A. Eat.) Clausen x Osmunda Claytoniana L. x x O. cinnamomea L. X X X Onoclea sensibilis L. X Dryopteris Thelypteris (L.) Gray var. pubescens (Lawson) Nakai X D. noveboracensis (L.) Gray X D. disjuncta (Ledeb.) C. V. Mort. X X D. Phegopteris (L.) Christens. X D. spinulosa (O. F. Muell.) Watt var. spinulosa X X D. spinulosa (O. F. Muell.) Watt var. intermedia (Muhl.) Underw. X D. spinulosa (O. F. Muell.) Watt var. americana (Fisch.) Fern. X X X X D. cristata (L.) Gray X Dennstaedtia punctilobula (Michx.) Moore X X Athyrium Filix-femina (L.) Roth var. Michauxii (Spreng.) Farw. X X X A. Filix-femina (L.) Roth var. Michauxii (Spreng.) Farw. f. laurentianum (Butters) Fern. X Polypodium virginianum L. X X X X Pteridium aquilinum (L.) Kuhn var. latiusculum (Desv.) Underw. X 88 Rhodora [Vol. 65 Taxus canadensis Marsh. x Abies balsamea (L.) Mill. var. phanerolepis Fern. x x x x Picea glauca (Moench) Voss X X X x P. rubens Sarg. x X X P. mariana (Mill.) BSP. x x Larix laricina (DuRoi) K. Koch x Thuja occidentalis L. X Ruppia maritima L. var. longipes Hagstr. X Triglochin maritima L. X Festuca rubra L. var. commutata Gaudin X X X X Puccinellia laurentiana Fern. & Weath. var. laurentiana x x x x P. paupercula (Holm) Fern. & Weath. var. paupercula x x x x P. paupercula (Holm) Fern. & Weath. var. alaskana (Scribn. & Merr.) Fern. & Weath. X X X X Glyceria canadensis (Michx.) Trin. X Poa annua L. X X X X X P. compressa L. x x P. pratensis L. x X X P. palustris L. X X X Agropyron trachycaulum (Link) Malte var. glaucum (Pease & Moore) Malte X A. trachycaulum (Link) Malte var. majus (Vasey) Fern. X A. repens (L.) Beauv. f. aristatum (Schum.) Holmb. X X X A. repens (L.) Beauv. f. trichorrhachis Rohlena X x Elymus arenarius L. var. villosus Mey. x x X x Deschampsia flexuosa (L.) Trin, var. flexuosa X x x X D. flexuosa (L.) Trin. var. montana (L.) Ledeb. x x X X Danthonia spicata (L.) Beauv. X Calamagrostis canadensis (Michx.) Nutt. var. robusta Vasey X X X X x C. canadensis (Michx.) Nutt. var. scabra (Presl.) Hitche. x Agrostis alba L. var. palustris (Huds.) Pers. X X X x x A. tenuis Sibth. X X A. scabra Willd. X X A. perennans (Walt.) Tuckerm. var. aestivalis Vasey x x 1963] Wolf Islands — Pike and Hodgdon Cinna latifolia ( Trev.) Griseb. Phleum pratense L. Alopecurus pratensis L. Anthoxanthum odoratum L. Hierochloé odorata (L.) Beauv. Eleocharis palustris (L.) R. & S. E, halophila Fern. & Brack. E. elliptica Kunth Scirpus validus Vahl. var. creber Fern. S. atrocinctus Fern. Eriophorum spissum Fern. E. tenellum Nutt. E. angustifolium Honckeny E. virginicum L. Carex stipata Muhl. . disperma Dew. . trisperma Dew. Mackenziei Krecz. canescens L. var. canescens canescens L. close to var. disjuncta Fern. canescens L. var. subloliacea Laestad. brunnescens (Pers.) Poir. var. sphaerostachya (Tuckerm.) Kukenth. . cephalantha (Bailey) Bickn. . angustior Mackenz. scoparia Schkuhr hormathodes Fern. . leptalea Waslenb. Emmonsii Dew. . paleacea Wahlenb. . crinita Lam. var. crinita . nigra (L.) Reichard var. nigra . nigra (L.) Reichard var. strictiformis (Bailey) Fern. . limosa L. . paupercula Michx. var. irrigua (Wahlenb.) Fern. debilis Michx. var. Rudgei Bailey viridula Michx. pauciflora Lightf. intumescens Rudge var. Fernaldii Bailey . rostrata Stokes var. rostrata . rostrata Stokes var. utriculata (Boott) Bailey oo6onoon anmaanaaaan ooooon nn X OX XX NX KK KKK KKK KKK KX XXXXXx XXX x X XXXXXX X 90 Rhodora Juncus bufonius L. J. bufonius L. var. halophilus Buchenau & Fern. J. tenuis Willd. J. effusus L. var. Pylaei (Laharpe) Fern. & Wieg. J. balticus Willd. var. littoralis Engelm. J. brevicaudatus (Engelm.) Fern. Luzula multiflora (Retz.) Lejeune var. acadiensis Fern. Clintonia borealis (Ait.) Raf. Smilacina stellata (L.) Desf. S. stellata (L.) Desf. var. crassa Vict. S. trifolia (L.) Desf. Maianthemum canadense Desf. Streptopus amplexifolius (L.) DC. var. americanus Schultes S. roseus Michx. var. perspectus Fassett Sisyrinchium montanum Greene var, crebrum Fern, Iris Hookeri Penny I. versicolor L. Cypripedium acaule Ait. Habenaria clavellata (Michx.) Spreng. var. ophioglossoides Fern. H. obtusata (Pursh) Richards. Goodyera repens (L.) R. Br. var. ophioides Fern. Listera cordata (L.) R. Br. Corallorhiza trifida Chatelain var. verna (Nutt.) Fern Salix rigida Muhl. S. Bebbiana Sarg. var. capreifolia Fern. 5. pedicellaris Pursh var. hypoglauca Fern. S. discolor Muhl. S. humilis Marsh. Populus tremuloides Michx. Myrica Gale L. var. Gale Betula lutea Michx. f. B. populifolia Marsh. B. papyrifera Marsh. var. papyrifera B. papyrifera Marsh. var. cordifolia (Regel) Fern. Alnus crispa (Ait.) Pursh var. mollis Fern. XxXxXXXKX X «XX X x x X X KX XK xX X KX xX X XK x x x [Vol. 65 1963] Wolf Islands — Pike and Hodgdon A. rugosa (Du Roi) Spreng. var. americana (Regel) Fern. Urtica gracilis Ait. Rumex pallidus Bigel. R. domesticus Hartm. R. Acetosella L. Polygonum aviculare L. P. aviculare L, var. vegetum Ledeb. P. aviculare L. var. littorale (Link) W. D. J. Koch P. sagittatum L. P. Convolvulus L. Rheum Rhaponticum L. Chenopodium album L. Atriplex patula L. var. patula A. patula L. var. hastata (L.) Gray A. glabriuscula Edmondston Suaeda sp. Spergularia marina (L.) Griseb. var. leiosperma (Kindb.) Gurke Sagina procumbens L. S. nodosa (L.) Fenzl Arenaria lateriflora L. Stellaria media (L.) Cyrillo S. graminea L. S. calycantha (Ledeb.) Bong. var. isophylla Fern. Cerastium vulgatum L. Ranunculus Cymbalaria Pursh R. acris L. Thalictrum polygamum Muhl. var. polygamum Coptis groenlandica (Oeder) Fern. Capsella rubella Reut. Cakile edentula (Bigel.) Hook. Raphanus Raphanistrum L. Sarracenia purpurea L. Drosera intermedia Hayne D. rotundifolia L. D. rotundifolia L. var. comosa Fern. Sedum Rosea (L.) Scop. Mitella nuda L. Ribes hirtellum Michx. var. hirtellum R. hirtellum Michx. var. calcicola Fern. x? X X XXXx XXxXXXXX x X Xxx X Xo KX KK XK XK XK SX x x KAXXA XXXXXX X 91 x 92 Rhodora [Vol. 65 R. lacustre (Pers.) Poir. X R. glandulosum Grauer x x X X Spiraea latifolia (Ait.) Borkh. X Pyrus melanocarpa (Michx.) Willd. X P. decora (Sarg.) Hyland X X X X Amelanchier laevis Wieg. X X X A. laevis Wieg. X Bartramiana (Tausch) Roemer X x x A. Bartramiana (Tausch) Roemer x Fragaria virginiana Duchesne var. virginiana x X X F. virginiana Duchesne var. terrae-novae (Rydb.) Fern. & Wieg. X Potentilla palustris (L.) Scop. X P. norvegica L. X X X X X x x x P. simplex Michx. var. calvescens Fern. P. anserina L. P. Egedei Wormsk. var. groenlandica (Tratt.) Polunin X Geum rivale L. f. virescens Lilja x Rubus pubescens Raf. var. pubescens x x X X R. pubescens Raf. var. pilosifolius A. F. Hill X R. idaeus L. var. aculeatissimus Regel & Tiling R. idaeus L. var. strigosus (Michx.) Maxim. X X x R. idaeus L. var. strigosus (Michx.) Maxim. f. tonsus Fern. R. idaeus L. var. canadensis Richards. R. vermontanus Blanch. R. canadensis L. R. allegheniensis Porter Rosa nitida Willd. R. virginiana Mill. Prunus pensylvanica L. f P. virginiana L. Trifolium pratense L. T. repens L. Lathyrus japonicus Willd. var. pellitus Fern. L. palustris L. var. linearifolius Ser. L. palustris L. var. pilosus (Cham.) Ledeb. Oxalis montana Raf. Empetrum nigrum (L.) Ilex verticillata (L.) Gray Nemopanthus mucronata (L.) Trel. X XX XXXXXXXxxX X X XXX X XX XXX X x xX xX XxX XK xX x x 1963] Wolf Islands — Pike and Hodgdon Acer spicatum Lam. X A. pensylvanicum L. A. rubrum L. Impatiens capensis Meerb. Hypericum canadense L. H. virginicum L. var. Fraseri (Spach) Fern. Viola cucullata Ait. var. microtitis Brainerd V. septentrionalis Greene V. Mackloskii Lloyd subsp. pallens (Banks) Baker V. incognita Brainerd var. incognita V. incognita Brainerd var. Forbesii Brainerd Epilobium angustifolium L. var. angustifolium E. leptophyllum Raf. E. glandulosum Lehm. var. adenocaulon (Haussk.) Fern. E. glandulosum Lehm. var. occidentale (Trel.) Fern. Oenothera perennis L. Circaea alpina L. Aralia nudicaulis L. Carum Carvi L. Ligusticum scothicum L. Coelopleurum lucidum (L.) Fern. Conioselinum chinense (L.) BSP. Cornus canadensis L. C. stolonifera Michx. Moneses uniflora (L.) Gray Monotropa uniflora L. M. Hypopithys L. Ledum groenlandicum Oeder Rhododendron canadense (L.) Torr. Kalmia angustifolia L. K. polifolia Wang. Cassandra calyculata (L.) D. Don Gaultheria hispidula (L.) Bigel. Vaccinium angustifolium Ait. var. laevifolium House V. Vitis-Idaea L. var. minus Lodd. V. Oxycoccos L. V. Oxycoccos L. var. ovalifolium Michx. V. macrocarpon Ait. XXX XXX X XX xX X x XX x x x XXX x XXX XX KX KKXXKXXKX XXX XK XK XK XK XK XX X X XXX X XXX XXXXXXX X X X xX xX XX 94 Rhodora Trientalis borealis Raf. Glaux maritima L. Lomatogonium rotatum (L.) Fries f. americanum (Griseb.) Fern. Menyanthes trifoliata L. var. minor Raf. Convolvulus sepium L. var. sepium C. sepium L. f. malachophyllus Fern. Mertensia maritima (L.) S. F. Gray Scutellaria epilobiifolia A. Hamilton Galeopsis Tetrahit L. var. bifida (Boenn.) Lej. & Court. Lycopus uniflorus Michx. Euphrasia americana Wettst. E. Randii Robins. var. Randii E. Randii Robins. var. Farlowii Robins. E. canadensis Townsend Rhinanthus Crista-galli L. var. fallax (Wimm. & Grab.) Druce Plantago major L. P. juncoides Lam. var. decipiens (Barneoud) Fern. P. juncoides Lam. var. glauca (Hornem.) Fern. P. juncoides Lam. var. laurentiana Fern. P. oliganthos R. & S. var. fallax Fern. Galium triflorum Michx. G. trifidum L. G. tinctorium L. G. tinctorium L. var. subbiflorum (Wieg.) Fern. G. labradoricum Wieg. Diervilla Lonicera Mill. Lonicera villosa (Michx.) R. & S. var. Solonis (Eat.) Fern. L. villosa (Michx.) R. & S. var. tonsa Fern. L. canadensis Bartr. Linnaea borealis L. var. americana (Forbes) Rehd. Viburnum cassinoides L. Sambucus pubens Michx. Campanula rotundifolia L. C. rotundifolia L. f. albiflora Rand and Redf. Solidago macrophylla Pursh S. sempervirens L. X X X XX XX X X X XX XXX x x x X x x X XxX XK XK xX xX [Vol. 65 x xX xX xX x XX 1968] Wolf Islands — Pike and Hodgdon S. uliginosa Nutt. var. terrae-novae (T. & G.) Fern. S. rugosa Ait. var. villosa (Pursh) Fern. S. canadensis L. S. graminifolia (L.) Salisb. var. Nuttallii (Greene) Fern. Aster radula Ait. . foliaceus L. var. arcuans Fern. . novi-belgii L. . nemoralis Ait. . acuminatus Michx. . acuminatus Michx. f. discoideus Ktze. or f. virescens Vict. & Rousseau A. umbellatus Mill. f. intercedens Fern. A. umbellatus Mill. var. pubens Gray Anaphalis margaritacea (L.) C. B. Clarke var. subalpina Gray Gnaphalium uliginosum L. Bidens frondosa L. Achillea borealis Bong. A. lanulosa Nutt. Matricaria matricarioides (Less.) Porter Chrysanthemum Leucanthemum L. Senecio vulgaris L. S. sylvaticus L. Cirsium vulgare (Savi) Tenore C. muticum Michx. C. arvense (L.) Scop. Leontodon autumnalis L. Taraxacum erythrospermum Andrz. T. officinale Weber Sonchus arvensis L. S. asper (L.) Hill Prenanthes trifoliolata (Cass.) Fern. var. nana (Bigel.) Fern. P. altissima L. f. integra Rousseau Hieracium aurantiacum L. H. floribundum Wimm. & Grab. > Pr > > xX XX xX X XXX Xo xK KK xX x x x x We have made an attempt to get a practically complete list of taxa from each island to determine the relationship between the area of land involved and the number of taxa. Table I shows the estimated relative areas of the Wolf 96 Rhodora [Vol. 65 Islands with the numbers of taxa we have now recorded from each. Undoubtedly there will be some slight changes in the totals, as previously undetected plants come to our attention. The present figures show a substantially correct picture of the relationship of the numbers present to the areas of the individual Wolf Islands. The areal scale was established by regarding Gull Rock, the smallest, as a unit of one and the others as multiples of this unit. Thus with Gull Rock as one, we gave Flat Wolf, the next largest, which is about three times as large, the factor of three and so on up to forty-two for East Wolf, the largest of the islands. Table I Total Taxa recorded 314 East South Fat Flat Gull Wolf Wolf Pot Wolf Rock Comparative areas of islands 42 12 6 3 1 Taxa recorded 283 163 115 95 29 Ratio of taxa to area 6.7 13.5 18.1 31.6 29* *Weather conditions prevented a final collection from Gull Rock which undoubtedly would have added several species to the list. The larger islands (East and South Wolf) present greater habitat possibilities not only because larger areas provide more ecological niches such as bogs, Swamps, barrier beaches, etc., but also because nearly all of the disturbed areas are and have been on these islands. On the other hand, the three smaller islands, particularly Gull Rock, have bird nesting colonies which may strongly influence the presence and absence of various taxa. Regardless of the other factors involved, the number of taxa present bears a fairly distinct relationship to the area involved, with a des- cending ratio to area as the total area increases. The com- parison is an interesting one and should be further tested as more island studies are carried on. DEPARTMENTS OF HORTICULTURE AND BOTANY, UNIVERSITY OF NEW HAMPSHIRE, DURHAM. Volume 65, No. 761 including pages 1-96, was issued March 29, 1963 ARLOW rr - ~ f: REFERZNCE LIBRARY vua 1 1963 SAS PERE OE ERRETEN Riodova 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 Associate Editors Vol. 65 April-June, 1963 No. 762 CONTENTS: The Widening Panorama in Medical Botany. Richard Evans Schultes ............. eere eerte 97 Helianthus laetiflorus and Helianthus rigidus — Hybrids or Species? Sarah Clevenger and Charles B. Heiser, Jr. .... 121 A New Feature in Bamboo Rhizome Anatomy. F. A. McClure |. tent nn een nto tn status senses stein en ntn n ntn 134 Pollen Morphology As A Taxonomic Tool in Linum. K. S. Xavier and C. Marvin Rogers ..........,."A enn 137 Phytogeographic Notes on Rottboellia, Paspalum, and Miscanthus (Gramineae). Richard W. Pohl .................. 146 (Continued on Inside Cover) The Neo England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. CONTENTS: — continued Life-Forms in the Plains Flora of Southern Mackenzie, Northwest Territories. John W. Thieret .. es 149 A Revision of the North American Genus Uvularia (Liliaceae). Robert L. Wilbur sese 158 The New York Botanical Garden Wild Flower Books. H. W. Rickett essere een 189 Phoradendron rubrum in Florida. George R. Cooley ............ 190 Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 65 April-June, 1963 No. 762 THE WIDENING PANORAMA IN MEDICAL BOTANY? RICHARD EVANS SCHULTES “And as there are discouvered new Regions, new Kingdoms, and new Prouinces by our Spaniards, so they have brought unto us new Medicines, and newe Remedies, wherewith they do cure many informities, which, if we did lacke them, would bee incurible, and without any remedie . . . for which cause I did pretend to treate, and to write of all things that they bring from our Indias, apperteyning to the Arte and Use of Medicine, and the remedie of hurtes and diseases that we doo suffer and endure . . ." Nicholas Monardes [transl. John Frampton] “Joyfull Newes out of the New-found Worlde" (1956) IL Introduction From an Andean mountain top overlooking the endless forest of the Amazon, dawn can be beautiful. One sees — one feels — the tropical panorama widen as the solid blanket of morning mist below first breaks into little openings, then, as the sun climbs higher, slowly lifts to bring the foreground into detailed view and the distance into clearer perspective. I have often experienced such a widening panorama dur- ing my years of work in South America. And I have just as often thought of how similar has been our recent experience in medical botany. The widening panorama of the Plant Kingdom as a field of exploration for new physiologically active substances — this panorama has already started to bring the foreground into detailed view and the distance into clearer perspective. For the first time perhaps in all history ‘Lecture given at the Medical Center Lecture Series, University of West Virginia, Morgantown, West Virginia, October 29, 1962. 97 98 Rhodora [Vol. 65 we know in a small way where we are going in our search for plants of promise in man's fight against the ills of his mind and body. At the Botanical Museum of Harvard University, we have a course entitled “Plants and human affairs", the oldest course in economic botany in this country — going back to 1876. When I was a student in this course in 1935, Professor Oakes Ames spoke somewhat nostalgically when he took up the medicinal plants, because, one by one, the old plant medi- cines were being synthesized or replaced by coal-tar products. Professor Paul C. Mangelsdorf, who took over the teaching of the course in 1941, felt that, in the future, he might have to discuss plant drugs solely from the historical point of view. Now, in this same course, I divide my lectures on medical botany into Medicines of the Past, Medicines of the Present and Medicines of the Future. Such an amazing turn-about has taken place in a quarter of a century that it is not easy for us fully to comprehend the far-reaching implications of the change. Perhaps it is even more difficult for the scientist who is close to the Plant Kingdom — especially a botanist who was fortunate enough to spend many years in the field in intimate association with rich tropical floras. He cannot help being thoroughly over- whelmed with the vastness, the boundless variation, the intricate adaptations, the potentialities of undiscovered con- stituents of the Plant Kingdom. Familiarity with such a welter of complexity can, of itself, relegate into a truly minor place the relationship between man's ills and the grandeur and universality of the vegetal world. When I was a graduate student, seeking material for my doctoral thesis amongst Indians isolated in the mountains of Oaxaca, Mexico, I could find no pharmaceutical house in the United States with the time or interest to investigate witch- doctors' plants, and I was forced to send my material to Sweden to the late Dr. C. G. Santesson who, since he was retired, could study whatever it pleased him to investigate. Now every pharmaceutical house of any size and serious- ness has become newly conscious of the Plant Kingdom. 1963] Medical Botany — Schultes 99 Colleges of pharmacy turn to the vegetal world for prob- lems for doctoral students. Government agencies keep an alert eye on promising openings for research into plant drugs. And this forward step has opened up almost limitless vistas for both academic and practical research programmes. Yet everything indicates that we are still far from grasp- ing the potentialities which lie before us. We are hesitant to grasp the reality, perchance, just because we can see no bounds. It is this opportunity — never before given to man and fraught with such imponderable promise — that I want to consider. Perhaps it may be interesting in the beginning to review a few simple figures to bring into higher relief this great opportunity of which I speak. I looked back into my college notes the other day and found that, in 1935, of the ten drugs then in most common use, only two (digitalis and codeine) were of vegetal derivation. The percentage of drugs of plant origin recognized by the United States Pharmacopoeia was steadily declining, whilst that of chemical drugs was steadily increasing. In 1820, our first Pharmacopoeia listed 223 plant drugs; a century and a quarter later, in 1946, the Pharma- copoeia and National Formulary enumerated 244. This would seem to indicate a slight increase, but it is not an indication of their relative importance. For, in 1820, 82% of the medi- cines listed were of plant origin, 15% of chemical, 3% of animal; in 1946, 38% were vegetal, 56% chemical, 6% ani- mal. To-day, only about 100 of the original 223 are still in use, but these include the “blue-chips” that have paid good dividends in health since earliest times. Perhaps, in passing, it might be interesting to compare this total of 100 plant drugs now in use with the 3800 species recommended medici- nally by John Parkinson, the last British herbalist, in his “Theatrum Botanicum” of 1640.° Now, since we are using but 100 of the original 223 drugs of plant origin, how can we be sure that we should delve so deeply and with such great hopes into the Plant Kingdom *Parkinson, John: “Theatrum Botanicum. The theatre of plantes” (1640). 100 Rhodora [Vol. 65 for new medicines? And when we realize that from 1900 to 1940 only about five new plant drugs were added to our list: Strophanthus, agar, chaulmoogra, Ephedra and Psyllium — what should we think then? But from 1940 to 1960, a minor miracle has come to pass. As the result of new methods of analysis and refined tech- niques, many new drugs have been discovered and new uses have been found for some of the older drugs. To-day, 75 to 80% of the drugs most frequently prescribed are recent additions to our pharmacopoeias ; most of them, in fact, were unknown in 1940. As a direct result of this rash of discovery, our tempo of research has been steadily accelerated, yet the frontiers for discovery have hardly been touched. The organic chemist might differ with our predictions, since he may believe that it is only a question of time before everything will be synthetic and that chemistry alone will dominate the field of new drug discovery. No one can deny that great strides have been made in organic synthesis. Many plant drugs have been produced synthetically (witness camphor and quinine) or closely allied substances, some- times more effective and efficient, have been made (such as aspirin vs. oil of wintergreen, novocaine vs. cocaine). Alka- loids, the active principles of so many drugs, have long baffled the chemist. It took 134 years to synthesize mor- phine, and production is still in the pilot-plant stage. Fifty- eight years elapsed after atropine was first synthesized be- fore it could be done on a large scale. All of this means that the opium poppy and the belladonna plant are still very much with us. Yet it is not that we would look to the Plant Kingdom as a commercial supply of our drug compounds, unless it be more economical to isolate them from plants than to produce them synthetically. We should look upon the Plant Kingdom as an almost untapped arsenal of ready-made compounds which, once isolated and understood, can serve us in at least three ways: 1) directly as medicinal agents; 2) as starting points for the elaboration of more complex compounds of therapeutic value; and 3) as academically interesting or stimulating exercises. 1963] Medical Botany — Schultes 101 Il. The Plant Kingdom Since the vegetal world does present this widening pano- rama, let us start at the beginning and take an appraising look at the Plant Kingdom. Just what does the Plant King- dom offer? The most diverse of organisms make up the Plant King- dom: the Bacteria, Algae, Fungi, Bryophytes, Pteridophytes and Spermatophytes. All told, there may be as many as 800,000 species in the Plant Kingdom. It is hard to appreci- ate how extensive a field this represents, but it is easy to realize how truly marginal our phytochemical knowledge of this vast assemblage must be. Estimates, naturally, are subject to appreciable variation, and calculations of the number of species can, of course, be made with greater precision in groups of plants which have been thoroughly studied taxonomically. This means that we are in a better position to caleulate how many species of spermatophytes there are than bacteria. Estimates vary, again, with the outlook of the taxonomist, but when we consider the Plant Kingdom as a whole, the personal equa- tion — what sundry taxonomists accept as species limita- tions — is more or less balanced out. We do not have a clear idea of how many species of Bac- teria there are. This is partly because they have been less thoroughly studied than most other groups of plants and partly because they are often known more from their physio- logical effects than from structural characteristics. Modern estimates? give approximately 1500 species in about 200 genera. The bacteria are, of course, of intense importance in medicine as the causative agents of many ills, but a disap- pointingly small number of therapeutically promising com- pounds have been isolated from them. Estimates for the Fungi have varied from approximately 30,000 to 85,000 species. One of the most recent workers*, *Thimann, K. V.: “The life of bacteria — their growth, metabolism and relationships" (1955). i ‘Martin, G. W.: “The numbers of fungi" in Proc. Iowa Acad. Sci. 58 (1951) 175. 102 Rhodora [Vol. 65 however, has written that 100,000 may be highly conserva- tive and that the total may well be over 200,000. Another contemporary worker? states that “it seems reasonable to predict that, as more facts are accumulated . . . the fungi will eventually be recognized as bearing the same relation- ship to the Plant Kingdom (on the basis of numbers of species) as insects bear to the Animal Kingdom." Now the fungi are extraordinarily important to man; but, as sources of medically active compounds, they have, until recently, been of strangely minor significance. The ergot alkaloids come immediately to mind as very old fungal medicines. But the rise of the antibiotics, beginning with the development of penicillin in the early 1940's, has brought into sharp relief as a source of new drugs not only the fungi but the related actinomycetes. The very recent work with hallucinogenic mushrooms and the isolation from them of phosphylated indole derivatives never known to occur in plants has turned our eyes towards this much neglected section of the fungi and may, one day soon, produce spectacular and highly prac- tical medical results. Furthermore, the study of fungi as allergens is still in its infancy and promises many new avenues of approach in research. I cannot tell you how many species of fungi have as yet never been subjected to chemical study; but, if we accept the estimate of 100,000 species for this group of plants, we can easily appreciate the widening panorama here awaiting the attention of our research scien- tists. That most varied of plant groups, the Algae, numbers about 19,000 species*. Here is a vast and hardly touched field for phytochemical investigation. Since most of the algae are aquatic, many of them marine, the problems as well as the opportunities awaiting us must be both numer- ous and unique. Only a beginning has been made in the search for medically promising chemicals from the algae. When we come to that interesting symbiotic group, the Lichens, we find that startling advances have likewise been "Gray, William D.: “The relation of fungi to human affairs” (1959). *Bold, H. C.: “The plant kingdom" (1960). 1963] Medical Botany — Schultes 108 registered in recent years. The newest estimates give the lichens some 450 genera and 20,000 species’. Since 1944, the bacteria-inhibiting properties of a number of lichens have been noted. In fact, about one-half of the temperate zone lichens have this property, due to lichen acids which can inhibit gram-positive bacteria and even the tuberculosis bacilli and some fungi. Lichen antibiotics are now used com- mercially in medicinal salves in northern Europe, especially in Finland, Russia and Germany, and there is every reason to believe that the lichens may supply more as well as more diverse antibioties as research continues. The Japanese are extremely active at the present time in the investigation of chemical components of the lichens. The Bryophyta, characterized as “a constellation of di- vergent groups rather than as a homogeneous division or phylum", has been severely neglected in phytochemical in- vestigation. Comprising some 14,000 species and being world-wide in distribution, the bryophytes promise interest- ing results from a concerted study of their chemical constit- uency. When we come to the so-called “higher” plants, the Pteri- dophyta and the Spermatophyta, we see again the potentiali- ties of a widening panorama, for here we have a significantly large and diversified group of plants. The pteridophytes — ferns and fern allies — have given us a few folk remedies, but they have been disappointingly parsimonious in contributing to modern medicine. Here is a field, however, which, since it has not been granted the inten- sive phytochemical efforts that have characterized other groups of the higher plants, may hold hidden surprises. The 250 genera and nearly 10,000 species of ferns and fern allies? deserve closer chemical scrutiny, especially those which are known to have outstanding folk uses in primitive societies. ; We come now to the seed plants or Spermatophyta, the ‘Lamb, I. M.: Personal communication. *Steere, W. C.: “Bryophyta” in “The eneyclopedia of the biological sciences” [Ed. P. Gray] (1961) 177-179. "Tryon, R.: Personal communication. 104 Rhodora [Vol. 65 dominant land flora of our present geological epoch. The two Spermatophyte groups, the gymnosperms and the angi- opserms, show extreme disparity both in their size and in what they have given to medicine. There are some 65 genera and 700 species of gymno- sperms,'" from which, in the United States, we use only about two dozen official and unofficial drugs“ — and these mainly for their volatile oils or resins. Would not a renewed phytochemical attack, utilizing improved techniques, upon this ancient group of spermatophytes seem to be warranted ? It is the angiosperms, above all other groups of plants, which have occupied man's attention from earliest times in his search for medicines. This is easy to understand. The angiosperms are not only numerous, but they are conspicu- ous and, even to-day, take a position of primacy in the pop- ular concept of the Plant Kingdom. Most of our plant reme- dies of the past and a very substantial number of the newer advances in medical botany have been made from the angio- sperms. Since it is my belief that the potentialities have only been superficially explored, even in this best known part of the Plant Kingdom, I shall discuss somewhat more fully the width of the panorama which the angiosperms present. It may surprise those who are not taxonomic botanists to hear that no one knows really how many angiosperms there are. Even though the angiosperms are better understood than other groups of plants, estimates vary significantly, and one sometimes wonders: May not even our highest esti- mates be seriously deficient? Estimates vary greatly. but the figure usually accepted is in the neighbourhood of 200,000 species'"."* in some 300 families and some 10,500 genera. The monocotyledons usually are credited as comprising about one quarter as many species as the dicotyledons. Having spent more than a decade in field work in the Amazon basin and in the northern Andes — one of the rich- est floristic areas of the world — I have now for some time "Lawrence, G. H. M.: *Taxonomy of vascular plants" (1951). "Youngken, H. W.: “A text book of pharmacognosy", Ed. 4 (1936). "Ames, O.: “Economic annuals and human cultures" (1939). 1963] Medical Botany — Schultes 105 given serious thought to this estimate. I am led to the con- clusion that we are greatly underestimating the species- wealth of the angiosperms. I once began to formulate my ideas about the number of species in the flora of the Republic of Colombia. After mar- shalling many facts and opinions and screening the results of past and current plant exploration, I was obliged to place my census of Colombia's flora in the neighbourhood of 50,000 species of higher plants. Even though Colombia is recog- nized as one of the two or three richest phytogeographic areas of the world, this calculation at first frightened me. Nevertheless, I published it, together with my reasons for arriving at such a figure. The reaction from botanists was, in general, very favourable. Now, if we hold to a figure of only 200,000 species of angio- sperms, Colombia would then have one quarter of all the world's flowering plants. As rich an area as Colombia is, it cannot be that rich. Then, let us look at it from another vantage point: the Orchidaceae is the largest phanerogamic family, with an estimated 25,000 to 30,000 species; and the next largest family is the Compositae with some 20,000 spe- cies. If two of the 300 angiosperm families add up to about 50,000 species, must we not alter our estimate of the total number of flowering plant species? In other words, do these two families comprise one quarter of all the angiosperms ? And let us not forget that every exploration beings back species new to science. I have, for myself at least, made the indicated statistical alteration and have felt for a long while that we must allow somewhere near half a million species, instead of 200,000 for the flowering plants. The future — and the very near future, I believe — will justify this point of view. III. Plant Constituents Perhaps the alkaloids are, to medicine, the most impor- "Schultes, Richard Evans, “La riqueza de la flora comombiana" in Rev. Acad. Col. Cienc. Exact. Fisico-Quim. Nat. 8 (1951) 230.— “Hacia un censo de la flora de Colombia" in Univ. Nac. Col., no. 23 (1958) 77. 106 Rhodora [Vol. 65 tant constituents — and certainly they are one of the most widespread — in the plant world, even though we cannot forget the glycosides, essential oils, gums, mucilages, tan- nins, fatty oils, colouring matters, resins and other types of chemical substances which man has sometimes found useful therapeutically. There is no question that the Plant King- dom has yielded an amazing variety of products in this fieid and, as chemical methods and laboratory techniques become more sophisticated, we can see no end to the useful materials available from the thousands of species as yet untouched. Notwithstanding the vast amount of research that remains to be done, perhaps more chemical and medical work has been carried out on alkaloids than on any other kind of plant con- stituent. A recent and most thorough compilation of alka- loids'* enumerates 3671 species of plants (including crypto- gams) in which alkaloids have been found. Not all of these have been structurally identified. Some of them are undoub- tedly “duplicates” — alkaloids which may be shown to be identical to others. In the flowering plants, certain families are known to be rich in alkaloids, but even in these families the figures indicate only partial study. Two good examples are the Leguminosae, a family of at least 15,000 species, in which 1525, or about 10%, are listed as having alkaloids; and the Solanaceae, with upwards of 2500 species, where 252, or again only 10%, are alkaloid-bearing. About 10% of the known alkaloids have been recorded from one family, the Apocynaceae, as a result of the effort initiated by Rauwolfia investigations during the past ten years. From the chemist's point of view, there are other families in which a concentra- tion of work, if it could be justified on other than purely academic grounds, would likewise yield large numbers of compounds. And probably an equally large variety of struc- tures would appear as well. The potentialities are obvious. To-day our techniques of alkaloidal detection — as well as tests for other types of constituents — are greatly improved. A recently perfected "Williams, J. J. and Bernice G. Schubert: *Alkaloid-bearing plants and their contained alkaloids", U. S. D. A. Techn. Bull. no. 1234 (1961). 1963] Medical Botany — Schultes 107 spot test for alkaloids," which can be applied to fresh plant material in the field or to fragments from herbarium speci- mens, promises rapidly and widely to extend our knowledge of the distribution of alkaloids in the higher plants. It will not, of course, tell us what kind of alkaloid is present; that must await more detailed examination. But it will tell us whether or not an alkaloid is present. The potentialities offered by this simple spot test can hardly be over-empha- sized, for it opens up for quick and easy preliminary study all the spermatophytes and some other plant groups. One American pharmaceutical house has tested some 15,000 to 20,000 species for alkaloids.?* On a random basis, about 15% would have been found to be alkaloid posi- tive, but, since certain families (Solanaceae, Papaveraceae, Amaryllidaceae, etc.) with alkaloids of well known structure were eliminated, the percentage would run probably about seven or eight of positives. In any case, about half of the 15% are suspected of possessing alkaloids in such minute amounts as to preclude a practical study of their chemistry or pharmacology. About half of the remainder contain some previously recorded alkaloid. Thus, we are left with about 4% of the species examined with possibly new alkaloids, hence potentially new medicinal agents. This means at least 8000 new alkaloids, more than three times the number now known, remain to be discovered. I have stressed the alkaloids, but we must bear in mind that alkaloids are not the only plant compounds of interest to medicine. There are another 3000 non-alkaloidal plant principles of known structure, many of which have or have had some application in or bearing upon medical problems. These include about 150 cardiac glycosides of the types which have been used as starting material for modern syntheses of the steroid hormones. 1V. Methods of Investigation Now, this brings us face to face with the question: How SRaffauf, Robert F.: “A simple field test for alkaloid-containing plants" in Econ. Bot. 16 (1962) 171. *Raffauf, Robert F.: Personal communication. 108 Rhodora [Vol. 65 can we take fullest advantage of this widening panorama? There is only one answear: Exploration. Exploration to hasten the widening of our panorama in medical botany may be done in the literature, both the ancient and the modern; in the herbarium; and in the field. And exploration for our present purposes had best be done along all three lines simultaneously. The literature of the ancients, the herbals of mediaeval Europe and the writings of modern anthropologists, travel- lers and missionaries must still be treated as repositories of much uninvestigated information. We should never pass quick judgment on statements in any of this literature con- cerning plant uses simply because they seem to be ridiculous. To do so might cause the loss to science, at least for many years, of plant medicines of supreme importance. This has happened frequently. Had we seriously studied the Egyptian papyri, we might many years ago have found a hint to the anti-bacterial activity of certain actinomycetes or fungi. And, had the ethnobotanical references in the chroniclers of post-conquest Mexico been seriously studied, we would not have had to wait until this past quarter century for a knowl- edge of the hallucinogenic mushrooms and morning glories. It is, nonetheless, true that this literature must be employed with wise restraint, since much of it may be scientifically unsound. To base an entire programme of research on litera- ture alone, as some pharmaceutical houses are doing, ap- pears to me to be a highly suspect modus operandi. Recently, our herbaria have come into their own as sources of ethnobotanical observations made in the field by plant collectors of the past. These reports have several advan- tages. Unlike much of the literature, they are, in great part, firsthand; they are attached to an actual plant specimen, and there can be, therefore, no problem concerning the pro- per identification of the plant; the ethnobotanical data are anchored down, through the information on the specimen label, to a definite locality and oftentimes to specific peoples who employ the plant. The vastness of our herbarium re- sources has hitherto not fully been recognized. The Harvard 1963] Medical Botany — Schultes 109 University Herbarium, for example, has a total of some 2,200,000 sheets. There is now underway a project consist- ing of a sheet by sheet search for records of medicinal uses amongst primitive peoples." It appears probable, on the basis of the first six families studied, that we may find in the entire collection in the neighbourhood of some 3700 notes of interest. If we vouchsafe that about half of these may be unpublished or new to science, there remain for investiga- tion about 1800 reports. We know that about 40% of the notes — or about 720 — are specifically medical. And we may be justified in assuming that about half of these — or 360— might prove to be of some real interest to pharmaceu- tical science. This may give us some inkling into the pano- rama as seen from the point of view of our herbarium re- sources. The herbarium may assume an even greater “exploratory” importance with the application of the spot test for alkaloids which can be made upon very small fragments. Here we have assembled in dried form material of most of the higher plants and from the farthest corners of the globe, and a drop of reagent on a few scraps can, in a few minutes and in the quiet of our own laboratories, tell us whether or not there are alkaloids in a tree from the Tibetan Himalayas, a shrub from driest Australia or a liana from the muddy banks of the Amazon. But field work still offers our very best opportunity of discovering new physiologically active substances. Because I am emphasizing the ethnobotanical aspects of field work, I must mention the rather generalized tendency in both popular and scientific circles to over-emphasize the importance of folk medicines. Notwithstanding the fact that primitive peoples do possess valuable understanding of plant properties, their knowledge has been optimistically exagger- ated in the past and is far from being complete. It, therefore, behooves us to carry out our own phytochemical studies of the flora in general in the field, and this is probably best done ~ "von Reis, S.: “Herbaria: sources of medicinal folklore” in Econ. Bot. 16 (1962) 283. 110 Rhodora [Vol. 65 along two paths: 1) intensive examination of families and genera known to be rich in active principles and 2) a system- atic examination, species by species, of a random sampling of floras. I cannot take up this most interesting and promising type of exploration for lack of space but I do want to keep its fundamental importance in mind throughout our considera- tion of other aspects of the medico-botanical panorama. Perhaps the most satisfying way of advancing our knowl- edge of medicinal plants is direct investigation amongst primitive peoples, and I shall discuss this aspect of our research, endeavouring the while to avoid exaggeration of its importance or of suggesting that it has any special exclusivity or primacy in ethnobotanical research. The discovery of some drug plants has been attended by exciting adventure. Knowledge of others was gained from travel incidental to “conquest, colonization or religious con- version".* Still others have been found through the efforts of individuals or groups specifically sent out to learn of their identity and use. There is a tendency now to send out expeditions exclusive- ly to discover new drug plants. There is historical precedence for this procedure. Shortly after the conquest of Mexico, Philip II of Spain sent his personal physician, Dr. Francisco Hernández, to the new realm to study its medicinal plants, animals and minerals. After field work from 1570 to 1575, he had finished 16 folio volumes containing an unbelievable wealth of native medical lore. What has been published of his work is still replete with uninvestigated opportunities for modern scientists.?? This represents undoubtedly the first offiicial purely scientific expedition in history, and it was sent out for the sole purpose of trying to capitalize on the medical knowledge of the conquered peoples of a country rich in natural history. In 1714, Peter the Great ordered the establishment of the Apothecaries' Garden in Russia and commissioned the Ger- "Cheney, Ralph H: and B. L. Milana: *Medicine and plant explora- tion" in Am. Journ. Pharm. 119 (1947) 323. "Hernández, Francisco: *Rerum medicarum Novae Hispaniae the- saurus . . ." (1651). 1963] Medical Botany — Schultes 111 man botanist Messerschmidt to collect the medicinal flora of Russia between 1720 and 1727. We could cite other examples in historie times where the search for drug plants was the unique purpose of an expedition. There were other expeditions, however, which set out to study the flora in general, the useful plants as well as those which were not employed. Most of the truly outstanding explorations were of this kind. The work of the 17th Cen- tury Dutch botanist, Rumphius, basic to natural history research in the East Indies, could be cited, for he wrote about the native uses of more than 700 plants. We could mention the three botanical expeditions sent out in the late 1700's by the King of Spain: Sessé & Mocifio to Mexico; Mutis to Colombia; Ruíz & Pavón to Peru. All of these expe- ditions gave very special attention to folk-medicines while studying general floristics. Although we have returned to the custom of sending out expeditions exclusively to search for new drug plants, I am not at all convinced that this is usually the most efficient way of working in the field. It may be the quickest and most di- rect way, when we are after a specific drug, the identity of which we know and about which we have a large fund of ethnobotanical knowledge. But to visit an area in the hope that a brief sojourn amongst natives will uncover their plant remedies — no; we should not be so sanguine. Many of our official drugs have come incidental to the work of individual botanists busy with some larger project, and I am convinced that most of the new discoveries will be made by botanists, ethnobotanists or anthropologists engaged in leisurely fashion in their own research rather than by expeditions sent out “to find new drugs". And we must not minimize or overlook the role which the layman has played in this work, both in the past and in the present. Perhaps because my own expeditions were carried out in this way, I hold out greater hope for success for resident work amongst native peoples, which does not unduly empha- size the search for drug plants. Few botanists, I realize, are See De Wit, H. C. D. [Ed.]: “Rumphius memorial volume” (1959). 112 Rhodora [Vol. 65 fortunate enough to spend an almost uninterrupted twelve years in the field. Long residence in one region most certain- ly enhances opportunity for ethnobotanical discovery through the creation of a rapprochement with natives, a fa- miliarity with their languages and customs and an intimate knowledge of the flora itself. Few botanists, I realize fur- ther, are fortunate enough to work in such an ethnobotani- cally rich and untouched area as the northwest Amazon. And I realize, finally, that, in such a virgin area as the northwest Amazon, any naturalist of long residence could not but un- cover both botanical and ethnobotanical rarities and nov- elties. Like most primitive peoples, the Amazon Indian cannot comprehend any purely academic interest in plants. His only understanding is of their utilitarian or magical value. The botanist who works amongst them is easily accepted, for here is à man who spends all of his time collecting plant specimens. 'The native lives intimately with his plant en- vironment. Consequently, from the start, the botanist has a common interest with the native. During my twelve years in the Amazon, I made some 24,000 collections of the Amazon flora. Only a small frac- tion were reputedly medicinal. If an Indian helper asked, during my collecting activities, why I wanted a certain plant, the only explanation which he could understand was that I needed it for a remedy. Now, the very fact that he asked about that particular plant, and no other at the moment, indicated that perhaps he had a medicinal use for it and wanted to see if his use and my reason for wanting it coin- cided. In such a case, I would contrive the next day or so to collect the same species farther up stream, followed a few days later with another collection — all the while saying nothing about the plant. If it were really a plant of utilitari- an importance to the native, he would most certainly, after seeing me collect it several times, begin to discuss it. This "cat-and-mouse" technique uncovered many ethnobotanical facts which, had I pressed impatiently at the start, would not have been divulged. When I first went into the Colombian Putumayo in 1941, 9 1963] Medical Botany — Schultes 113 I was anxious to identify botanically “yoco”, a plant em- ployed by the Indians as the source of a strong stimulant known for years only by its vernacular name. Persistent research for nearly a year failed to uncover a flowering or fruiting specimen of the lana which clambers through the crowns of 100-foot jungle trees. I had alerted Indians far and wide of my need for flowers of yoco. Finally, after eight months, serious leg ulcers from work in inundated forests forced me to go to a small Colombian naval base on the Putu- mayo River to await a hydroplane to civilization. The clean accommodations offered me on a river gun-boat by the offi- cers were so pleasant that when an Indian, who had paddlec downstream in search of me, reported that he had seen the liana in flower, I was reluctant to leave. Yet intuition told me that I should, so back I went two days upstream and a day through flooded swamp-forest. Locating the liana, the ground under it strewn with minute white flowers, we had to fell seven trees to bring it down. But we were rewarded in being able to establish the identity of the elusive yoco, later shown to be rich in caffeine, as an undescribed species of the sapindaceous genus Paullinia.” This experience is indicative of the interest and loyalty which I found at all times amongst the Indians of the region. I cannot subscribe to the widespread belief that he regards his plant lore as something secret zealously to be guarded and that civilized man must pry it from him by ingenious duplicity. A sympathetic understanding and tolerance of his beliefs and ways and a participation in his customs do more than anything else to win the Indian’s respect and confidence. I naturally learned to chew his coca and, finding it to be not only a pleasant but a most helpful custom, used it for eight years myself. This may explain perhaps why certain Ma- kunas of the Apaporis River, where I spent a total of some three years, told me of a remote and isolated group of Tani- mukas which prepared a superior type of coca. Finally making a trip to investigate this report, I learned of a most "Schultes, Richard Evans: “Plantae Colombianae II” in Bot. Mus. Leafl. Harvard Univ. 10 (1942) 301. 114 Rhodora [Vol. 65 ingenious method of infusing into coca powder the acrid incense of the resin of the tacamahaca tree (Protium hepta- phyllum), one of the few variations ever found in the prepa- ration of this widespread narcotic masticatory, the source of the alkaloid cocaine.** Anthropological writings indicate (but we now know that they are in error) that the narcotic snuff yopo, prepared from the seeds of Piptadenia, is employed throughout the upper Orinoco and much of the Amazon basin. I was puzzled, however, by my failure to encounter a single tree of this plant in the northwest Amazon. Yet the witch-doctors em- ployed a highly narcotic snuff which was not tobacco. What could it be? Had I been insistent, I might never have known. After nine years, one of my Puinave boys, himself the son of a medicine-man, suddenly said one day: “This is the tree that gives yakee snuff.” Controlling my excitement, I showed only mild interest. We decided to prepare snuff from it. From strips of bark, the boy scraped off a reddish exudate, mixed it with water, boiled it down in four hours to a thick syrup which he allowed to sun-dry. The resulting solid, pul- verized and mixed with bark-ashes of a wild cacao tree, gave us the snuff. Since I believe in experimenting with these plant products personally in the field, I took one-quarter the dose normally snuffed by a medicine-man to produce the psychotomimetic effects essential for his divination and di- agnosis of disease. I was ill in my hammock for several days, so strong was the snuff. Strangely enough, the source of yakee snuff belongs to the myristicaceous Virola and is, therefore, related to our nutmeg, which has itself been em- ployed as a narcotic.” The moral from this experience is that patience will usu- ally pay good dividends in this kind of work, but time for such patience is not to be had on the usual expedition set up with the purpose of looking for medicinals and nothing else. ?Schultes, Richard Evans: “A new method of coca preparation in the Colombian Amazon" in Bot. Mus. Leafl. Harvard Univ. 17 (1957) 241. ?Schultes, R. E.: “A new narcotic snuff from the northwest Amazon” in Bot. Mus, Leafl. Harvard Univ. 16 (1954) 241. 1963] Medical Botany — Schultes fa" | tr 115 A IA M y^ d j / i N UNI - 116 Rhodora [Vol. 65 I might go on in such a vein, telling how, without prying or seeking to outwit the Indian, knowledge of folk-medicines accumulated along with my general botanical work. I could relate the unexpected cireumstances which led me to infor- mation about plants employed as styptics, to treat conjunc- tivitis, as snake-bite remedies, in treating burns — not to mention the many everyday uses such as carminatives, febrifuges and purgatives. I could tell about the fascinating detective work with curares which uncovered for the first time the use as a basic arrow poison component a species of Thymelaeceae, Sehoenobiblus peruvianus. Or I might relate how participation in native dances and rituals when hal- lucinogens are taken gave me a deeper understandine of the extensive use of the numerous vision-producing narcotics of the region, some hitherto botanically unknown. I might ex- plain how there are still psychotomimetic agents lurking un- identified in these forests for future study. It might be in- teresting to discuss my work in Mexico on the sacred Mexi- can mushrooms and on the narcotic morning glory. Rivea corymbosa. It would be interesting to relate the circum- stances which led me to identify three species of aroids used in different tribes in various parts of the Amazon of Colombia as oral contraceptives. I could occupy many pages discussing the many poisonous plants, some of them species new to science, known and used by these people. But the panorama is too vast to allow a discussion of all of its details. V. Final Considerations This widening panorama — can we avail ourselves of it? The answer, at the present time, is: “No”. The reason is simple: we do not have the trained man-power for the effort. There is no reason why we cannot rectify this short-coming, and there are vaguely discernible signs that perhaps we are beginning to take steps in this direction. There is an urgency to our training of the needed corps of investigators. Civilization is on the march in many, if not “Schultes, R. E.: Tapping our heritage of ethnobotanical lore" in Econ. Bot. 14 (1960) 257. 1963] Medical Botany — Schultes 117 SCHOENOBIBL US eruianus c, tande. 118 Rhodora [Vol. 65 most, primitive regions. It has long been on the advance, but its pace is now accelerated as the result of world wars, extended commercial interests, increased missionary activ- ity, widened tourism. The rapid divorcement of primitive peoples from dependence upon their immediate environment for the necessities and amenities of life has been set in motion, and nothing will check it now. One of the first aspects of primitive culture to fall before the onslaught of civilization is knowledge and use of plants for medicines. The rapidity of this disintegration is frightening. Our chal- lenge is to salvage some of the native medico-botanical lore before it becomes forever entombed with the cultures that gave it birth. Though it is by no means an insurmountable task, it will not be the easiest accomplishment to prepare enough men thoroughly to cope with all the ramifications of the widen- ing panorama in medical botany. They may be basically anthropologists, botanists, medical or pharmaceutical scien- tists, but they must have an interdisciplinary training. Here is where all of us, as members of teaching and training insti- tutions, must contribute to progress in medical botany. We should watch for the student keen for this kind of research, appraise him of the sundry needs in the field and counsel him as to the best type of preparation and where it may best be acquired. This will demand on our part continual alertness and an imaginative flexibility in our counselling. We cannot now fail medical science, for never before has history offered us similar opportunities. The task before us is big, but never, I am certain, have we had the human material, scientific tools and financial support more adequate to a frontal attack and, cognizant as all of us must be of the virgin fields that lie ahead, we must utilize these resources conscientiously and efficiently. I cannot close with words more appropriate than those written in 1754 by the great Linnaeus himself in the preface to a museum catalogue. In expressing what has been aptly described as his creed, words most appropriate for us as we *See Krutch, J. W.: “The gardener’s world” (1959) 177. 1963] Medical Botany — Schultes 119 c / PE, | ^ DES ty A SE m yo SN ^ NC C1 2 t R LE AY oo d y D KE m SS iA EN " = Y N B \ AS he e N G e Zš NV =S, corymbosa GC) Tta tl fe 120 Rhodora [Vol. 65 stand upon the threshold of great new advances, he said in part: “Man, ever desirous of knowledge, has already explored many things; but more and greater still remain concealed; perhaps reserved for far distant generations, who shall prosecute the examination of their Creator's work in remote countries, and make many discoveries for the pleasure and convenience of life. Posterity shall see its increasing Mu- seums, and the knowledge of the Divine Wisdom, flourish together; and at the same time all the practical sciences . . . shall be enriched; for we cannot avoid thinking, that what we know of the Divine works are much fewer than those of which we are ignorant." BOTANICAL MUSEUM HARVARD UNIVERSITY HELIANTHUS LAETIFLORUS AND HELIANTHUS RIGIDUS — HYBRIDS OR SPECIES? SARAH CLEVENGER AND CHARLES B. HEISER, JR. The identification of perennial sunflowers has long been a persistent source of difficulty, resulting in a large part from the blurring of species lines by hybridization. Through field and experimental studies, it has been established that interspecifie hybridization between diploid members of the genus is common (see Heiser et al., 1962, for references). In the present paper four hexaploid taxa, Helianthus laeti- florus Pers., H. rigidus (Cass.) Desf., H. subrhomboideus Rydb. and H. tuberosus L. are considered. Although H. tuberosus has been accepted as a species by all who have worked on the genus, the other three taxa have been treated in a variety of ways. Watson (1929) recog- nized H. laetiflorus and H. rigidus as distinct species and considered H. subrhomboideus, a synonym of the latter; he also described two new species, H. suberbus and H. severus, which we feel should be referred to H. laetiflorus. Fernald (1946) recognized but a single species, H. laetiflorus and treated rigidus and subrhomboideus as varieties. Cronquist (1952) accepts but a single species, H. laetiflorus, and recog- nizes no varieties. We would like to extend thanks to the curators of the following herbaria: University of Arkansas, Duke Univer- sity, Gray Herbarium, Michigan State University, Missouri Botanieal Garden, New York Botanical Garden, University of Wisconsin, and the United States National Herbarium. The distributions shown in the maps have been compiled from the specimens in these herbaria and the ones at Indi- ana University. We would also like to thank Prof. J. Lean- dri of the Museum National d'Histoire Naturelle, Paris, who furnished us with the photograph of the type of H. laeti- florus. Type material of H. subrhomboideus in the herbari- um of the New York Botanieal Garden has also been ex- amined, but we have not yet been able to secure authentic material of H. rigidus and our interpretation rests on Cas- 121 122 Rhodora [Vol. 65 sini’s original description. A grant to the senior author from the American Philosophical Society made it possible to examine certain European types. This research was also aided by a grant from the National Science Foundation to the junior author. DISTRIBUTION: Helianthus subrhomboideus (Fig. 1) is a prairie plant widely distributed in Canada from Alberta east to Quebec and extending southward through the plains to western Texas and New Mexico. Helianthus rigidus (Fig. 2) is found in the central United States and in the northern and western portion of its range is sympatric with the pre- ceding species. These two species may be found occasionally in the eastern United States as adventives or escapes from cultivation. The third taxon, H. laetiflorus (Fig. 3), has a scattered distribution in the central United States where it is usually associated with both H. rigidus and H. tuberosus and has a spotty distribution in the east, where it has been widely cultivated as a garden ornamental. Many plants of H. laetiflorus from the east are seed sterile probably indi- cating that they came from members of a single clone. The last species, H. tuberosus, (Fig. 4) has a wide distribution in central and eastern North America, broadly overlapping the areas of the previous two species, and extending into the range of H. subrhomboideus in the north central states. In general, it grows in somewhat wetter habitats than do the others. As the “Jerusalem artichoke,” it has been wideiy cultivated and since it also readily escapes, it is practically impossible to determine its prehuman distribution. All of these species reproduce vegetatively from rhizomes or tubers and are quite aggressive, frequently beoming established as escapes. MORPHOLOGY: In their extreme forms, the four taxa con- sidered here are quite distinct, but intergradations are fre- quent in nature. The principal features of each are given in Table I. Many characters, such as leaves, are extremely variable and, hence, it is difficult to give precise measure- ments. The phyllaries (Fig. 5) seem to offer the most stable 123 Clevenger & Heiser, Jr. Helianthus 1963] 'uopaer) [P9oruejog Linossi]y ey} pue "if ‘uospooM "M "d “Ad JO Ásojinoo əy} qu3noiqgj pesn dew eseg -c5nso42qm] 'H pue snaoyijon] `H ‘SNP ‘H ‘ssnapoquoyaqns *Hg jo uornnqugsip əy} Summoys sdeW "p-T amns ^ 4 Nes ES snyo41 2V37 M { P )We I 124 Rhodora [Vol. 65 Figure 5. Leaves and phyllaries of Helianthus subrhomboideus (A) from Heiser 3111; H. rigidus (B), Steyermark 9064; H. laetiflorus (C), Breitung 6154; and H. tuberosus (D), Ownbey 1166. Leaves, X 2/5. Phyliaries, slightly enlarged. charac ers, and perhaps are of the greatest value in delimit- ing the taxa. From the table it can be seen that H. subrhomboideus and H. tuberosus represent the extremes. Helianthus laetiflorus has no unique features and is largely intermediate between H. rigidus and H. tuberosus. Helianthus rigidus, on the other hand, approaches either H. subrhomboideus or H. lael- iflorus or is intermediate between them. A scatter diagram (Fig. 6), based on herbarium material, illustrates this situ- ation for phyllary length and leaf length. Artificial Hybrids: Reciprocal hybrids, excepting the 1963] Helianthus — Clevenger & Heiser, Jr. 125 PHYLLARY LENGTH IN MM. Pa r bt Pus. a.a. e O A 2 mg e. PD, E ^ gg E Spg . a H -e * H a $5.4 o oH x < x x S Z x x < o S E = x= Kn x m d = x < 2 See > on mM x o o Bu oes xX x X ® x = di r ME. « x ST ER m x « o r 4 x NUS «d T x < om x n B èS = = o È S - a m y o ES g oe z mw " - ER S i P s x o EE s ES PR. CE yt - = Th © $ 2 o SN E o o ao - D g a < S gc TP ct [eee ri y- < combination, H. subrhomboideus X H. laetiflorus, have been made involving the four taxa discussed above. All of the nybrids obtained were vigorous and fertile and largely in- termediate morphologically with the exception of those so indicated below. Hybrids have also been made between sister plants and between different races of H. subrhom- boideus and H. tuberosus, and all gave progeny closely resembling their parents. Since all of these species are normally self-incompatible, it has been impossible to obtain 126 Tuber Stem height Leaves surface length color shape petiole Peduncles Disk-flowers Phyllaries shape pubescence margin surface length Rhodora TABLE 1 [Vol. 65 COMPARISON OF CERTAIN MORPHOLOGICAL FEATURES OF FOUR HEXAPLOID SUNFLOWERS H. subrhomboideus H. rigidus lacking usually simple 0.5-1.0 m, opposite scabrous below 5-12 em. grey-green to light-green linear to rhombic or ovate very short long, leafless dark, red-brown tightly appressed elliptical to oblong-ovate conspicuously ciliate glabrous shorter than disk 5-10 mm. lacking simple to branched 0.8-2.0 m. opposite to alternate scabrous below 8-27 cm. grey-green to light-green lanceolate to ovate short to medium long, nearly leafless red, rarely predominately yellow tightly appressed oblong-ovate conspicuously ciliate glabrous shorter than disk, 5-12 mm. H. laetiflorus occasionally present branched 1.5-2.5 m. upper alternate scabrous below 12-25 em. grey-green to dark-green lance-ovate medium medium long with reduced leaves yellow or only slightly red appressed oblong- lanceolate conspicuously ciliate glabrous to pubescent equalling the disk, 5-12 mm. H. tuberosus frequently present branched 1.8-2.6 m. upper alternate glabrous to puberulent or tomentulose below 14-23 cm. dark-green lance-ovate to ovate long short, leafy yellow loose, sometimes reflexed lanceolate inconspicuously ciliate frequently pubescent longer than disk, 11-18 mm. selfs. The small population size of most of the hybrid popu- lations resulted from failures of seed germination which is not uncommon among perennial sunflowers. Herbarium specimens of the parents and representatives of the hybrid populations are deposited at Indiana University. H. subrhomboideus X H. tuberosus (P29 x P59). The 15 1963] Helianthus — Clevenger & Heiser, Jr. 127 reciprocal F,’s were largely intermediate, but with a leaf shape nearer H. subrhomboideus and phyllaries with acute tips similar to that of H. laetiflorus. Seventeen F. plants were secured which showed a variety of leaf shapes ap- proaching, but not reaching the parental types. All had phyllaries similar to H. laetiflorus and the amount of stain- able pollen varied from 70 to 95%. Most of the F, plants would have been identified as H. laetiflorus had they been collected in the wild. Meiosis in the F; showed mostly pairs with a few multivalents, which is similar to the pairing found in the parental species. An F, generation between P29 x P21 gave plants essentially similar to the above. In an F, of 29 plants of this combination most of the plants were nearer to H. rigidus, although some broad leaf and yellow disk segregation appeared. Thirteen of these plants showed considerable weakness and died before maturity. H. subrhomboideus X H. rigidus (P79 x P147A). The seven F,’s were for the most part intermediate in morpholo- gy and fertile. H. rigidus X H. tuberosus (P147A x P21). The three F. plants obtained were intermediate except for the leaves which were nearer to those of H. tuberosus. These plants were all fairly good matches for H. laetiflorus. An F, of 24 plants was grown and these were similar to the F, plants or closely approached H. tuberosus in appearance. H. rigidus X H. laetiflorus (P147A x N3). The F, gener- ation of twenty plants was fairly uniform, and the plants resembled typical H. rigidus except for the more leafy ped- uncles and variations in leaf shape and size. H. rigidus X H. laetiflorus (P147A x B). Seeds for both of the parent plants came from the same population in White County, Illinois. Only three F, plants were obtained. Two of these were very similar to P147A, but the third had yellow disks and leaves resembling those of H. tuberosus in shape. In a second cross (P147 x Ar1) all of the 14 plants secured were more like the H. rigidus parent but one plant was completely sterile, two showed low pollen stainability (45 and 59%), and three were quite weak. A third combi- nation (N3 x 147A) gave 11 plants which while showing 128 Rhodora [Vol. 65 some plant to plant variation were essentially similar to H. rigidus and were all fertile. H. laetiflorus X H. laetiflorus (P147B1 x 2). Although this cross was made several times, only one sced was ever obtained. The plant secured from it was nearly sterile and had leaves resembling H. tuderosus, but with shorter pet- ioles and a dense pubescence quite unlike that found in any of the species under consideration here. It is unlikely that contamination could have occurred, and moreover it is diffi- cult to visualize any species that could have crossed with H. laetiflorus to produce such a hybrid. It is possible that the plant represents a type of recombinant which is not success- ful in the wild. More than 50 seeds were obtained in a cross involving two different accessions of H. laetiflorus (N3 x Arl) but only one seed germinated giving rise to a fertile plant, approaching H. rigidus in appearance. H. laetiflorus X H. tuberosus (P147B x P22). Thirty- seven F,’s, were secured, all of which appeared very similar to the H. tuberosus parent except for the slightly shorter phyllaries. Some of the plants showed reduction in seed set when intercrossed or backcrossed, but when open pollinated gave good seed set. Fifty F, plants were grown which gave segregates approaching the parental types as well as the plants similar to the F,. A second cross of these two species (H408 x Arl) gave only five seeds, one of which germinated to give rise to a plant with 93% pollen stainability and show- ing some features of both parents. These crosses indicate that hybrids involving these four taxa are readily obtained, and that with the exception of certain crosses involving H. laetiflorus are fertile and vigor- ous. Moreover, it is clear that hybrids between either H. subrhomboideus or H. rigidus with H. tuberosus give plants resembling H. laetiflorus. It is unfortunate that more hy- brids of H. laetiflorus X H. laetiflorus were not secured, but it does appear significant that neither of the two plants ob- tained from this cross resembled their parent. PROGENY TESTS: P68. H. laetiflorus. Three plants were secured which showed considerable variation, particularly in regard to leaf shape and size. Plants of H. tuberosus are 1963] Helianthus — Clevenger & Heiser, Jr. 129 known to occur in the area and could have served as the pollen parent. L1. The parent plant was somewhat intermediate be- tween H. tuberosus and H. laetiflorus. One of the three off- spring resembled H. rigidus and the other two approached H. tuberosus. L2. H.laetiflorus. The 11 offspring produced little pollen but the stainability ranged from 60 to 9096 with a mean of 79%. The plants all showed various combinations of char- acters of H. laetiflorus and H. tuberosus. Although the pollen parent is unknown for all of these plants, they are of significance in showing either that natur- al eross pollination takes place between species or that H. laetiflorus shows segregation. DISCUSSION: Although the four hexaploid sunflowers, H. subrhomboideus, H. rigidus, H. laetiflorus, and H. tubero- sus, are quite distinct in their extreme form, they intergrade so freely that it is difficult to describe limits to these taxa. This blurring of the species boundaries could be explained by natural hybridization, and it is possible that two of the taxa, H. laetiflorus and H. rigidus, are of hybrid origin. Plants which have been identified as H. laetiflorus could represent F, hybrids or hybrid derivatives of H. tuberosus x H. rigidus (or perhaps H. subrhomboideus). The evidence for this conclusion rests on the following: (a) H. laetiflorus has no unique characters, but combines various features of its putative parents; (b) artificial hybrids between the post- ulated parents can be obtained and would be classed as H. laetiflorus; (c) H. laetiflorus apparently does not breed true from seed; and (4) this taxon is generally found in areas where the two supposed parents grow. The species was described in 1807 from plants grown in Europe, and al- though only one of its putative parents was known there at this time, we nevertheless feel that it is most likely that the type is based on a hybrid (Heiser, 1960). Helianthus rigidus might be postulated to have a hybrid origin from H. subrhomboideus X H. tuberosus. Helianthus rigidus shows no character which could not have come from 130 Rhodora [Vol. 65 the supposed parents, although it does occur in areas outside of the range of one of the “parents” (Fig. 2). If it were to have originated from introgression of H. tuberosus into H. subrhomboideus, however, selection for certain introgressive types might have resulted in constancy and the introgres- sants might have been able to invade new areas or effectively replace H. subrhomboideus. Since the species are all polyploids, certain other explana- tions for the intergradation must be explored. Although at present we cannot definitely identify all the diploid progeni- tors, it is likely that the hexaploids are partly or entirely allopolyploid in origin. Kostoff (1939) on the basis of analy- sis of the hybrid between H. annuus and H. tuberosus has suggested the genomic formula A4, A, A» Ag B. B, for H. tuberosus with the B genome being nearly homologous with that of H. annuus. The 4. genomes could come from one of the three closely related tetraploids of the eastern United States, H. hirsutus, H. strumosus, or H. decapetalus which themselves may be of alloploid origin (Smith, 1961). Most hybrids of the diploid perennials exhibit good pairing (Heiser, et al., 1962) which could account for the similarity of the two A genomes postulated by Kostoff for H. tubero- sus. Let us therefore, for the present study, assume a gen- omic formula for H. tuberosus of A,A,A.A.BB. Helianthus laetiflorus might conceivably be A,A.A,A,BB ; H. rigidus, A.A,A,A,BB ; and H. subrhomboideus A,A,A,A,BB, where A,, A., etc. represent genomes from closely related peren- nial diploids. Such a hypothesis would explain the essential- ly good pairing and interfertility between all the hexaploid species as well as the morphological similarities and differ- ences. Segregation in such polyploids might give results re- sembling interspecific hybridization. Stebbins (1950) has pointed out that segmental allopolyploids, in contrast to strict alloploids, may give segregates approaching one or the other of the parents. Smith (1961) found plants of the tetraploid H. strumosus apparently showing introgression from H. hirsutus, even though the latter did not grow in 1963] Helianthus — Clevenger & Heiser, Jr. 181 the same area, and he suggested the possibility of allotetra- ploid segregation. Although we have no evidence that allo- polyploid segregation is occurring in the hexaploids, it nevertheless offers an alternative hypothesis to interspecific hybridization. It is, of course, feasible that we may have various combi- nations of the possibilities discussed above — part of the variability may result from allopolyploid segregation, part from hybridization. Considering all the possibilities, how- ever, we feel that there is fairly good evidence that H. laeti- florus is a hybrid. On the other hand, while H. rigidus may owe its origin to hybridization, we do not feel that any defi- nite decision can be reached at present. In regard to taxonomic treatment, this group of taxa offers certain difficulties inherent in many plants of poly- ploid origin which have no barriers to interbreeding. Heli- ant’ us tuberosus, in spite of some intergradation with other species, deserves recognition as a species. On the other hand, if H. laetiflorus is a collection of hybrids or hybrid deriva- tives, it does not deserve designation as a species. Although in general, we are opposed to giving distinct names to hybrids, in the case of a well established binomial, it is perhaps best to continue its use with indication that it is a hybrid.’ The authors are not in agreement as to the treat- ment of the other two taxa. One feels that H. subrhomboid- eus should be treated as a species and H. rigidus as a hybrid, whereas the other thinks that they should be considered races of a single species. If the latter course is adopted, H. subrhomboideus becomes a subspecies of H. rigidus. SUMMARY: The four hexaploid sunflowers, H. subrhom- boideus, H. rigidus, H. laetiflorus, and H. tuberosus form a morpholcgical series with the first and last named species representing the extremes. Artificial hybrid combinations of all the species, except H. subrhomboideus X H. laetiflorus, were largely fertile. Certain of the hybrids of H. tuberosus x H. subrhomboideus and H. rigidus strongly resemble H. "The name then becomes H. x laetiflorus Pers. (pro. sp.) Syn. 2: 476. 1807. 132 Rhodora [Vol. 65 TABLE II SOURCE OF PLANTS USED IN THE CROSSES AND PROGENY TESTS Number Name Locality and Collector Arl H. laetiflorus Cult. Monroe Co., Ind. (Heiser) H408 H. tuberosus Darke Co., Ohio (R. T. Neher) L1 H. laetiflorus-tuberosus Auburn, Mass. (B. N. Gates) L2 H. laetiflorus Worcester, Mass. (B. N. Gates) N3 H. laetiflorus White County, Ind. (Heiser) P21 H. tuberosus Polk Co., Iowa (Heiser) P29 H. subrhomboideus Cass Co., N. D. (O. A. Stevens) P59 H. tuberosus Hennepin Co., Minn. (Gerald Ownbey) P68 H. laetiflorus Cult. Monroe Co., Ind. (Heiser) P79 H. subrhomboideus Lawrence Co., Ind. (Heiser)’ P147A H. rigidus White Co., Ill. (Heiser) P147B H. laetiflorus White Co., Ill. (Heiser) laetiflorus. On the basis of the study of morphology, geo- graphical distribution, and the hybrids, it is suggested that H. laetiflorus is of hybrid origin and that H. rigidus possibly represents an introgressant. On the other hand, it is also possible that allopolyploid segregation is occurring in these species giving results suggestive of hybridization. BEREA COLLEGE, BEREA,KENTUCKY, INDIANA INDIANA UNIVERSITY, BLOOMINGTON, INDIANA LITERATURE CITED CRONQUIST, A. 1952. Compositae in H. A. Gleason, The New Britton and Brown Illustrated Flora. Vol. 3. FERNALD, M. L. 1946. Technical Studies on North American Plants. Rhodora 48: 65-81. HEISER, C. B. 1960. Notes on the origin of two ornamental sun- flowers, Helianthus multiflorus L. and H. laetiflorus Pers. Baileya 8: 146-149. , WILLIAM MARTIN, AND D. M. SMITH. 1962. Species crosses in Helianthus. I. diploid species. Brittonia 14: 137-147. — ?Plants from this same colony (Kriebel 3965) were originally identi- fied as H. silphioides Nutt. by Fernald (1946) who later admitted his misidentification in a letter to the junior author. So far as is known H. silphioides does not occur in Indiana. 1963] Helianthus — Clevenger & Heiser, Jr. 133 KosTorF, D. 1939. Autosyndesis and structural hybridity in F, hy- brid Helianthus tuberosus L. x Helianthus annuus L. and their sequences. Geneties 21: 285-300. SMiTH, D. M. 1961. Variation in the tetraploid sunflowers, Helian- thus decapetalus, H. hirsutus and H. strumosus. Recent Advances in Botany 1: 878-881. U. of Toronto Press. STEBBINS, G. L. 1950. Variation and Evolution in Plants, Columbia University Press. : Watson, E. E. 1929. Contributions to a monograph of the genus Helianthus. Pap. Mich. Acad. 9: 305-475. A NEW FEATURE IN BAMBOO RHIZOME ANATOMY! F. A. MCCLURE? In the course of my search for vegetative characters that effectively differentiate the two currently recognized species of Arundinaria native to the United States I discovered that, in cross section, the rhizomes of A. tecta (Walt.) Muhl. show well-developed peripheral air canals (fig. 1),° a feature of bamboo anatomy apparently not hitherto reported. In both living plants and dried herbarium specimens, this feature was present in all rhizomes associated with flowering ma- terial identified as belonging to this species. I have not, however, found air canals in the rhizome of any plant or specimen typical of A. gigantea (Walt.) Muhl. (fig. 2) the only other currently recognized species native to this coun- try. This anatomical feature is presumed to originate in the same manner as similar spaces found in the stems of some herbaceous grasses of the genera Leersia, Oryza, Sacciolepis, ete. (cf. C. R. Metcalfe, Anatomy of the Monocotyledons, I. Gramineae. The Clarendon Press, Oxford. Ixi and 731 p., illus., bibl. 1960.) Anatomical details shown by cross sections of bamboo rhizomes are illustrated by Shibata (1900, fig. 2 & 3) for A. nipponica (as Bambusa nipponica) and for A. japonica; by Takenouchi (1931, pl. 1, fig. 2-4) for 3 species in 2 gen- era; and by Takenouchi (1932, fig. 7-10) for 7 species in 4 genera. No illustration or mention of the presence of air canals in the bamboo rhizome appears in any of these works. No other illustration or discussion of bamboo rhizome anato- my has come to my attention. This paper was prepared with the assistance of a grant from the National Science Foundation. The facilities made available by the Department of Botany, U. S. National Museum, Smithsonian Institution have been indispensable. "Honorary Research Associate, Smithsonian Institution. 3Rhizome sections and the photographs on which the illustrations are based were prepared by Professor Yong-no Lee, on leave from the Department of Biology, Ewha Woman's University, Seoul, Korea, using facilities made available by the Department of Botany, Yale University. Fig. 1. Cross section of a young rhizome of A. tecta, showing air canals, ca. X 25. Study material taken from living plants originally native at Stony Run, Anne Arundel Co., Md., cultivated under MBG 2762 at 5507 Charles Street, Bethesda, Md. re $ INNS Plate 1884 Fig. 2. Cross section of a young rhizome of A. gigantea, showing absence of air canals, ca. X 25. Study material taken from living plants originally native at Leban- on, Warren County, Ohio, cultivated under MBG 2792 at 5507 Charles Street, Bethesda, Md. 136 Rhodora [Vol. 65 Although the presence of air canals in their rhizomes may be presumed to have certain adaptive advantages to plants growing in waterlogged soil or on frequently inun- dated sites, its presence or absence apparently is not, in our native bamboos, influenced in any way by the ecological conditions under which the plants happen to be growing. This is presumed, therefore, to be an inherited, rather than an ecological manifestation. The results of exploratory studies in field and herbarium indicate that the presence or absence of air canals in the rhizomes can be used with confidence to differentiate plants typical of these two species, even in the sterile condition. This feature was originally discovered by means of a 9 X hand lens, and the presence or absence of air canals can be diagnosed easily without the use of a compound microscope. Air canals in the rhizomes of Arundinaria tecta are typically continuous, longitudinally, through the nodes and inter- nodes, and are typically distributed uniformly around the periphery of the rhizome axis. However, two specimens have been found, in which their distribution is discontinu- ous, both peripherally and longitudinally: Biltmore 1405 [leg. C. D. Beadle ?], and Radford & Wood 6879-A, both from North Carolina. LITERATURE CITED SHIBATA, K. 1900. Beiträge zur Wachstumsgeschichte der Bam- busgewüchse. Jour. Coll. Sci. Imp. Tokyo 13:427-496. pl. 22-24 (fig 1-61). TAKENOUCHI, YOSHIO. 1931. Systematisch-vergleichende Morpholo- gie und Anatomie der Vegetations-organe der japonischen Bam- bus Arten. Taihoku Imp. Univ. (Formosa) Fac. Sci. Mem. v. 3, no. 1 (Bot. no. 2) p. 1-60. (illus). 1932. [Studies of Bamboo] [Tokyo] 291 p., 181 text fig., 31 tables. Bibl. (Text in Japanese). ' POLLEN MORPHOLOGY AS A TAXONOMIC TOOL IN LINUM! K. S. XAVIER AND C. MARVIN ROGERS Among modern palynologists, Wodehouse (1935) and Erdtman (1952) most successfully show the use of pollen characters in taxonomy. The latter brings the knowledge of pollen morphology in 98 families of angiosperms, including the Linaceae more or less up to date. He classifies and de- scribes the pollen of about 36 species, representing 14 genera of the family, including, however, only five species of Linum, all of which are of the Old World. Small (1907) has provided the most thorough taxonomic treatment of the genus in North America, recognizing 60-70 species (Linum and Cathartolinum), but the relationships of these to one another and to species in other parts of the world have never been too clearly known. It was hoped that knowledge of the comparative morphology of the pollen might provide additional information and, for that reason, preliminary studies were undertaken, of which the present paper gives some results. Since these are sufficiently en- couraging, a more extensive survey of the genus is now under way. Fourteen species and six varieties of Linwm were ex- amined, principally eastern North American taxa, but including certain selected western North American and cul- tivated species. Some of the pollen grains were collected from living plants, but most have come from herbarium specimens, since such grains are not only intact and perfectly usable, but scarcely differ from those taken from living plants of the same species. The pollen was prepared according to Brandt’s method as described in Wodehouse (1935) except that, in order to view the surface sculpturing better, most of it was left unstained. The grains were removed from the anther, placed on a slide, ‘Contribution No. 96 from the Department of Biology, Wayne State University. 138 Rhodora [Vol. 65 cleared in 95% alcohol and embedded in pure glycerin jelly. At ordinary temperatures, these slides are permanent. The shape as well as the external features of the pollen may be easily observed when the grains are floated in 95% alcohol in a deep well slide. Two significant facts became quickly apparent. One, the pollen grains within most species show comparatively little variation, even when the plants have come from widely sep- arated localities. Two, there is a rather remarkable amount of diversity among the different species. Because of their variation and the clearness with which they can be seen, the pollen characters found to be most helpful in differentiating species are the size and shape of the grain, number, size and shape of the germ pores or fur- rows, the thickness of the exine and the nature of its sculp- turing. Diameter of the grain varies from 38.5-112 u. Shape ranges from triangular to subspherical. Triangular grains have three meridionally extended germ furrows which may be acute or blunt at the ends and vary in width and length in different species. Spherical grains have about 20 more or less circular pores while intermediate types have the sur- face undulated due to the protruded germ pores which are six to twelve in number and oblong in shape. Thickness of the exine ranges from 2-9 ,. The sculpturing of the exine consists of verrucoid (wartlike) exerescences, which may be mono-, di- or polymorphic and range from 0.5-5.25 p in diameter. They may be rounded or many sided as viewed from above and be rounded or irregularly truncate at the summit. The excrescences vary in abundance and on the furrow membrane they may be absent, few, or as dense as on the other sporoderm surfaces. Though some species are more clearly distinct than others, it is possible to construct a key based on pollen characters alone, by which most of the species studied may be identified. The sizes given in the key and in the descriptions which fol- low were obtained by making three to five series of measure- ments of each specimen. In using the key, one must take care to choose only polar views of fully expanded, unbroken grains. 1963] Pollen Morphology — Xavier & Rogers 139 Plate 1285 Fig. 1-7 Pollen grains of Linum. Fig. 3 and T X 600; others X 750. 1. L. lewisii; 2. L. striatum; 3. L. usitatissimum; 4. L. lewisii; 5. L. bahamense var. bracii; 6. L floridanum var. floridanum; T. L. rigidum var. rigidum. 140 Rhodora [Vol. 65 L.rigidum L.sulcatum GROUP IV | L. intercursum L. westii L. floridanum GROUP Ill L.medium vor. medium | L.medium var. texanum L. striatum, L.virginianum i L. lewisii L. usitatissimum L.rupestre L. catharticum GROUP | L. arenicola L.bahamense GROUP |l Fig. 8. The species of Linum studied, arranged in four groups on the basis of pollen morphology, with Group I being most primitive and Group IV most advanced. KEY TO THE SPECIES Germ furrows three. Germ furrows pointed at the ends (Group I) (Fig. 1). Excrescences 0.75-3 u in diameter; intine broadly rounded at the furrows (Fig. 4). Diameter of grains 38.5-45.5 u. 1. L. catharticum Diameter of grains 52-73.5 u. Germ furrows 27 u wide. 2. L. rupestre Germ furrows 10.5-21 u wide. Excrescences 0.5-2 u in diameter, 84-90 per 10 w (Fig. 5). 3. L. arenicola and 4. L. bahamense Excrescences 1.5-3 & in diameter, 28-30 per 10 x’ (Fig. 4). 5. L. lewisii Excrescences 0.5-0.75 u in diameter; intine subacute at the fur- rows (Fig. 3). 6. L. usitatissimum Germ furrows rounded at the ends (Group II) (Fig. 2). Exine 2 u thick; excrescences 70-72 per 10 x’. 7. L. striatum 1963] Pollen Morphology — Xavier & Rogers 141 Exine 2.75-3.5 u thick; excrescences 40-53 per 10 z. 8. L. virginianum and 9. L. medium var. texanum Germ pores six to twenty. Germ pores six to twelve; surface of grain undulated (Group III) (Fig. 6). Germ pores six. Germ pores tapering toward the ends, 14 » wide; excrescences in a central patch on the pore membrane. 10. L. westii Germ pores subcircular, 21 4 in diameter; excrescences uniform- ly sparse on the pore membrane. Excreseences with central pits, 40-44 per 10 p. 9. L. medium var. medium Excrescences without central pits, 67-75 per 10 4. 11. L. floridanum var. chrysocarpum Germ pores ten to twelve. 11. L. floridanum var. floridanum and 12. L. intercursum Germ pores about twenty; surface of grain not undulated (Group IV) (Fig. 7). Exine 5.25-7.5 u thick. Grains 70-87.5 » in diameter; largest excrescences 3 u in di- ameter. 13. L. suleatum Grains 84-112 uw in diameter; largest excrescences 5.25 mw in diameter. 14. L. rigidum var. rigidum Exine 8.75-9 u thick. 14. L. rigidum var. carteri DESCRIPTIONS OF THE SPECIES 1. L. catharticum L. Compressed triangular, diam 38.5-45.5 4; colpi 3, very long, narrow, pointed, 7-10.5 u wide; exine 3.5-4.75 u thick; ex- crescences monomorphic, diam 1.75 y, many sided, sparse, absent on furrow membrane. 2. L. rupestre Engelm. Compressed, subtriangular, diam 70-73.5 4; colpi 3, pointed, 27 4 wide; exine 3 uw thick; excrescences polymorphie, diam 0.75-2.5 u, rounded, with central pits, very sparse on furrow membrane. 3 L. arenicola Small. Oblate triangular, diam 56-66.5 u; colpi 3, long and pointed, 17.5-21 & wide; exine 3 u thick; excrescences polymorphie, diam 1-2 4, rounded, with minute central pits, very sparse on the slightly bulging germinal furrows. 4. L. bahamense Northrop, including var. bracii (Small) Rogers and var. corallicola (Small) Rogers. Oblate triangular, diam 52.5-66.5 H; colpi 3, long and pointed, 10.5-21 4 wide; exine 2.25-3 u thick; excres- cences rounded, di- or polymorphie, diam 0.5-1.5 u, very sparse on the furrow membrane. 142 Rhodora [Vol. 65 5. L. lewisii Pursh. Oblate triangular, diam 63-73.5 u; colpi 3, 21 u wide, pointed; exine 3-4.5 u thick; excrescences di- or polymorphic, diam 1.5-3 4, rounded or many sided with central pits, very sparse on furrow membrane. 6. L. usitatissimum L. More or less oblate triangular, diam 52.5-59.5 u; colpi 3, short, pointed, 7-10.5 u wide; exine 3.5-4.25 u thick; excres- cences dimorphic, diam 0.5-0.75 4, absent on furrow membrane. 7. L. striatum Walt. Oblate triangular with blunt angles, diam 42-49 #; colpi 3, oblong, 14-24.5 u wide; exine 2 y thick; excrescences mono- morphic, diam 1.5 a, many sided, with central pits, fewer on the fur- row membrane. 8. L. virginianum L. Oblate triangular with blunt angles, diam 45.5- 59.5 u; colpi 3 (one specimen with 4), elliptic, 21-28 u wide; exine 2.75 u thick; excrescences monomorphic, diam 1.5 4, many sided, with cen- tral pits, sparse on furrow membrane. 9. L. medium (Planch.) Brit. var. medium. Oblate triangular but tending toward spheroidal, diam 52.5-70 a; pori 6, elliptic, 21 4 wide; exine 3-3.75 mw thick; excrescences monomorphic, diam 1.5-1.75 u, many sided, with central pits, sparse on pore membrane. L. medium var. texanum (Planch.) Fern. Compressed triangular, diam 42-56 4; colpi 3, elliptic, 28-32 4 wide; exine 3.25 4 thick, excres- cences monomorphic, diam 1.5-2.5 u, fewer on the furrow membrane. 10. L. westii Rogers. Spheroidal with undulate surface, diam 63-66.5 4; pori 6, oblong, with slightly tapering but blunt ends, 14 4 wide; exine 3 u thick; excrescences monomorphic, diam 1 4, rounded, without central pits, arranged on the furrow membrane as a central longitudi- nal patch. 11. L. floridanum (Planch.) Trel. var. floridanum. Spheroidal with undulate surface, diam 50.5-59.5 4; pori about 10, small, rounded, diam 12-14 4; exine 3.5 u thick; exerescences monomorphic, diam 1.25 u, rounded, without central pits, somewhat fewer on the bulging pore membrane. L. floridanum var. chrysocarpum Rogers. More or less similar, sometimes appearing four-sided, diam 52.5-63 u; pori usually 6, diam 21 u; excrescences monomorphic, diam 1.75 4, many sided, sometimes with central pits, fewer on the pore membrane. 12. L. intercursum Bickn. Spheroidal with undulate surface, diam 52.5-66.5 u; pori about 12, rounded, diam 17.5 4; exine 2.25 u thick; excrescences monomorphic, diam 1.75 u, many sided, somewhat fewer on furrow membrane. 13. L. sulcatum Ridd. including var. harperi (Small) Rogers. Sub- spheroidal, diam 70-87.5 u; pori about 20, circular, diam 10.5-14 4; exine 5.25-7.5 u thick; excrescences polymorphic, diam 1.5-3 4, eylindri- cal with rounded summit, nearly as dense on the pore membrane. 14. L. rigidum Pursh var. rigidum. Subspheroidal, diam 84-112 y; pori about 20, circular, diam 10.5-12.5 4; exine 5.25-7 uw thick; excres- 1963] Pollen Morphology — Xavier & Rogers 143 cences polymorphie, diam 1.5-5.25 u, cylindrical, rounded at the sum- mit, equally abundant on the pore membrane. L. rigidum var. carteri (Small) Rogers. Similar, diam 91-108.5 4; exine 8.75-9 u thick. It is generally thought that pollen grains with many circu- lar pori and with excrescences on the pore membrane are more specialized than those with few, slender, pointed fur- rows with naked membranes. Correlated with these ad- vanced traits in Linum are large excrescences and thick exine. Certainly additional species should be examined before definite conclusions can be drawn, but on the basis of those thus far studied, the pollen features thought to be primitive or advanced can be shown as follows: PRIMITIVE FEATURES ADVANCED FEATURES Germinal furrows three Germinal pores about twenty Grain triangular Grain spherical Furrows slender, with Pores circular pointed ends Pore membrane with Furrow membrane naked excrescences Excrescences small Excrescences large Exine thin Exine thick Upon taking these characteristics into consideration, it is possible to segregate the species studied into four groups, the members of each group resembling one another in their important features. These groups are shown in Fig. 8. It is not to be implied that the species studied or the groups shown constitute a single line of evolution, for infor- mation from pollen morphology alone is too scanty and data from all other sources possible must contribute to our knowl- edge. However, the species in Group I combine the greatest array of primitive features, those of Group IV, the largest number of advanced features, while Groups II and III are intermediate. ! On the basis of observations on the general morphology of the various species, it would appear that Group I is à heterogeneous one in which, though evolution has proceeded in other directions, the pollen has remained rather un- specialized. Linum arenicola, L. bahamense and L. rupestre have very similar pollen. That of L. lewisii and L. usitatissi- 144 Rhodora [Vol. 65 mum is rather different and the relationship of these species to each other and to the other species in the group is not clarified here. Groups II and III, from their gross morphology, would be judged to be quite closely related to one another. The pollen morphology of the wide ranging species, L. medium var. texanum, L. striatum and L. virginianum, is very similar, as is that of L. floridanum and L. intercursum. Linum westi, L. floridanum var. chrysocarpum and L. medium var. medi- wm appear to be intermediate between members of Groups II and III. The present studies indicate that the systematic position of the varieties of L. medium needs clarification. The species of Group IV possess pollen that is significant- ly different from that of the other groups and the relation- ship of members of this group to those of the others needs further study. Conclusions from other data (Rogers, 1963) that some members of Group I are most primitive and that Groups II and III and Group IV may be natural assemblages of succes- sively more highly evolved species is given considerable sup- port by the results of the study of pollen morphology. Many workers in palynology have found it almost impos- sible to separate species within a genus and sometimes even different genera within a family on the basis of pollen mor- phology alone. In the species of Linum studied, generally speaking, though the pollen characters within a species are without much variation, one can distinguish many species rather readily. The large amount of variation, involving several characters, especially since it appears to be corre- lated with differences in gross morphology, may indicate that the usual treatment of Linum as a single genus is too conservative. The examination of other species, both Old and New World, may provide further information as to the proper systematic treatment of the genus. In any case there is little question but that the use of pollen characters will constitute a valuable additional tool in the study of the tax- onomy of the genus. WAYNE STATE UNIVERSITY, DETROIT, MICHIGAN. 1963] Pollen Morphology — Xavier & Rogers 145 LITERATURE CITED ERDTMAN, G. 1952. Pollen morphology and plant taxonomy. Angio- sperms. The Chronica Botanica Co., Waltham, Mass. Rocers, C. M. 1963. Studies in Linum. The yellow flowered species of eastern North America. Brittonia 15. (in press). SMALL, J. K. 1907. Linaceae in North Amer. Flora 25: 67-84. WODEHOUSE, R. P. 1935. Pollen grains. McGraw Hill, N. Y. PHYTOGEOGRAPHIC NOTES ON ROTTBOELLIA, PASPALUM, AND MISCANTHUS (GRAMINEAE)' RICHARD W. POHL Paspalum fimbriatum H. B. K. Additional records from Florida. This tropical species occurs in northern South America and the West Indies but has not been previously reported from the continental United States. It is abundant in orch- ards north of Homestead, Florida, and occurs along the Keys Highway. The author has made the following collec- tions: FLORIDA: DADE CovNTY: In an avocado orchard west of Princeton, Sec. 27, R. 39 E, T. 56 S. Pohl 7589, June 22, 1957 (isc). MONROE Co.: Median strip of U. S. Hy. 1, Key Largo, 3 miles by road n. of Port Largo. Pohl 9182. April 13, 1962 (Isc). Rottboellia exaltata L.f.: A second American station. This Asiatic species was reported by Hitchcock and Chase (1950) from Miami. It grows in abundance on dry coral in vacant lots in Homestead, forty miles southwest. First seen there by the author in 1957, it had increased markedly by 1962. FLORIDA: DADE CovNTY: Weed in vacant lot on dry coral, Home- stead. Pohl 7584. June 22, 1957. (1sc); Homestead. Pohl 9177. April 12, 1962. (isc). Miscanthus sacchariflorus (Maxim.) Hack. in the Midwest. The Asiatic ornamental grass, M. sacchariflorus (Fig. 1) was first listed by Hitchcock and Chase (1950) from Clinton Co., Iowa, where it was an escape from cultivation. Steyer- mark, et al, (1957) gave several Illinois localities as well, and I have reported it from fourteen counties of eastern Iowa. On recent field trips in southeastern Iowa, I have noticed this species spreading aggressively along roadsides throughout this area. Apparently it is introduced as an Journal paper No. J-4484 of the Iowa Agricultural and Home Eco- nomics Experiment Station, Ames, Iowa. Project No. 1156. 146 1963] Rottboellia, Paspalum, and Miscanthus — Pohl 147 ornamental and later spreads by rhizomes from cultivated patches, particularly along roadsides and ditches. Spread by seed is evident in some areas, where small satellite colo- nies occur near original plantings. Many of the colonies observed during 1962 were remote from buildings and prob- ably arose from seed. M. sacchariflorus was also seen grow- ing as an ornamental in southern Minnesota. Four recent collections from Wisconsin now add that state to the range of this species. The writer is indebted to Dr. Hugh Iltis for the loan of two of the Wisconsin specimens cited below. The vigor and large size of M. sacchariflorus, coupled with its aggressive rhizomatous spread, make it a potentially dangerous weed, with many of the qualities which J ohnson grass exhibits in the South. WISCONSIN: LAFAYETTE Co.: Two colonies along n. side of Hy. 11, 1.5 mi. e. of South Wayne. Pohl 9330. August 6, 1962. (ISC, MIL). This colony has existed for at least ten years, according to the author's observations. GRANT Co.: Sec. 21, T 1 N, R 2 W, Jamestown Twp. Roadside and field on east side of Wis. Hy. 35; one colony about 8 ft. in diameter. R. W. Freckmann, (F).62-158. Oct. 13, 1962 (1sC) ;T 1 N, R 2 W, Sec. 28. H. H. Iltis 18478. Aug. 10, 1961 (WIS). WAUSHARA Co, T8 N, R 9 E, Sec 30. Road bank. K. Pochmann 49. Sept. 5, 1957 (WIS). DEPT. OF BOTANY, IOWA STATE UNIVERSITY, AMES, IOWA. LITERATURE CITED HITCHCOCK, A. S. AND AGNES CHASE. 1950. Manual of the Grasses of the United States, U. S. Dep. Agr. Misc. Publ. 200, 2nd Ed. Pourn, RrcHARD W. 1959. Introduced weedy grasses in Iowa. Proc. Iowa. Acad. 66:160-162. STEYERMARK, J., F. A. SWINK, AND J. THIERET. 1957. Additions to the flora of Illinois. Rhodora 59:31. 148 Rhodora [Vol. 65 Fig. 1. Miscanthus sacchariflorus. Inflorescence and rhizome. LIFE-FORMS IN THE PLAINS FLORA OF SOUTHERN MACKENZIE, NORTHWEST TERRITORIES JoHN W. THIERET This paper presents a life-form analysis of the 558 species of seed plants that are reported to be native or naturalized in the Great Plains of southern District of Mackenzie. The life- form classification used here is that of Raunkiaer (1934), which is based on the kind and degree of protection given to the perennating buds during the unfavorable season. Of the several life-form systems that have been proposed, that of Raunkiaer is most widely used, in large measure because of its simplicity and ready applicability. Apparently the only life-form studies previously made on Canadian floras are those of Raunkiaer (1934), which were first published in 1908, and those of Scoggan (1950). Raun- kiaer based his work on such floristic lists as were then avail- able for the regions he chose to analyze. These lists, some of which had been compiled a number of years before Raun- kiaer made use of them, were for the most part quite incom- plete accounts of the various floras. As a result, the life- form studies based on them present statistics and conclu- sions that may not be entirely valid. Scoggan's studies, in contrast, were based on his own and other relatively com- plete and recent accounts of the eastern Canadian floras con- cerned. If the Raunkiaer system of classification of life- forms is to be tested for boreal America, many additional life-form analyses are needed. The present paper is offered as a contribution to this end. The Great Plains of southern Mackenzie extends from the Laurentian Plateau on the east to the Mackenzie Mountains on the west and from the Mackenzie-Alberta border (609N) on the south to latitude 62°N on the north. The area en- compassed is about 45,000 square miles. Elevation varies from about 520 feet at Great Slave Lake to 1600 feet in the Mackenzie Lowlands east of the Liard River. The region lies wholly within the boreal forest and principally in the 149 150 Rhodora [Vol. 65 Mackenzie Lowlands. The climate can be described as north- ern continental, with short, dry, and relatively warm sum- mers and long cold winters. (see climatological data in Table 1). The region was chosen for life-form analysis because of three main considerations. First, the flora of the Great Plains of southern Mackenzie is better known than that of any other large portion of the district and is on record in Table 1. Climatological data for Fort Simpson, Hay River, and Fort Smith, N.W.T. (from Anonymous, 1954). Fort Simpson Hay River Fort Smith Mean annual precipitation in inches 12.13 12.02 12.63 Mean annual rainfall in inches 7.61 7.84 7.97 Mean June, July, and August rainfall in inches 5.46 4.06 5.05 Mean January temperature in degrees F —15 —12 —13 Mean July temperature in degrees F 62 60 61 relatively few publications (Cody, 1957, 1961; Jeffrey, 1961; Raup, 1947; and Thieret, 1961, 1962, in press). Second, the over-all climate of the region appears not to vary markedly either from east to west or from north to south, as can be seen by examining climatological data (Table 1) from Fort Simpson, Hay River, and Fort Smith, three widely separated settlements. Third, the author has carried out field work during four seasons of study of flora and vegetation along the Great Slave Lake Highway and has observed in the field and collected about 80 percent of the species considered in the preparation of this paper. Too often, life-form studies are based almost entirely upon herbarium specimens and not upon all-important field observations on the plants concerned. Raunkiaer's life-form system is completely independent of the usual classification of plants into species, genera, etc., and recognizes five principal classes, which may be charac- terized as follows. Class I. PHANEROPHYTES (Ph) bear their perennating 1963] Life-Forms — Thieret 151 buds well above the ground and they are almost all trees and shrubs. Because the buds are elevated and exposed to the full impact of the environment, and because the severity of conditions increases with height above the ground, the phan- erophytes are logically subdivided into height classes for life-form analysis: megaphanerophytes (Mg), over 30 meters tall; mesophanerophytes (Ms), between 8 and 30 meters tall; microphanerophytes (Mc), between 2 and 8 meters tall; and nanophanerophytes (N), between 25 centi- meters and 2 meters tall. As a class, the phanerophytes are predominant in humid tropical floras and tend to decrease in proportion to other life-forras in regions with climates less favorable to plant growth. Class II. CHAMAEPHYTES (Ch) have their buds above the eround but lower than 25 centimeters, so that in the un- favorable season they may receive some protection from fallen snow and leaves or from the dense growth of the plant itself. The buds of this class are obviously less exposed to the impact of the environment than are those of phane- rophytes. In a general way, the percentage of chamaephytes in a flora tends to increase with increasingly high altitude or latitude or both. An especially high proportion of chamae- phytes appears to characterize floras of arctic and alpine regions. Class III. HEMICRYPTOPHYTES (H) have their buds in the surface of the soil and are thus even better protected than chamaephytes. This class tends to be dominant in floras of temperate regions and often constitutes half or more of the species of an area, especially in grasslands and in deciduous forests; they are also common in tundra except under ex- treme conditions (Cain, 1950). Raunkiaer distinguished three principal subtypes: the non-rosette or protohemicrypt- ophytes (Hp), the semirosette (Hs), and the rosette types (Hr). The first type is without basal rosettes of leaves; the second has both basal and stem leaves; and the third has its leaves in a compact basal rosette. Class IV. CRYPTOPHYTES (Cr) have their buds beneath the surface of the soil, in water, or in the substratum under the 152 Rhodora [Vol. 65 water. The buds are manifestly much better protected than those of plants whose bud-bearing shoots are in or above the surface of the soil. Raunkiaer recognized three principal sub- divisions: geophytes, helophytes, and hydrophytes. Geo- phytes are land plants, and their perennating structures are commonly bulbs, corms, rhizomes, stem-tubers, or root- tubers. Helophytes grow in soil saturated with water, or in the water itself, but their vegetative shoots are emergent. The hydrophytes include those aquatics that are free-floating and those that root in the substratum beneath the water but whose vegetative shoots are submerged. In the present study, as in most other life-form analyses, the hydrophytes and helophytes are combined into one class (HH). Crypto- phytes appear not to be the dominant life-form of any par- ticular climate. Class V. THEROPHYTES (Th) are annual plants, which survive the unfavorable season in the form of seeds. They are particularly abundant in desert floras and in the weedy communities that develop where native vegetation is dis- turbed. The “life-form spectrum" of a particular flora shows the percentage-distribution of the five life-form classes in that flora. Such a spectrum can be used, in comparison with spec- tra of other floras, to reflect phytoclimatic differences be- tween regions and can give an indication of the type of phytoclimate (i.e, whether phanerophytie, chamaephytic, hemicryptophytic, or therophytic) of the region concerned. Raunkiaer’s “normal spectrum" was developed by him as the result of 1000 random samplings of the world flora. It may or may not represent accurately the flora of the world as a whole, but it does serve as a useful standard for comparison. Every regional spectrum will have at least one class whose percentage is higher than that of the normal; this class can be taken as an indicator of the phytoclimate of the region (Oosting, 1956). The data used in the present study were obtained from field work and from supplementary observations on dried specimens in the herbaria of Chicago Natural History Mu- 1963] Life-Forms — Thieret 153 seum, the University of Minnesota, and the University of Southwestern Louisiana. The field work was supported by Chicago Natural History Museum (1958, 1959) and by a grant from the National Science Foundation (1961, 1962). A few species that have been reported to occur in the south- ern Mackenzie Great Plains were not considered during the compilation of the data for this paper because these plants are seemingly waifs and not truly naturalized members ot the flora. Table 2. Life-form distribution of the southern Mackenzie Great Plains flora. Data obtained from the study of 558 species. PHANEROPHYTES (Ph) total species percent of flora Megaphanerophytes (Mg) 1 [ 0.2 Mesophanerophytes (Ms) 7 1.2 Microphanerophytes (Mc) 16 2.9 Nanophanerophytes (N) 39 TE Total Phanerophytes 63 11.3 CHAMAEPHYTES (Ch) Herbaceous chamaephytes (Chh) 23 4.2 Woody chamaephytes (Chw) 18 3.2 Total Chamaephytes 41 7.8 HEMICRYPTOPHYTES (H) Protohemieryptophytes (Hp) 49 8.9 Semi-rosette hemieryptophytes (Hs) 214 38.9 Rosette hemicryptophytes (Hr) 37 6.7 Total Hemicryptophytes 800 53.8 CRYPTOPHYTES (Cr) Helo-Hydrophytes (HH) 41 7.8 Geophytes (G) Rhizome geophytes (Grh) 44 8.0 Stem-tuber geophytes (Gst) 6 1-1 Root-tuber geophytes (Grt) 3 0.5 Bulb geophytes (Gb) 8 0.5 Root-bud geophytes (Gr) 1 0.2 Root parasites (Gp) Ji 0.2 Total Cryptophytes 99 APTARI THEROPHYTES (Th) 55 9.8 154 Rhodora [Vol. 65 During the course of the field work, specimens for life- form analysis were collected toward the end of the growing season or even after the first frost in order to assure that the perennating structures would be as nearly as possible in winter condition. Colonies or individuals of the various spe- cies were located earlier in the season and were suitably marked for later study. For all collections, data were record- ed concerning the position of perennating structures in rela- tion to the soil surface. These field data are especially important in the case of hemicryptophytes and cryptophytes, which may be difficult to assign to the proper life-form class on the basis of herbarium study alone. Table 3. Life-form spectrum of the southern Mackenzie Great Plains flora, compared with the “normal spectrum” and with spectra of other North American regions. Ph Ch H Cr Th Southern Mackenzie Great Plains 11.5 7.3 53.8 17.7 9.8 "Normal Spectrum" 46.0 9.0 26.0 6.0 13.0 Ellesmere Island (Raunkiaer, 1934) — 23.5 65.5 110 — Baffin Island (Raunkiaer, 1934) 1.0 30.0 51.0 16.0 2.0 Canadian Eastern Arctic (Scoggan, 1950) 3.5 29.4 54.5 10.5 24 Sitka, Alaska (Raunkiaer, 1934) 11.0 7.0 60.0 17.0 5.0 Bic and Gaspe (Scoggan, 1950) 10.5 7.8 48.7 18.9 14,1 West and Central Quebec (Scoggan, 1950) 16.6 3.5 48.6 22.4 13.8 Indiana (McDonald, 1937) 15.8 1.7 50.3 19.6 13.0 Illinois (Hansen, 1952) 13.9 2.0 47.5 17.1 14.4 Kentucky (Gibson, 1961) 17.6 1.4 52.6 16.6 11.8 Connecticut (Ennis, 1928) 15.0 1.9 49.4 21.7 11.7 Olympic Peninsula, Washington (Jones, 1936) 11.0 6.0 52.0 22.0 9.0 A tabulation of life-form data for the southern Mackenzie Great Plains is given in Table 2. In Table 3 comparison is made between the “normal spectrum,” the spectrum of the Mackenzie plains, and the spectra of several other North American regions. The following excerpt from Raunkiaer (1934, p. 133) serves well as an introduction to a brief discussion of the data in tables 2 and 3. 1963] Life-Forms — Thieret 155 In the northern cold temperate and cold zones as we gradually go towards the north we find that the biologi- cal spectrum of the vegetation changes in a very definite manner. The Phanerophytes and the Therophytes de- crease and finally disappear. The Cryptophytes, too, which are well represented throughout most of the region, disappear entirely from the hostile regions of the extreme north. The percentage of Hemicrypto- phytes keeps fairly constant, being approximately double the percentage found in the whole world. The Chamaephyte percentage on the other hand gradually inereases towards the north; in the southern parts of the region it is a long way below the Normal Spectrum, but after reaching this figure it soon doubles it. Ulti- mately the Chamaephyte percentage becomes three times or more that of the Normal Spectrum. All these changes follow the same series everywhere, whichever meridian we follow. It is evident from the data in Tables 2 and 3 that the phytoclimate of the southern Mackenzie Great Plains is decidedly hemicryptophytic. The hemicryptophyte percent- age is about double that of the “normal spectrum,” in line with Raunkiaer’s assertion. Although Raunkiaer states that the percentage of hemicryptophytes keeps fairly constant in the northern cold temperate and cold zones, there seems to be a tendency for this class to increase somewhat, to a point at least, with increase in latitude and with accompanying decrease in phanerophytes and therophytes. The significance and constancy of this tendency cannot be known until many more life-form data are available than at present. The phanerophyte percentage is, as could be expected, somewhat lower than that of regions further south. The total percentage and the vhanerophyte-subdivision percent- ages provide statistical demonstration of the well-known phenomenon that the stature of woody plants is progressive- ly reduced as the continental tree-limit is approached. About 130 feet is the maximum height for trees in the southern Mackenzie Great Plains. The only megaphanerophyte is Picea glauca; all other trees in the region (Picea mariana, 156 Rhodora [Vol. 65 Pinus banksiana, Abies lasiocarpa, Larix laricina, Betula papyrifera, Populus tremuloides, and P. balsamifera) are mesophanerophytes here but may be megaphanerophytes further south. Other good examples of the change in life- form of woody plants with increase in latitude are seen in Prunus virginiana and P. pensylvanica, which in southern Mackenzie, at the northern edge of their range, are nano- phanerophytie, reaching about 5 feet in height. Further south each of these species may be mesophanerophytic, reaching at least 30 feet in height. The chamaephyte percentage is considerably above the percentage of this class shown by floras of parts of North America with a climate more favorable to plant growth. This relationship is in harmony with Raunkiaer's postulate that the chamaephyte percentage in floras tends to increase with increase in latitude. Of particular interest in the southern Mackenzie Great Plains flora is the preponderance of herbaceous forms among the chamaephytes. According to Raunkiaer, the eryptophytes are well repre- sented throughout most of the northern cold temperate and cold zones. They are well represented in the southern Mac- kenzie Great Plains flora, the percentage being neither greatly above nor greatly below the percentage for other studied North American floras, except those in the extreme north. In connection with the therophyte percentage in the southern Mackenzie Great Plains flora it should be noted that most of the introduced plants in the region are annuals and that these introduced annuals constitute one-half of the therophyte population. The subarctic climate is, of course, not too favorable to the therophytic habit; it may account for the low percentage (4.9) of native therophytes in the flora and may explain, in part at least, why more introduced annuals are not present. In summary, the data in this life-form study indicate that a climate most suitable to hemicryptophytes prevails in the southern Mackenzie Great Plains. Those data are generally in harmony with Raunkiaer's assertions concerning the 1963] Life-Forms — Thieret 157 change in life-form spectra of floras in cold temperature and cold zones with increase in latitude. UNIVERSITY OF SOUTHWESTERN LOUISIANA LAFAYETTE, LOUISIANA REFERENCES ANONYMOUS. 1954. Addendum to Volume 1 of Climatic Summaries for Selected Meteorological Stations in Canada. Average Values of Temperature and Precipitation. Meteorological Division, Can- ada Department of Transport, Toronto. Cain, S. A. 1950. Life-forms and phytoclimate. Bot. Rev. 16: 1-32. Copy, W. J. 1957. New plant records for northern Alberta and south- ern Mackenzie District, Can. Field-Nat. 70: 101-130. 1961. New plant records from the upper Mackenzie River Valley, Mackenzie District, Northwest Territories. Can. Field-Nat. 75: 55-69. ENNis, B. 1928. The life-forms of Connecticut plants and their sig- nificance in relation to climate. Conn, State Geol. and Nat. Hist. Surv. Bull. 43. Gipson, D. 1961. Life-forms of Kentucky flowering plants. Amer. Midl. Nat. 66: 1-60. HANSEN, C. E. 1952. The life-forms of the flowering plants of Illinois. Unpubl. Master's Thesis, Northwestern University. JEFFREY, W. W. 1961. Notes on plant occurrence along lower Liard River, N. W. T. Natl. Mus. Canada Bull. 171: 32-115. JoNES, G. N. 1936. A botanical survey of the Olympic Peninsula. Univ. Wash. Publ. Biol. 5. McDowarp, E. S. 1937. The life-forms of the flowering plants of Indiana. Amer. Midl. Nat. 18: 687-773. OosriNG, H. J. 1956. The Study of Plant Communities. 2nd ed. W. H. Freeman and Company, San Francisco. RAUNKIAER, C. 1934. The Life Forms of Plants and Statistical Plant Geography. Clarendon Press, Oxford. Raup, H. 1947 The Botany of Southwestern Mackenzie. Sargentia 6. ScogGAN, H. J. 1950. The Flora of Bic and the Gaspe Peninsula, Quebec. Natl. Mus. Canada Bull. 115. THIERET, J. W. 1961. New plant records for southwestern District of Mackenzie. Can. Field-Nat. 75: 111-121. 1962. New plant records from District of Macken- zie, Northwest Territories. Can. Field-Nat. 76: 206-208. in press. Additions to the flora of the Northwest Territories. Can. Field-Nat. A REVISION OF THE NORTH AMERICAN GENUS UVULARIA (LILIACEAE) ROBERT L. WILBUR" The endemic North American genus Uvularia is comprised of five species with one or more of these being frequently encountered from Nova Scotia south into northern Florida and west into Louisiana and eastern North Dakota. All of the species are very attractive members of the early spring flora and hence are widely known and often collected. As is befitting a small and uncomplicated genus, its history is a simple one. It was founded by Linnaeus in 1738 in his Hortus Cliffortianus but takes 1753 as its starting point for nomenclatural purposes. Linnaeus recognized three differ- ent species (U. amplexifolia, U. perfoliata and U. sessili- folia) of which only the second and third species, which possess capsular fruit, are now considered to belong to that genus, or by some to it and a closely related segregate. Lin- naeus’ U. amplexifolia, whose fruit is a berry, has been treated as a member of the genus Streptopus since the early nineteenth century. As is demonstrated by the extensive list of "excluded names" at the end of this paper, earlier botanists had a much broader concept of the genus which included members of the following genera: Disporum (in- cluding Prosartes), Fritillaria, Streptopus and Tricyrtis. However, at least by the last quarter of the nineteenth cen- tury, these extraneous elements had been removed and the "Grateful acknowledgment is made to the National Science Founda- tion (G-18799) and to the Duke University Research Council for grant- ing the funds that made this study possible. My indebtedness and gratitude ought also to be expressed to the curators of the herbaria indicated below by Lanjouw and Stafleu’s abbreviations, whose speci- mens, so kindly loaned, form the basis of this study: CHRB, DUKE, F, FLAS, FSU, GA, GH, IA, ILL, ISC, KY, MICH, MIN, MO, MSC, MISSA, NCSC, NCU, NY, 0S, PENN, PH, SMU, TENN, UARK, US, USF, WIS, WVA. My thanks are also due to Drs. W. H. Wagner, Jr. and Edward G. Voss of the Uni- versity of Michigan and Dr. R. K. Godfrey of Florida State University who secured pickled flowers and fruits which materially aided this investigation. 158 1963] Uvularia — Wilbur 159 genus (or the group of two closely related genera) was restricted to the same eastern North American species that we recognize today. Watson (Proc. Am. Acad. 14:221, 268, 269. 1879) segre- gated the two sessile-leaved species recognized by him from the perfoliate species, naming the first mentioned group Oakesia. In reviewing this paper, Asa Gray (Am. Jour. Sci. ser. 3. 18: 314. 1879) expressed his disinclination to ac- cept the segregate although admitting the several “good” differences upon which the separation was based. The ge- nus was taken up, however, in the sixth edition of Gray's Manual which Watson co-edited with Coulter. The segre- gate was also recognized by Small who, however, provided the substitute name Oakesiella since, under the American Code, which he more or less followed (as under the more recent versions of the International Code), a later homonyn could not be employed even if the first use of the name were treated as a synonym. In addition to Gray, mentioned above, the segregate has been rejected by Bentham and Hooker (Gen. Pl. 3:830. 1833), Engler (Nat. Pflanzf. II. 5: 24. 1888), Britton (Bull. Torrey Club 18: 272. 1891), Fernald (Gray’s Man. 8th ed. P. 428, 429. 1950), and Gleason Ill. Fl. 1:428. 1952), while Oakesia or Oakesiella has been taken up by Small (Fl. Se. U. S. 271, 272. 1903; Man. Se. FI. 299, 300. 1933), Robinson and Fernald (Gray's Man. 7th ed. p. 280, 285, 286. 1907), Macbride (Contr. Gray Herb. 53:5. 1918) and Rydberg (Fl. Plains & Prairies p. 219, 221, 222. 1932). INTERGENERIC RELATIONSHIPS Unfortunately, it is not possible to discuss meaningfully the generic affinities of Uvularia. The tribe Uvularieae has been variously interpreted since it was first published by Kunth (Enum. Pl. 4: 199-214. 1843). In addition to Uvular- ia it then included such American genera as Streptopus Michx. (including Hekorima Raf.), Disporum Salisb. (in- cluding Prosartes Don), and four other genera from the Old World. There is a high degree of correspondence as to the contents of the tribe in the works of Bentham and Hooker 160 Rhodora [Vol. 65 (Gen. Pl. 3:829-832. 1883), Engler (Nat. Pflanz. II. 5: 24- 27. 1888) and Hutchinson (Fam. Fl. Pl. 2:606. 1959). The last two authors however, restrict the tribe to capsuled genera, thus excluding the baccate Disporum which was relegated by both of them to the tribe Polygonateae. None of the other genera included in the tribe by Hutchinson possesses a range that is ordinarily to be expected for very close relatives of eastern North American plants. Engler placed the eight genera of his concept in his Melanthioi- deae — Uvularieae all of which, with the exception of Uvul- aria, are confined to the Old World continents of Africa, Asia and Australia. Only one of the Asian genera, the aber- rant, septicidally dehiscent Tricyrtis (considered by Hutch- inson as the tvpe of another small, ditypic tribe), reportedly from Japan, Korea, China and the Himalayas, possesses a range that is an expected one for close relatives of eastern North American plants. Until considerably more knowledge is acquired, the classifications of the Liliaceae, like most large taxa, will continue to be artificial, being largely based upon one or very few characters, and hence speculation as to the generic relatives of a genus such as Uvularia will remain most unproductive and unconvincing. MORPHOLOGY HABIT. All five species are perennial herbs whose aerial parts die back to the ground each winter. UNDERGROUND PARTS. The five species are divisible into two distinct groups upon the basis of their underground parts. Three species possess a very short (less than 1 em long) underground rhizome from which arise the annual aerial stems and from which descend a cluster of thickened roots. One of these species (U. perfoliata) spreads by means of a pair of slender, subterranean stolons, each of which forms buds and roots at its distal end, after which the con- necting portion of the stolon disintegrates. These struc- tures have been seldom observed but this is not surprising’ as considerable care must be devoted in securing them intact. Whether they are present or absent in the other two species (U. grandiflora and U. puberula) with a similar rhizome 1963] Uvularia — Wilbur 161 and root system is at present unknown but they have at least not been observed. Two species (U. sessilifolia and U. floridana) possess elongate (10 cm long or more), thick- ened rhizomes from which descend scattered fibrous roots and from the tips of which ascend the aerial stems. AERIAL STEMS. In all species flowering or fruiting stems appear to be once-branched if one ignores the short, flow- ering or fruiting branches. The branching pattern is sym- podial. The stems of the perfoliate species are terete and hollow while those of the sessile-leaved group are typically strongly angled (at least above) and solid. LEAVES. The genus is divisible into two distinct groups based upon whether the leaves are perfoliate or sessile. (The two groups however are not the same as would result from a division based upon the subterranean parts.) The upper leaves of the perfoliate group are merely sessile and strongly clasping or amplexicaul. The leaf margins of the perfoliate- leaved species are perfectly smooth while those of the ses- sile-leaved species are all very minutely papillose. INFLORESCENCE. The flowers are solitary and terminal although superficially appearing lateral or axillary and often so described (e. g. Dietz, p. 221). The pattern of growth is hence sympodial. The flowers of U. floridana, U. perfoliata, and U. grandiflora are borne on short branches bearing a small to large leaf (= bract) at the base of the pedicel. This feature has been rarely seen on specimens of U. puberula and it has never been observed in U. sessilifolia. It would be of considerable interest to know if there is any anatomical evidence as to whether those species lacking the flower-or fruit-subtending bract have merely lost the bract or have lost both bract and the supporting stem. INDUMENT. Uvularia perfoliata, U. floridana and U. ses- silifolia are completely glabrous. The lower leaf-surface of U. grandiflora is usually densely short-pubescent but the degree of pubescence varies from dense to almost, or rarely even completely, glabrous. Dietz (1952) has concluded that this variation is due to introgression from U. grandiflora into the glabrous U. perfoliata but, as areas of comparative 162 Rhodora [Vol. 65 glabrousness of what surely is U. grandiflora are to be found in Arkansas and Minnesota far from the influence of U. perfoliata, it would seem that another or an additional ex- planation is required. Uvularia puberula is exceedingly variable as to the presence and amount of pubescence, its distribution and even the length of the trichomes. The pubescence is usually to be found, when present, upon the angles of the upper portion of the stem but sometimes ex- tends onto the lower surface of the leaves along the elevated principal veins. In some specimens of this species no pube- scence is to be detected and this feature is characteristic of the so-called var. nitida. PERIANTH. The perianth is a typically liliaceous one con- sisting of 6 distinct tepals arranged in two imbricate series. The tepals of U. perfoliata are the most distinctive in the genus because of the presence of granular papillosities on their inner surface. There is a tendency for the size and shape of the tepals to be more or less characteristic of each species but the variation is so great that the tendencies are not too helpful. The tepals are rendered somewhat gibbous at base through the presence of a nectariferous pouch at the base of each perianth-segment. Small (1903, 1933) was in error in stating that the sessile-leaved species (Oakesiel- la) lack nectaries. Watson (1879) stated that the perfoliate species were characterized by the presence of a callus or ridge on either side of the nectary while the sessile-leaved group lacked the ridee. Study of pickled flowers of all five species indicated that a moderately to slightly elevated, thin flap was present on each side of the nectariferous de- pression in the two perfoliate-leaved species while only the slightly hardened wall of the nectary was present in the sessile-leaved species. This distinction is however, difficult to detect and hence of little taxonomic utility. STAMENS. The six stamens usually have been described as adnate to the very base of the tepals and this may be their actual condition but they appear to me to be attached to the receptacle slightly above the tepals. The slightly flattened filaments are about half-again as long (or even less) as the 1963] Uvularia — Wilbur 163 linear, laterally dehiscing, extrorse, adnate anthers. The stamens appear to be of slightly different lengths in two alternating series but this is the result of the slight differ- ence in attachment levels of the two series. The connective extends beyond the anther-sacs in all the species but is most obviously exserted in U. perfoliata. The degree of exser- tion is quite variable. PISTIL. The ovary is sessile or very nearly so in all species but U. sessilifolia, which possesses a stipitate ovary. The stipe appears to be merely the sterile base of the ovary. The ovary is in general rounded (although lobed) in the two perfoliate-leaved species but sharply angled in the three sessile-leaved species. The stigma is moderately to deeply cleft in the various species, being probably most deeply divided in U. puberula and U. grandiflora and the least so in U. floridana and U. sessilifolia. Fernald (Rhodora 37: 409. 1935) reported that U. puberula exhibits heterostyly (which, if true, would be a most unusual condition in the Liliaceae) but it appears from admittedly casual observa- tions that the variation in length of style and exsertion of the stigmatic branches beyond the anthers within a given species is almost completely a matter of age. The flowers are apparently protandrous. FRUIT. The loculicidally and tardily dehiscent capsules offer a ready means of identification but the distinguishing features are often lost in pressed specimens. The fruits of the perfoliate-leaved species are more or less truncate at the apex while those of the sessile-leaved species gradually taper to an acute (U. puberula and U. sessilifolia) or even to a long-beaked apex (U. floridana). The fruit of U. per- foliata is more or less angularly obpyramidal and each of the three lobes is deeply bifid, thus appearing 6-horned. The fruit of U. grandiflora is more roundedly obpyramidal and with very little or no evidence of deeply bifid or 2- horned angles. The fruits of the sessile-leaved species are much more sharply angled, with U. sessilifolia being at once distinguished by its long-stipitate base in contrast to the sessile or very nearly sessile bases of the other two species. 164 Rhodora [Vol. 65 The long-rostrate beak of U. floridana distinguishes that species from U. puberula as well as all other species in the genus. SEEDS. The seeds of the perfoliate-leaved species, which possess an apparent outgrowth of the raphe which at first is a swollen, balloon-like sac but at maturity deflates into a thin membrane nearly half-enveloping the seed, are at once distinguishable from the two turgid ridges which parallel the raphial slit in the seeds of the three sessile-leaved spe- cies. The nature of these outgrowths is unfortunately un- known. ANATOMY Holm (Bull. Torrey Club 18: 1-5. 1891) presented the only account known to me concerned with the internal anat- omy of any species of this genus. He compared the anatomy of two species: U. perfoliata and U. sessilifolia. The results of his anatomical comparison are presented in the chart below. U. sessilifolia U. perfoliata 1. Walls of endodermis Thin Thick 2. Number of fibro-vascular bundles in root 20 10 3. Number of layers of cells between epidermis of stem and mechanical tissue 1 2 4. Number of layers in mechanical tissue of stem 5 2 Eames (Chronica Bot. 14: 128, 129. 1953), although pre- senting no supporting evidence, made the following state- ment: “Anatomically there is no basis for the maintenance of Oakesia. There is more anatomical difference between U. perfoliata and U. grandiflora than between U. grandi- flora and Oakesia.” It is to be hoped that the evidence for such a sweeping statement will soon be presented, for it is diffieult to conceive of anatomical evidence that would show a closer relationship of U. perfoliata to a species such as U. sessilifolia than to U. grandiflora. 1963] Uvularia — Wilbur 165 Anderson and Hubricht (Am. Nat. 77: 285-287. 1943), by making celloidin peels of herbarium specimens, found that the “epidermal cells of Uvularia grandiflora tend to be larger, more irregular and more variable than those of U. perfoliata" and concluded that this tendency resulted in “a coarser, cruder leaf-texture". CYTOLOGY Anderson and Whitaker (1934) reported the chromosome number for three species (U. grandiflora, U. perfoliata and U. sessilifolia) to be n — 7. The chromosomes were found to be large, and each was identifiable by its distinctive mor- phology. They concluded from this study that “there is no evidence that chromosome duplication or chromosome inter- change are of phylogenetic importance in Uvularia.” Pre- viously Belling (Jour. Geneties 15: 245-266. 1925) demon- strated in U. grandiflora that under experimental conditions non-disjunction, fragmentation, lack of pairing and dupli- cation of all or part of the chromosome complement all occur. Belling also reported that meiosis of the pollen mother cells in his plants of U. grandiflora grown in a cool green- house took place in late February. Alden (Bull. Torrey Club 39: 439-446. 1907) reported for U. sessilifolia that mature microspore mother cells are to be found by the middle of September and that these divide by October and overwinter as mierospores. Meiosis within the ovule however did not occur prior to late April and even the megaspore mother cells were not formed prior to mid-March. Sato (Jap. Jour. Bot. 12: 76. 1942) reported that the U. sessilifolia studied by him (apparently grown from bo- tanical garden material) had 2» = 16 and not 2n = 14 as indicated by Anderson and Whitaker. The reason for this discrepancy has not been determined. Since the chromosome number of several Japanese species of Disporum including D. sessile (=Uvularia sessilis) is 2n = 16, suspicion is a- roused as to the identity of Sato's material. INTRAGENERIC RELATIONSHIPS A suggestion as to the probable relationships between the 166 Rhodora [Vol. 65 species is presented in the accompanying figure (fig. 1). Such a hypothetical scheme is of course highly subjective and hence open to criticism but for most taxa of plants we now have little choice but to attempt to visualize a hypotheti- cal ancestral type from which a suggested “phylogeny” for the group may be logically derived. PERFOLIATA FLORIDANA GRANDIFLORA SESSILIFOLIA PUBERULA Fig. 1. Probable relationships of the species of Uvularia An attempted reconstruction of the more immediate hypo- thetical ancestral taxa of these five species of Uvularia would result in a perhaps pubescent perennial with a very short rhizome or caudex, fleshy roots, sessile leaves and sympodial- ly arranged flowers borne on a leafy-bracted short branch. This plant would most closely resemble U. puberula which, however, only very rarely possesses a leafy bract. There would seem to be no doubt that sessile leaves are more primitive than perfoliate ones. (The presence of ses- sile leaves at the ends of the branches and the increasing perfoliateness of leaves are thought to be evidence of this.) There is certainly much less evidence that a very short rhi- zome is more primitive than an elongate rhizome or that the presence of pubescence is a primitive characteristic and its absence an indication of evolutionary advancement. But both of these assumptions are here made since the short- rhizomed, puberulent, sessile-leaved species, U. puberula, 1963] Uvularia — Wilbur 167 resembles the two perfoliate-leaved species in its under- ground parts. With its similar unornamented tepals, U. puberula is seemingly more closely related to the puberulent U. grandiflora than to the apparently more specialized papil- late-tepaled U. perfoliata. The presence of leafy bracts below the flowers or fruits of U. floridana, U. grandiflora and U. perfoliata is here interpreted to be a primitive char- acter. Those species lacking this bract are thought to have lost it through suppression. Several specimens of other- wise unmistakable U. puberula, but with a small subtending bract, were noted in this study. (It would be, however, of considerable interest to know if there is any anatomical evi- dence of the hypothesized suppressed leafy bract within the two species.) And finally the stalked fruit of U. sessili- folia is thought to be an indication of advancement. SUBGENERIC CLASSIFICATION Evaluation of the taxonomic merit of such a segregate as Oakesiella ( = Oakesia) is of course highly subjective. The two groups give every morphological evidence of being natural and yet whether the two taxa are better accorded generic or subgeneric rank is largely a matter of individual taste. It is believed however that criteria suggested by MeVaugh (Wrightia 1: 15, 16. 1945) for the evaluation of proposed generie segregates are met with by Oakesiella. I choose to recognize the two taxa as well-defined sections but others with whom I could not argue no doubt would consider that the differences warranted subgeneric status and still others would deem the degree of divergence repre- sented as meriting no more than the rank of series. And of course many would perhaps rightly consider it folly to erect a formal subgeneric classification within so small and relatively homogeneous a taxon. SYSTEMATIC TREATMENT Uvularia L., Sp. Pl. 304. 1753; Gen. Pl. 144. 1754. Anonymos Walt., Fl. Car. 122. 1788. nom. illegit. Art. 20. Oakesia S. Wats., Proc. Am. Acad. 14:221. 1879; not Oakesia 'Tuckerm., Hook. Lond. Jour. Bot. 1: 445. 1842 which — Corema G. Don, Edinb. 168 Rhodora [Vol. 65 New Phil. Jour. 2: 63. 1826. Oakesiella Small, Fl. Se. U. S. 271, 1328. 1903. Pubescent or glabrous, occasionally glaucous, perennial herbs with very short or elongate rhizomes and with thickened or fibrous roots. Stems erect, terete to strongly angled, and, excluding the flowers, but once branched, with several sheathing, papery cataphylls below the blade-bearing leaves. Branching-pattern sympodial. Leaves alternate, perfoliate or sessile, oblong-linear to oblong-ovate and longitudinally several- to many-nerved. Flowers solitary, terminal but appearing axillary, pendant, narrowly campanulate. Perianth 6-parted; the seg- ments multiveined, similar to one another in two imbricate series, dis- tinct, pale- to deep-yellow, linear-oblong, obtuse to acute, slightly gib- bose at base with a nectariferous depression, dropping off along with the stamens several days after pollination. Stamens 6, hypogynous, apparently completely free from the tepals, always shorter than the perianth segments; filaments slightly flattened and dilated below, rather short, being several times exceeded in length by the pollen sacs; usually noticeably but slightly alternating with one another in length with 3 longer and 3 shorter; anthers elongate, linear-oblong, extrorse, adnate, dehiscing by a longitudinal slit down the abaxial face; connec- tive slightly to conspicuously exserted, its apex acute to obtuse or even truncate. Pistil 3-carpellate; ovary 3-loculate, sessile to conspicuously stipitate, rounded- to sharply-triangular in cross-section, truncate to acute or even rostrate at apex, acutely tapering at the base; style united for about one-third to two-thirds the distance from ovary to the tip of the tripartite stigma, deciduous at maturity; stigmas linear, at first appressed together but arching outward at maturity and stigmatic along inner surface. The style and stigmatic lobes at first exceeded in length by the anthers but later exceeding them. Ovules anatropous, 2-6 per locule, horizontal from axile placentae. Fruit a greenish to stramineous (or in age becoming brownish) loculicidal capsule with walls smooth or pebbled within. Seeds globose, smooth, few (1-3) in each locule at maturity, with a very small embryo and a hard, white endosperm. LECTOTYPE: Uvularia perfoliata L. (First chosen by Britton and Brown, Ill. Fl. ed. 2. 2: 518. 1913, but indirectly determined by Watson's exclusion of U. sessilifolia in 1879.) KEY TO THE SECTIONS OF UVULARIA Leaves sessile, their margins scarious but also very minutely papillose- denticulate; seeds with a thick ridge-like crest bordering each side of the slit-like raphe; capsule acute at apex; the inner surface of fruit smooth or at least not densely pebbled. .. Section 1. OAKESIELLA. Leaves perfoliate, their margins scarious but smooth; seeds largely surrounded by a bladdery sac, this later becoming a flattened, mem- branous, loose envelope nearly half-enclosing the seed; capsule trun- 1963] Uvularia — Wilbur 169 cate at apex or at least not gradually tapering from near the middle; inner surface of fruit densely pebbled. ............ Section 2. UVULARIA. Sectipn 1. OAKESIELLA (Small) stat. nov. Oakesia S. Wats., Proc. Am. Acad. 14: 269. 1879; not Oakesia Tuckerm., Hook. Lond. Jour. Bot. 1: 449. 1842 = Corema G. Don, Edinb. New Phil. Jour. 2: 63. 1826. Oakesiella Small, Fl. SE. U. S. 271, 1328. 1903. Pubescent or glabrous herbs with elongate or very short rhizomes and with either scattered fibrous or clustered, thickened, fleshy roots. Stems triquetrous, at least above. Leaves sessile and, if sometimes appearing somewhat clasping at base, never perfoliate, margins minutely denticulate. Ovary conspicuously 3-angled in cross-section and tapering from middle to both apex and base. Capsule acute at apex; the inner surface of fruit smooth or slightly ridged but never densely pebbled. Seeds with two swollen, elevated ridges paralleling the raphial slit. LECTOTYPE: Oakesiella sessilifolia (L.) Small = Uvularia sessilifolia L. (Chosen by Small, Fl. Se. U. S. 1328. 1903). KEY TO THE SPECIES OF THE SECTION OAKESIELLA 1. Stigmatic lobes nearly equaling or at least not more than twice exceeded by the undivided style; upper stems and sometimes the lower surfaces of the leaves often moderately to densely puberulent but occasionally glabrous; aerial stem arising from a very short rhizome (less than 0.5 em. long) with clustered, thickened roots .... Le seen ei E 1. U. puberula. 1. Stigmatic lobes 3-5 times exceeded by the length of the undivided style; stem and lower surfaces of the leaves always glabrous; aerial stem arising from an elongate fleshy rhizome with scattered fibrous roots. 2. Short flowering-branchlet bearing a foliaceous bract stigmatic lobes 3 mm. long or longer; capsule sessile at base and con- spicuously rostrate at apex; connective of filament exserted 0.5 mm. or more beyond anther-sacs; tepals with an acumi- nate Apex ooo Erde dic 2. U. floridana. 2. Short flowering-branchlet lacking a bract; stigmatic lobes less than 3 mm. long; capsule clearly stipitate and lacking a ros- trate beak; connective of filament barely, if at all, exserted beyond anther-sacs; tepals rounded to acute at apex sss... a ee DM SER FU DLL Section 2. UVULARIA. Pubescent or glabrous herbs with very short rhizomes and with clustered, thickened, fleshy roots. Stem terete. Leaves perfoliate (except for the uppermost which is merely sessile) ; margins smooth. Ovary rounded in cross-section although usually with 3 principal lobes. 170 Rhodora [Vol. 65 and each of these bilobed to a lesser extent. Capsule truncate at apex; the inner surface of fruit densely pebbled. Seeds first partly enveloped by a large, swollen, balloon-like sac which at maturity becomes deflated and then forms a membrane which half envelopes them. | KEY TO THE SPECIES OF SECTION UVULARIA Perianth-segments glabrous within; leaves usually at least moderately puberulent beneath but rarely glabrate; connective of filament ex- tended less than 0.8 mm. beyond anther-sacs; principal capsule-lobes neither deeply lobed nor 2-horned; blade-bearing leaves below the lowest branch typically but one .................. eene 4. U. grandiflora. Perianth-segments conspicuously granular-papillate within; leaves glabrous beneath; connective of filament extended more than 0.8 mm. beyond anther-sacs; principle capsule-lobes deeply 2-lobed or 2- horned; blade-bearing leaves below the lowest branch typically 8-4 RR MEEEMS 5. U. perfoliata. 1. Uvularia puberula Michx., Fl. Bor.-Am. 1: 199. 1803. (Type pre- sumably at Paris, but not seen. Description is such that no other species could have been meant.) Oakesia puberula (Michx.) S. Wats., Proc. Am. Acad. 14: 269. 1879. Oakesia sessilifolia var. (?) nitida Britt., Trans. N. Y. Acad. Sci. 9: 13. 1899. (Type! NY, Iso- types GH, CHRB.) U. sessilifolia var. nitida (Britt.) Morong, Mem. Torrey Club 5: 111. 1894. U. nitida (Britt.) Mackenzie, Tor- reya 8: 14. 1908. Oakesiella puberula (Michx.) Small, Fl. Se. U.S. 272, 1328. 19083. Oakesiella nitida (Britt.) A. A. Heller, Muhlen- bergia 6: 83. 1910. U. puberula var. nitida (Britt.) Fern., Rhodora 37: 407. 1935. U. pudica var. nitida (Britt.) Fern., Rhodora 41: 536. 1939. Plant usually moderately to densely short-puberulent on the upper portion of the stem, especially at the nodes and occasionally sparingly so on the lower surface of the leaves but rarely glabrous or very nearly so throughout; the stem conspicuously angled above and often triquet- rous, about (2.0-) 2.5-4.5 (-5.0) dm. tall with several stems arising in a cluster from a very short rhizome and with a cluster of descending fleshy roots. Leaves dark green above and lighter green beneath, glab- rous or very sparingly short-puberulent below, especially near the base of the principal veins, mostly (3.5-) 4.5-7.5 (-8.5) em. long and (1.0-) 1.7-3.2 (-4.0) em. wide, oblong-elliptic to broadly elliptie in general out- line with a broadly rounded, sessile to subclasping base, and an acute to even abruptly short-acuminate apex; the leaf subtending the first branch about (3.7-) 4.5-6.5 (-8.0) cm. long and (0.9-) 1.0-2.0 (-2.5) cm. wide. Blade-bearing leaves below the first branch typically but one. Leaf-margin very minutely but abundantly papillose-denticulate and very slightly scariously edged. Flower borne solitary and termi- nating the stem but appearing as if lateral and borne opposite the leaf; its downwardly arching peduncle lacking a leaf or bract and in flower 1963] Uvularia — Wilbur TIL about 0.5-1.4 (-2.0) cm. long and glabrous to densely short-puberulent. Perianth segments about (1.0-) 1.5-2.5 (-2.7) cm. long and (1.5-) 2.0- 4.0 (-5.0) mm. wide; each tepal more or less tapering to a rounded apex. Tepals fleshy-thickened for about the basal 1-1.5 mm. with an indistinet, shallow, wedge-shaped, nectariferous depression about 1 mm. long and 0.5-0.8 (-1.2) mm. wide; the nectary scarcely bordered on either side by a short ridge that is but little elevated above the sur- face of the tepal. Filaments smooth, oblong to linear and scarcely if at all tapering from middle to either apex or base, about (1.5-) 2.0-4.0 (-5.5) mm. long and about 0.8-1.2 mm. wide. Connective extending beyond the anther-sacs for about 0.6-0.8 mm. and about 0.5 mm. wide, very slightly inwardly arching and very conspicuously truncate. Anther-sacs about (5.0-) 7.0-12.0 (-13.0) mm. long. Ovary conspicu- ously triquetrous (except through abortion of one locule), tapering distinctly to both apex and base, hence appearing shortly stipitate; the three sides of the ovary appearing slightly concave but not at all medially lined or grooved, about 0.5-0.8 em. high and each face about 25.3.5 mm. wide. Style united for about one-half the distance to the stigmatic tips which somewhat reflex with age; style and stigmas together about 8.0-14.0 mm. long. Capsule typically conspicuously three-winged, triquetrous and roundedly tapering to both apex and base, its sides somewhat concave and each face appearing broadly elliptic in outline; its base scarcely stalked (at most 1-1.5 mm.) ; about (1.5-) 1.8-2.7 (-3.5) em. long and (1.0-) 1.5-2.0 (-2.2) em, wide, dotted reddish-brown. Seed globose or nearly so, about 3.0-5.0 mm. in diame- ter, reddish-brown at full maturity but straw-colored when it first reaches full-size and with a thickened mound-like crest extending along either side of the slit-like raphe and about half encircling the seed. Usually in moist but well-drained wooded sites from New York (Long Island) and New Jersey, then in the Coastal Plain and lower Piedmont of southeastern Virginia south into Georgia and in the mountains and upper Piedmont from southern Pennsylvania south to west-central Georgia (Fig. 2). P There is considerable variation in the amount of pube- scence exhibited by individual plants of this species, rang- ing from glabrous to very densely puberulent. The repre- sentatives of the species found in New Jersey are apparently all of the glabrous type, but elsewhere within the range of the species equally glabrous plants are to be found. And even more frequently encountered are individuals that are only moderately to sparsely puberulent. The glabrous ten- dency has been treated as a variety by Britton, Morong, and Fernald and as a species by Mackenzie. Fernald stressed, in addition to the glabrousness of this taxon, its thinner and 172 Rhodora [Vol. 65 +, is N 7 U.PUBERULA S | Fig. 2. Distribution of Uvularia puberula less prominently reticulate leaves and gave its range as Long Island, New Jersey and southeastern Virginia. I have found equally glabrous plants (often included on the same sheet as densely puberulent ones) in the mountains of Vir- ginia, West Virginia and the Carolinas but have not con- sidered that the variation warranted formal taxonomie re- cognition. Although glabrous plants are more frequent in the Coastal Plain, puberulent individuals are also encoun- tered there. Recognition of a glabrous variety largely re- stricted to the Coastal Plain would appear somewhat more plausible if the individuals were either glabrous or densely puberulent instead of ranging fully between the two ex- tremes. The vegetative parts of this species are thicker and more sturdy than those of any other species and its leaves and fruit persist in a green and apparently vigorous condition until frost or for several months later than its congeners. As discussed in the section on “Dubious and Excluded Names”, I now believe that I was too hasty in proposing a 1963] Uvularia — Wilbur 173 new combination, based upon Gmelin's Erythronium carol- inianum, as the proper name of this species. 2. Uvularia floridana Chapm., Fl. S. U. S. 487. 1860. (Type not seen. Original description thought to be conclusive.) Oakesia floridana (Chapm.) Macbride, Contr. Gray Herb. n.s. 53: 5. 1918. Oakesiella floridana (Chapm.) Small, Fl. Se. U. S. 272, 1328. 1903. Plants completely glabrous and the upper stem and lower surfaces of the leaves slightly glaucous; the stem about 2.5-4.0 (-4.5) dm. high and triquetrous at least above and arising from a branching, whitish, fleshy rhizome about 3-5 mm. in diameter. Leaves glabrous, narrowly to broadly elliptic with a tapering to rounded base and an acute to rounded apex, mostly 4.0-7.0 (-8.5) em. long and (1.2-) 1.8-3.0 cm. wide; the leaf subtending the lowest branch mostly about 4.0-7.5 em. long and 1.2-2.8 em. wide. Blade-bearing leaves below the lowest branch 1-2. Leaf-margin scarious and very minutely but densely denticulate. Fertile branchlet composed of a strongly 3-angled, slightly hyaline- winged stem about (0.8-) 1.2-1.6 (-2.0) cm. long; an ovate to lanceo- late foliaceous bract with a rounded base and an acute apex mostly 1.0-2.0 (-3.0) em. long, and a terete peduncle about 2-8 mm. long and ar i A y U.FLORIDANA i Fig. 3. Distribution of Uvularia floridana 174 Rhodora [Vol. 65 one flower. Perianth-segments about (1.8-) 2.2-3.0 em. long and about 3-4 mm. wide, tapering to an acuminate apex, perfectly smooth, both internally and externally, very pale yellowish-white. Tepals narrowing basally into a thickened, somewhat fleshy, channeled base about 1.0-1.8 mm. long with a well-defined, narrow, nectariferous depression about 0.8-1.4 mm. long and 0.5-1.0 mm, wide with more or less parallel sides and thickened walls but lacking an elevated flap or ridge. Filaments smooth, white, narrowly oblong to linear and scarcely, if at all, taper- ing to either apex or base, about 3-5 mm. long and about 0.8-1.0 mm. wide. Connective bluntly rounded, whitish, and extending 0.5-2.0 mm. beyond the anther-sacs. Anther-sacs bright yellow, about 6-9 mm. long. Ovary conspicuously triquetrous and very indistinctly, if at all, stipi- tate; tapering from its widest point at the middle to both apex and base; the three sides each slightly concave and at least not noticeably medially grooved, about 3-4 (-5) mm. high and each face about 1.5-2.5 mm. wide. Style united for at least four-fifths of the distance from the ovary to the stigmatic tips which at maturity are widespreading and somewhat reflexed; style and stigmas together about 1.0-1.5 «m. long with the stigmatic lobes about 3.5-5.0 mm. long. Capsule strongly 3-winged, triquetrous and rather abruptly tapering to apex and base so that its rather concave sides appear rhombic in cutline, its base sessile and the apex with a conspicuous rostrate beak about 4-7 mm. long and 2 mm. wide; the body of the fruit about 1.5-3.0 em. long, excluding the beak, and 1.2-2.0 em. wide. Seed (known only from one imperfect example) globose to subglobose, about 4 mm. in diameter, reddish-brown at maturity with a thickened mound-like ridge or crest extending along either side of the slit-like raphe and perhaps 1/3- (1/2?) encircling the seed. Moist and often poorly drained bottomland hardwood forests along the Coastal Plain from South Carolina into northern Florida and west into Mississippi (Fig. 3). This little-collected species is known only from the bot- tomlands and floodplains of the southeastern Coastal Plain. It is readily distinguishable from its nearest relative U. sessilifolia, which also possesses sessile leaves and an elong- ated rhizome, by its long-beaked but sessile fruit and the cordate bract which subtends its flower or fruit. It seems desirable to cite a specimen of this infrequently collected species from each county in which it is known. REPRESENTATIVE COLLECTIONS: — SOUTH CAROLINA: BARNWELL co.: Batson & Kelley 30 Mar. 1953 (UNC); BERKELEY CO.: north of Goose Creek Reservoir. Hunt 2771 (GA); CALHOUN CO.: 4 miles south- east of Lone Star. Radford 9366 (UNC); DORCHESTER CO.: vicinity of Four Holes Swamp. Ahles 21906 (UNC). GEORGIA: BURKE CO.: 5 miles west of Waynesboro. Harper 2074 (F, GH, MSC, MO, NY, US); 1963] Uvularia — Wilbur 175 CALHOUN CO.: near Ichawaynochaway Creek, west of Leary. Thorne & Muenscher 7994 (GA, GH, IA, NY); EARLY CO.: along Odum Creek, west of Blakely. Thorne & Muenscher 8598 (GA, IA) ; HOUSTON CO.: between Grovania and Beech Haven. Harper 1966 (F, GH, MO, MSC, NY, US) ; LAURENS CO.: Dublin. Harper 1363 (MO, NY, US) ; WEBSTER CO.: 2 miles east of Preston. Thorne & Muenscher 9339 (1A). FLORIDA: — JACK- SON CO.: along the Chipola River, between Mariana and Mariana Cav- erns State Park. Godfrey 56319 (DUKE, FLAS, FSU, IA, NCSC, NCU, MO, NY, TENN, USF, WVA) ; LEON CO.: about 6 miles west of Tallahassee. Godfrey 61910 (DUKE). ALABAMA: MONTGOMERY CO.: wooded bot- toms of Catoma Creek. Harper 83 (GH, MO, NY, vs); MISSISSIPPI: OKTIBBEHA CO.: State College. Eckles 7 April 1937 (MISSA). 3. Uvularia sessilifolia L., Sp. Pl. 304. 1753. (Type in Linnean Her- barium. Phototype seen.) Oakesia sessilifolia (L.) S. Wats., Proc. Am. Acad. 14: 269. 1879. Oakesiella sessilifolia (L.) Small, Fl. Se. U. S. 272. 1903. Plants completely glabrous but stems and lower surfaces of the leaves glaucous; the stem about (1.7-) 2.5-4.0 (-4.5) dm. high and triquetrous at least above and arising from a branching, whitish, fleshy, subterranean rhizome about 3-5 mm. in diameter. Leaves glab- rous, narrowly to broadly elliptic in general outline with a gradually tapering to broadly rounded base and an acute to even short-acuminate apex, mostly (3.7-) 4.5-6.5 (-8.2) em. long and (1.2-) 1.7-2.7 (-4.0) em. wide; the leaf subtending the lowest branch mostly about (3.0-) 5.0-7.0 (-7.2) em. long and (0.8-) 1.0-1.5 (-2.5) em. wide. Blade-bearing leaves below the lowest branch 1-2. Leaf-margin scarious and very minutely denticulate. Fertile branchlet leafless and hence consisting of no more than the terete peduncle partially fused basally to the leafy axillary stem and about (5.0-) 7.0-14.0 (-18.0) mm. long and one flower. Peri- anth-segments about (1.2-) 1.5-2.2 (-2.8) em. long and about 2.0-4.5 mm. wide, gradually tapering to the more or less rounded apex; per- fectly smooth both externally and internally, very pale yellow. Tepals narrowing basally into a thickened, somewhat fleshy, channeled base about 1.2-1.8 mm. long with a well-defined, narrow, nectiferous depres- sion about 0.8-1.0 mm. long and about 0.5-0.8 mm. wide with more or less parallel sides and thickened walls but lacking an elevated flap or ridge. Filaments smooth, narrowly oblong to linear and scarcely, if at all, tapering to either apex or base, about (2.0-) 2.5-5.0 (-6.5) mm. long and 0.6-0.8 mm. wide. Connective bluntly rounded just above the anther-saes and barely, if at all, extended beyond them. Anther-saes about 5.0-10.0 (-13.0) mm. long. Ovary conspicuously triquetrous and clearly stipitate, tapering from the widest point at the middle to both apex and base; the three sides slightly concave and at least not con- spicuously medially grooved, about 3.0-5.0 mm. high excluding the 0.8- 1.2 mm. long stipe and each face about 1.5-3 mm. wide. Style united for at least four-fifths of the distance to the stigmatic tips which some- 176 Rhodora [Vol. 65 what reflex in age; style and stigmas together about 1.0-1.5 em. long and the stigmatic lobes about 1-2 mm. long. Capsule conspicuously three-winged, triquetrous and roundedly tapering to both apex and base; its sides somewhat concave and each face appearing broadly ellip- tic in outline; its base conspicuously stalked with the stalk usually about 2-4 (-6) mm. long; the fruit about (1.2-) 1.5-2.7 (-3.2) em. long (excluding the stipe) and (0.8-) 1.0-1.5 (-1.8) em. wide. Seed globose or nearly so, about 3-4.5 mm. in diameter, reddish-brown at maturity with a thickened mound-like ridge or crest extending along either side of the slit-like raphe which half encircles the seed, Typically found in moist hardwoods and especially in bottomlands but (particularly in the northern midwest) also encountered in more xeric sites including even jack pine woods; ranging from Nova Scotia south into northern Florida and west into North Dakota and Louisiana (Fig. 4). Ü U.SESSILIFOLIA "D T ^j Fig. 4. Distribution of Uvularia sessilifolia 4. Uvularia grandiflora J.E. Sm., Exot, Bot. 1: 99. t. 51. 1804, (Type not seen. Description and plate seemingly conclusive.) U. perfoli- ata a major Michx., Fl. Bor.-Am. 1: 199, 1803. (Type. not seen Description thought to be conclusive.) U. grandiflora f. latifolia Louis-Marie, La Revue d'Oka 14: 153. 1940. (Type not seen.) U. 1963] Uvularia — Wilbur 177 grandiflora f. variegata Louis-Marie, La Revue d'Oka 14: 153. 1940. (Type not seen.) Plant glabrous except for the lower surface of the leaves; the stem terete and about (2.0-) 4.0-7.0 (-8.0) dm. tall. Leaves sparsely to very densely pubescent beneath (or very rarely glabrous); the trichomes extremely short to rather long, ranging in length from about .05-0.5 mm. long; mostly (4.0-) 7.0-12.0 (-13.5) cm. long and (2.0-) 3.0-5.0 (-5.5) cm. wide, ovate-oblong to somewhat elliptie in general outline with a broadly rounded base and an acute to short-acuminate apex; the leaf subtending the lowest branch mostly about (6.5-) 9.0-12.0 (313.5) em. long and (2.0-) 2.5-3.5 (-5.5) em. wide. Blade-bearing leaves below the lowest branch typically but one. Leaf-margin smooth and very narrowly scariously rimmed. Fertile branchlet bearing one perfoliate leaf and one flower on a peduncle about (1.0-) 1.2-2.0 (-2.5) cm. long. Perianth segments about (2.0-) 2.5-4.5 (-5.0) em. long and about (2.5-) 3.0-7.0 (-9.0) mm. wide, acute to acuminate, perfectly smooth both externally and internally, pale yellow. Tepals fleshy- thickened at base with a shallow nectariferous depression about 2 mm. long and 1 mm. wide bordered distally by a very slightly elevated ridge or mound about 0.1 mm. high. Filaments smooth, more or less oblong but noticeably tapering towards both base and apex from a point approximately two-fifths the distance from base, about (2.0-) 3.0-6.0 (-7.0) mm. long and 0.5-1.2 (-1.6) mm. wide. Connective extending beyond the anther in an inwardly arching, obtusely conical beak about 0.2-0.7 mm. long and 0.4 mm. wide and not at all indented at apex. Anther-sacs elongate, about (8-) 12-15 (-20) mm. long. Ovary oblong- cylindric in general outline, obtusely and moderately 3-lobed with each segment slightly longitudinally grooved medially, about 2.5-4 mm. high, sessile. Style united for about one-fifth to one-third or rarely two- fifths the distance from the base towards the stigmatic tip; stigmatic tips slightly reflexing with age; style and stigmas together about 9-12 mm. long. Capsule with a rounded to nearly truncate apex and more or less obpyramidal in general outline but moderately 3-lobed and each lobe itself very slightly grooved; mostly about 0.8-1.5 em. high and (1.0-) 1.2-1.7 (-2.2) em. wide at its broadest point. Seed nearly globose, but often partially compressed out of shape, about 3-5 mm. in diameter, reddish-brown at maturity, about one-third to one-half enveloped by a thin membranous covering. Rich and usually moist hardwood- forested slopes and bottomlands from southern Quebee south along the mountains into Alabama (and even as far as southern Mississippi) and west into Arkansas, the eastern Dakotas and southern Ontario (Fig. 5). The possibility that this binomial may be superceded by U. lanceolata Ait. is discussed under that binomial with the Doubtful and Excluded Names at the end of this paper. 178 Rhodora [Vol. 65 Dietz (1952) has discussed introgression in both this and the following species under which I have presented a few comments. Most of the specimens that have proved difficult to identify resemble U. grandiflora in most particulars ex- cept that they are glabrous or very nearly so. $ |. ——9 PU | N X \ U.GRANDIFLORA ES | Fig. 5. Distribution of Uvularia grandiflora 5. Uvularia perfoliata L., Sp. 11. 304. 1753. (Type in the Linnean Herbarium. Phototype seen.) U. perfoliata B minor Michx., Fl. Bor.-Am. 1: 199. 1803. (Type at Paris; not seen, deseription seem- ingly definite.) Uvularia flava J. E. Smith, Exot. Bot. 1: 97. 1804. (Type not seen. Plate and description seemingly adequate for de- termination.) Plant completely glabrous; the stem slightly to heavily glaucous and about (1.5-) 2.0-3.5 (-6.0) dm. tall with (1-) 2 very slender, subter- ranean, slightly geniculate stolons up to 1.5 dm. long (apparently pres- ent only early in growing season). Leaves glaucous beneath, mostly 1963] Uvularia — Wilbur 179 (3-) 4-8 (-12) em. long and (1.5-) 2-4 (-6) cm. wide, ovate to ovate- oblong in general outline with a broadly rounded base and an acute to abruptly short-acuminate apex; the leaf subtending the first branch mostly about 4-8 (-10) cm. long and 2-5 cm. wide. Blade-bearing leaves below the first branch typically 3-4 but occasionally fewer. Leaf-margin smooth and very narrowly scariously rimmed. Fertile branchlet bear- ing one perfoliate leaf and one flower on a peduncle about (0.6-) 1.0-1.8 (-2.4) cm. long. Perianth segments (1.5-) 2.0-3.0 (-3.7) em. long and about 3-5 (-6) mm. wide, acute, densely papillose-tuberculate within, pale yellow but typically of an orange color in the area of the rounded mound-like papillosities; the rounded, half globose to oblong papillose beads varying greatly in size on each tepal, ranging from about 0.2- 0.5 (-1.0) mm. long. Tepals fleshy-thickened at base with a shallow but pronounced nectariferous depression about 1.5 mm. long and abcut 1 mm. wide and with an inconspicuous ridge bordering each side of the nectary. Filaments smooth, more or less oblong but noticeably taper- ing to both base and apex from a point about one-half the distance from base towards the apex, about 3-4 mm. long and 1.0-1.5 mm. wide, rather turgid. Connective extending beyond the anther into an in- wardly arching acute beak about 1 mm. long and about 0.5 mm. wide, very slightly indented at apex. Anther-sacs about 7-10 mm. long. Ovary oblong-cylindrical to obpyriform in general outline, obtusely and deeply 3-lobed with each lobe again slightly indented medially, about 3 mm. high, very slightly elevated at base by receptacular bulge but sessile and not stipitate. Style united for about half to two-thirds the distance to the stigmatic tip; stigmatic tips reflexing with age; style and stigmas together about 8-10 mm. long. Capsule nearly trun- cate at apex and obpyramidal in general outline but strongly 3-lobed and each lobe itself deeply 2-lobed with the lobes appearing apically rostrate, mostly about 0.8-1.3 cm. high and 1.2-1.6 cm. wide at apex. Seed subglobose to plumply reniform, about 3-6 mm. long, reddish- brown at maturity, one-third to one-half enveloped by a thin membran- ous, previously balloon-like covering originating from the funiculus and outer integument. Usually in open to densely forested deciduous slopes or well-drained uplands but occasionally in floodplain or swampy forests and more rarely growing in piney woods or even in meadows or along road or railroad embankments; ranging from southern New Hamp- shire and southern Ontario south into northern Florida and westward into central Ohio and Louisiana (Fig. 6). Dietz (1952, p. 243) has suggested that U. flava J. E. Smith was a hybrid (“probably an introgressive hybrid, or perhaps even an F,”) between U. perfoliata and U. grandi- flora stating that it resembled U. perfoliata “but with deeper yellow flowers and perianth smooth within — floral char- acteristics which suggest U. grandiflora.” Dietz also con- 180 Rhodora [Vol. 65 y UPERFOLIATA w | Fig. 6. Distribution of Uvularia perfoliata cluded that "U. flava apparently disappeared gradually, for the manuals of the day reflect increasing uncertainty about it until it finally was dropped into synonymy with U. per- foliata and was removed entirely from the literature." Smith, however, in both the plate and the original accom- panying description clearly indicated that the tepals were tubereulate-roughened within and that the anthers were pointed. He admitted that although he felt U. flava to be distinct from U. perfoliata “it is very difficult to express a specific distinction." The flower of U. flava was described as being "remarkably different, larger, more taper and elongated, with narrower sharper petals, an inch long, yel- low, with orange-coloured tubercles on the inside" and the point of the anthers are “also longer and more conspicu- ous," In contrast the tepals of U. perfoliata were described 1963] Uvularia — Wilbur 181 by Smith to be “scarcely half an inch long, of a very pale greenish buff-colour", and with “their inner side rough with yellowish protuberances." 'There is no evidence in the original aecount that would indicate that U. flava was any- thing other than a more orange-tinted race of U. perfoliata; there certainly is nothing to indicate that Smith was de- scribing either an F, or introgressant between U. perfoliata and U. grandiflora. Watson (Proc. Am. Acad. 14: 268. 1879) placed U. flava in the synonymy of U. grandiflora, a disposition which caused Gray (Am. Jour. Sci. ser. 3. 18: 313. 1879) to express surprise. Anderson (1954) and his student Dietz (1952) have both agreed that much of the variability observed within the perfoliate species of Uvularia is the result of introgressive hybridization. Amazingly, Anderson, on the basis of a de- tailed study of only ten herbarium specimens "selected at random" of each of the two perfoliate species, was able to conclude that 1) the specimens of U. grandiflora. formed a “coherent group", 2) those of U. perfoliata were “not so coherent" and 3) the variation within the last mentioned species was due to introgression from U. grandiflora. Dietz's (1952) more lengthy investigation convinced him that U. grandiflora was introgressing into U. perfoliata and that any introgression of U. perfoliata into U. grandiflora would be obscured by the possibility of undetected introgression from any of the sessile-leaved taxa. These sessile-leaved species, Dietz felt, were more closely related to U. perfoliata than that species was to U. grandiflora. In fact Dietz indi- cated that the sessile-leaved taxa would be somewhat inter- mediate in morphology and hence apparently in taxonomic position to the two perfoliate species but closer in their affinities to U. perfoliata. This conclusion as to relationship is in my opinion highly erroneous and demonstrates aptly the dangers inherent in weighing all characters and even tendencies as of equal importance in what certainly must be judged as a futile effort to achieve objectivity. Science will be advanced further by the more normal taxonomic proce- dure of attempting to evaluate the significance of the char- 182 Rhodora [Vol. 65 acters and basing classification upon the results of these considered judgments. Only a distorted picture can be ex- pected if one treats rather ill-defined tendencies such as the number of fruit per plant and general leafiness as being equally significant in determining relationships as the un- varying characters with strong morphologic bases such as the nature of the underground parts, the manner of attach- ment of the medial cauline leaves, fruit shape, presence or absence of ornamentation upon the tepals and the very dif- ferent raphial outgrowths of the seeds. Anderson and Whitaker (1934) point out that taxono- mists tend to “think of species in terms of the precise dif- ferences which permit their ready classification" and that *to them the really essential differences between U. grandi- flora, and U. perfoliata. will be those few discontinuous ones which are ordinarily used in identifying the species." An- derson and Whitaker contrasted this taxonomic viewpoint with that of those whose interest was “in the biological makeup of the units which are being classified" and stated this latter group found that “the many slight overlapping differences between species are more characteristic than the few sharply discontinuous ones." A taxonomist can be expected to be biased towards the advantages of the out- look held by his own discipline and my own study of Uvul- aria certainly did not present evidence that would convince me of the validity of this so-called “biological” viewpoint. It seems reasonable to conclude that the “variants more cr less difficult to classify” found in the area of overlap between the ranges of the two perfoliate species might well be the result of introgression between those two species and my experience with the genus emphatically confirms Dietz's conclusion “of the relative rarity of introgression as a fac- tor in the perfoliate Uvularias." The most useful character in distinguishing the two perfoliate species is the presence or absence of papillae upon the inner surface of the tepals. No specimen possessing tepals proved to be a problem in identification ; those lacking them occasionally did. Dietz's study was based upon mature specimens which, of course, 1963] Uvularia — Wilbur 183 lack tepals. A taxonomist attempts to identify specimens in all stages of maturity but he would consider it a pointless handicap not to utilize those characters which experience has shown to be the most reliable and instead to insist upon basing the study upon specimens in those stages most diffi- cult to distinguish. Anderson (1953) attempted to refute those who have felt that at least part of the variability encountered within species or even variability shared between segments of two supposedly distinct taxa might be due to common ancestry (i.e., a common “store of variability" or the “genetic pool hypothesis"). It still would appear that those features, thought by Anderson or Dietz to be indications of introgres- sion, such as the number of fruit borne on each plant, gen- eral leafiness, position of the longest internode, ratio of leaf-width to leaf-length, are all variable to the degree of overlapping between the two species. The noted variation is hence not necessarily the result of hybridization. Actually, Dietz admits the possibility of plants of “good” U. perfoliata with more than one fruit and leaves more numerous than nine but in spite of this both of these features are used as evidence of introgression from U. grandiflora when coupled with longer internodes and a higher ratio of leaf width to length since “where introgression is a factor, there is a strong tendency for the sum of the measured characters to vary together." To one who is not a student of introgres- sion, it would seem only proper to rule out as evidence of introgression those features that normally vary sufficiently in one species to match the condition characteristic of the other species. Characteristic of Dietz's approach are the comments made about a population of the midwestern perfoliate species from Missouri in which was noted *the complete absence of plants with more than one fruit." This was followed immediately by the statement that U. sessilifolia, which is found in Mis- souri, “is typically a single-fruited plant." A population sample from northern Minnesota “suggests strongly an in- trogression from U. sessilifolia" by the “presence of some 184 Rhodora [Vol. 65 sessile leaves on the upper portions of occasional plants", a more strongly divergent primary branch “as in U. sessili- folia" and *the gross appearance of the plant." Suspicion of miscegenation was so great that it was thought to be “noteworthy that a Uvularia cross apparently did not occur at Hawley [Pennsylvania] where two species occur togeth- er." Throughout most of the generic range two or more species grow either together or more characteristically in close proximity to one another. Evidence of hybridization between any of these species is certainly slight and between most species non-existent. Much of the published "evidence" is indeed flimsy and highly circumstantial and hence is any- thing but convincing. It would be of the greatest interest to know if it were possible for crosses to occur between spe- cies of the two sections recognized in this account of the genus and in spite of the considerable time and effort that might be required to attempt such experimental crosses this would seem to be the only certain way to demonstrate or even to disprove partially the possibility of intersectional crosses. Certainly evidence of the circumstantial sort pre- viously presented cannot be accepted. The judicial Anglo- Saxon precept of *innocent until proven guilty" ought to be extended by botanists to plants. Baker's (Nature 159: 221- 223. 1947) recommendation that “wherever possible, arti- ficial erosses between the forms involved . . . should precede investigations of natural populations" has much merit in spite of Dietz's excuse that "experimental data would in- volve a prohibitively long investigation." These lengthy in- vestigations appear to be the only ones that will provide the desired information. DOUBTFUL AND EXCLUDED NAMES Uvularia acutifolia Raf., Am. Monthly Mag. 359. 1817. nom. nud. Uvularia amplexicaule Mill., Gard. Dict. ed. 8. No. 1. 1768. — Strep- topus amplexifolius (L.) DC., Fl. Fr. 3: 174. 1805. Uvularia amplexifolia L.. Sp. Pl. 3804. 1753. — Streptopus amplexifolius (L.) DC., Fl. Fr. 3: 174. 1805. Uvularia Betua Buch.-Ham. ex D. Don, Trans. Linn. Soc. 18: 517. 1841. nom. illegit. (Art. 34.4). Published in synonymy of Disporum Hamil- tonianum (Wall.) D. Don, le. 1963] Uvularia — Wilbur 185 Uvularia calcarata Wall., Cat. n. 5087. 1831-32. = Disporum. calcara- tum (Wall) D. Don, Trans. Linn, Soc. 18: 516. 1841. Uvularia caroliniana (Gmel.- Wilb., Rhodora 63: 39. 1961. ( = Anony- mos pudica, Walt., Fl. Car. 123. 1788. nom. illegit. Art. 43. Erythron- ium carolinianum Gmel., Syst. Nat. 2(1) : 546. 1791. U. pudica (Walt.) Fern., Rhedora 41: 536. 1939. nom. illegit. Art. 43.) I now belatedly concur with Dietz's conclusions (1952, p. 226), if not completely with his reasoning, that Walter's account, the basis for all the above-mentioned synonyms, is too uncertain as to the species (or species) described to warrant the abandonment of Michaux's clearly-defined and certain binomial, U. puberula. It now would appear that both Fernald and I relied too heavily upon the note made by Asa Gray after examining the specimens in Walter's her- barium in 1839 that “Anonymos (Erythronio aff.) pudica! = Uvularia puberula.” Even though Gray may well have seen a specimen of what is here called U. puberula in Wal- ter's herbarium, it does not follow that the specimen seen by Gray should be accepted as the type of Anonymos pudica in view of the conflicting statements in the original ac- count in Walter's Flora. The fruit and seed were described by Walter in the generic description as follows: Per. capsula turbinato-triangularis, angulis bifidis, tri- loeularis, trivalvis. Som. bina in singulo loculamento depresso-globosa, ab uno latere cincta membrana vesiculari. The distinguishing features mentioned here are “capsula turbinato-triangularis, angulis bifidis . . ." and“... ab uno latere cincta membrana vesiculari". This best describes U. perfoliata which possesses an inversely pyramidal capsule deeply lobed at each of the three apical angles and a bladdery membrane half-encompassing the seed. These features cer- tainly exclude from consideration any member of section Oakesiella, all of which possess hornless, triquetrous cap- sules tapering to both apex and base and seeds with a har- dened, double ridge extending half-way around each in a manner resembling the crest on a Roman helmet. In the diagnosis of Walter's species, the only even vaguely delimiting term is "folis amplexicaulibus". Michaux de- 186 Rhodora [Vol. 65 scribed the leaves of U. puberula as sub-amplexicaul, and, although any such term seems to be an overemphasis of the sessile leaves of this species, it still would appear to be a better description of it than of the perfoliate leaves of U. perfoliata. And of course Gray, as reported by Fernald (Rhodora 41: 537. 1939), did identify a specimen that he took to be Anonymos pudica as U. puberula in Walter's col- lection. In view of the certainty of the identity of the plant in- cluded in Walter's generic description as being other than the species to which the name has been applied and the in- conclusiveness of the one word in the original diagnosis for the species, which might be considered at all diagnostic, it would seem best to treat Gmelin's binomial, based completely as it was upon Walter's account, as a nomen dubium and to take up again Michaux's unequivocal U. puberula. Uvularia chinensis Ker-Gawl., Curtis Bot. Mag. t. 916. 1806. = Dis- porum. cantoniensis (Lour.) Merr., Phil. Jour. Sci. 15: 229. 1919. Uvularia cirrhosa Thunb., Fl. Jap. 136. t. 2. 1784. = Fritillaria. Uvularia distorta (Michx.) Pers., Syn. Pl. 1: 360. 1805. = Streptopus amplexifolius (L.) DC. var. americanus Schultes, Syst. Veg. 7: 311. 1829. ( = Streptopus distortus Michx., Fl. Bor.-Am. 1: 200, 1803). It is very doubtful if Persoon actually should be charged with having published this combination for it appears under the genus Uvularia merely as “7. amplexifolia (distorta), . . ." Uvularia grandifolia Pursh, Fl. Am. Sept. 1: 231. 1814. [typographical error for U. grandiflora J. E. Sm.] Uvularia Hamiltoniana Wall., Cat. n. 5088. 1831-32. = Disporum cal- caratum (Wall.) D. Don var. Hamiltonianum (Wall.) Baker, Jour. Linn, Soc. Bot. 14: 589. 1875. Uvularia hirta Thunb., Fl. Jap. 136. 1784. = Tricyrtis hirta (Thunb.) Hook., Bot. Mag. t. 5355. 1863. Uvularia lanceolata Soland. in Ait., Hort Kew. 1: 434. 1789. A strong suspicion exists that Solander here for the first time provided a binomial for the plant that is now called U. grandiflora. Aiton's account is presented below in full. lanceolata. 2. U. foliis perfoliatis ovato-lanceolatis acuminatis. Polygonatum ramosum flore luteo minus Corn. canad. 40. tab. 41. Spear-leav'd Uvularia. Nat. of North America. Introd. 1785, by Mr. Archibald Menzies. Fl. July H. 4. 1963] Uvularia — Wilbur 187 Cornut's plate is anything but diagnostic but there is a slight hint of U. grandiflora in the leaves depicted as there is also in the diagnos- tic phrase of Aiton and the suggested common name. Pursh (Fl. Am. Sept. 1: 231. 1814) equated Aiton's name to U. grandiflora “according to specimens in the Herbarium of A. B. Lambert, Esq." while Baker (Jour. Linn. Soc. Bot. Lond. 17: 462. 1879) includes U. lanceolata in the synonymy of U. perfoliata noting that it “est forma mera angusti- folia". The name obviously should remain unassigned until authentic specimens are discovered. Uvularia lanuginosa (Michx.) Pers., Syn. Pl. 1: 360. 1805. = Dis- porum lanuginosum (Michx.) Nicholson, Dict. Gard. 2: 485. 1884. Uvularia lanuginosa B major Hook., Fl. Bor.-Am. 2: 174. 1838. = Dis- porum trachycarpum (S. Wats.) Benth. & Hook. f., Gen. Pl. 3: 832. 1883. Uvularia Leschenaultiana Wall., Cat. n. 5089. 1831-32. = Disporum Leschenaultianum (Wall.) D. Don, Trans. Linn. Soc. 18: 518. 1841. Uvularia multiflora Reinw. ex Kunth, Enum, Pl. 4: 207. 1843. nom. illegit. (Art. 34.4). Published by Kunth in the synonymy of Dis- porum multiflorum Don, l.c. Uvularia oppositifolia Schlecht. ex Kunth, Enum, Pl. 4: 254. 1843. nom, illegit. (Art. 34.4). Published by Kunth in the synonymy of Fritillaria camtschatcensis (L.) Ker., Bot. Mag. sub t. 1216. 1809. Uvularia parviflora Wall, Asiatic Res. 13: 378. 1820. = Disporum parviflorum (Wall.) D. Don, Prodr. Nep. 50. 1825. Uvularia Pitsutu Buch.-Ham. ex D. Don, Prodr. Fl. Nep. 50. 1825. nom. illegit. (Art. 34.4). Published by Don in the synonymy of Dis- porum. Pitsutu Buch.-Ham. ex D. Don, l.c., = Disporum cantoniensis (Lour.) Merr., Phil. Jour. Sci. 15: 229. 1919. Uvularia puberula J. E. Sm., Rees. Cycl. 37: No. 5. 1818. — Disporum Smithii (Hook.) Piper, Contr. U. S. Nat. Herb. 11: 201. 1906. Uvularia pudica (Walt.) Fern., nom. illegit. see discussion under U. caroliniana. Uvularia rosea (Michx.) Pers, Syn. Pl. 1: 360. 1805. = Streptopus roseus Michx., Fl. Bor.-Am. 1: 201. 1808. Uvularia sessiliflora J. F. Gmel. Syst. Nat. ed. 13.2: 1642. 1792. Uvularia sessilis Thunb., Fl. Jap. 135. 1748. = Disporum. The com- bination D. sessile has been attributed to D. Don (Fl. Nep. 50. 1825) but it was not made there according to the strictures of Art. 33. The only reference there by Don is made following the generic de- scription and is as follows: “Obs, Huc Uvularia chinensis Bot. Mag. et Uvularia sessilis Thunb. Jap. cui folia, sessilia, et flores sub- solitarii albi.” But since Thunberg included the binomial U. ses- silifolia L. as a synonym, it would appear that the legitimacy of the name itself ought to be investigated thoroughly. Uvularia Smithii Hook., Fl. Bor. Am. 2: 174. t. 189. 1838. — Disporum Smithii (Hook.) Piper, Contr. U. S. Nat. Herb. 11: 201. 1906. 188 Rhodora [Vol. 65 Uvularia umbellata Wall., Asiatice Res. 13: 379. 1820. = Disporum cantoniensis (Lour.) Merr., Phil. Jour. Sci. 15: 229. 1919. Uvularia ? viridescens Maxim., Prim. Fl. Amur. 273, 1859. — Disporum smilacinum A. Gray, Perry Jap. Exped. 2: 321. 1856. DEPT. OF BOTANY, DUKE UNIVERSITY, DURHAM, NORTH CAROLINA REFERENCES CITED ANDERSON, E. 1953. Introgressive hybridization. Biol. Rev. 28: 280- 307. 1954. Efficient and inefficient methods of measuring specifie differences. pp. 93-106 in O. Kempthorne et al. Statistics and mathematies in biology. Iowa State College Press, Ames. . AND T. W. WHITAKER. 1934. Speciation in Uvularia. Jour. Arnold Arb. 15: 28-42. Dietz, R. A. 1952. Variation in the perfoliate Uvularias. Ann. Mis- souri Bot. Gard. 39: 219-247. SMALL, J. K. 1903. Flora of the southeastern United States. (Uvu- laria and Oakesiella pp. 268, 271 & 272, 1328). Lancaster, Pa. 1933. Manual of the southeastern flora. (Uvularia and Oakesiella pp. 295, 299 & 300, 1500.) New York. WATSON, S. 1879. A revision of the North American Liliaceae. Proc. Am. Acad. 14: 213-288. THE NEW YORK BOTANICAL GARDEN WILD FLOWER BOOKS The New York Botanical Garden has undertaken the preparation of a series of illustrated books on the wild flow- ers of the United States (excluding Alaska and Hawaii). The books will be directed to the amateur with little or no botanical training, as well as to the professional scientist in non-botanical fields, and will present wild flowers without the embellishment of technical descriptive terms found in manuals and floras. At the same time the descriptions will be scientifically accurate and the coverage complete within the usual concept of “wild flowers" (excluding grasses, sedges, and unattractive weeds, and, of course, all woody plants). Most species will be illustrated in color, and the very best available color photographs are now being sought. Line drawings will be used to distinguish related species whenever color is inadequate for the purpose. Means of tracing the plant to its family, genus, and species will be included. But the dichotomous keys so useful to the taxono- mist and so terrifying to the layman will be conspicuously absent. It is planned to treat the United States in five regions, a volume being devoted to each: the northeastern states, the southeastern states, the central plains and mountains (in- cluding the Rocky Mountains and the Great Basin), the Southwest, and the Pacific Northwest. A committee headed by Mrs. David Rockefeller has undertaken to obtain the subsidy necessary to ensure publi- cation. The work will be centered at The New York Botani- cal Garden, under the direction of Harold William Rickett, Senior Botanist. An artist, a photographer, and a secretary will devote all or part of their time to the work. William C. Steere will serve as General Editor. For outstanding illus- trations in color the cooperation is sought of amateur and professional photographers of wild flowers in all parts of the country. The first volume will deal with the northeastern states, and color photographs (transparencies) of the wild flowers of this region are now required. Amateurs are urged to send their best “shots” at once to Dr. Rickett. H. W. Rickett, The New York Botanical Garden. 189 PHORADENDRON RUBRUM IN FLORIDA Dr. Frank C. Craighead, on low-flying flights over the northern keys of Monroe County, Florida, observed on ma- hogany trees, Swietenia Mahagoni Jacq., a parasite which appeared to kill gradually the over-mature hosts. In the company of Doctors John Popenoe and Carl W. Campbell, of the Subtropical Agricultural Experiment Station, Home- stead, Dr. Craighead, searching on North Key Largo, found a tree on which the parasite was growing low enough to be reached. The plant proved to be Phoradendron rubrum (L.) Grisebach, a species known from Cuba and the Bahamas, where it has been found only on Pithecellobium and Bursera, according to labels on specimens in the Gray Herbarium. The mahogany trees on which Dr. Craighead has ob- served Phoradendron rubrum have been practically always large specimens, two to four feet in diameter and twenty or more feet to the lowest limbs. On the higher branches the mistletoe forms large masses two to three feet in diameter, very conspicuous in winter when the hosts shed their leaves. Dr. Craighead reports that recently-dead trees, seen from a plane fifty feet above them, were covered with dead or dying mistletoe, indicating that the parasite may play a part in killing the trees. He writes, “There are very few of these big mahoganies left on Key Largo. Some occur on Rhodes and Sand Keys just to the north. The mistletoe is here also from what I could make out in flying over. These mahoga- nies are the remnants of a fine stand that has been depleted over the past thirty to fifty years. Most of them are defec- tive and for that reason were left. They occur in the ham- mocks of these keys and on a narrow coastal strip from Fla- mingo to US highway 1 on the tip of the peninsula. On the keys the mahoganies grow on coral rock; on the mainland on marl soil or oólitic limestone, the largest trees on the latter.” Specimens of Phoradendron serotinum (Raf.) M. C. John- ston (P. flavescens) in the Gray Herbarium indicate Lee County, Florida, as the southern limit of its range, a loca- tion one hundred miles northwest of Key Largo. It can be distinguished quickly from P. rubrum by its whitish berries 190 1963] Phoradendron Rubrum — Cooley 191 and terete internodes and, upon more critical examination, by the lack of cataphylls. Phoradendron rubrum (L.) Grisebach, Fl. Br. W. Ind. 314. 1864. Viscum rubrum L., Sp. Pl. 1023. 1753. Parasitic, evergreen shrub; stems and branches ligneous, furrowed in drying; sections of the branches with pairs of basal, broad and spreading, often connate cataphylls, and rarely with one or two addi- tional pairs upward, particularly on new growth; these sections 8 to 32 mm., furrowed when dried, terete at base, two-edged and flattened upward with a 90? tortion from base to apex; leaves elliptical, spatu- late to obovate, mostly cuneate at the base, short-petioled, caducous on drying and at death of plant; veins few, prominent, pinnate, branch- ing; inflorescence axillary, spicate; spikes elongating through a series of cup-like braets, monoecious, with trimerous flowers, the staminate flowers above, berries globose, glabrous, about 4 mm. in diameter, lemon to light orange in color. Specimens cited: FLORIDA. Key Largo, Monroe County, growing on mahagoni in hammocks, berries orange, 19 April 1941, J. M. Crevasse (FLAS); South end of Key Largo, Monroe County, on large Swietenia Mahagoni in hammock, berries orange, 6 February 1944, S. J. Lynch (FLAS); Key Largo, Monroe County, on Swietenia, berries pink to orange, March 1953, Ray Garrett (FLAS, GH); North Key Largo, Mon- roe County, on Swietenia Mahagoni, 2 February 1963, Frank C. Craig- head, Carl W. Campbell and John Popenoe (GH, USF). GEORGE R. COOLEY, GRAY HERBARIUM,HARVARD UNIVERSITY. EDITOR'S NOTE Two scientifie articles of considerable length on South American plants have appeared recently in Rhodora. This does not represent a new departure since Rhodora in the past has had papers dealing with the continent south of us. At this time we would like to define in a general way the geographical coverage of our journal in order to assist our contributors and facilitate the flow of good manuscripts in our direction. We would like to emphasize the fact that suitable papers on South American botany will be welcome insofar as space will permit their printing. We feel that it is important for Rhodora to be in the main stream of tax- onomic research. As indicated by Dr. Schultes in his ac- companying article in this number, South America is un- usually rich in numbers of species of Angiosperms. In our hemisphere inevitably most of the descriptive taxonomy of the future will relate to the American tropics. In our opinion Rhodora should play a part in this great pioneering venture. Also, many members of the New England Botanical Club and some of the editorial board of Rhodora are very active in South American Taxonomy, and Rhodora certainly should make its pages available for their contributions. As has been customary, in order to maintain some geo- graphical identity, we shall continue not to accept papers dealing chiefly with plants of the Old World. A. R. HODGDON. Volume 65, No. 762 including pages 97-192, was issued June 28, 1963. 192 K:RLOW REFERENCE LIBRARY 75 R7 QCT 2 107? Todora 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 Associate Editors Vol. 65 July-September, 1963 No. 763 CONTENTS: The Genus Reverchonia (Euphorbiaceae). Grady L. Webster and Kim I. Miller ..................... erret 193 An Annotated List of Vascular Plants from Cape Sabine, Alaska. Stamwyn G. Shetler ..ccccccorrccccercreccersssceseccccccsescccnctoonses 208 Chromosome Studies in Mexican Compositae. D. C. D. De Jong and E. K. Longpre .iisssimsesssssesseossascsoconeasoss 225 | Habenaria conopsea in North America. Donovan S. Correll i misssssiiearissnisehsaseerant diokyia pintaa s e resi RR 241 A Companion Volume to the New Britton and Brown Illus- trated Flora. (Review). Reed C. Rollins .............................. 248 (Continued on Inside Cover) The Nef England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. CONTENTS: — continued Further Considerations in Stylosanthes (Leguminosae). Robert H. Mohlenbrock ......... eene nns — 245 Nomenclatural Changes in the Family Palmae. S. F. GASSMAN wo.eesecccsccsessessscsssessesseecesesssassecscessseseacsesscaccacecenenens 259 Hybridization of Rubus hispidus and R. setosus. Frederic Steele and A. R. Hodgdon ............ eee 262 Hedyotis australis in Georgia. Kenton L. Chambers ............ 271 Drosera filiformis in Connecticut. William R. Linke, Jr. ........ 273 Epilobium X wisconsinense hybr. nov. Donald Ugent .......... 274 The Identity of Bumelia lacuum Small. Olga Lakela ............ 280 John Lyon, Nurseryman and Plant Hunter (Review). Albion R. Hodgdon ................ ——— —— 283 New Records of Grasses from the Chicago Region and Lower Michigan. S. F. Glassman ........... eee 284 Notes on Great Wass Island, Maine. Radcliffe B. Pike and A. R. Hodgdon ............. tes 285 Note on Primula laurentiana in Maine. Radcliffe B. Pike ...... eee ttt tnter 286 QTRbooora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 65 July-September, 1963 No. 763 THE GENUS REVERCHONIA (EUPHORBIACEAE)! ? GRADY L. WEBSTER AND KIM I, MILLER Reverchonia, a monotypic genus of the subfamily Phyl- lanthoideae, has been placed in the subtribe Phyllanthinae adjacent to Phyllanthus (Pax and Hoffmann, 1931). The single species, R. arenaria A. Gray, is a highly specialized desert annual (fig. 2) found in disjunct sand-dune areas in the southwestern United States and northeastern Mexico. Gray, in the original description (1880), noted that “the relationship of this plant to Phyllanthus is so close, that, were it not for a combination of characters, it might be taken for an aberrant Phyllanthus.” This character-com- plex, which has sufficiently impressed subsequent authors so that all have followed Gray’s lead and maintained Rever- chonia as distinct, includes features of both habit and repro- ductive organs. The aspect of the fruit-bearing plant, due Research was supported by a grant from the National Science Foun- dation (G23604). The authors are indebted to Barbara Webster and Lillian Miller for assistance, and to Mr. Chester M. Rowell for his diligence in obtaining materials. *We would like to express our thanks to the curators of the following herbaria for loaning specimens or furnishing locality data: University of Arizona (ARIZ); Chicago Natural History Museum (F); Gray Herbarium (GH); Missouri Botanical Garden (MO); Uni- versity of New Mexico (UNM); Oklahoma State University (OKLA) ; University of Oklahoma (OKL); Southern Methodist University (SMU); University of Texas (TEX); Texas Technological College (TTC) ; United States National Museum (US) ; Utah State Agricultural College (UTC) ; and University of Utah (UT). Voucher specimens for research described in this paper are de- posited at the Kriebel Herbarium, Purdue University (PUL). 193 194 Rhodora [Vol. 65 Fig. 1. Flowers of Reverchonia (Warnock 10723). A. Male flower, with one calyx-lobe removed. B. Male calyx-lobe, adaxial view. C. Disk of male flower (as seen from above, partially enclosing the filament bases). D. Female flower, with 2 calyx-lobes removed. 1963] Genus Reverchonia — Webster and Miller 195 to the disproportionately large capsules hanging from slen- der lateral branches, is distinctive and unlike any other taxon of American Phyllanthoideae. Especially prominent reproductive features (fig. 1) include the dark reddish flow- ers with unusually shaped calyces (the male vaguely sug- gesting the flowers of certain Fumariaceae), the introrse stamens and central disk of the male flower, the large seeds with a subchalazal invagination, and the embryo with linear cotyledons. As duly noted by Gray, the most anomalous character of Reverchonia is the narrow cotyledon shape, which would technically place the genus in the “series” Stenolobeae as conceived by Mueller Argoviensis (1866). The Stenolobeae, as delimited by Mueller and later by Pax (1890) and Gruen- ing (1913), comprise a number of genera with more or less ericoid habit which are restricted to Australia and New Zealand. Except for the narrow cotyledons, however, most of these taxa seem to have little in common with Rever- chonia. Micrantheum, which would probably key out the nearest, differs in having foliate stipules, extrorse anthers, a pistillode in the male flower, and (like most other Steno- lobeae) carunculate seeds. A search has been made among genera of American Phyl- lanthoideae other than Phyllanthus for any which might show similarities to Reverchonia. The only group in which any degree of resemblance can be detected is T'etracoccus. As recently revised by Dressler (1954), it comprises 4 taxa of desert shrubs which grow in southwestern desert areas adjacent to the range of Reverchonia. Although the species of Tetracoccus are completely dissimilar in overall habit (being intricately branched bushes), the leaves (when en- tire) have an aspect suggesting those of Reverchonia. The flowers show a considerable superficial resemblance, especi- ally in the central male disk and the dilated style-tips. The likeness of the male disk is especially striking, since it is in this particular character that Reverchonia diverges most strongly from Phyllanthus. In order to evaluate possible relationships of Reverchonia 196 Rhodora [Vol. 65 with Phyllanthus, Tetracoccus, and other Phyllanthoideae, we have carried out palynological, cytological, and anatomi- eal studies. Punt (1962), in a valuable detailed study of Euphorbiaceous pollen, has described the microspores of most of the taxa in question. He shows that in most of the taxa of Stenolobeae, together with certain other Austral- asian taxa possessing broad cotyledons, the pollen grains are of a very characteristic echinate, non-colporate type (“Aristogeitonia configuration"). The microspores of Tet- racoccus ilicifolius, with their 6-7 small colpi and prominent spines 3.5 , long, definitely belone in this grouping. In contrast, the pollen grains of Reverchonia are prolate, tri- colporate with a small colpus transversalis, and psilate (the tectate exine essentially smooth). Punt's illustration (his Plate II, fig. 7) and that of Erdtman (1952: fig. 97D) agree well with our own observations (on microspores from Miller & Miller 1322, Webster 4615) except that the exine reticu- lation is much fainter than ind'eated in Erdtman's drawing. Punt classified the grains of Reverchonia in his Phyllanthus pentaphyllus subtype along with those of several herbaceous species of Phyllanthus and Savia erythroxyloides. In the species of Phyllanthus sect. Phyllanthus examined by us the tectate exine is distinctly finely pitted (Punt describes the exine of P. niruri as "intra-reticulate"), and Reverchonia differs only in its somewhat more obscure ornamentation. However, despite the palynological similarity, the herbace- ous species of Phyllanthus sect. Phyllanthus do not appear very similar to Reverchonia, as they are highly specialized vegetatively (i.e. with phyllanthoid branching) and dis- similar in floral details (e. æ., male disk of distinct segments, anthers extrorse, seeds differently ornamented). Rever- chonia shows a certain approach to the condition of phyl- lanthoid branching (as defined by Webster. 1956: 104) in that flowers are borne only on the lateral determinate axes; however, it differs in the lack of reduction of leaf-blades on the main axis and in the spiral rather than distichous phyl- lotaxy of the lateral branchlets. Chromosome counts have been made by the junior author 1963] Genus Reverchonia — Webster and Miller 197 from aceto-carmine squashes of root-tips and immature leaves. The chromosome complement of Reverchonia proves to be Zn — 16 (fig. 3). This is a very striking result, in view of the fact that within the subfamily Phyllanthoideae this Fig. 2. Habit of Reverchonia, sand dunes near Kermit, Texas, X 1/10 (photograph by C. M. Rowell). Fig. 3. Mitotic chromosomes of Reverchonia, showing 2n — 16, X 2000 ( K. and L. Miller 1322). Fig. 4. Longitudinal section of ovary, Reverchonia, showing ovule and associated parts (O, obturator; N, tip of nucellus), X 60 (K. and L. Miller 1322). Fig. 5. Longitudinal section of ovary, Argithamnia mercurialina, x 60 (K. and L. Miller 1174). Plate 1286 198 Rhodora [Vol. 65 number has been found only in Phyllanthus polygonoides, belonging to sect. Isocladus (Webster & Ellis, 1962). In most of the other Phyllanthoideae studied, the base number is 13; and in Tetracoccus it appears to be 12 (Perry, 1943). The cytological evidence, therefore, strongly supports a closer relationship between Reverchonia and sect. Isocladus. The pollen grains of taxa in this section certainly would not negate such a relationship, since they show considerable similarity to Reverchonia. As observed in P. platylepis and P. polygonoides, they differ mainly in being less prolate and in having a less elongated colpus transversalis. However, all known species of sect. Isocladus are strongly dissimilar in certain floral details, including extrorse anthers, male and female disks of separate segments, stamens united by the filaments, female calyx-lobes herbaceous, stigmas subcapi- tate, columella persistent, and seeds small with verruculose ornamentation and lacking a hilar invagination. Studies on the gynoecium in taxa of Phyllanthoideae have yielded most interesting results which suggest particular relationships between Reverchonia and other putatively re- lated genera. Gray (1880) had noted that Reverchonia was unusual in having amphitropous ovules. Since Pax and Hoffmann (1931) flatly categorize the ovules of the Euphor- biaceae as anatropous, the situation in Reverchonia might seem unusual indeed. However, anatomical researches in progress indicate that amphitropous ovules occur in a con- siderable number of Phyllanthoideae, as was clearly pointed out long ago by Baillon (1858: 164). The ovule of Rever- chonia, as seen in longitudinal section (fig. 4), resembles that in Phyllanthus and allied genera in having 2 well-devel- oped integuments, a nucellar beak extending through the micropyle and in contact with the obturator, and a funicle which departs from the placenta below the middle of the locule. The funicle is definitely attached to the ovule toward its base. In the sense of Goebel (1933: 2003), the ovule of Reverchonia would be classed as “hemitropous.” In contrast, the ovule in most other Euphorbiaceae is definitely anatrop- ous (fig. 5), with the funiculus departing from the upper 1963] Genus Reverchonia — Webster and Miller 199 half of the locule and adnate to the ovule for most of its length (cf. Schweiger, 1905). These ovular differences are usually also apparent in the seed, despite various ontogenetic changes in proportion. In the usual anatropous Euphorbiaceous seed the hilum is near the micropylar end and the raphe (funicle scar) traverses the length of the seed before ending at a chalazal area often marked by a depression. In the seed of Reverchonia and certain other Phyllanthoideae, the hilum is below the middle of the seed (subchalazal) and there is no definite raphe. In such seeds, as well as some anatropous ones, the chalazal pit may be the most conspicuous external mark on the seed- coat, and it has been described as the hilum by some in- vestigators. Vindt (1953), for example, refers to the sub- micropylar attachment of the funicle as the “hile apparent" and its chalazal ending as the “hile vrai." A survey of ovular configurations in the Euphorbiaceae, although not yet complete, indicates that the two kinds of ovules are correlated with definite systematic groups. Am- phitropous ovules are found in Reverchonia, Phyllanthus, and a number of other genera in the Phyllanthoideae, while anatropous ovules occur in other Phyllanthoideae and in all of the Crotonoideae and Stenolobeae. Tetracoccus has car- unculate anatropous seeds which, as pointed out by Croizat (1942), resemble those of Petalostigma and certain other Phyllanthoideae of Australasia. Thus the palynological, cytological, and anatomical data all speak strongly against any close relationship of Tetracoccus with Reverchonia; any similarities must be due to convergent evolution. For the same reasons Tetracoccus cannot be related to Secur- inega (in the usual sense) either, as suggested by Croizat. Securinega (including the closely related Fluggea) has tri- colporate reticulate pollen grains, amphitropous ovules, and (at least in Fluggea virosa) a haploid chromosome number of 13 (Webster and Ellis, 1962). Certain species of Secur- inega, in fact, show a definite resemblance to Reverchonia; in the mediterranean S. buxifolia the male disk-segments are fused in a manner suggestive of the disk in Reverchonia. On 200 Rhodora [Vol. 65 the other hand, all the species of Securinega have a definite pistillode in the male flower, and all are shrubby plants dis- similar in habit. Furthermore, as shown above, the base chromosome number in Seeurinega is 13 rather than 8 as in Reverchonia and Phyllanthus sect. Isocladus. On the basis of the evidence in hand, it appears that Reverchonia definitely belongs in the subtribe Phyllanthinae, where it has much in common with both Securinega and Phyllanthus sect. Isocladus. Although the chromosomal evi- dence suggests a closer relationship to /socladus, the chro- mosome numbers in subtribe Phyllanthinae are too poorly sampled for this to be considered conclusive. It is possible that the closest affinity of Reverchonia may turn out to be with some Old World taxon of Phyllanthinae rather than with any of the American groups considered in this paper. Tetracoccus, which should be excluded from the Phyllan- thinae, may share with Reverchonia a similar migrational history; both genera appear to be relict groups surviving from an ancient dispersal of Phyllanthoideae through tropi- cal or subtropical desert regions. Johnston and Warnock (1963) have questioned the status of Reverchonia as a genus distinct from Phyllanthus, on the grounds of its lack of diagnostic morphological char- acters. It is true that except for the male disk there are no diagnostic features which would immediately separate the two genera. Gray's recognition of Reverchonia on the basis of its particular ensemble of characters still seems the most reasonable solution. Because of the isolated position of Reverchonia, its inclusion in Phyllanthus would extend fur- ther the boundaries of that already vastly diversified genus. For those who still prefer to base classification on phylogeny, inclusion in Phyllanthus would seem unwise in view of the possibility that Reverchonia may be more closely related to Securinega. SYSTEMATIC TREATMENT REVERCHONIA A. Gray, Proc. Amer. Acad. Arts and Sci. 16:107. 1880. Annual herbs; phyllotaxy spiral on all axes, branches persistent; leaves entire, stipulate, petiolate. Monoecious; flowers pedicellate, in 1963] Genus Reverchonia — Webster and Miller 201 dense axillary clusters (cymules) on lateral branches; cymules typic- ally androgynous, with one central female and several lateral male flowers. Male flower: calyx-lobes 4, biseriate, inflated, constricted above the middle, the distal portion flaring abaxially; disk central, deeply 4-lobed, partially surrounding the bases of the stamens; stamens 2, opposite the outer calyx-lobes; filaments free; anthers introrse, dehiscing longitudinally and vertically; pollen grains prolate, tricol- porate, tectate. Female flower: calyx-lobes 6, not inflated as in the male; disk subentire or angular; carpels 3, styles fused below, stigmas bilobed; ovules two in each cell, collateral, amphitropous. Fruit cap- sular, dehiscent, 6-seeded; columella usually deciduous; seeds trigon- ous, with a conspicucus subchalazal invagination; embryo slightly curved, cotyledons very narrow. Type species: Reverchonia arenaria A. Gray. Reverckonia arenaria A. Gray, Proc. Amer. Acad. 16: 107. 1880 Glabrous annual herb becoming 2-5 dm, high, with sparsely branch- ing taproot; main stem subterete, smooth, glaucous-white; lower later- al branchlets 2-3 dm. long (upper ones shorter), mostly 1.5-2 mm. thick. Leaves: stipules reddish, papery, lanceolate, acuminate, per- sistent, entire or irregularly toothed, (0.7-) 0.9-1.7 (-2.3) mm. long; leaf-blades elliptic to narrowly oblong-elliptic or nearly linear, thickish, c. (15-) 20-40 (-45) mm. long, (1.8-) 2.5-8 (-9) mm. broad, apiculate at the tip, narrowed at the base, veins more or less obscure; petiole 1-3 mm. long. Flowers in reduced bracteolate cymules axillary to the leaves on lateral branchlets (never on main stem), each cymule typically pro- ducing 1 central female and 4-6 lateral male flowers. Male flower: pedicel slender, 1.5-2.5 mm. long; calyx-lobes 4, ovate-oblong, sub- medianly constricted, 1-veined, obtuse, purplish or pinkish with a narrow central stomatiferous greenish strip, c. 1.5-2.5 mm, long, 0.7- 1.5 mm. broad; disk of 4 roundish lobes continuous across the center of the flower (between the stamens), with the outline of an I-beam; stamens 2, opposite the outer calyx-lobes; filaments free, subterete, 0.7-1.2 mm. long; anthers erect, oblong, c. 0.5-0.75 mm. long, dehiscing vertically; pollen grains very finely tectate-reticulate, with prominent colpus transversalis. Female flower: pedicel stout, c. 1.5-2 mm. long at anthesis, becoming (2.5-) 3.2-6.5 (-8.7) mm. long in fruit; calyx- lobes 6 (rarely 5), oblong, colored and distally constricted as in the male but not especially inflated, becoming (1.3-) 1.5-2.5 (-2.9) mm. long; disk flat, rather thin, 1.1-1.8 mm. in diameter, roundish or 6- angled in outline; ovary oblate-spheroidal, smooth, grooved; styles erect, 0.5-0.8 mm. high, united halfway or less, stigmas somewhat dilated, emarginate or slightly bilobed. Capsule oblate-spheroidal, smooth, stramineous, 7-9.8 mm. in di- ameter; columella usually deciduous. Seeds trigonous, dark- or reddish- brown, smooth on the back (tangential face), papillate on the lateral 202 Rhodora [Vol. 65 (radial) faces, (4.4-) 4.7-6.2 (-6.6) mm. long; hilum subchalazal, deep- ly invaginated, with a raised thickened rim. Cotyledons (measured on seedlings) linear, c. 20-30 mm. long, 1.2-1.6 mm. broad. Type: Texas, Baylor Co., sandy island in the Brazos River near Seymour, September 1879, Reverchon (GH, lectotype; F, MO, isolecto- types). Gray did not cite a collection number, but the duplicate sheets at Chicago and St. Louis bear the number 876. Gray also cited the collection made by Bigelow in 1853; this was apparently the first dis- covery of the genus. Judging from the map and itinerary of Whipple’s exploring party (Gorman, 1941) Bigelow collected the plant along the bed of the Canadian River in northeastern Hutchinson County, Texas, between Spring Creek and the Roberts County line. DISTRIBUTION : OKLAHOMA. coTTON CO.: along Red River, Burkburnette Bridge, Wood 15 (OKL, OKLA). ELLIS CO.: shinnery sand hills, Engleman (OKL, TEX). HARMON CO.: drifting sand along Buck Creek, 4 mi. W and 6.5 mi. S of Hollis, Waterfall 8340 (OKL, OKLA, TEX). WOODS CO.: dunes along Cimarron River near Waynoka, Goodman 4942 (OKL), Goodman and Waterfall 4520 (GH, OKL, TEX), Hansen (US), Rice (OKL), Water- fall 8169 (OKL, OKLA, PUL, SMU, TEX), 10372 (OKLA, SMU), 12317 (GH, TEX). TEXAS. ANDREWS CO.: 17 mi. SE of Andrews, McVaugh 10767 (MO, US). BAYLOR CO.: island in Brazos River near Seymour, Reverchon 876 (F, GH, MO). CHILDRESS CO.: dunes along Red River 9 mi. N of Chil- dress, Gould and Thomas 7726 (SMU). COCHRAN CO.: 2 mi. W of Bled- soe, Cory 16524 (GH). CRANE CO.: 13 mi. N of Imperial, Warnock 15505 (TEX). EL PASO CO.: El Paso to Monument 53, Wagner 994 (US); dunes E of El Paso, Hershey (SMU); deep sand c. 20 mi. E of El Paso, Hinckley 4795 (US); 15 mi. E of El Paso, Warnock 10901 (SMU); Hueco Mts., 17 mi. E of El Paso, Waterfall 3899 (GH, MO). HARDEMAN CO.: 4.8 mi. N of Romero, York and Rodgers 309 (SMU, TEX). HOCKLEY co.: sand-dunes north of Anton [possibly in LAMB co.], Reed 3446 (US). HUTCHINSON CO.: Canadian River bottoms, N side of Borger, Shinners 8091 (SMU); Bugbee Creek, dunes in floodplain, 9 mi. E of Stinnett, Thornton 52-435 (TEX). LOVING CO.: between Mentone and Wink, Warnock 10723 (PUL, SMU). OLDHAM CO.: 13 Aug. 1891, Carle- ton 415 (US). WARD CO.: dunes 3-5 mi. E of Monahans, Muller 8515 (SMU), Miller and Miller 1308 (PUL), Rowell 60-064 (PUL, SMU), War- nock 7877 (PUL, SMU, TEX), Webster 4615 (F, PUL, SMU). WHEELER CO.: S side of N fork of Red River, 3.5 mi. N of Shamrock, Cory 50257 (GH, SMU, US, UT). WILBARGER CO.: dunes S of Red River, Round Timbers Ranch, Tharp and Miller 51-156 (TEX). WINKLER CO.: dunes c. 9 mi. E of Kermit, Correll 15183 (us); dunes 6-11 mi. N and E of Kermit, Lewis and Rowell 8234, Miller and Miller 1322 (PUL), Rowell 8263, 60-047 (PUL, TTC), 60-074 (PUL, SMU, TTC). WICHITA CO.: Red River above Burkburnett, Tharp 606 (TEX). 1963] Genus Reverchonia — Webster and Miller 203 NEW MEXICO. cHAVES co.: Arroyo Ranch, near Roswell, Griffiths 5694 (MO, US); shinneries E of Roswell, Goodding 6541 (ARIZ); sandy soil near Acme (c. 25 mi. NE of Roswell), Williams 9588 (UNM). DONA ANA CO.: Jornada Range Reserve, Hurtt 49 (US) : between Strauss and Anapra, Stearns 396 (US). OTERO CO.: dunes S of Alamagordo, Hershey 3653 (UNM); 18 mi. S of Alamagordo, Johnston 2727 (SMU). QUAY co.: sandy roadside, 4.8 mi. W of Glenrio, Shinners 21077 (SMU). SOCORRO CO.: 7 mi. W of the atom bomb crater, Dunn 4851 (UNM) ; 12.5 mi. S of junction W of Carthage, red sand dune area, Dunn and Lint 5011 (UNM); north of Lava, Wooton (TEX, us); W of Bingham, Shin- ners 9589 (UNM). ARIZONA. COCONINO CO.: Leupp Indian Reservation, with Hilaria and Sacaton, Casteter (UNM), Oakley 373 (ARIZ). NAVAJO CO.: Moki (= Moenkopi?) Reservation, and Little Colorado River, Hough 39 (us) ; Second Mesa, Hopi Reservation, Whiting 756 (ARIZ). UTAH. KANE co.: dunes WSW of Kanab, Harrison 11080 (US) ; 10 mi. N of Kanab, Hinckley (ARIZ) ; dunes 6 mi. N of Kanab, Hitch- cock, Rethke, and van Raadshooven 4536 (GH, UTC) ; dunes N of Kanab, Milner 8949 (UT). MEXICO. CHIHUAHUA: dunes, LeSueur Mex-287 (F, GH, SMU, TEX), 765 (F, TEX); 38 mi. S of Juarez, sandhills in mesquite desert, Gentry 8207 (GH, US); 40 mi. S of Juarez, dunes, Gentry 17900 (US); sand hills near Samalayuca, Pringle 3044 (F); dunes 6 mi. S of Samalayuca, Waterfall 12475 (US) ; sandhills near Paso del Norte, Pringle 792 (F, GH, US). A very dubious collection — Texas, Tarrant Co., Fort Worth, Mar. 1890, Bodin (US) — has not been mapped, as the plant has not been recollected within 100 miles of Ft. Worth, and it never flowers as early as indicated on Bodin's label. Shinners (1952) has noted that in Texas and Oklahoma the distribution of Reverchonia is remarkably parallel to that of Euphorbia carunculata, which was originally de- scribed by Waterfall from the Waynoka sand dunes in north- ern Oklahoma. Shinners also records Reverchonia from the state of Durango, Mexico, but we have not been able to con- firm this and suspect that the mention of Durango was a slip for Chihuahua. All of the Mexican records of Rever- chonia seem to come from the same area of dunes south of Samalayuca; and judging from the map of Chihuahuan veg- etation presented by LeSueur (1945), dune habitats suita- ble for Reverchonia occur only in the northeastern corner 204 Rhodora [Vol. 65 of the state, adjacent to El Paso and Hudspeth counties in Texas. It seems unlikely, therefore, that Reverchonia will be found much further south in Mexico. The spotty records Fig. 6. Distribution map of Reverchonia. Large dots, exact locali- ties; small dots, county records or inexact localities. from Arizona and western New Mexico are more difficult to interpret; it is possible that the apparent rarity of the plant there is an artifact of the inadequate collection records. The area occupied by Reverchonia is fundamentally dis- continuous due to its very strong preference for (or restric- tion to) dune habitats. It covers a considerable spread of altitudes, from around 1000 ft. along the Red River in Texas- Oklahoma to between 5500 and 6000 ft. in northern Arizona and Utah. According to Dr. Robert Vickery (in litt.) the plant in Utah grows on brilliant red dunes surrounded by pine forest (north of Kanab) ; this suggests very different conditions from the dunes of pale sand with shinnery oak (Quercus havardii) and Prosopis where Reverchonia occurs in west Texas (in Crane and Ward counties). The differ- 1963] Genus Reverchonia — Webster and Miller 205 ences in altitude and precipitation (varying from 25 to less than 10 inches per year) which occur within the range sug- gest that Reverchonia possesses some degree of adaptability. In its flowering pattern, Reverchonia behaves as a long- day plant. The earliest flowering specimen seen was col- lected on May 28 (McVaugh 10707) and the latest on Sep- tember 25 (Gould and Childress 7726); fruiting begins by mid-July (at least in Texas and Oklahoma) and continues into October. Germination of the large seeds is rapid and the first internodes elongate greatly; the conspicuous nar- row cotyledons may persist on certain plants until they be- gin to flower. It seems possible that the failure of Rever- chonia to extend westward into the Californian and Sonoran deserts might be correlated with the different seasonal dis- tribution of precipitation there (i.e., very few summer rains). Along the other boundaries of the species range it is impossible to suggest correlations with any one climatic Table 1. MORPHOLOGICAL VARIATION IN REVERCHONIA' Character N Range (mm.) K s(mm.) C. V. seed length 60 4.4-6.6 5.34 0.48 8.98 8 4.8-5.9 5.43 0.38 6.94 length fruiting pedicel Gi 2.5-8.7 4.49 1.08 23.99 21 2.5-5.8 4.11 0.65 15.82 capsule diameter 16 7.0-9.8 8.43 0.91 10.81 8 7.8-9.3 8.8 0.52 5.95 stipule length 107 0.7-2.3 1-27 0.27 21.5 22 1.2-2.1 1.48 0.35 24.47 leaf length 97 16-44 27.4 6.1 22.25 21 22-42 29.05 5.2 17.9 leaf width 96 1.7-9.1 4.78 1.58 33.14 21 2.5-8.8 5.12 1.45 28.28 'Parameters are based on one measurement of each character per specimen. The upper row for each character gives data based on meas- urements of specimens throughout the range of the species; the lower row is based on a single population sample from Winkler Co., Texas (Miller and Miller 1322), except that for seed and pedicel length and fruit diameter specimens were added from a nearby collection (Rowell 60-074). 206 Rhodora [Vol. 65 factor, which indicates that a complex of interrelationships is probably involved. Comparison of specimens from all portions of the range indicates that there is relatively little geographic variation. A population sample from Winkler Co., Texas, shows a range of variation quite close to that of the species as a whole (Table 1). The only character with some suggestion of geographical differentiation is the length of fruiting pedi- cel, which tends to be somewhat longer in some of the Chi- huahuan specimens than from other localities. However, although the pedicels are over 8 mm. long in Gentry's col- lection, they are within the usual range of variation in plants collected by Pringle and Waterfall in the same general area. The seeds of the Chihuahuan specimens may average some- what longer than those of most populations, but the avail- able samples are not large enough to be decisive. In any event, it is fair to say that on the whole Reverchonia aren- aria is a rather homogeneous species, even though there is considerable random — i.e. non-geographic — variability (as indicated by the high coefficients of variability for the characters in Table 1). We suspect that this lack of geo- graphic differentiation may be related to the fact that the plant is sufficiently well adapted for cross-country dispersal (perhaps by travelling along sandy stream-beds) so that the populations do not remain isolated. DEPARTMENT OF BIOLOGICAL SCIENCES, PURDUE UNIVERSITY, LAFAYETTE, INDIANA LITERATURE CITED BAILLON, H. E. 1858. Etude générale du groupe des Euphorbiacées. 684 pp. Victor Masson, Paris. CROIZAT, LEON. 1942. New and critical Euphorbiaceae chiefly from the southeastern United States. Bull. Tcrrey Club 69: 445-460. DRESSLER, R. L. 1954. The genus Tetrecoecus (Euphorbiaceae). Rhodora 56: 45-61. ERDTMAN, GUNNAR. 1952. Pollen Morphology and Plant Taxonomy. 539 pp. Almqvist & Wiksell, Stockholm. GOEBEL, K. 1933. Organographie der Pflanzen. 3 Aufl. 3 Teil. Samen- pflanzen. Fischer, Jena. 1963] Genus Reverchonia — Webster and Miller 207 GORMAN, GRANT. 1941. A Pathfinder in the Southwest. 298 pp. Uni- versity of Oklahoma Press, Norman. Gray, ASA. 1880. Contributions to North American botany. Proc. Amer. Acad. 16: 78-108. GRUENING, G. 1913. Euphorbiaceae — Porantheroideae et Ricinocar- poideae. Pflanzenr. IV. 147 (Heft 58) : 1-97. JOHNSTON, M. C., AND B. H. WARNOCK. 1963. Phyllanthus and Rever- chonia (Euphorbiaceae) in Far Western Texas. Southw. Nat. 8:15- 22. 1963. LESUEUR, HARDE. 1945. The ecology of the vegetation of Chihuahua, Mexico, north of Parallel 28. Univ. Tex. Publ. 4521: 1-92. MUELLER, JEAN. 1866. Euphorbiaceae, in DC. Prodr. 15(2): 189- 1286. Pax, FERDINAND. 1890. Euphorbiaceae, in Engler and Prantl, Natürl. Pflanzenfam. 3(5) : 1-119. , AND K, HoFFMANN. 1931. Euphorbiaceae, in Eng- ler and Prantl, Natiirl. Pflanzenfam. 2 Aufl. 19c: 11-233. PERRY, B. A. 1943. Chromosome number and phylogenetic relation- ships in the Euphorbiaceae. Amer. Jour. Bot. 30: 527-543. PuwT, WiLLEM. 1962. Pollen morphology of the Euphorbiaceae with special reference to taxonomy. 116 pp. North-Holland, Amster- dam. SCHWEIGER, JOSEPH. 1905. Beitrüge zur Kenntnis der Samenentwick- lung der Euphorbiaceen. Flora 94: 338-379. SuiNNERS, L. H. 1952. Addenda on Texas Chamaesyce (Euphorbi- aceae). Field & Lab. 20: 24-26. VINDT, JACQUES. 1953. Monographie des Euphorbiacées du Maroc. Premiére partie. 220 pp. Trav. Inst. Sci. Cherifien 6. WEBSTER, G. L. 1956. A monographic study of the West Indian species of Phyllanthus. Jour. Arnold Arb. 37: 91-122. , AND J. R. ELLIS. 1962. Cytotaxonomic studies in the Euphorbiaceae, subtribe Phyllanthinae. Amer. Jour. Bot. 49: 14-18. AN ANNOTATED LIST OF VASCULAR PLANTS FROM CAPE SABINE, ALASKA STANWYN G. SHETLER INTRODUCTION While making field studies of Campanula (Campanu- laceae) in western North America during the summer of 1959, Mr. Karl Stone and I had an opportunity to spend several days (July 11-14) in the vicinity of Cape Sabine, a small cape on the Arctic Coast of northwest Alaska (ca. 68° 55' N. lat., 164° 30’ W. long.). At this cape, located about 65 km. east of Cape Lisburne, the westernmost land-point of the Arctic Slope, the Pitmegea River enters the Arctic Ocean. This small river flows off the North Slope of the Brooks Range, and its drainage basin lies, physiographically, very near the western limits of the Arctic Slope as defined by Spetzman (Fig. 4). The Arctic Slope of Alaska was rather poorly collected prior to 1941 when Hultén’s monumental Flora of Alaska and Yukon began appearing, and the many large gaps on his range maps have undoubtedly been responsible in part for stimulating the intensive floristic research conducted in Arctic Alaska in recent years. Many new areas have been visited and collected with the result that a reasonably accur- ate picture of distribution patterns on the Arctic Slope is emerging. No small part of the stimulus for this research has come from the Arctic Research Laboratory at Point Bar- row. Through its facilities field work that hitherto was impossible has become routine. Now, virtually no spct on the Arctic Slope is inaccessible. The floristic research of the last two decades was summarized in 1962 in the much-needed volume by Wiggins and Thomas, A Flora of the Alaskan Arctic Slope. What is at once evident from this work is that despite all the collecting on the Arctic Slope in recent years many areas still remain relatively unknown. Collections from certain areas, such as Cape Lisburne and Point Barrow, are very extensive and must be virtually complete, whereas many other areas have been little collected. Continuing pub- 208 1963] Plants, Cape Sabine — Shetler 209 lication of floristic contributions from the Arctic Slope of Alaska seems, therefore, amply justified. Cape Sabine is not far from Cape Lisburne, but it has been passed over by many collecting parties for the more alluring Cape Lisburne. From a phytogeographic standpoint, how- ever, Cape Sabine should be interesting, because this general area combines physiographically features of both the coastal plain and the foothills of the Brooks Range (cf. Spetzman). Although several previous parties have made recent collec- tions in this area, their results have not been made known and their records have not, regrettably, been reported by Wiggins and Thomas. I think it worthwhile, therefore, to bring our collection to the attention of other arctic botanists, despite its incompleteness. In the very short time available to us, the equivalent perhaps of two full days of collecting, we were able to collect 143 species. This number surely must represent at least half and possibly as much as two-thirds of the local vascular flora (cf. Spetzman, p. 52). Even so, one new record for North America and several new ones for the Arctic Slope were established (cf. later discussion). Pub- lication of this list was first suggested by Dr. Henry Childs, my camping partner, at that time working under the direc- tion of Dr. Frank Pitelka, University of California, studying mammal populations in the Pitmegea valley as a part of a longterm ecological study supported by the Arctic Institute of North America. Dr. Childs felt that they very much need- ed specific information on the vegetation of the Pitmegea valley for their study. EXPLANATORY NOTES We collected at two locations in the Pitmegea valley. Mr. Karl Stone, with a camping partner, collected in the immedi- ate vicinity of the Cape, while I collected inland about 11 km. (7 mi). Henceforth, the two sites will be called the “Coastal” and 'Sevenmile" sites, respectively. Few if any vascular plant collections have been made previously at Sevenmile. All numbers given in the catalog, following the Latin names and synonyms, are joint Shetler-Stone collection numbers. Those in italics represent Coastal collections, while 210 Rhodora [Vol. 65 all others are Sevenmile collections. An asterisk (*) follow- ing a number indicates a unicate collection. The first and only complete set of specimens is deposited at the University of Michigan Herbarium. Partial sets can be found at the following institutions: Arctic Research Laboratory (Point Barrow), Komarov Botanical Institute (Leningrad), Nation- al Museum of Canada (Ottawa), Riksmuseum (Stockholm), U. S. National Museum, and the University of Alaska. The genera are arranged as in Hultén's Flora of Alaska and Yukon, and the species are arranged alphabetically. Certain critical groups were sent to A. E. Porsild (Nat. Mus. Canada), who kindly determined or verified their identities for me. He is credited accordingly in the catalog. All deter- minations not otherwise credited were made by me. In June 1962, after the collections had all been determined, E. Hultén kindly looked at our specimens while examining all arctic collections at the University of Michigan, and his opinions have been duly incorporated and credited in the catalog. Un- less otherwise noted, he concurred with my determinations. Hultén's Flora of Alaska and Yukon has served as the point of reference throughout, and all unmodified page cita- tions to Hultén are to this work. Where I have used a name other than that used by him, I have given his name in paren- thesis. Few other synonyms are given. A wide range of publications was consulted, including the Flora of the USSR, and where for compelling reasons I have departed from Hul- tén's interpretations the appropriate citations and explana- tory notes are given. Insofar as our material permitted, I have attempted to indicate, by Latin epithets or critical com- ment, the nature of the infraspecific variation exhibited. I regret that Wiggins' and Thomas' flora appeared too late for me to consider it in the original preparation of this paper. In a cursory way, however, I have checked my manuscript against their nomenclature, descriptions, and records, and wherever possible I have taken their point of view into con- sideration. PHYTOGEOGRAPHICAL COMMENTS The Cape Sabine region is really in the foothills of the Brooks Range (cf. Spetzman, fig. 4), and, as already noted, 1963] Plants, Cape Sabine — Shetler 211 physiographically this area has certain features of both the foothills and the coastal plain. There are many low ridges and domes which give the terrain the gently rolling aspect of the lower foothills, but the extensive flat meadows, some very poorly drained, and numerous small lakes in the Seven- mile area suggest typical coastal plain. Certainly the climate is more coastal than montane, because of the proximity of the sea, although the amount of relief is atypical for coastal plain. Between the Cape and Sevenmile the land rises from sea level to nearly 350 m. This is a rather sharp rise for a distance of only about 11 km. Thus, more habitats than are typical for coastal plain are afforded in this area, and the flora might be expected to be more diverse than at other coastal stations eastward. The following list of 16 species collected by us which are rare or unknown on the Arctic Coast of Alaska suggests a strong montane element in the Cape Sabine flora: Bromus inermis pumpellianus, Carex atrofusca, C. lugens, Juncus albescens, Luzula wahlenbergti, Anemone narcissiflora, Smelowskia calycina, Chrysosplenium wrightii, Saxifraga davuriea, S. eschscholtzii, Potentilla biflora, P. ledebouriana, Primula tschuktschorum, Pedicularis oederi, Senecio hyper- borealis, Taraxacum phymatocarpum. Comparisons with the published record (cf. Hultén, F!. AI. & Yuk; Spetzman; Wiggins and Thomas) suggest that at least 19 of our collections are significant contributions to present knowledge of plant distribution on the Arctic Slope of Alaska. These 19 significant records have been indicated in the catalog by prefixing the name of the species with an asterisk. Further comment and explanation can be found under each. It should be noted here, however, that of these 19 species Braya siliquosa apparently is a new record for North America, and the following four seem to be new to the Arctic Slope, although in several cases taxonomic opinion is involved, and the records prove less significant if a differ- ent taxonomic interpretation is espoused: Draba bellii, Erigeron compositus, Primula sibirica, Salix stolonifera. 212 Rhodora [Vol. 65 ANNOTATED LIST Equisetum variegatum Schleich. in Weber et Mohr ssp. variegatum. 3230*. Alopecurus alpinus Sm, var. alpinus. 3156. Arctagrostis latifolia (R. Br.) Griseb. s.l. 3237-A*, 3237-B, 3237-C*. Tussocks of this grass were very abundant in a niggerhead meadow about 1 km. west of Sevenmile camp. Nine pieces from several tus- socks were collected, and examination of these pieces revealed two rather distinct types and one more or less intermediate type. The first, 3237-A, represented by two pieces, has a relatively short culm, quite purple spikelets that are 4-6 mm long, and purple anthers averaging 2.5 mm long. These pieces would seem to match Hultén's typical A. latifolia (p. 144). The second, 3237-B, represented by six pieces, has a taller culm than A, spikelets that are less purple and 3.5-4.5 mm long, and yellow anthers mostly 1.5-2.0 mm long. Some of these anthers have purple tips, and on one piece they are distinctly longer than 2.0 mm. Perhaps these specimens belong to var. arundinacea (Trin.) Griseb., but the yellow anthers apparently are not typical for this variety. Porsild (1951, p. 82, under A. arundinacea) commented, “The anthers that in early anthesis are purple become yellow in age. . . ." Our specimens vary from preanthesis to immature fruit, but I find no evidence that anther color varies with age. Rather, the color difference seems constant with age, suggesting that it is genetic. The third, 3237- C, represented by one piece, is quite similar to A, but differs by having some yellow suffused in the purple anthers and by having less purple spikelets that are about 4 mm. long. In these latter characters it is more similar to B. All three types have rather congested and strict panicles and quite variable scabrosity on the pedicels. In general, the pedicels are sparsely if at all scabrous. As Hultén commented (p. 147), this seems to be an extremely vari- able species in Alaska. Careful field study is needed to determine the validity of segregating several varieties within it. At Sevenmile, two rather distinct types are growing side-by-side, along with some appar- ent intermediates, suggesting hybridization. Peschampsia caespitosa (L.) Beauv. ssp. caespitosa var. glauca (Hartm.) Sam. 3252. Trisetum spicatum (L.) Richt. 3285. Poa arctica R. Br. 3234*, 3269; verified by Porsild. No. 3234 comprises one immature specimen, and the racial relation- ship could not be determined. No. 3269 represents ssp. arctica. P. glauca Vahl. 2778; verified by Porsild. *P. komarovii Roshew. 3325. Our specimens are not altogether typical of this species and suggest P. lanata Scribn. et Merr. somewhat. Both Porsild and Hultén have seen these specimens. Porsild expressed some doubt about my deter- mination but without suggesting any other identity, while Hultén 1963] Plants, Cape Sabine — Shetler 213 concluded that my determination is probably correct. In favor of P. komarovii are the low habit (cf. Anderson, p. 77), general resemblance to P. alpina L., presence of many dried basal sheaths, and the short (less than 2 mm.) anthers. In favor of P. lanata are the very greenish- purple spikelets, relatively narrow leaves, and prominent lanatum on the lower part of the lemmas and on the keel and lateral nerves to above the middle. Possibly our specimens belong to some other Asiatic species, but for the present it seems best to refer them here. Poa komarovii is an Amphi-beringian species, which in Alaska is distributed primarily in the Aleutians and on the islands in the Bering Sea. It is, in Hultén's opinion (1937, p. 39), a radiant from so-called "Southern Beringia." Previous records from the Arctic Slope are all from the general area of Point Barrow (Hultén, p. 212; Wiggins and Thomas, p. 364). Our specimens were collected very near the shore. It might be added here that more typical specimens of this species were collected by us also at Atkasuk along the Meade River southwest of Point Barrow. Porsild concurred fully with my determination of those Specimens. Apparently, this typically more southern species has invaded the Arctic Coastal Plain and migrated at least as far east as Point Bar- row. Arctophila fulva (Trin.) Anders. 3260. Dupontia fischeri R. Br. ssp. psilosantha (Rupr.) Hult. 3290. Festuca rubra L, s.l. 3214. Bromus inermis Leyss. ssp. pumpellianus (Scribn.) Wagnon var. arcti- cus (Shear) Wagnon. (B. pumpellianus Scribn., incl. vars. arcticus (Shear) Porsild and villosissimus Hult.) 3215. Our specimens, all collected on the same gravel bar, are extremely variable with respect to pubescence and seem to fit best the views of Wagnon (1950), regarding taxonomic treatment of plants of this affinity. Eriophorum angustifolium Honck. ssp. angustifolium. 3265. *E. brachyantherum Trautv. 3199; det. by Hultén. Apparently this is the second record for the Arctic Slope. Hultén (p. 279) did not report it, and Wiggins and Thomas (p. 91) assign only one collection to this species and “with some reluctance.” They report a Spetzman collection from Lake Noluck, in the foothills of the Brooks Range about 200 km east of Cape Sabine, as this species. E. scheuchzeri Hoppe. 3300. E. vaginatum L. ssp. vaginatum. 3301. The anthers of our specimens are 2.0-2.5 mm. long, most of the scales are not conspicuously darker in the center, and the spikes are more or less oblong (cf. Hultén's key, p. 275). This subspecies is unreported from the Arctic Slope, but is not unexpected in western Alaska. Sub- species spissum is the common North American race, whereas ssp. vaginatum is Eurasian. 214 Rhodora [Vol. 65 Carex aquatilis Wahlenb. s.l. 3289. Our specimens are intermediate between C. aquatilis and C. stans Drej., but approach C. aquatilis more closely (cf. key of Krechetovich, Fl. SSSR 3: 201-202, 1935). Hultén (pp. 339-340) included C. stans in C. aquatilis, but generalized that all specimens of the Arctic Coast belong to the C. stans type. The records of Wiggins and Thomas (p. 104) are in agreement with this generalization, but our collection would seem to be an exception. C. atrofusca Schk. 3341. *C. krausei Boeckeler ssp. krausei. 3250. The strictly gynoecandrous condition of the terminal spikes immedi- ately places this collection in C. krausei, according to Hultén (p. 348). But, according to the recent cytotaxonomic revision of Carex Section Capillares by Lóve, Lóve, and Raymond (pp. 744-745), the identity of our specimens is not so clearcut. Interesting as their study is from a biosystematic point of view, it also indicates the difficulties of achiev- ing a practical taxonomic treatment that faithfully reflects underlying cytological differences. If one accepts the taxa that they have recog- n'zed in this group on cytological grounds, one is hard-pressed to dis- tinguish them morphologically with their key. All characters used are highly variable, and the key is replete with numerous overlaps and relative distinctions. Despite what I would call “clavate-linear” termi- nal spikes that do overtop the lower ones and despite culms up to 20 cm. high, I think our specimens can only belong to C. krausei ssp. krausei as delimited by them. On the whole our plants are rather low — mostly 15 em. or less. Our record appears to extend the known range of this species con- siderably westward on the Arctic Slope (cf. Wiggins and Thomas, p. 112). *C. lugens Holm. 3238-A, 3238-B, 3238-C*, 3238-D*. This species is unreported from the western sector of the Slope (cf. Wiggins and Thomas, p. 106). C. membranacea Hook. 3155, 3287. C. physocarpa Presl. 3288. Our specimens are not entirely typical. They have predominantly just one staminate spike, the perigynia are almost black, only some of the scales have hyaline tips, and the leaves are narrow, mostly 3 mm. wide or less. C. rariflora (Wahlenb.) Smith. 3268. C. scirpoidea Michx. 3249. Juncus albescens (Lge.) Fern. (J. triglumis sensu Hultén, in part). 3204*. J. balticus Willd. var. alaskanus (Hult.) Porsild. (J. arcticus Willd. ssp. alaskanus Hult.) 3209. J. castaneus Smith s.l. 2757, 3208. Luzula confusa Lindb. 3224-A*. 1963] Plants, Cape Sabine — Shetler 215 L. tundricola Gorodk. (L. nivalis var. latifolia sensu Hult. Fl. Al. & Yuk., Wiggins and Thomas; cf. Hult., 1962, p. 10). 3179, 3224-B; det. by Hultén. *L. wahlenbergii Rupr. 3267. This essentially circumpolar species has been reported previously for the Arctic Slope only from Umiat (Spetzman, p. 43) and Chandler Lake (Wiggins and Thomas, p. 128). Lloydia serotina (L.) Rchb. 3257. Salix alaxensis (Anders.) Cov. var. alaxensis. 2340; det. by Argus and Raup. S. arctica Pall. 3282, 3330; det. by Argus and Raup. S. brachycarpa Nutt. ssp. niphoclada (Rydb.) Argus. (S. niphoclada Rydb.) 3302*; det. by Argus. S. farrae Ball ssp. walpolei (Cov. et Ball) Hult. 3296; det. by Argus and Raup. S. glauca L. s.l. 3256*, 32778, 3298; det. by Argus. I have deferred here entirely to Dr. George Argus, who has just com- pleted a revision of the S. glauca complex, and have made no attempt to evaluate our specimens according to Hultén's treatment. Argus be- lieves that infraspecific taxa cannot be maintained meaningfully in this extremely variable species. He recognizes instead several “phases.” He determined our collections as his so-called *Beringian Phase." (Personal conversation, Dec. 1960). S. glauca X S. brachycarpa ssp. niphoclada. 3256-A*, det. by Argus. S. phlebophylla Anders. 3223. *S. polaris Wahl. ssp. pseudopolaris (Flod.) Hult. 3222, 3276*, 3328; det, by Hultén. In my opinion these plants approach ssp. polaris more closely than ssp. pseudopolaris. Our collections extend the known range of S. polaris considerably westward on the Slope (cf. Wiggins and Thomas, p. 149). S. reticulata L. 3275, 3329. S. richardsonii Hook. 3299*; det. by Raup. *S. stolonifera Cov. 3332; det. by Argus and Raup. This would appear to be a most significant record, since S. stoloni- fera is a species of the Pacific Coast of Alaska (Hultén, p. 523). The one previous record for the Arctic Slope (Raup, 1959, p. 52) is also from Cape Sabine, based on a specimen collected by Cantlon and Gillis and determined by Raup. Several other species of the Ovalifolia group are known from the Slope, however, and, since this group is much in need of study, I am inclined to doubt the significance of our record. I suspect that what Raup has here designated S. stolonifera may prove to be only a variant of some other species of the group common to the Arctic Slope. Betula glandulosa Michx. var. sibirica (Ledeb.) Blake. (B. nana L. ssp. exilis (Sukatch.) Hult.) 3281. 216 Rhodora [Vol. 65 American opinion (Porsild, 1951, p. 152; Wiggins and Thomas, p. 152) seems to favor regarding this taxon as a variety of B. glandu- losa, while Hultén is convinced that it represents a race of the Eurasi- an B. nana. This question clearly needs careful study. Our specimens fit B. exilis quite well (cf. Kuzeneva, Fl. SSSR. 5: 271, 1986), and I find Hultén's view not without merit. Rumex arcticus Trautv. 3196. Polygonum bistorta L. ssp. plumosum (Small) Hult. 3244. P. viviparum L. 3277*. Stellaria longipes Goldie. 3236; det. by Hultén. Cerastium beeringianum Cham. et Schlecht. 3286. Arenaria arctica Stev. (Minuartia arctica (Stev.) Aschers. et Graebn.) 3175, 3251, 3338. No. 3175 and 3338 approach A. obtusiloba (Rydb.) Fern. in habit and in length and ciliation of leaves. Unfortunately, none of the collec- tions has seeds present, and this rather important character could not be checked. A. rossii R. Br. (includ. Minuartia elegans sensu Hultén) ; 3180*, 3225, 3337. No. 3180 has the closely tufted habit and short pedicels of typical A. rossii, while nos. 3225 and 3337 have a much more diffuse habit and longer pedicels, resembling M. elegans, sensu Hultén (p. 681). Silene acaulis L. var. exscapa ( All.) DC. 3183, 3233. Melandrium apetalum (L.) Fenzl ssp. arcticum (Fr.) Hult. 3203, 3321. One specimen of 3203 has corolla limbs considerably exceeding the calyx and less emarginate than on the specimens with short limbs, but otherwise this specimen is similar to the others. Caltha palustris L. var. arctica (R. Br.) Hutch. 3284. Tolmatchev (1955, p. 149) regards this race as a good species, but the differences separating it from C. palustris are quantitative and tenuous. It is not clear to me why both Hultén (pp. 712-713) and Porsild (cf. 1955, pp. 113-114, and 1957, p. 180, map 157) treat this taxon as a variety nomenclaturally when they seem to be of the firm opinion that it is a geographic race. Delphinium brachycentrum Ledeb. 3187*. Aconitum delphinifolium DC. ssp. paradoxum (Rchb.) Hult. 3169. Anemone narcissiflora L. s.l. 3226. Hultén (pp. 732-736) and Wiggins and Thomas (pp. 186-187) dis- tinguish subspecies in this species, but I incline to the position of Porsild (1951, pp. 177-178) and Raup (1947, p. 178) that this variable species is best regarded in the broad sense. Our plants fit ssp. sibirica (L.) Hult., as would be expected on geographic grounds, but I find, as Porsild does, that the distinction between this subspecies and ssp. interior Hult. is indeed tenuous. A. parviflora Michx. 3229*, 3242. Ranunculus nivalis L. 3228. 1963] Plants, Cape Sabine — Shetler 217 R. pallasii Schlecht. 3259. R. sulphureus Sol. 3313*. Papaver macounii Greene. 3291, 3316-B*. P. radicatum Rottb. s.l. (includ. P. alaskanum Hult.) 3316-A. These specimens seem to fit Hultén's segregate species P. alaskanum. Corydalis pauciflora (Steph.) Pers. 2168. Cochlearia officinalis L. ssp. arctica (Schlecht.) Hult. 3173. Cardamine digitata Richards. (C. richardson?i Hult.) 3254. * Contrary to Hultén (p. 838), there is no earlier homonym of this binomial (cf. Shetler, 1961). *Draba bellii Holm. (possibly D. macrocarpa sensu Hultén — cf. his remarks regarding D. bellii, p. 868) ; 3202; det. by Porsild. Neither Spetzman nor Wiggins and Thomas reports this species from the Arctic Slope, but the systematics of arctic Draba are so complex that one hesitates to attribute any great phytogeographical significance to this segregate species of the D. alpina L. complex. I have deferred here to Porsild, however, whose wide experience with this genus in the Arctic convinces him that plants exhibiting the characteristics of our specimens should be distinguished as the separate species D. bellii. If our plants are placed in D. alpina s.l., then our record loses significance. D. lactea Adams. 2172-A ; verif. by Porsild. D. longipes Raup. 3170, 3172-B* ; verif. by Porsild. D. nivalis Liljebl. var. nivalis. 3327*; verif. by Porsild. D. pilosa Adams ex DC. 3191*; det. by Porsild. Smelowskia calycina (Steph.) C. A. Mev. ssp. integrifolia (Seemann) Hult. 3150, 3326. Erysimum pallasii (Pursh) Fern. 3148. *Braya siliquosa Bunge. 3201; det. by Porsild. Although the taxonomic situation in this genus is very complex, Porsild believes (letter, Dec. 1960) that our plants can only belong to B. siliquosa, a small Asiatic species which apparently is unknown to North America. This species is distinguished from other Eurasian species by its narrow linear siliques, 10-15 mm. long by about 1 mm. wide (cf. Vassilezenko, Fl. SSSR 8: 70-71, 1939). The siliques of our specimens have a maximum length of 10 mm. and a width of about 1 mm. These dimensions fit B. siliquosa, although the siliques are on the short side. The Cape Sabine record represents a most remarkable range extension. According to Vassilezenko, this species is widespread in Asia from western Siberia (Altai) to eastern Siberia and Okhotsk in the Far East, and it inhabits alpine meadows and slopes. He also reported that it is found in North America from 52°-57° N. lat., but I have found no American confirmation of this. Popov (1: 545-546) called it an alpine zone species of certain parts of the Sayan Mountain forest region and considered that it radiated westward from alpine areas of the Okhotsk region. Hultén (1937, pp. 16-17, pl. 3) regarded it as a boreal Eurasiatic species that has radiated from the Amur 218 Rhodora [Vol. 65 region and is widely distributed in central and eastern Asia, but has not reached either Europe or extreme northeastern Siberia. That B. siliquosa is a boreal, not arctic, species in Asia makes the Cape Sabine record all the more significant. Our plants were collected on a sandflat along the Pitmegea River at Sevenmile, the site more definitely in foothills topography, and quite possibly this species has only recently “washed down” from alpine areas in the Brooks Range. It should also be noted, however, that Cape Sabine lies in a general area that has not been glaciated and presumably, therefore, has served as an impor- tant refugium for plants during the Pleistocene. Perhaps prior to Pleistocene glaciation, B. siliquosa was more widespread in Alaska. Parrya nudicaulis (L.) Regel s.l. 3255, 2320. No. 3320 fits Hultén's ssp. interior, but 3255 is so variable in the width and toothing of the leaves as to defy racial identity. My own limited experience with this variable species leads me to question Hultén's races. Saxifraga bronchialis L. ssp. funstoni (Small) Hult. 3160. S. cernua L. 3195*, 3263. S. davurica Willd. ssp. grandipetala (Engler et Irmscher) Hult. 3280, 3310. S. eschscholtzii Sternb. 3182. S. flagellaris Willd. ex Sternb. ssp. flagellaris. 3219, 3323. Phytogeographically, our plants should belong to this subspecies, although morphologically they tend toward ssp. platysepala (Trautv.) Pors. (cf. Porsild, 1955, pp. 136-138). S. foliolosa R. Br. 3264. S. hieracifolia Waldst. et Kit. var. rufopilosa Hult. 3270. S. hirculus L. 3189. S. oppositifolia L. 3198, 3307*, 3336*. S. punctata L. ssp. nelsoniana (D. Don) Hult. 3193, 3311. Chrysosplenium tetrandrum (Lund) Th. Fries. 3194*. C. wrightii Franch. et Sav. 3342. Parnassia kotzebuei Cham. et Schlecht. 3211. Rubus chamaemorus L. 3190. Potentilla biflora Willd. ex Schlecht. 3186, 3232. P. hyparctica Malte. (P. emarginata Pursh s.l.) ; 3231*; verif. by Por- sild. P. ledebouriana Porsild. (P. uniflora Ledeb.) ; 3319; det. by Porsild. P. palustris (L.) Scop. 3266. Dryas integrifolia M. Vahl. 3177*, 3253, 3331*; det. by Porsild. D. octopetala L. 2776, 3221; det. by Porsild. No. 3176 is a very silvery canescent form designated by Porsild as "f. canescens-argentea.” D. integrifolia X octopetala. 3283-A*; det. by Porsild. Lupinus arcticus S. Wats. 3218, 3317*. Astragalus alpinus L. s.l. 3279, 3335. 1963] Plants, Cape Sabine — Shetler 219 Our plants seem to fit what Hultén (pp. 1081-1085) called the “main type," but the question of races in this species needs further study. * A. polaris Seemann in Benth. 3274, 3339; det. by Porsild. This attractive little Astragalus species is an Alaskan endemic of gravel bars, known only from a few widely scattered localities. Wig- gins and Thomas cite only one specimen from the Arctic Slope (Utukok R., cf. W. and T., p. 263). A. umbellatus Bunge. 3206, 3312. Oxytropis leucantha sensu Hultén. 3239; det. by Porsild. Porsild annotated this collection as follows: “This is Oxytropis leucantha sensu Hult. Fl. Al. & Yuk. not Pall. It is close to O. glu- tinosa Pors. but differs in several characters nor is it O. viscida or O. viscidula. It is probably an undescribed species of which I have cther collections from W. Alaska [Dec. 1960]." O. nigrescens (Pall.) Fisch. 3154, 3185, 3205*, 3324; det. by Porsild. Our collections exhibit several rather distinct types. Nos. 3184 and 3205 are very pulvinate and gray-pilose and seem to fit Hultén's ssp. pygmaea (cf. Hultén, pp. 1102-1105) fairly well, whereas no. 3185 is not particularly pulvinate and is less gray-pilose. No. 3324 includes plants of both types. The less pulvinate plants approach ssp. bryophila (Greene) Hult. Hedysarum alpinum L. ssp. americanum (Michx.) Fedtsch. 3294. H. mackenzii Richards. 3295, 3334. Epilobium latifolium L. 3304. Hippuris vulgaris L. 3258 Bupleurum americanum Coult. et Rose. 3146. These plants are very low, in some cases almost acaulescent, and have relatively few rays (3-6) per umbel. Leaf-width however is quite variable. The cauline leaves mostly are narrowly lanceolate or linear, but they range up to nearly a centimeter in width and in some cases are distinctly clasping at the base. Thus in general habit they are strikingly different from the large plants of Interior Alaska, and there is no doubt that they are closely related to the Asiatic B. triradiatum Adams ex Hoffm. Quite possibly they should be referred to ssp. arcti- cum (Regel et Tiling) Hult. of the latter species, but until a compre- hensive study of the Alaskan forms is available, including extensive comparisons with Asiatic material, it seems best to follow Hultén (pp. 1166-1168) in relegating this dwarf arctic form to the endemic Ameri- can species, B. americanum (cf. also, Hult. Fl. Kamtch., pp. 157-158; Lincheviskii, Fl. SSSR 16 : 301-303, 1950). Conioselinum cnidiifolium (Trucz.) Pors. 3143. The relationship of this species to the Asiatic C. vaginatum (Spreng.) Thell. s.l. needs careful scrutinv. Although C. enidiifolium is reported to occur in eastern Asia (Hultén, p. 1177; Polunin, 1959, p. 328), Schischkin makes no mention of it in Flora SSSR (17: 1-10, 220 Rhodora [Vol. 65 1951). Apparently, he referred all material of this affinity to C. vagi- natum. Pyrola grandiflora Radius. 3271*. Ledum palustre L. ssp. decumbens (Ait.) Hult. 3273. Cassiope tetragona (L.) D. Don. 3217, 3322*. *Primula borealis Duby var. borealis. 3171. This Amphi-beringian species, apparently restricted to the Arctic seashores (cf. Wiggins and Thomas, p. 295), was collected on a moist turfy slope right on shore at Cape Sabine. Although rare on the Arc- tie Slope, its occurrence here is not unexpected, since it has been col- lected before on the coastal plain northeast of Cape Sabine and at Cape Lisburne to the west. *P. egaliksensis Wormskj. 3200. According to Wiggins and Thomas (p. 296), this species is *rare in the foothills of the Brooks Range and on the Coastal Plain." *F. sibirica Jacq. 3207*. This Eurasian species has a very disjunct range in Alaska and Yukon, and is previously unreported north of the Brooks Range. Ac- cording to Hultén (1937, p. 92), it is an *Arctic-montane" radiant from eastern Asia. *P. tschuktschorum Kjellm. 3752. Although known from a widespread group of localities south of the Brooks Range, the only previously known locality north of the Range was Cape Lisburne (cf. Wiggins and Thomas, p. 295). Androsace chamaejasme Host. ssp. lehmanniana (Spreng.) Hult. 3241, 3315. Dodecatheon frigidum Cham. et Schlecht. 2153, 3305. Phlox sibirica L. ssp. borealis (Wherry) stat. et comb. nov. ( P. borealis Wherry, The Genus Phlor, p. 126, 1955; P. sibirica sensu Hult.) 3149. In Wherry's opinion (op. cit.), Alaskan plants of the affinity of P. sibirica L. should be segregated into a separate taxon because they differ consistently from Asiatic plants of this species in being “smaller in stature and in sizes of parts." He proposed to call the Alaskan taxon a distinct species, P. borealis Wherry. While he seems to have a good case for segregating the Alaskan plants, the differences he cites are all variable and quantitative, and it seems best to regard these plants as nothing more than a geographical race of P. sibirica L. Polemonium acutiflorum Willd. in Roem. et Schult. 3235. Myosotis alpestris F. W. Schmidt ssp. asiatica Vestergr. in Hult, 2147. Lagotis glauca Gaertn. s.l. 3188*. In the absence of a convincing statistical demonstration of the validity of var. stelleri (Cham. et Schlecht.) Trautv. (cf. Hultén, FI. Kamtch. 4: 103-105, 1930), I am inclined to consider L. glauca broadly, even though both Hultén (pp. 1384-1385) and Wiggins and Thomas (p. 307) refer all Arctic Slope plants to this variety. The question is 1963] Plants, Cape Sabine — Shetler 2231 further complicated by the problem of L. minor (Willd.) Standley (cf. Vikulova, Fl. SSSR 22: 501, 1955). Whether it is still a third closely related taxon or simply the oldest name for var. stelleri, as suggested by Vikulova's synonymy, can only be resolved by comparing Eurasian and American specimens statistically. Castilleja pallida (L.) Spreng. s.l. 3216, 3333. According to Pennell's key (p. 522), our specimens belong to the typical subspecies, ssp. pallida (ssp. typica of Pennell). Phytogeo- graphically, however, they should belong to ssp. caudata Pennell, and Hultén has determined our 3216 as this subspecies. In point of fact, the distinctions drawn by Pennell between these two subspecies are questionable (cf. Hultén, pp. 1391-1392), and I refrain for the present from recognizing them. Pedicularis capitata Adams. 3181*, 3306. F. langsdorfii Fisch. ex Steven. s.s. (exclud. P. arctica R. Br.) 3220*, 3318". *P. oederi Vahl. 3240*. This species appears to be very rare on the Arctic Slope, although south of the Brooks Range it is rather widespread and reasonably common (cf. Hultén, map 1056, p. 1472). Spetzman (p. 49) records it only from Anaktuvuk Pass. Wiggins and Thomas (pp. 309-314), curiously, make no mention of this species, although the specimen on which they based the presence of P. flammea L. on the Slope was Spetz- man's no. 1747, which he determined as P. oederi. This is undoubtedly the specimen on which Spetzman based his Arctic Slope record, be- cause it is from Anaktuvuk Pass. I have examined this collection (1747) at the U. S. National Herbarium and have compared it with both P. flammea and P. oederi. It does fall somewhat on the low side of the size range for P. oederi, giving some room for doubt perhaps, but there is no question but that it is this species and not P. flammea, as Wiggins and Thomas decided (p. 314), albeit with some doubt. This means that P. flammea should be removed from their list cf species for the Arctic Slope, and P. oederi should be added. *P. pennellii Hult. 3261. Not previously recorded this far west on the Arctic Slope (cf. Wig- gins and Thomas, p. 311). P. sudetica Willd. ssp. albolabiata Hult. (cf. Hultén, 1961); 3164; det. by Hultén. F. sudetica ssp. pacifica Hult. (cf. Hultén, 1961); 3212; det. by Hultén. P. verticillata L. 3262*. Valeriana capitata Pall. ex Link. 3166, 3192. Campanula uniflora L. 3154, 3158*, 3167, 3174. Blue and white-flowered forms seemed to be equally common and to occur intermixed in the populations. Solidago multiradiata Ait, 3308*. Aster sibiricus L. s.l. (includ. A. subintegerrimus (Trautv.) Ostenf. et 222 Rhodora [Vol. 65 Resvoll, and A. richardsonii Spreng.; cf. Tamamshan, Fl. SSSR 25: 77-110, 1959, and Hultén, pp. 1493-1496). 3210. *Erigeron compositus Pursh var. glabratus Macoun. 2144. This is the first record of the species for the Arctic Slope and repre- sents a very significant range extension northwestward. Chrysanthemum integrifolium Richards. 3293. Artemisia arctica Less. (A. norvegica ssp. saxatilis (Bess.) Hall et Clements, Wiggins and Thomas, p. 341). 3165, 3303. * A. richardsoniana Bess. (A. borealis sensu Hultén, in part: cf. Fl. AI. & Yuk., p. 1556, under A. aleutica Hult.; cf. also, Porsild, 1955, pp. 185-186;? A. trifurcata Steph. ex Spreng., Wiggins and Thomas, p. 340.) 3297; det. by Porsild. Porsild (loc. cit) segregates this species from the circumpolar A. borealis Pall. as an endemic of the Canadian Arctic Archipelago, while Hultén (loe. cit.) and Polunin (1959, p. 430) unite these two species under A. borealis. Our collection represents a significant range exten- sion only if the interpretation of Porsild is accepted. Owing to the absence of discussions and to the minimal synonymy in Wiggins’ and Thomas' treatment, I am hard-pressed to relate it to the treatments of Porsild and Hultén. Petasites frigidus (L.) Fries. 3227. Arnica louiseana Farr ssp. frigida (Meyer ex Iljin) Maguire. 3162, 3247. Senecio atropurpureus (Ledeb.) Fedtsch. ssp. frigidus (Richards.) Hult. 3161, 3197*, 3272, 3309. No, 3309 approaches ssp, atropurpureus. S. fuscatus (Jord. et Fourr.) Hayek. 2159*, S. hyperborealis Greenm. 3145. It seems to me highly doubtful that this species is really distinct from S. conterminus Greenm. and/or S. resedifolius Less. Polunin (1959), p 460) unites it with the latter species, a relatively widespread species on the Arctic Slope, although Hultén, Porsild, and Wiggins and Thomas maintain it. S. lugens Richards. 3163, 3245. S. resedifolius Less. 3246. Saussurea angustifolia ( Willd.) DC. 3292*. *Taraxacum phymatocarpum J. Vahl. (T. lyratum (Led.) DC., in part, Wiggins and Thomas, p. 353) ; 2151*, 3243; det. by Porsild. From the specimen citations it appears that Wiggins and Thomas referred all specimens of this species to T. lyratum, although this is made clear neither by discussion nor synonymy. I am not able to compare our specimens against their key at present and think it best to accept Porsild's determination. Whether or not our record from Cape Sabine is phytogeographically significant depends entirely upon one's taxonomic view of the complex of forms relegated to T. lyratum. If one segregates the apparently New World T. phymatocarpum from 1963] Plants, Cape Sabine — Shetler 223 it, then our record is reasonably significant. Only one previously pub- lished report for the Arctic Slope appears to exist (Spetzman, p. 51). Crepis nana Richards. 3213*. UNITED STATES NATIONAL MUSEUM, WASHINGTON, D. C. BIBLIOGRAPH Y ANDERSON, J. P. 1959. Flora of Alaska and adjacent parts of Canada, integrated and indexed by Richard W, Pohl. Ames, Iowa State Univ. Press. 543 p. FLORA SSSR [Flora of the USSR]. 1934-1960. Acad. Sci. Leningrad. Vols. 1-25, 30. [In Russian] HULTÉN, ERiC. 1927-30. Flora of Kamtchatka and the adjacent islands. Kungl. Sv. Vet. Akad. Handl., Ser. 3. 1135 p. (4 pts.) . 1937. Outline of the history of arctic and boreal biota during the Qua[r]ternary Period. Stockholm. 168p. 1941-50. Flora of Alaska and Yukon. Lunds Univ. Arsskr. N. F. Avd. 2. 1902 p. (10 pts.) . 1961. Two Pedicularis species from NW. America, P. albertae n. sp. and P. sudetica sens. lat. Sv. Bot. Tidskr. 55(1): 193-204. 1962. The circumpolar plants, I: vascular crypto- gams, conifers, monocotyledons. Stockholm, Almqvist & Wiksell. 275 p. LÖVE, ASKELL, Doris LÓVE, AND MARCEL RAYMOND. 1957. Cytotax- onomy of Carex Section Capillares. Canad. Journ. Bot. 35: 715- 761. PENNELL, FRANCIS W. 1934. Castilleja in Alaska and northwestern Canada. Proc. Acad. Nat. Sci. Phila.86: 517-540. POLUNIN, NICHOLAS. 1959. Circumpolar arctic flora. London, Oxford Univ. Press. 514 p. Popov, M. G. 1957. Flora of central Siberia. Moscow. 2 vols. [In Russian ] PoRsILD, A. E. 1951. Botany of southeastern Yukon adjacent to the Canol Road. Bull. Nat. Mus. Canada 121: 1-400. . 1955. The vascular plants of the western Canadian Arctic Archipelago. Bull. Nat. Mus. Canada 155: 1-226. . 1957. Illustrated flora of the Canadian Arctic Archi- pelago. Bull. Nat. Mus. Canada 146: 1-209. Raup, HuGH M. 1947. The botany of southwestern Mackenzie. Sar- gentia 6: 1-275. . 1959. The willows of boreal western America, Contr. Gray Herb. Harvard Univ. 185: 1-95. SHETLER, STANWYN G. 1961. Application of the names Cardamine digitata, Dentaria pentaphylla, and D. heptaphylla. Taxon 10(9): 264-267. 224 Rhodora [Vol. 65 SPETZMAN, LLOYD A. 1959. Vegetation of the Arctic Slope of Alaska. Washington, U. S. Govt. Print, Office, iii, 19-58. (U. S. Geol. Surv. Prof. Pap. 302-B). TOLMATCHEV, A. 1955. Nota de Caltha Orientis extremi. Not. Syst. Leningrad 17: 144-159. WAGNON, H. KEITH. 1950. Nomenclatural changes in Bromus. Rho- dora 52: 209-215. WHERRY, EpGAR T. 1955. The genus Phlox. Lancaster (Pa.), Wick- ersham Print. Co. 174 p. WIGGINS, IRA L., AND JOHN H. THoMas. 1962. A flora of the Alaskan Arctic Slope. Toronto, Univ. Toronto Press, 425 p. (Arctic Insti- tute of North America Special Publ. No. 4) ACKNOWLEDGMENTS The fieldwork for this paper was supported by the Arctic Institute of North America (Contract 51), and the paper was prepared while I held Cooperative Graduate Fellowships from the National Science Foundation. I am much indebted to Karl Stone for assisting in the field; to Max Brewer, Director of the Arctic Research Laboratory (Point Barrow), for making the trip to Cape Sabine possible; to W. S. Benninghoff, University of Michigan, for encouraging the prepara- tion of this report; and to the several specialists, duly credited in the catalog, who so kindly determined or verified critical collections. Special thanks are also tendered to Erie Hultén, Riksmuseum (Stock- holm), A. E. Porsild, National Museum of Canada, Rogers MeVaugh and Edward Voss, both of the University of Michigan, for critically reviewing this paper and offering numerous helpful and important suggestions. CHROMOSOME STUDIES IN MEXICAN COMPOSITAE: D. C. D. DEJoNG AND E. K. LONGPRE The chromosome numbers reported in this paper are, for the greater part, based upon specimens collected during the summer of 1961 along with material for the authors' separ- ate monographic studies of the genera Astranthium (Com- positae-Astereae) and Sabazia (Compositae-Heliantheae). Bud collections were obtained from plants growing in their native habitats and were placed in vials containing a fresh- ly mixed Carnoy solution of 6 parts 95% ethyl alcohol, 3 parts chloroform, and 1 part glacial acetic acid. The vials were subsequently placed in a foam-plastic cooler with bulk ice and kept refrigerated during the entire field trip; upon return to East Lansing they were stored at 1° C until used. Slides were made by the aceto-carmine smear technique. Chromosomes were drawn with the aid of a Zeiss drawing apparatus at an initial magnification of ca. 4000X and are here reduced to ca. 1300X. A complete list of the taxa studied is contained in Table 1; the tribal sequence is that of Hoffmann (1897), whereas the genera and species have been placed in alphabetical order. All counts are documented by voucher specimens in the Herbarium of Michigan State University. We are in- debted to Dr. J. H. Beaman for aid in the identification of certain specimens and in the preparation of the manuscript, and to Dr. Rogers McVaugh for the use of herbarium facili- ties at the University of Michigan, Ann Arbor. Dr. Arthur Cronquist kindly identified some species of Erigeron and Conyza; Drs. B. L. Turner and M. C. Johnston determined the species of Aphanostephus, Dyssodia, and Schkuhria. Al other determinations are our own. 'Supported by National Science Foundation Grant G-9045 and by funds for field work partially supplied by the Department of Botany and Plant Pathology, Michigan State University, East Lansing, Michigan. 225 226 Rhodora [Vol. 65 TABLE 1. SUMMARY OF THE COLLECTIONS STUDIED. GAMETIC TAXON LOCALITY CHROMOSOME NUMBER EUPATORIEAE Eupatorium greggii DURANGO: 1 mi. s. of La Zarca. Gray De Jong & Longpre 951. 10 (Fig. 1) Kuhnia chlorolepis CHIHUAHUA: 11 mi. e. of Majal- Woot. & Standl. ca. De Jong & Longpre 924. 9 (Fig. 2) ASTEREAE Achaetogeron forreri DURANGO: at railroad crossing Greene near Hacienda Coyotes. De Jong & Longpre 1003. 27 (Fig. 3) A. griseus Greenman DURANGO: 30 mi. w. of Ciudad Durango, along Hwy 40. De Jong & Longpre 974. 9 A. griseus DURANGO: 31 mi. w. of Ciudad Durango, along Hwy 40. De Jong & Longpre 978. 9 (Fig. 4) Aphanostephus ramos- CHIHUAHUA: along Hwy to issimus DC. Buenaventura, 7 mi. w. of Hwy 45 turnoff. De Jong & Longpre 904. 4 (Fig. 5) A. ramosissimus CHIHUAHUA: 15 mi. w. of Buen- aventura. De Jong & Longpre 916. 4 A. ramosissimus DURANGO: 1 mi. s. of La Zarca. De Jong & Longpre 961. 4 A. ramosus (DC.) Gray PUEBLA: 1.5 mi. w. of Chachapa. Beaman 3614. 4 A. ramosus MICHOACAN: meadow along road to Cerro San Andres, 2815 m. alt. De Jong 758. 4 A. ramosus MICHOACAN: At Km. 43.5, Hwy 120 to Uruapan. De Jong 762. 4 A. ramosus PUEBLA: 3.4 mi, s. of Tlachichu- pa on road to Zuapam. Beaman 3621. 4 Aster pauciflorus DURANGO: ca. 12 mi. n. of Don- Nutt. ato Guerra. De Jong & Longpre 964. 9 (Fig. 6) A. exilis Ell. var. JALISCO: 13 mi. s. of Guadala- australis Gray jara. De Jong & Longpre 1018. 5 Astranthium mexicanum STATE OF MEXICO: At Km. 75, (Gray) Larsen Amecameca—Popocatépetl road, 3235 m. alt. De Jong 566. 18 1963] A. mexicanum A. mexicanum A. mexicanum A. mexicanum A. orthopodum (Robins. & Fern.) Larsen A. purpurascens (Robins.) Larsen A. xanthocomoides (Less.) Larsen A. xanthocomoides A. xylopodum Larsen Baccharis glutinosa Pers. Conyza canadensis (L.) Cronq. C. aff. confusa Cronq. Erigeron coronarius Greene E. aff. coronarius E. delphinifolius Willd. E. delphinifolius Willd. aff. subsp. neomexicanus (Gray) Cronq. E. delphinifolius Willd. subsp. neo- mexicanus var. neomexicanus Mexican Compositae — De Jong and Longpre FEDERAL DISTRICT: near railroad overpass at La Cima. De Jong 647. FEDERAL DISTRICT: on slope along road to La Cima. De Jong 653. MICHOACAN: Cerro San Andres, ca. 3100 m. alt. De Jong 757. MICHOACAN: East slope of Cerro Tancitaro, 3080 m. alt. De Jong 1068. DURANGO: 38 mi. w. of Ciudad Durango, along Hwy 40. De Jong 984. HIDALGO: at village limit of Co- brecito. De Jong 1227. PUEBLA: 5.3 mi. sw. of San Sal- vador El Seco. De Jong 630. TLAXCALA: 2.2 mi. s. of Puebla State Line, just n. of Tlaxco. De Jong 1195. JALISCO: Sierra del Halo, s. of Tecalitlan. De Jong 1028. CHIHUAHUA: 11 mi. e. of Majal- ca. De Jong & Longpre 925. CHIHUAHUA: along Hwy to Buenaventura, 7 mi. w. of Hwy 45 turnoff. De Jong & Longpre 9035. DURANGO: 1 mi. e. of La Ciudad. De Jong & Longpre 1011. DURANGO: 38 mi. w. of Ciudad Durango, along Hwy 40. De Jong & Longpre 994. DURANGO: 14 mi. w. of Ciudad Durango, along Hwy 40. De Jong & Longpre 973. STATE OF MEXICO: 4.5 kms. s. of Tlalmanaleo. Beaman 4525. DURANGO: ca. 12 mi. n. of Donato Guerra. De Jong & Longpre 967. DURANGO: 14 mi. w. of Ciudad Durango, along Hwy 40. De Jong & Longpre 972. 227 18 18 18 18 3 (Fig. 5 (Fig. 9 (Fig. 9 (Fig. 7) 9) 10) 11) 228 E. delphinifolius Willd. subsp. neo- mexicanus var. oreophilus (Greenman) Cronq. E. divergens T. & G. E. pubescens HBK. Grindelia oxylepis Greene G. oxylepis G. oxylepis G. sublanuginosa Steyermark Haplopappus spinulosus (Pursh) DC. subsp. scabrellus (Greene) Hall H. spinulosus (Pursh) DC. subsp. scabrellus Leucelene ericoides (Torr.) Greene Machaeranthera tanace- tifolia (HBK.) Nees M. gymnocephala (DC.) Shinners Psilactis asteroides Gray Xanthocephalum gymno- spermoides (Gray) Benth. & Hook. ex Rothrock in Wheeler X. gymnospermoides X. sericocarpum Gray Rhodora CHIHUAHUA: 1 mi. e. of Majal- ca. De Jong & Longpre 934. CHIHUAHUA: 11 mi. e. of Majal- ca. De Jong & Longpre 922. [Vol. 65 9 (Fig. 12) HIDALGO: 3 mi. s. of Cuyama- 35 (Fig. 13) univalents loya. De Jong & Longpre 1226. CHIHUAHUA: 6 mi. w. of Cuaoh- temoc. De Jong & Longpre 940. CHIHUAHUA: 22 mi. s. of Hidal- go del Parral. De Jong & Long- pre 948. DURANGO: 1 mi. s. of La Zarca. De Jong & Longpre 956. JALISCO: Lago Chapala, 49 mi. s. of Guadalajara. De Jong & Longpre 1026. CHIHUAHUA: along Hwy to Buenaventura, 7 mi. w. of Hwy 45 turnoff. De Jong & Longpre 901. CHIHUAHUA: 17 mi. n. of Ciudad Camargo. De Jong & Longpre 919. CHIHUAHUA: 17 mi. n. of Ciudad Camargo. De Jong & Longpre 920. CHIHUAHUA: along Hwy to Buenaventura, 7 mi. w. of Hwy 45 turnoff. De Jong & Longpre 900. DURANGO: 1 mi. s. of La Zarca. De Jong & Longpre 955. CHIHUAHUA: 6 mi. w. of Cuaoh- temoc. De Jong & Longpre 944. CHIHUAHUA: meadows at Majal- ca. De Jong & Longpre 930. CHIHUAHUA: 6 mi. w. of Cuaoh- temoc. De Jong & Longpre 943. CHIHUAHUA: 9 mi. s. of V. Mat- amoros. De Jong & Longpre 950. 6 6 (Fig. 8 (Fig. 4 (Fig. 4 (Fig. 14) 15) 16) 18) 1963] INULEAE Gnaphalium lavanduli- folium (HBK.) Blake HELIANTHEAE Berlandiera lyrata Benth. Calea palmeri Gray C. scabra (Lag.) Robins. Cosmos palmeri Robins. var. palmeri Chrysanthellum mexi- canum Greenman Parthenium hysteropho- rus L. P. hysterophorus Sabazia humilis (HBK.) Cass. Sanvitalia procumbens Lam. Verbesina tetraptera Gray V. callilepis Blake Ximenesia encelioides Cav. Zexmenia palmeri Greenman in Jones HELENIEAE Dyssodia cancellata (Cass.) Gray D. hartwegii (Gray) Robins STATE OF MEXICO: Nevado de Toluea, ca. 4090 m. alt. De Jong & Longpre 1116. CHIHUAHUA: along Hwy to Buenaventura, 3.6 mi. e. of Ri- cardo Flores Magon. De Jong & Longpre 910. MICHOACAN: 12 mi. e. of Zacapu. Longpre 113. JALISCO: Sierra del Halo, s. of Tecalitlan. De Jong 1033. DURANGO: 33 mi. w. of Ciudad Durango, along Hwy 40. De Jong & Longpre 983. JALISCO: 13 mi. s. of Guadala- jara. De Jong & Longpre 1020. DURANGO: ca. 12 mi. n. of Don- ato Guerra. De Jong & Longpre 963. JALISCO: 13 mi. s. of Guadala- jara. De Jong & Longpre 1021. PUEBLA: between Kms. 65 and 66, Hwy 190, se. of Mexico City. Longpre 350A. DURANGO: ca. 12 mi. n. of Don- ato Guerra. De Jong & Longpre 960. MICHOACAN : 12 mi. e. of Zacapu. Longpre 127. ca. DURANGO: 1 mi. w. of La Ciudad. De Jong & Longpre 1012. CHIHUAHUA: along Hwy to Buenaventura, 9.2 mi. w. of Ri- cardo Flores Magon. De Jong & Longpre 911. MICHOACAN: 12 mi. e. of Zacapu. Longpre 128. ca. CHIHUAHUA: 9 mi. s. of V. Ma- tamoros. De Jong & Longpre 949. DURANGO: 1 mi. s. of La Zarca. De Jong & Longpre 957. Mexican Compositae — De Jong and Longpre 14 15 16 17 17 13 26 229 (Fig. (Fig. (Fig. (Fig. (Fig. (Fig. (Fig. (Fig. (Fig. (Fig. 19) 20) 21) 22) 26) 27) 28) 230 Rhodora [Vol. 65 ON CV QM X. 077 471,9 | ON rd ni f á 4 27 l Sty ex stale ud 5 6 T 8 p 9$ ot Lra D t e. že IO a a I | É |2 fu Pads I OR ve Y n: | 14 I5 I6 II g y sÑ sS fa $ L4 Te 2p 6 > A UI qtr Cow €" 1963] Mexican Compositae — De Jong and Longpre 231 Gaillardia pinnatifida CHIHUAHUA: 17 mi. n. of Ciudad Torr. Camargo. De Jong & Longpre 914. IT Psilostrophe gnaphalodes DURANGO: 1 mi. s. of La Zarca. DC. De Jong & Longpre 954. 32 (Fig. 29) P. tagetina CHIHUAHUA: along Hwy to (Nutt.) Greene Buenaventura, 7 mi. w. of Hwy 45 turnoff. De Jong & Longpre 905. 16 (Fig. 30) Schkuhria anthemoidea JALISCO: Lago de Chapala, 49 (DC.) Coult. var. mi. s. of Guadalajara. De Jong wislizeni (Gray) & Longpre 1024. 10 (Fig. 31) Heiser Tagetes lucida Cav. CHIHUAHUA: 3 mi. w. of Guer- rero. De Jong & Longpre 947. 11 (Fig. 32) ANTHEMIDEAE Achillea lanulosa CHIHUAHUA: meadows at Majal- Nutt. ca. De Jong & Longpre 932. 18 SENECIONEAE Cacalia sinuata DURANGO: 33 mi. w. of Ciudad Llav. & Lex. Durango, along Hwy 40. De Jong & Longpre 982. 30 MUTISIEAE Chaptalia dentata DURANGO: 14.5 mi. w. of Hacien- (L.) Cass. da Coyotes. De Jong & Longpre 1008. 16 (Fig. 33) CICHORIEAE Stephanomeria pauci- CHIHUAHUA: 17 mi. n. of Ciudad fiora (Torr.) A. Nels. Camargo. De Jong & Longpre 915. 8 (Fig. 34) S. tenuifolia (Torr.) CHIHUAHUA: 11 mi. e. of Majal- Hall ca. De Jong & Longpre 927. 8 (Fig. 35) Fig. 1-20. Meiotic chromosomes of Mexican Compositae, X ca. 1300. Fig. 1. Eupatorium greggii (n = 10) — Fig. 2. Kuhnia chlorolepis (n = 9) — Fig. 3. Achaetogeron forreri (n = 27) — Fig. 4. Achae- togeron griseus (n = 9) — Fig. 5. Aphanostephus ramosissimus (n = 4) — Fig. 6. Aster pauciflorus (n = 9) — Fig. 7. Astranthium ortho- podum (n — 3) — Fig. 8. Astranthium xanthocomoides (n — 8) — Fig. 9. Astranthium xylopodum (n = 5) — Fig. 10. Conyza aff. confusa (n = 9) — Fig. 11. Erigeron delphinifolius (n = 9) — Fig. 12. Eriger- on divergens (m = 9) — Fig. 18. Erigeron pubescens (m = 35 unival- ents) — Fig. 14. Grindelia sublanuginosa (m = 6) — Fig. 15. Leuce- lene ericoides (n = 8) — Fig. 16. Machaeranthera gymnocephala (n = 4) — Fig. 17. Xanthocephalum gymnospermoides (n —6) — Fig. 18. Xanthocephalum sericocarpum (n = 4) — Fig. 19. Gnaphalium lavan- dulifolium (n = 14) — Fig. 20. Calea scabra. (n = 16). 232 Rhodora [Vol. 65 e 4» 425 . + 9» M Q9 NI A P » Cee c 4 2| 22 23 24 e ve € 4 e - J > ef E p’ p ge & -— j os r e è " ae wv ^ * 4 Tes o5 9*" 26 ?7 28 Py 9 » ¢ $345» (2 oes 9> a 9 e ety ne fo . Ge are oO ,, E 29 d 30 3 2 U a? a Q e? > «4. sw 1963] Mexican Compositae — De Jong and Longpre 233 DISCUSSION EUPATORIEAE — Eupatorium. Most species counted to date are members of sect. Eximbricata. E. greggii (n = 10) is a member of sect. Conoclinium, in which another species, E. betonicum, has also been counted as n = 10 by Turner, Powell, and King (1962). Eupatorium sect. Conoc- linium has close affinities with Ageratum which is unibasic on x = 10. Kuhnia chlorolepis (n — 9). This species was also report- ed as n — 9 by Turner (1959). ASTEREAE — Achaetogeron (x — 9). The chromosome numbers reported for A. forreri (n = 27) and A. griseus (n —9) are the first counts for this primarily Mexican genus. In the description of A. wislizeni, the species upon which Achaetogeron was based, Gray (1849) noted it to be“... entirely like a true Erigeron . . . except as to the pappus which refers it to the Bellideae . . ." Thus Achaetogeron was established by Gray to accommodate a species charac- terized by a simple pappus of small setae and without tne inner pappus bristles found in Erigeron. In addition to A. wislizeni, 20 species have been described in Achaetogeron by subsequent authors. In Achaetogeron we can distinguish three groups of spe- cies on the basis of pappus characters. One group, including the type species of the genus, has a simple pappus of short setae or squamellae. The second is epappose; the third has a double pappus of inner bristles and an outer crown of setae Fig. 21-35. Meiotic chromosomes of Mexican Compositae, X ca. 1300. Fig. 21. Cosmos palmeri var. palmeri (n — 17) — Fig. 22. Chrysanthel- lum mexicanum (n = 8) — Fig. 23. Sabazia humilis (n = 4) — Fig. 24. Sanvitalia procumbens (n = 8) — Fig. 25. Verbesina callilepis (n = 18) — Fig. 26. Ximenesia encelioides (n = 17) — Fig. 27. Dys- sodia cancellata (n = 13) — Fig. 28. Dyssodia hartwegii (n = 26) — Fig. 29. Psilostrophe gnaphalodes (n = 32) — Fig. 30. Psilostrophe tagetina (n = 16) — Fig. 31. Schkuhria anthemoidea var. wislizeni (n — 10) — Fig. 32. Tagetes lucida (n — 11) — Fig. 33. Chaptalia dentata (n = 16) — Fig. 34. Stephanomeria pauciflora (n = 8) — Fig. 35. Stephanomeria tenuifolia (n — 8). 234 Rhodora [Vol. 65 or squamellae. Most of the species of this third group were described by Larsen (1948). Since the double pappus is also found in the majority of species of Erigeron and since this third group of species is like Erigeron in all other charac- ters, there seems to be no reason to retain these species in Achaetogeron. To this effect it should be noted that we have found Achaetogeron fisheri Larsen to be conspecific with Erigeron delphinifolius. On this basis we hesitate at pres- ent to effect the transfers of these species without a careful study of the Mexican species of Erigeron. With respect to the remaining species of Achaetogeron, their generic status has been questioned by several workers. In fact, Greene (1891) considered the genus to be artificial and transferred the species then known to Erigeron. Our studies have so far indicated that Achaetogeron is somewhat heterogeneous and that some species have such close rela- tives in Erigeron, that they may be placed with these species in Erigeron, without affecting the naturalness of that genus. The basic chromosome number of Achaetogeron (x = 9) further supports the close relationship with Erigeron which has also 2x —9; further morphological and cytological studies will be carried out on the Mexican species of both genera. The chromosome numbers listed for Aphanostephus and Aster agree with the basic numbers reported for these genera by other workers. The count for Aster pauciflorus (n = 9), a member of sect. Orthomeris, is a first report. Astranthium (x = 3, 4, 5). The counts for A. orthopod- um (n — 3), A. vxylopodum (n = 5), and that reported by Baldwin (1941) for A. integrifolium (n = 4) establish & = 3, 4, and 5 as the basic chromosome numbers of this predom- inantly Mexican genus which is a member of the subtribe Bellidinae. Of the eleven genera placed in this subtribe by Hoffmann (1897), five (excluding Keerlia which is now in Chaetopappa) are not yet known chromosomally. Of the remaining genera, Achaetogeron, Bellis, and Lagenophora have x = 9; the Australian Brachycome likewise has x = 9 (De Jong, unpublished). The genus Aphanostephus, like 1963] Mexican Compositae — De Jong and Longpre 235 Astranthium, is tribasic with « = 3, 4, and 5 (Turner in Raven et al., 1960). A. mexicanum (n = 18). Diploid and tetraploid counts were published previously for this species by Beaman, De Jong, and Stoutamire (1962). A. mexicanum is found at alpine and subalpine elevations in the transvolcanic belt of South-Central Mexico and extends southward into the State of Oaxaca. The species is anomalous in the genus in that it has morphological characters commonly found in Achaeto- geron, while, cytologically, it agrees with the basic number of that genus. A. xanthocomoides (n = 8). A count of n = 8 was also obtained for this species by Turner, Beaman, and Rock (1961). These authors likewise reported a second collection from Nuevo Leon to have n = 8, but this collection, Beaman 2697, should be regarded as an undescribed species of Achaetogeron. In view of the basic chromosome number of this genus, x = 9, it seems that a recount is in order for collection 2697. A. xanthocomoides, as understood at pres- ent, has its northern limit in the Sierra de Pachuca in the State of Hidalgo (De Jong, unpublished). Erigeron (x = 9). The chromosome numbers of the spe- cies listed in Table 1 are consistent with the basic chromo- some number of the genus (cf. Montgomery and Yang, 1960). E. delphinifolius (n. = 9). This collection (Beaman 4525 ) was obtained from near the type locality of Achaetogeron fisheri (cf. discussion of Achaetogeron). E. pubescens. We found 35 univalents in most cells ex- amined, whereas cells in which a few bivalents were formed were rare. Pollen was found to be extremely variable in size, with a high percentage of the grains aborted. Turner, Beaman, and Rock (1961) reported another collection of this species with 36 univalents. E. divergens (n — 9). We have found this species to be diploid, whereas E. divergens var. cinereus was reported by Montgomery and Yang (1960) to have 2n = 27. The counts listed for Grindelia are consistent with the 236 Rhodora [Vol. 65 basic number of the genus. The chromosome number of G. sublanuginosa (n = 6) has not been reported before. Haplopappus. Raven et al. (1960) indicated a collection of H. spinulosis subsp. scabrellus from Arizona to have n = 4. We have found n = 4 and n = 6 in two Chihuahuan col- lections of this subspecies. Jackson (1957a) likewise ob- tained n = 4 and n —6 in H. spinulosis subsp. cotula. Leucelene (x — 8). L. ericoides (n — 8); a count of n — 16 was published (cited as Aster hirtifolius) by Raven et al. (1960). This widespread, weedy perennial apparently has diploid and tetraploid races, which may account for the polymorphic nature of the species. We have followed Shin- ners (1946) in recognizing this monotypic genus which seems to have closer affinities with Chaetopappa than with Aster. Xanthocephalum (x = 4, 6). The count here reported for X. sericocarpum (n = 4) brings to 4 the number of species in the genus that have this chromosome number. On the other hand, X. gymnospermoides has n = 6, as reported by various authors (Raven et al., 1960; Turner, Powell, and King, 1962; see also Table 1). Three of the 8 species recog- nized by Solbrig (1961) are not yet known cytologically. The chromosome numbers listed for species of Baccharis, Conyza, Machaeranthera, and Psilactis are consistent with the basic numbers reported for these genera (cf. Raven et al., 1960 ; Turner, Powell, and King, 1962). INULEAE — Gnaphalium (x — T1). The count (n = 14) for G. lavandulifolium, a suffrutescent alpine species, has not been previously reported. Another alpine species, G. vulcanicum, was also found to be tetraploid by Beaman, De Jong, and Stoutamire (1962). HELIANTHEAE — Berlandiera lyrata (n = 15). Our count agrees with that published by Turner, Powell, and King (1962). Calea (x = 16, 18). We have found both C. palmeri and C. scabra to have n = 16. Previous counts in the genus were reported by Turner, Powell, and King ( 1962), but an un- equivocal count was only obtained for C. trichotoma (n =- 1963] Mexican Compositae — De Jong and Longpre 237 18) by these authors. Bentham (1873) observed that some species of Calea were difficult to distinguish from species of Sabazia. Indeed, the herbaceous C. palmeri closely resembles Sabazia in floral and vegetative characters and may well be a member of that genus. Our count for C. scabra (n — 16) adds a second basic number to Calea. Cosmos (x = 12, 17). Of the 27 species recognized in the genus by Sherff (1955), only 4 species have so far been counted. From all reported chromosome numbers, the genus has appeared to be unibasic with x — 12. We have found C. palmeri var. palmeri to have n = 17 and are considering it to be diploid on a base of x — 17. Chrysanthellum (x — 8). The chromosome number of C. mexicanum (n = 8) is the first report for this genus of 2 or 3 small annual species which Hoffmann (1897) placed in the subtribe Coreopsidinae. A species of Heterospermum, a genus which Bentham (1873) considered to be closely re- lated to Chrysanthellum, was counted as n = 25 by Turner, Beaman, and Rock (1961). Parthenium. Our counts for two collections of P. hystero- phorus (n — 17) agree with those reported by Rollins (1950) for the same species. On the other hand, Thombre (1959) found P. hysterophorus to have n — 18. Since some other species in the genus have been reported as n — 18, the problem here appears to be taxonomic rather than cytologi- cal. Sabazia (x — 4). Our count for S. humilis (n = 4) is consistent with that of Turner and Johnston (1961). 5. humilis is an annual species and has the lowest chromosome number so far found in the genus (Longpre, unpublished). Although Hoffmann (1897) considered Sabazia to be a mem- ber of the subtribe Verbesininae, we agree with Turner and Johnston (1961) that the genus is better placed in the sub- tribe Galinsoginae. Turner and Johnston thought the genus to have affinities with Tridax; we are of the opinion that Sabazia is also close to Galinsoga, morphologically as well as chromosomally. Verbesina. The chromosome number of V. callilepis, n = 238 Rhodora [Vol. 65 18, has not been previously reported. The species is a mem- ber of sect. Pterophyton. Ximenesia (x — 17). X. encelioides (n — 17). Previous counts of n = 17 were reported by Carlquist (1954, as Ver- besina encelioides) and Turner and Ellison (1960). X. en- celioides is the only species so far counted in this small genus which Hoffmann (1897) considered to be a section of Ver- besina. Zexmenia. Jones (1905) recognized 42 species in this genus, the greater number of which is found in Mexico and Central America. Turner, Powell, and King ( 1962) consid- ered the genus to be multibasic with x — 10, 11, and 14. Since our count for Z. palmeri is an approximation, (n = ca. 17), no conclusion is warranted concerning its bearing on the basic chromosome numbers of the genus. HELENIEAE — Psilostrophe (x = 16). P. gnaphalodes (n — 82) seems to have diploid and tetraploid races, since Turner, Beaman, and Rock (1961) counted the species às n = 16. Our count for P. tagetina (n = 16) agrees with that reported for the same species by Jackson (1957b), and Raven and Kyhos (1961). Schkuhria (x — 10, 11). S. anthemoidea var. wislizeni (n — 10). Turner, Powell, and King (1962) reported this variety as tetraploid. The chromosome numbers listed in Table 1 for species of Dyssodia, Gaillardia, and Tagetes are consistent with those previously obtained for the same species (cf. Johnston and Turner, 1962 ; Raven and Kyhos, 1961). MUTISIEAE — Chaptalia dentata (n = 16). Chromosome numbers of 2n = 48 have been reported by Baldwin and Speese (1947) for C. nutans var. nutans and C. integrifolia, and Turner (1959) obtained n = 24 for C. nutans var. tex- ana. These authors considered x = 12 to be the basic num- ber of the genus. Although the few genera studied in the Mutisieae have high basic chromosome numbers, our count for C. dentata suggests that x — 8, rather than 12, is the basic number of Chaptalia. CICHORIEAE — Stephanomeria (x — 8). Our counts for S. pauciflora and S. tenuifolia, both with n = 8, agree with 1963] Mexican Compositae — De Jong and Longpre 239 the counts obtained by Stebbins, Jenkins, and Walters (1953) for the same species. SUMMARY This paper reports chromosome numbers in 74 collections, representing 57 taxa of Mexican Compositae; the chromo- some numbers of 21 of these are reported for the first time including first counts for the genera Achaetogeron and Chrysanthellum. The generic status of Achaetogeron is discussed on the basis of morphological and cytological evi- dence; the genus is considered as provisionally distinct. Basic chromosome numbers in addition to those previous- ly obtained by other authors are reported for the genera Astranthium, Calea, and Cosmos; the basic number of Chap- talia is thought to be x = 8 rather than 12 as previously reported. Chromosome counts of Haplopappus spinulosis subsp. scabrellus (n= 6), Leucelene ericoides (n=8), Psilostrophe gnaphalodes (n = 32), and Schkuhria anthe- moidea var. wislizeni (n = 10) differ from those published earlier for these taxa. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY, MICHIGAN STATE UNIVERSITY, EAST LANSING, MICHIGAN. LITERATURE CITED BALDWIN, J. T., JR. 1940. Cytophyletic analysis of Astranthium in- tegrifolium. Bull. Torr. Bot. Club 68: 615-617. , AND B. M. SPEESE. 1947. Chaptalia nutans and C. integrifolia: their chromosomes. Bull. Torr. Bot. Club 74: 283- 286. - BEAMAN, J. H., D. C. D. DE JONG, AND W. P. STOUTAMIRE. 1962. Chromoscme studies in the alpine and subalpine floras of Mexico and Guatemala. Am. Jour. Bot. 49: 41-50. BENTHAM, G. 1873. Notes on the classification, history, and geo- graphical distribution of Compositae. Jour. Linn. Soc. Bot. Lon- don 13: 335-577. CaRLQUIST, S. 1954. Im Doeumented chromosome numbers of plants. Madroño 12: 210. Gray, A. 1849. Achaetogeron. In Pl. Fendl. Mem. Am. Acad. 4: 72. GREENE, E. L. 1891. New or noteworthy species. Pittonia 2: 161-173. HoFFMANN, O. 1897. Compositae. In Engler, A., and K. Prantl. Die natürlichen Pflanzenfamilien 4: 87-891. 240 Rhodora [Vol. 65 JACKSON, R. C. 1957a. New low chromosome number for plants. Science 126: 1115-1116. JACKSON, R. C. 1957b. In Documented chromosome numbers of plants, Madrono 14: 111. JOHNSTON, M. C., AND B. L. TURNER. 1962. Chromosome numbers of Dyssodia (Compositae-Tagetinae) and phyletic interpretations. Rhodora 64: 2-15. JONES, W. W. 1905. A revision of the genus Zexmenia. Proc. Amer. Acad. 41: 143-167. LARSEN, E. L. 1948. New species of Achaetogeron. Jour. Wash. Acad. Sci. 38: 199-201. MONTGOMERY, F. H., AND S. YANG. 1960. Cytological studies in the genus Erigeron. Can. Jour. Bot. 38: 381-386. RAVEN, P. H., O. T. SorBRIG, D. W. KvHos, AND R. SNow. 1960. Chromosome numbers in Compositae. I. Astereae. Am. Jour. Bot. 47: 124-132. RAVEN, P. H., AND D. W. KvHos. 1961. Chromosome numbers in Com- positae. II. Helenieae. Am. Jour. Bot. 48: 842-850. Rouuins, R. C. 1950. The guayule rubber plant and its relatives. Contr. Gray Herb. 172: 1-73. SHERFF, E. E. 1955. Cosmos. In SHERFF, E. E., AND E. J. ALEXAND- ER. Compositae-Heliantheae-Coreopsidinae. N. Am. Flora II. 2: 1-190. SHINNERS, L. H. 1946. Revision of the genus Leucelene Greene. Wrightia 1: 82-89. SoLBRIG, O. T. 1961. Synopsis of the genus Xanthocephalum. (Com- positae). Rhodora 63: 151-164. STEBBINS, G. L., JR., J. A. JENKINS, AND M. S. WALTERS. 1953. Chro- mosomes and phylogeny in the Compositae, Tribe Cichorieae. Univ. Calif. Publ. Bot. 26: 401-430. THOMBRE, M. V. 1959. Chromosome numbers in some common flow- ering plants. Curr. Sci. 28: 206-207. TURNER, B. L. 1959. Meiotic chromosome counts for 12 species of Texas Compositae. Brittonia 11: 173-177. TURNER, B. L., AND W. L. ELLISON. 1960. Chromosome numbers in the Compositae. I. Meiotie chromosome counts for 925 species of Texas Compositae including 6 new generic reports. Texas Jour. Sci. 12: 146-151. TURNER, B. L., AND M. C. JOHNSTON. 1961. Chromosome numbers in the Compositae, III. Certain Mexican species. Brittonia 13: 64-69. TURNER, E. L., J. H. BEAMAN, AND H. F. L. Rock. 1961 Chromosome numbers in the Compositae. V. Mexican and Guatemalan species. Rhodore 63: 121-129. TURNER, B. L., M. PowELL.AND R. M. KING. 1962. Chomosome num- bers in the Compositae. VI. Additional Mexican and Guatemalan species. Rhodora 64: 251-271. HABENARIA CONOPSEA IN NORTH AMERICA Among some miscellaneous collections of orchids that were recently sent to me for determination from the Her- barium of the University of North Carolina is a sheet that contains a solitary flowering plant collected by W. C. Coker (s.n.) “In damp peaty soil, Abisco, Labrador, July 19, 1921." Although the inflorescence of this plant was unfortunately embedded in a generous amount of glue, it was possible to extricate enough of two flowers to make a dissection for study. This plant belongs to the Habenaria conopsea (L.) Benth. — H. odoratissima (L.) Franchet complex of Eura- sia. Although the plant is unquestionably dwarfed, and the solitary linear leaf is more like those attributed to typical H. odoratissima, in my opinion the floral characters place this plant in H. conopsea. In 1913, Britton and Brown included this species (as Gymnadenia conopsea (L.) R. Br.) in their Illustrated Flora. of the Northeastern United. States and. Adjacent Can- ada. (n. 553) with the following note: *. .. otherwise known only from the Old World, [it] has been collected at Litch- field, Connectieut [in 1887]. The flower has a broad 3- lobed lip and a slender spur much longer than the ovary." I have been unable to locate in any herbarium a specimen from the above locality. In 1952, in Native Orchids of North America, North of Mexico (p. 116), I considered the report by Britton and Brown to represent non-persistent waifs in Connecticut since the species, in the meantime, had not been observed again in Connecticut or anywhere else in this hemisphere. It now appears, however, that this orchid should be con- sidered as an element of the flora of North America, whether it be indigenous or as a naturalized species introduced from. Europe. It is most likely that the species occurs elsewhere in eastern North America and it has just been overlooked, possibly because it is either most uncommon or is to be found only in relatively isolated or inaccessible areas. A description of the solitary plant found in Labrador is as follows: 241 242 Rhodora [Vol. 65 Plant 12 em. tall, rigidly erect, glabrous; rootstock thick, palmately divided (in our area similarly to that of Habenaria viridis (L.) R. Br. var. bracteata (Willd.) A. Gray); stem with several tubular sheaths at the base and two narrow somewhat foliaceous bracts above; only one leaf present that arises from within the sheaths at base of stem, linear, narrowly obtuse, apparently conduplicate, 6.5 cm, long and 3 mm. wide; inflorescence with about 15 densely placed flowers that are not all open, 2 cm. long; floral bracts ovate-lanceolate, long- acuminate, up to at least 8 mm. long; flowers small; dorsal sepal broadly elliptic, obtuse, about 4 mm. long and 2 mm. wide; lateral sepals similar to the dorsal one but slightly oblique; petals elliptic, obtuse, about 3.5 mm. long and 2 mm. wide; lip suborbicular-flabellate in outline, broadly cuneate below the middle, shallowly 3-lobed above with the lobes bluntly rounded, about 4 mm. long; dise with 5 veins, with only the central vein unbranched and extending to the apex of the mid-lobe; spur cylindric, curved, 8-10 mm. long, nearly twice as long as the pedicellate ovary. DONOVAN S. CORRELL, TEXAS RESEARCH FOUNDATION, RENNER, TEXAS. A COMPANION VOLUME TO THE NEW BRITTON AND BROWN ILLUSTRATED FLORA.’ The northeastern part of North America is fortunate to have a newly prepared manual condensed from the large three volume Illustrated Flora by H. A. Gleason. The book under review was prepared by Arthur Cronquist and is based squarely on Gleason's revision of the earlier Britton and Brown Illustrated Flora, first published in three vol- umes in 1896, 1897 and 1898. A second Britton and Brown edition came in 1913 and the Gleason edition was published in 1952. Much of the material in the present volume goes back to the Gleason edition and in this respect it is not new. On the other hand, Cronquist has incorporated some of the results of more recent studies and has modified the descrip- tive matter and nomenclature accordingly. In relation to the area covered by Gleason's Illustrated Flora this manual pertains to an area reduced in the extreme northeast to exclude the Gaspé Peninsula and in the south- west to exclude the area south of the Missouri River in the state of Missouri. Thus, the area covered is somewhat less than that accounted for by Gray's Manual, ed. 8, and an otherwise obvious direct comparison of these two manuals for their included numbers of taxa cannot readily be made. However, it is satisfying to note that the nomenclature of these two manuals is closer to being the same than in earlier editions. This offers the hope that a high degree of stability of nomenclature is in sight for the plants found in north- eastern North America, at least. Some botanists perhaps will find the treatments of such difficult genera as Rubus and Crataegus to be over simplified and unsatisfactory for identifying their plants. But I am personally glad to see that the complications introduced by hybridization, polyploidy, aneuploidy and apomixis are frankly recognized and some deference is given to the fact 'Manual of Vascular Plants of Northeastern United States and Adjacent Canada by Henry A. Gleason and Arthur Cronquist. li & 1- 810. D. Van Nostrand Co., Inc., Princeton, N. J. 1963. $11.75. 243 244 Rhodora [Vol. 65 that the effects of these phenomena on the taxonomy of these two genera have not as yet been workd out. It is doubtful whether a sensible classification of Rubus and Crataegus can be produced without much more information than is now available. I have not used this manual to identify any plants and that is the only way to test its utility to the student and the field botanist. The book has a flexible cover and is of a size that will be handy in the field. Certainly, it is an important ad- dition to our literature on the vascular plants and the au- thors are to be congratulated for bringing it into published form. REED C. ROLLINS, GRAY HERBARIUM, HARVARD UNIVERSITY. FURTHER CONSIDERATIONS IN STYLOSANTHES (LEGUMINOSAE) ROBERT H. MOHLENBROCK Since publication of the writer's *A Revision of the Genus Stylosanthes" in 1957, additional collections and further study have resulted in an increase in number of species in the genus. Nooteboom (1961), after studying considerable Malaysian material, has pointed out some errors in this writer's revision. These are corrected in this paper. New illustrations of the loments of each species are provided. These have been prepared by Miss Miriam Hope Wysong. The species of stylosanthes have been divided into two sections from the time of Vogel (1838). It is unfortunate in choosing the type species for the genus from Swartz’ S. vis- cosa and S. procumbens ( — S. hamata (L.) Taub.) that I selected the latter, rather than S. viscosa which belongs to Vogel's Section Eu-Stylosanthes. Vogel's sectional names are valid, and should be recognized. The correct nomen- clature of the sections follows: SECTION STYLOSANTHES Sect. Hu-Stylosanthes Vog. in Linnaea 12:63. 1838. Asty- posanthes Herter, in Rev. Sudamer. Bot. 7:209. 1943. Sect. Astyposanthes (Herter) Mohl. in Ann. Mo. Bot. Gard. 44:327. 1957. None of the flowers subtended by an axis rudiment ; inner bracteole usually 1. Type: Stylosanthes viscosa Sw. SECTION STYPOSANTHES Vog. in Linnaea 12:68. 1838. Sect. Stylosanthes sensu Mohl. (1957). Each flower, or at least the lower, subtended by an axis rudiment; inner bracteoles usually 2. Type: Stylosanthes hamata (L.) Taub. Although Taubert described S. sundaica in 1891 in Sec- tion Styposanthes, I reduced it to S. humilis (a species with similar loments in Section Stylosanthes) because I could find no trace of an axis rudiment in the scanty material at 245 246 Rhodora [Vol. 65 my disposal Nooteboom (1961), with ample material for study, reports the presence of an axis rudiment, albeit caducous, so that S. sundaica should be recognized as a valid species belonging to Section Styposanthes. Five species of Stylosanthes are recognized in this paper which were not recorded in the author's revision in 1957. These are S. ingrata, S. mucronata, S. suborbiculata, S. suf- fruiticosa, and S. sundaica. KEY TO THE SECTIONS OF STYLOSANTHES A. Each flower, or at least the lower flowers, subtended by an axis rudiment; inner bracteoles usually 2 (1 in S. sericeiceps) ceee deveecccssscscccsscnccssoacsesensccessessconsosnscsssossscoscosssasoneneeesss Section Styposanthes AA. None of the flowers subtended by an axis rudiment; inner brac- teole usually 1l ............... een Section Stylosanthes KEY TO THE SPECIES OF SECTION STYPOSANTHES A. Bracts (8-) 10-15 mm. broad, with 11-21 conspicuous and usually broad, colored nerves. B. Fertile articulations of the loment often 2, the upper glabrous, the beak less than one-third as long as the upper articulation, glabrous or with a very few short stiff hairs; nerves of the bracteal sheath usually 15; stems usually with scattered bristles. Brazil, Venezuela. (Fig. 1) ............- 1. S. capitata BB. Fertile articulation of the loment 1, pilose, the beak only slight- ly shorter than the upper articulation, pilose; nerves of the bracteal sheath usually 21; stems pilosulous to villous. Brazil, Paraguay. (Fig. 2) ....... eene 2. S. bracteata AA. Bracts at most 10 mm. broad, with fewer than 15 nerves which are inconspicuous. C. Beak of the loment straight or only slightly curved; leaflets glabrous on both surfaces (rarely with marginal cilia); teeth of the upper stipules mostly longer than the sheath. Florida, Central America, Mexico, Bahamas, Cuba. (Fig. 3) .....595- NEMMMNMMMMNMMMMWWRMMMMRMMMWAMWRWAAWAMEMMEIEMMM 3. S. calcicola CC. Beak of the loment uncinate or circinate; leaflets usually pubes- cent, at least on the lower surface (occasionally glabrous in S. erecta of Africa and S. hamata of the West Indies) ; sheath of the upper stipules mostly longer than the teeth. D. Loment completely glabrous or with some pubescence on the beak only or occasionally on the nerves in S. mexicana. E. Stems sericeous or bristly; leaflets sparsely but conspicu- ously bristly-ciliate; loment often green, with both articu- lations usually fertile; inflorescence obovoid. Mexico, Ven- ezuela, Bolivia. (Fig. 4) .......... m 4. S. mexicana 1963] Stylosanthes — Mohlenbrock 247 EE. Stems glabrous or puberulent above; leaflets without brist- ly cilia; loment brown, with only one articulation usually fertile; inflorescence often narrow. West Africa. (Fig. 5) ERE E i ivaceevest els GENS RI 5. S. erecta DD. Loment pubescent on the body and usually on the beak. F. Stem and bracts bearing tuberculate-based hairs (some- times merely with tubercles); lower surface of leaflets usually villous with interspersed tuberculate bristles; beak of the loment usually shorter than the upper articulation (except in S. sundaica and sometimes in S. fruticosa, S. macrocarpa, and S. nervosa). G. Beak of the loment half to one-third the length of the upper articulation. H. Braets shortly scabrous-hispid; inflorescence usually nearly as broad as long; beak of the loment short- bristly. I. Leaflets often punctate beneath, obtuse; stem bearing short dark setae, the whole aspect of the plant dingy brown; fertile articulation usually 1, pubescent throughout. Brazil, Ecuador, Venezuela, Colombia, Hohvis. (Bie 8] osi TESI 6. S. scabra II. Leaflets not punctate beneath, acute; stem setose or puberulent to densely pilose, not dingy brown; fertile articulations often 2, pilosulous only on the ribs. Bahamas, Cuba, Colombia, Venezuela, Peru. (Fig. 7) TE bpbiinieasie E E E en 7. S. tuberculata HH. Bracts villous or with long tuberculate bristles; inflor- escence often 2-3 times longer than broad; beak of the loment usually rufous-pilose. J. Braets softly villous and rarely with tuberculate bristles; inflorescence 2-3 times longer than broad; loment 1.0-1.5 mm. broad, the beak rufous-pilcse. Ecuador, Peru, Galápagos Islands. (Fig. 8) .............. E EE E UNE M eos 8. S. sympodialis JJ. Bracts with tuberculate bristles; inflorescence about as broad as long; loment 1.5-2.5 mm. broad, the beak puberulent. K. Beak of loment 1.5-3.0 mm. long; stems evenly pubescent. Africa. (Fig. 9) ........ 9. S. mucronata KK. Beak of loment 3.5-4.0 mm. long; stems unilater- ally pubescent. Ceylon, India, (Fig. 10) ................ E Us quee S iuo pne rh oc 01 éco Ep EORR 10. S. fruticosa GG. Beak nearly equaling to exceeding the upper articulation. L. Beak of the loment strongly circinate. M. Beak longer than body of loment; stems not seri- ceous. East Indies. (Fig. 11) ........ 11. S. sundaica 248 Rhodora [Vol. 65 MM. Beak and body of loment about equal in length; stems sericeous. Honduras, Mexico. (Fig. 12) ........ sovsouosscseusacesesasgosssconsenesesibsoposbibsessanesesesoey 12. S. subsericea LL. Beak of the loment curved to strongly uncinate; brac- teal sheath and stem with tuberculate bristles but not sericeous. N. Upper articulation and beak combined 7.5-8.5 mm. long, the beak about equaling the upper articulation ; plants to 0.2 mm. long. Mexico. (Fig. 13) .............. MRMMMMMMMMMMMMMAMWMWMWMWWMWWMMWAAWWMWW——mRQUEMMS 13. S. macrocarpa NN. Upper articulation and beak combined 5.0-7.5 mm. long, the beak sometimes slightly shorter than the upper articulation; plants to 1 m. tall. O. Fertile articulations mostly 2; bracteal sheath usually bearing long tuberculate-based bristles; leaflets elliptic, obtuse to sub-acute, usually pub- erulent. P. Beak of loment 1.5-3.0 mm. long; stems evenly pubescent. Q. Terminal leaflet 8-10 mm. long. British Gui- ana. (Fig. 14) .................... 14. S. suffruticosa QQ. Terminal leaflet 15-25 mm. long. Africa. (Fig. 9)... 9. S. mucronata PP. Beak of loment 3.5-4.0 mm. long; stems uni- laterally pubescent. Ceylon, India. (Fig. 10) saccosssosnoonsenssasnsassnspssesninauswesesbsepsviguniaeé 10. S. fruticosa OO. Fertile articulation usually 1; bracteal sheath short-hispid to densely ciliate; leaflets usually oblanceolate, acute to acuminate, glabrous or oc- casionally hispid beneath. Venezuela, Peru, Bo- livia, Argentina. (Fig. 15) ............ 15. S. nervosa FF. Stem and bracts pilose, villous, or nearly glabrous, lacking tuberculate bristles; lower surface of leaflets pilose or appressed-villous or glabrous, never with tuberculate bristles; beak of the loment equaling or exceeding the upper articulation (except S. sericeiceps). R. Beak of the loment equaling or exceeding the upper articulation; pubescence of the bracteal sheath whitish; stem usually not pubescent throughout. Florida, Ba- hamas, Cuba, Central America, Colombia, Venezuela. (Fig. 16) cercsseseseonecesssnsrnecszesonasvsconceensoanasnssarsnt¥s 16. S. hamata RR. Beak of the loment about one-half as long as the upper articulation; pubescence of the bracteal sheath tan or rufous; stem often pubescent throughout. S. Loment pubescent throughout; leaflets glabrous above, 15-26 mm. long; bracteal sheaths tan-pilose. Venezuela. 1963] Stylosanthes — Mohlenbrock 249 (Fig 105) ue cutis =e 17. S. sericeiceps SS. Loment pubescent above, glabrous below; leaflets minutely pubescent above, 20-40 mm. long; bracteal sheaths rufous-pilose. Peru, Ecuador, Galapagos Islands. (bip. 8) 5... en eee 8. S. sympodialis KEY TO THE SPECIES OF SECTION STYLOSANTHES A. Beak of the loment minute, at most about one-fifth as long as the upper articulation, the fertile articulation 1. B. Fertile articulation shortly hairy throughout or rarely glabrous, terete, 2.5-5.0 mm. long. Eastern United States. (Fig. 18) ... Dc M M LM MM EN NE 18. S. biflora BB. Fertile articulation minutely tuberculate near the apex or rarely with a few scattered appressed white hairs, flattened, 1.5-3.0 mm. long. Bahamas, Central America, Mexico, South America. (Fir. 19) |..— nere 19. S. guyanensis AA. Beak of the loment from one-fourth as long to exceeding the upper articulation, the fertile articulations 1 or 2. C. Leaflets 0.5-2.0 mm. broad; inflorescence very narrow and elongate; beak of the loment strongly uncinate, often 2-3 times as long as the pubescent upper articulation. Brazil, Guianas. Gurt dae Es c SOPORE NR OE Balle cue 20. S. angustifolia CC. Leaflets 2-6 mm. broad; inflorescence usually capituliform, globose to ovoid or obovoid; beak of the loment various. D. Loment glabrous (occasionally pubescent only on the beak). E. Loment with 2 fertile articulations, the beak straight or uncinate; bracteal leaflets stalked. F. Beak straight or nearly so, about equaling the upper articulation, beak and upper articulation together about 5-6 mm. long; bracts, leaflets, and stems densely short- bristly. Brazil, Uruguay, Paraguay. (Fig. VI WW eeeede ru Mrd aD E Ru AMT IM Mc TE 21. S. leiocarpa FF. Beak uncinate, one-third as long as the upper articula- tion, beak and upper articulation together 3.0-3.5 mm. long; bracts and sometimes the leaflets and the stems with scattered long bristles, often glabrous. French Giana. (Fig. 22)... rene 22. S. cayennensis EE. Loment with one fertile articulation, often two in S. ingrata the beak strongly uncinate or circinate; bracteal leaflets sessile or subsessile. G. Loment conspicuously reticulate-nerved, about as broad as long (excluding the beak); inflorescence 4- to 8-flowered. Uruguay, Argentina. (Fig. 23) ........ 23. S. hippocampoides GG. Loment obscurely nerved, a little longer to twice as long as broad (excluding the beak); inflorescence 2- to 4- flowered. [Vol. 65 Rhodora 250 Fig. 1. Stylosanthes capitata, X 10. Fig. 2. Stylosanthes bracteata, X 10. Fig. 3. Stylosanthes calcicola, X 10. Stylosanthes — Mohlenbrock 251 1963] Fig. 4. Stylosanthes mexicana, »« 107 Fig. 5. Stylosanthes erecta, X 10. Fig. 6. Stylosanthes scabra, 10. Fig. 7. Stylosanthes tuberculata, X 10. [Vol. 65 Rhodora 252 5 s ET SS [c ` Ñ: Fig. 8. Stylosanthes sympodialis, X 10. mucronata, X 10. Fig. 10. Stylosanthes fruticosa, X 10. Fig. 9. Stylosanthes Fig. 11. Stylosanthes sundaica, X 10. 1963] Stylosanthes — Mohlenbrock 253 Tr re EUN z Si Fig. 13. Stylosanthes macrocarpa, X 10. Fig. 14. Stylosanthes suffruticosa, X 10. Fig. 15. Stylosanthes nervosa, X 10. 254 Rhodora [Vol. 65 18 v Fig. 16. Stylosanthes hamata, X 10. Fig. 17. Stylosanthes sericeiceps, X 10. Fig. 18. Stylosanthes biflora, X 10. Fig. 19. Stylosanthes guyanensis, X 10. 1963] Stylosanthes — Mohlenbrock 255 FS "e A AC pex 1 io 4 rs UV Y FATE US: R i nb, Vig My PHU Ed 20 2 | 23 Fig. 20. Stylosanthes angustifolia, X 10. Fig. 21. Stylosanthes leiocarpa, X 10. Fig. 22. Stylosanthes cayennensis, X 10. Fig. 23. Stylosanthes hippocampoides, 10. 256 Rhodora [Vol. 65 Fig. 24. Stylosanthes suborbiculata, X 10. Fig. 25. Stylosanthes ingrata, X 10. Fig. 26. Stylosanthes macrosoma, >< 10. 1963] Stylosanthes — Mohlenbrock 257 Bigs. 27. Stylosanthes viscosa, X 10. Fig. 28. Stylosanthes figueroae, X 10. YN PROS. STA RIS VA ^t s^ lm V tie 2 x: EXE ; Pai thee 3 — Fig. 29. Stylosanthes montevidensis, X 10. Fig. 30. Stylosanthes humilis, X 10. 258 Rhodora [Vol. 65 H. Beak of loment about equalling body in length; leaflets suborbicular. Somaliland. (Fig. 24) ........................ ee puesssbsbabusis taseubusyssvenbecavessssvusstazesdiuvenssoueassss 24. S. suborbiculata HH. Beak of loment about one-half as long as the body; ieaf- lets lanceolate to elliptic. I. Upper articulation and beak together 3.0-4.5 mm. long; leaflets lanceolate. British Honduras. (Fig. 25) ............ suneoubsusssessusesseesassascecstevesanasseseacdecsscnssesseneeceseessaea 25. S. ingrata II. Upper articulation and beak together 5.0-7.5 mm. long; leaflets elliptic. Paraguay. (Fig. 26) ................ ee DD. Loment pubescent (if nearly glabrous, the stem viscid). J. Beak shorter than or nearly equaling the upper articu!a- tion; stem often viscid. K. Stems with viscid hairs or short-hispid; leaflets usually punctate beneath; fertile articulations 1 or 2. L. Fertile articulations usually 2, the upper decidedly widest above the middle, the beak circinate, about one- third the length of the upper articulation. Bahamas, Cuba, Central America, Mexico, South America. (Fig. Pro RR 27. S. viscosa LL. Fertile articulation usually 1, broadest at or below the middle, the beak nearly straight or slightly uncin- ate, one-half to nearly equaling the upper articulation. Colombia. (Fig. 28) ...........1........ 28. S. figueroae KK. Stems without viscid hairs; leaflets not punctate; fertile articulation 1. Brazil, Paraguay, Uruguay, Bolivia, Argentina, Colombia. (Fig. 29) ...... 29. S. montevidensis JJ. Beak nearly 2-4 times longer than the upper articulation; stems not viscid. Central America, Mexico, Bahamas, Cuba, Colombia, Venezuela, Brazil. (Fig. 30) .......... 30. S. humilis SOUTHERN ILLINOIS UNIVERSITY, CARBONDALE. REFERENCES MOHLENBROCK, R. H. (1957): A Revision of the Genus Stylosanthes. Ann, Mo. Bot. Gard. 44:299-355. (1960): Recent Studies in the Leguminous Genus Stylosanthes. Rhodora 62:340-343. NOoOTEBOOM, —. (1960): In C. G. G. J. van Steenis’ Malaysian Papi- lionatae. I. Reinwardtia 5:446-450. TAUBERT, P. (1890): A Revision of Stylosanthes. Verh. Bot. Brand. 32:1-32. VOGEL, K. (1838): De Hedysareis Brasiliae. Linnaea 12:62-71. NOMENCLATURAL CHANGES IN THE FAMILY PALMAE S. F. GLASSMAN In order to standardize the nomenclature for the forth- coming revision of Dahlgren's *Index of American Palms" (Field Mus. Nat. Hist. Bot. 14: 1-456. 1936) the following changes are necessary. All names of Pyrenoglyphis, not already transferred, are being included under the genus Bactris. Since there already exists a Bactris microcarpa Spruce, based on another type (Spruce 31), it is necessary to give P. microcarpa Burret (based on Jenman 7725) a new name. Bactris Burretii Glassman, nom. nov. Pyrenoglyphis mi- crocarpa Burret, Fedde Rep. 34: 250. 1934. All species described under the genus Yuyba are being transferred to Bactris. Yuyba, as defined by Bailey (Gent. Herb. 8: 173. 1949.), is not a clear-cut genus. Its species apparently represent a specialized branch of the large genus Bactris in which the spathes and spadices have been con- siderably reduced in size and the spines have been reduced to small prickles, or are completely absent. Bactris dakamana (Bailey ex Maguire) Glassman, comb. nov. Yuyba dakamana Bailey ex Maguire, Bull. Torrey Club 75: 108, fig. 9. 1948. Bactris essequiboensis (Bailey ex Maguire) Glassman, comb. nov. Y. essequiboensis Bailey ex Maguire, l.c. 108, fig. 10. B. Gleasonii (Bailey) Glassman, comb. nov. Y. Gleasonii Bailey, Gent. Herb. 8: 174, fig. 71. 1949. B. Schultesii (Bailey) Glassman, comb. nov. Y. Schultesii Bailey, l.c. 174, fig. 71. B. Stahelii (Bailey ex Maguire) Glassman, comb. nov. Y. Stahelii Bailey ex Maguire, Bull. Torrey Club. 75: 106, fig. 8. 1948. B. trinitensis (Bailey) Glassman, comb. nov. Y. trinitensis Bailey, Gent. Herb. 7: 416. fig. 189. 1947. In 1941 and 1942, Gregorio Bondar described several new 209 260 Rhodora [Vol. 65 species under the genus Cocos. Since that time all of these taxa have been transferred to other genera. After studying the type specimens of these species at the Chicago Natural History Museum, I have decided to make the following changes: Syagrus Campos-Portoana (Bondar) Glassman, comb, nov. Cocos Campos-Portoana Bondar, Field Mus. Nat. Hist. Bot. 22: 460. 1942. Arecastrum Campos-Portoanum (Bondar) Hawkes, Arq. Bot. S. Paulo II: 175. 1952. This taxon is more closely related to species of Syagrus than Arecastrum. Although the seed cavity is slightly ir- regular, it is mostly smooth inside; and the seed itself is only slightly gibbous and has a pointed tip. In the genus Cocos, only one species, C. nucifera L., is being recognized at the present time. Syagrus Getuliana (Bondar) Glassman, comb. nov. Cocos Getuliana Bondar, Bol. Inst. Centr. Fom. Econ. Bahia 9: 35, fig. 8-9. 1941. Barbosa Getuliana (Bondar) Hawkes, l.c. 177. This species does not belong to Barbosa, a doubtful segre- gate of the genus Syagrus, because the fruit is without a distinct operculum and the endosperm of the seed is homo- geneous rather than ruminate. Syagrus X mataforme (Bondar) Glassman, stat. nov. Cocos mataforme Bondar, Field Mus. Nat. Hist. Bot. 22: 459. 1942. Syagrus mataforme (Bondar) Hawkes, l.c. 178. This taxon is apparently a hybrid between Syagrus vagans (Bondar) Hawkes, a trunkless palm, and S. coronata (Mart.) Becc. It grows alongside of these two species in the Municipio of Santa Teresinha in the state of Bahia. S. X mataforme has the trunk characters of S. coronata, but shows a similarity in leaves and inflorescences to the other species. The fruits seem to be intermediate in size (both length and diameter) and shape, and the endocarp is inter- mediate in thickness between the two species. Both S. X mataforme and S. vagans have short, dentate spines on the margins of the lower half of the petiole, whereas in S. 1963] Palmae — Glassman 261 coronata the petiole has long, narrow, flat spine-like ap- pendages along most of its length. Syagrus Ruschiana (Bondar) Glassman, comb. nov. Cocos Ruschiana Bondar, Bol. Inst. Centr. Fom. Econ. Bahia 9: 45, fig. 10-18. 1941. Arikuryroba Ruschiana (Bondar) To- ledo, Arq. Bot. S. Paulo II: 6. 1944. This taxon is more closely related to species of Syagrus. The endosperm is only ruminate externally and the petioles have smooth margins, whereas in the questionable genus Arikuryroba the endosperm is both externally and internally ruminate and the petiole margins are spiny. Syagrus Tostana (Bondar) Glassman, comb. nov. Cocos Tostana Bondar, Field Mus. Nat. Hist. Bot. 22: 458. 1942. Arikuryroba Tostana (Bondar) Hawkes, Arq. Bot. S. Paulo II: 175. 1952. The endosperm of this species is not ruminate and there- fore should not be included in the genus Arikuryroba. Syag- rus Tostana appears to be closely related to S. coronata. UNIVERSITY OF ILLINOIS, CHICAGO HYBRIDIZATION OF RUBUS HISPIDUS AND R. SETOSUS! FREDERIC STEELE AND A. R. HODGDON In an earlier paper (Hodgdon and Steele 1962) it was pointed out that Rubus allegheniensis is highly variable in its glandularity thus casting doubt on the validity of using the presence or absence of glands on the primocane as a character to separate species in the Section Alleghenienses. It is presumed that this and other characters such as the nature of the inflorescence, shape of leaf or leaflet, amount of pubescence, etc., may show a similar degree of intraspe- cific variability in many species of the Subgenus Eubatus. Part of the taxonomic confusion in the blackberries un- doubtedly has resulted from the failure of students to ap- preciate the range of variability in the better known and more widely distributed species. An added dimension of variation in Rubus Subgenus Eu- batus is produced by hybridization. With reference to cros- sing in the group, Bailey (1941) presented a singularly crit- ical commentary on those taxonomists such as Rydberg, Brainerd and Bicknell who had postulated or accepted hy- bridity as an important factor in the American blackberries. Bailey was unwilling to accept hybridization as of impor- tance in the group, for he stated (p. 7) “It is to be noted that even after all these years of assumption of miscellaneous crossing in the American brambles we do not yet have a satisfactory demonstration of the problem in nature . . ." and later on, “one may find in the field what are apparently real hybrids but they appear to be no more common or any more puzzling than in other large genera; and in such cases a scrap on an herbarium sheet would not be evidence. The hybridity postulate cannot explain the pomological blackberries.” In order to understand Bailey’s thinking about Rubus bet- ter it may be well to quote further. Toward the end of page ‘Published with approval of the Director of the New Hampshire Agricultural Experiment Station as Scientific Contribution No. 314. 262 1963] Rubus — Steele and Hodgdon 263 7 he gave the following criteria for recognizing hybrids, *(1) the presence of the two parents in the vicinity; (2) occurrence usually in small numbers, as if incidental or ex- ceptional to the main population; (3) characters that appear to belong only to the parents in various degrees of combina- tion." From the context in which he wrote, it is evident that Bailey's rejection of hybridity was a reaction to the common but unfortunate taxonomic practice of calling plants hy- brids without sufficient evidence, but his thinking about hy- brids is hardly acceptable as a generalization in the light of present knowledge. Hybridization between species may oc- cur infrequently or not at all in many groups, although it is a frequent process in many genera and it is certainly com- mon in Rubus Section Eubatus in which hybrids may be found between many of the common well-known species whenever these occur together in some abundance in dis- turbed areas. Fernald (1950) had quite aecepted hybridization in the blackberries. But it is one thing to accept hybridization as an operating mechanism in a group and another to apply it successfully in working up the taxonomy of that particular group. Fernald's procedure in the Manual was to retain all but the most obviously overlapping taxa. He would have had to be highly arbitrary to do otherwise for all of the hun- dreds of "species" that had been described were set apart on the basis of supposed morphological differences that seemed to make each distinct and in the majority of cases there was no available information to show that one was any better as a species than another. He did perform the valuable serv- ice of eliminating a very large number of very dubious taxa in the group. Because he did retain those taxa which he be- lieved to be most distinct morphologically, his treatment does provide an excellent point of departure for further work. In the “Manual” therefore is to be found the nucleus of any sound taxonomic study of blackberries in North- eastern America. Gleason & Cronquist (1963) have accepted the fact that 264 Rhodora [Vol. 65 hybrids are common in blackberries; they have also reduced the number to a very few and at times arbitrarily selected “collective species" some of which are too broad to represent biological realities and others are trivial and based on in- consistent distinctions. While attractive to the non-specialist who wants to give some name to his collections, their treat- ment does have shortcomings. If an enormous assemblage of taxa is to be reduced to few, it is only by an intensive and specialized approach that the accurate dilineation of all the basic species can be accomplished. This paper presents evidence that hybrid and introgres- sing populations between 2 species of Rubus Section Eubatus do occur rather frequently and further that three widely recognized, so-called, species exhibit taxonomic character- isties that are shared by various of these evident hybrids. Thus perhaps for the first time we present evidence from field studies that hybridization is an important factor in blackberry taxonomy. Quite early Brainerd and Peiterson (1920) suggested hy- bridization as an important factor in the Rubus problem, and proposed a rather elaborate scheme of hybrids. Al- though they made a number of collections and did some experimental work, they published no direct evidence of hybridization and their results were not accepted by Bailey or Fernald. Nevertheless an examination of specimens in the Pringle Herbarium at the University of Vermont has con- vinced us that some of Brainerd and Peiterson's conclusions were valid. However, each supposed hybrid must be con- sidered separately and appropriate tests applied. We have used the following criteria for hybridization: 1. that the putative parents be present in the area ; 2. that the supposed hybrids occupy a disturbed environment; 3. that the sup- posed hybrids show some degree of intermediacy between the two parents. Blackberries frequently grow in disturbed environments, such as gravelly edges of roads or lumbered areas so condition 2 is easily satisfied. It was realized that many of the populations would show introgression with one of the parents. 1963] Rubus — Steele and Hodgdon 265 Although we have similar evidence that many species of Rubus hybridize, this paper will be concerned with R. his- pidus and R. setosus. R. hispidus is typically a blackberry of dry, open, or shady habitats with rather poor soil. It also grows in boggy areas. It is most easily recognized by its prostrate habit, lack of sharp prickles, and three lustrous coriaceous primocane leaflets. R. setosus is typically an in- habitant of alluvial meadows and poorly drained soil, but will grow in a variety of open habitats. It is a low erect blackberry with numerous soft bristles and poor fruit. The first indication that these species might hybridize a- rose when the senior author noticed a mongrel population of blackberries along the gravelly edge of a newly construct- ed state road. A number of clones were scattered in the bare gravel, each somewhat different from the others, and all clearly having existed for only two years. Both R. his- pidus and R. setosus grew in the area so it seemed quite pos- sible that this was a hybrid population. Collections were made and investigations started on the possibility of fre- quent hybridization and introgression of these two species. To analyze the situation it was decided to use the hybrid index method as described by Stebbins (1950). For each colony of blackberries investigated, 8 plants, consisting of both primocane and floricane from the same rootstock were collected. Each plant was tagged at the time of collection with data as to the growth habit. It was noticed that new primocanes often started off as erect plants, and may not have developed trailing tendencies until after flowering time so no collections were made until after the first of July. In the case of a long primocane, the midportion with leaves that seemed to be typical of the whole plant were secured; for the floricane, care was taken to secure a vigorous branch of the inflorescence with remains of flowers. Fruits often do not develop. Samples were taken from the colony at regular intervals in the case of a roadside population, or else in such a way as to indicate the range of variation. In the preparation of the index all characters except growth habit were rated 0, 1 or 2. 266 Rhodora [Vol. 65 TAXONOMIC CHARACTERS USED IN STUDY GROWTH HABIT — varies from prostrate through low and high doming, to erect. This is an excellent character and it was rated from 0 to 4. Good R. hispidus is always prostrate, but in certain situations may trail on vegetation to some height above the ground. Floricanes of R. setosus are often reclining or lodged because of the weight of snow; if the primocane is erect the plant is considered erect. It should be emphasized that growth habit can only be accurately rated if notes are made at time of collection. A number of other- wise excellent herbarium specimens lack this essential data. ARMATURE PER DECIMETER — a good character but some- what subject to environmental modification. The count in- cluded prickles, bristles, and glands. Although R. setosus typically has a large number of bristles, they may be quite sparse near the base of the stem; their number is also af- fected by shade. LENGTH OF THE LONGER BRISTLES — runs from .25 cm. or less for R. hispidus to .4 cm. or greater for R. setosus. It is not always consistent. CHARACTER OF PRIMOCANE LEAF — R. hispidus coriaceous and lustrous under good light; R. setosus dull and char- taceous. NUMBER OF PRIMOCANE LEAFLETS — R. hispidus typically 3; R. setosus 5 or sometimes 3 with 2 of them partly divided. LENGTH OF CENTRAL PRIMOCANE LEAFLET — R. hispidus noticeably smaller, usually less than 5 cm.; R. setosus 6.5 cm. to 11 em. This character is subject to some environ- mental modification. Position of broadest part of leaf was expressed as the ratio of the distance of the broadest part of the leaf to the center of the leaf divided by half the length. This ratio varies from .1 to .3 for R. hispidus; thus the leaves tend to be obovate. In R. setosus the ratio is usually 0 with leaflets broadest at the middle. CHARACTER OF TEETH OF LEAFLET — R. hispidus has teeth rounded with an abrupt point; R. setosus has teeth triangu- lar or accuminate. 1963] Rubus — Steele and Hodgdon 267 LEAF TIP OF CENTRAL PRIMOCANE LEAFLET — R. hispidus is usually rounded and abruptly pointed ; R. setosus typically has an accuminate tip. DIAMETER OF PRIMOCANE — R. hispidus 2mm. or less; R. setosus usually 3 mm. or more but occasionally less in shady situations. GLANDULARITY OF FLORAL AXIS — R. hispidus has either no glands or rather sparse glands of uniform length; R. setosus has abundant glands of varying length. Table I. List of characteristics used in Hybrid index with values assigned. R. hispidus R. setosus Growth habit |Prostrate = 0 Doming = 2 Arching to erect — 4 Armature per |0-2000 = 0 2001-2999 = 1 3000 or more = 2 dm. Bristle length |3 mm. or less = 0|3.1-3.9 = 1 4 mm. or more — 2 Character of |Lustrous Intermediate — 1 |Dull Chartaceous leaf coriaceous — 0 =2 No. primocane |3 — 0 3-5 — 1 —2 leaflets Length of leaflet/0-5 cm. = 0 5-6.5 — 1 6.5 or more — 2 Mid ratio 3-.1 = 0 0=2 Teeth R. hispidus Intermediate = 1 |R. setosus type — 0 type = 2 Tip of leaflet Abruptly pointed —— Acuminate = 2 =0 Stem-diameter |2 mm. = 0 —— 3 mm. or more — 2 Glands in infl. |0-few = 0 —— Abundant — 2 Discussion In order to provide standard material for effective com- parisons, we made collection s from characteristic Rubus hispidus and R. setosus colonies. The results of the analysis of these collections provide the information about the two species shown in figure 1. A total of 11 hybrid or intro- gressing populations were analyzed of which only 6 could be shown in the diagram. In general these were chosen to show the range of intermediacy in these populations. It is 268 Rhodora a-MADISON 724 62 3 2 l O 5 10 15 20 THORNTON 7 I2 62 [Vol. 65 SANDWICH 7 15 62 CHOCORUA 715 62 _PITTSFIELD 95 62 -MADISON 7 24 62 Tee MAARA SARA "ma SANDWICH 8 4 62 _SANDWICH 8 I5 62 b kh 1 LOUDON 9 5 62 SANDWICH 8 15 62 p TAMWORTH 8 22 61 _CHOCORUA 8 II 6I TRENT „TAMWORTH 7 I6 61 d Figure 1. -CHOCORUA 8 3 6I L : NORTH CONWAY | | SANDWICH 9 7 62 GROVETON 7 7 6I | | J Frequency Distribution Diagram: Rubus hispidus popu- lations at left. R. hispidus X R. setosus in center and R. setosus at right. The hybrid index values are on the horizontal axis, the numbers of individuals measured on the vertical axis (scale given in upper left of diagram). 1963] Rubus — Steele and Hodgdon 269 to be noted that in some of the populations certain plants may grade into one or both parents though the majority fall into a position between them. The limit of anything we would tend to call R. hispidus would be not more than 7 in value while R. setosus would be not less than 16 on the scale. During this investigation it has become apparent that several controversial “species” are hybrids or introgressants intermediate in character between Rubus hispidus and R. setosus. Those most clearly intermediate are R. adjacens Fernald, R. jacens Blanchard and R. trifrons Blanchard. Two others, R. spiculosis Fernald and R. tholiformis Fernald display many intermediate features and may belong in this category or may have a more complicated heredity. In the folders of R. adjacens in the Harvard Herbaria one finds a varied assortment of material varying from specimens close to R. setosus to strongly intermediate plants. The same in general holds true for R. jacens while R. tri- frons has more frequently been confused with R. hispidus. To lend clarity to this problem we have investigated the type specimens of these three and have graded each according to our hybrid index values. Although information is lacking about the habit of growth of the type of R. trifrons, it is possible nonetheless, to get a fairly good idea of the probable relationship of these three from the total values. The hybrid index value of 11 for R. adjacens shows it to be clearly inter- mediate between R. hispidus and R. setosus as does the value of 10 for R. jacens. R. trifrons, as might be suspected from its frequent confusion with R. hispidus, has a slightly lower value of 9. We conclude that hybridization and introgression of R. hispidus and R. setosus occur frequently especially in dis- turbed areas giving rise, in addition to the above mentioned recognized taxa, to various other forms difficult to classify. We wish to thank the curators of the Gray Herbarium, the Arnold Arboretum Herbarium and that of the New Eng- land Botanical Club for permission to examine their col- lections of Rubus. We also extend thanks to Dr. H. W. Vogelmann and L. Charette for the loan of specimens and 270 Rhodora [Vol. 65 for permitting us to examine the Rubus collections in the Pringle Herbarium at the University of Vermont. Voucher specimens are deposited in the herbarium of the University of New Hampshire. ST. MARY’S-IN-THE-MOUNTAINS, LITTLETON, NEW HAMPSHIRE, AND UNIVERSITY OF NEW HAMPSHIRE, DURHAM. Literature Cited BAILEY, L. H. 1941. Species Batorum. The Genus Rubus in North America. Gent. Herb. 5: 6-8. BRAINERD, EZRA and A. K. PEITERSEN. 1920. Blackberries of New England — Their classification. Vt. Ag. Exp. Sta. Bull. 217: 84 p. FERNALD, M. L. 1950. Grays New Manual of Botany, 8th ed. New York: 822. GLEASON, H. A. and ARTHUR CRONQUIST. 1963. Manual of Vascular Plants. Princeton, N. J.: 810 p. Hopepon, A. R. and FREDERIC STEELE. 1962. Glandularity in Rubus Allegheniensis Porter. Rhodora 64: 161-168. STEBBINS, G. L. 1950. Variation and Evolution in Plants. Columbia: 265-270. HEDYOTIS AUSTRALIS IN GEORGIA KENTON L. CHAMBERS The recently completed merger of Houstonia with Hedy- otis (W. H. Lewis, Rhodora 63: 216-223, 1961, and other authors) provides yet another epithet for the hapless Small Bluet whose name is confused among Houstonia patens Ell., H. minima Beck, and H. pusilla Schoepf in the two principal manuals of the northeastern flora. In Hedyotis, the species must be called H. crassifolia Raf., and current opinion favors the placement of all three of the above names in synonymy (see, for example, L. H. Shinners, Field & Lab. 18: 100, 1950, and Spring Flora of the Dallas-Fort Worth Area, 359- 360, 1958). On March 23, 1963, I found Hedyotis crassifolia, growing on grassy roadsides in Ocmulgee National Monument, Ma- con, Georgia, and with it an “obviously different" species Figure 1. A. Hedyotis australis Lewis & Moore (K. L. Chambers 1973); B. Hedyotis crassifolia Raf. (K. L. Chambers 1974); both X 2/8. Plate 1287 271 272 Rhodora [Vol. 65 having much smaller, white corollas and a distinctive growth form (fig. 1). The white-flowered species is Hedyotis aus- tralis Lewis & Moore (Southwestern Naturalist 3: 208-211, 1959) based on H. crassifolia var. micrantha Shinners. The chromosome numbers of H. crassifolia and H. australis have been reported to be 2n = 16 and 2n = 32, respectively (W. H. Lewis, Amer. Journ. Bot. 49: 855-865, 1962) ; and Lewis and Moore (op. cit.) report five distinctive characteristics by which the two species can be distinguished. A clear mor- phological hiatus was observed between the two populations growing intermingled at Macon, which supports the as- sumption by Lewis and Moore that a barrier to crossability exists. In view of the intraspecific polyploidy reported in Hedyotis caerulea (L.) Hook. (W. H. Lewis and E. E. Ter- rell, Rhodora 64: 313-323, 1962) and other species, one might not wish to predict that H. australis and H. crassifolia will be found to differ consistently in chromosome number throughout their area of geographical overlap. However, an additional barrier to interbreeding is provided by the pre- dominant self-fertilization which must occur in these spe- cies, both of which have monomorphic flowers with the stig- ma closely surrounded by the anthers. Hedyotis australis is said by Lewis (op. cit., 1962) to be the most highly evolved species in subgenus Edrisia ( = Houstonia) ; in the speci- mens from Georgia the corolla-tube is only 1.3 mm. long, about 2/3 the length of the sepals, and the anthers are in- serted 0.3 mm. from the base of the tube. In habit, the plants differ from H. crassifolia in their less divaricate branching, their erect rather than nodding floral buds, and their well distributed cauline leaves. The Ocmulgee population of H. australis represents a disjunction three states eastward from the range reported for the species by Lewis and Moore (Louisiana to Arkansas and Texas). However, as the plant becomes familiar to collectors, it will probably be found through the intervening area. Collection data for this report are as follows: K. L. Chambers 1973; grassy roadside flat, 100 yards east of the Great Temple Mound, Ocmulgee National Monument, Bibb 1963] Drosera Filiformis — Linke 240 County, Georgia; associated species, Krigia oppositifolia Raf., Viola rafinesquii Greene, Phacelia dubia (L.) Trel. Specimens are to be deposited at the herbaria of Oregon State University, Southern Methodist University, and Har- vard University. DEPARTMENT OF BOTANY, OREGON STATE UNIVERSITY, COR- VALLIS. DROSERA FILIFORMIS IN CONNECTICUT. — In a dry bog in the state leased area “Latimer Brook" just east of Route #161 in East Lyme, in a space of approximately 500 square feet grow at least seventy-five plants of Drosera. filiformis Raf. Among the sundews are Drosera intermedia, Vaccinium macrocarpon, a Rubus of the type hispidus and the moss, Polytrichum commune in two comparatively bare spots in contrast to the grassy or shrubby cover of most of the bog. Other plants in the immediate vicinity are Iris prismatica, Myrica gale, Kalmia angustifolia, Calopogon pulchellus, Chamaedaphne calyculata var angustifolia, Gaultheria pro- cumbens and Vaccinium angustifolium var. laevifolium. On June 23, 1963, the Connecticut Botanical Society held one of its field trips at this site and at least 40 people saw the plants. At this meeting two plants were taken and trans- ferred to the eastern part of Guilford, Conn. in a bog where it is hoped they may persist for some time. Mr. J. J. Neale reports that there is a sheet of this species in the herbarium of the Connecticut Botanical Society, from the collection of Anna E. Carpenter, labelled “July 29, 1905, South Glastonbury, Mrs. F. W. Starmer". Connecticut botanists who have seen this specimen have thought it un- likely to have been native there. Though South Glastonbury is some thirty miles northwest of East Lyme, this alleged record now perhaps appears a little less improbable. Specimens have been deposited in the Gray Herbarium and the herbaria of the New England Botanical Club, the Connecticut Botanical Society, and the University of Con- necticut. — WILLIAM R. LINKE JR., NEW LONDON. EPILOBIUM X WISCONSINENSE, HYBR. NOV. DONALD UGENT The following hybrid between Epilobium coloratum Bieh- ler and E. glandulosum var. adenocaulon (Haussk.) Fernald was discovered during the preparation of a treatment of the Onagraceae of Wisconsin (Ugent, 1962: 105-110). The recent series of “Preliminary Reports on the Flora of Wis- consin" reflects the efforts of Hugh H. Iltis and his students to achieve a Flora of Wisconsin, a goal set by the late Nor- man C. Fassett. Epilobium X wisconsinense, hybr. nov. Caules tetragoni, laxe dumosi-ramosi, constanter patuli- glandulosi et minute pilosi. Folia caulina petiolata, laminis longioribus 3-6.5 em. longis, 8-21 mm. latis, anguste lanceo- latis, apice acuminatis, margine excisiserrulatis vel ser- rulatis; petiolis 1-3 mm. longis. Flores solitarii, supra- axillares. Gemmae orbiculatae, sepalis apice ereflexis, aut reflexis, saepe intermediis. Calyces 3-4.2 mm. longi. Petala purpurea vel lilacina, 3.5-6 mm. longa. Capsulae maturae 1-3.5 (-4.5) em. longae. Semina maxime sterilia, quibus maturis 1-1.2 mm. longis, 0.2-0.5 mm. latis, striati-papil- losis ; comae fuscae vel fulvae, quibus immaturis albis. HOLOTYPUS: Polk Co., Wisconsin, edge of road in wet swamp, not common, West Sweden Tp. sec. 56, Johnson s. n. (WIS). Stems loosely bushy-branched, spreading glandular pube- scent as well as minutely pilose, the sides as well as the angles pubescent. Larger leaves 3-6.5 cm. long, 8-21 mm. wide, narrowly lanceolate, acuminate, closely and irregularly serrulate, on petioles 1-3 mm. long. Flowers solitary in the upper leaf axils; buds rounded, or with divergent sepal-tips, often somewhat intermediate. Calyx 3-4.2 mm. long. Petals purple or lilac, 3.5-6 mm. long. Mature capsules 1-3.5 (-4.5) em. long. Seeds mainly aborted, the mature ones 1-1.2 mm. long, 0.2-0.5 mm. wide, striate-papillose; the coma brown or tawny, white in inmature capsules. Southern and northwestern Wisconsin, very sporadic in 274 1963] Epilobium — Ugent 215 disturbed sedge-goldenrod peat marshes, spring-saturated sedge meadows, and along wet swampy roadsides, river banks, and railroad tracks. Flowering from early July to early September, and fruiting from mid-August to early September. This hybrid resembles Epilobium coloratum in the sharply E. COLORATUM SEEDS ALL NORMAL CAPSULE ABORTED AND NORMAL SEEDS EPILOBIUM WISCONSINENSE F1 g. | ILLUSTRATION OF TYPE SPECIMEN 276 Rhodora [Vol. 65 WISCONSINENSE M lou i Mf am d N b Ng COLORATUM Fig. e STEMS, BUDS, AND SEEDS OF SOME WISCONSIN EPILOBIUMS 1963] Epilobium — Ugent 277 and irregularly serrulate leaves, the bushy-branched in- florescence (with the lower branches longer and less crowded than usual), the papillose nature of the seeds, and in the brown or tawny hairs of the coma. The pubescence of the stem and the striate character of the seeds are features definitely associated with Epilobium glandulosum (Fig. 2). Unlike either parent, the hybrid has short capsules contain- ing many aborted seeds (Fig. 1). The shortness and slen- derness of the hybrid capsules are no doubt due to the abnormally high amount of seed abortion, which on different plants may vary from 28% to 96% (Fig. 3). NUMBER OF SEEDS % SPECIMEN CAPSULE ABORTED | MATURE | ABORTION 1 49 10 | Hale sine loc. 2 83 14 | 85 3 55 7 $—--1 .[.A10--3 T Bachman & Patrick Dunn Co. 2 84 b 94 3 98 6 ED NEC EN ZZ. Burger Jefferson Co. 2 20 45 2 3 16 60 | Johnson Polk Co. un UE ee Nu ree | od. Benner Polk Co. n 1 | ECT 4 E 95 TOTAL | 719 197. 3 — 98 FIGURE 3. Seed abortion in Epilobium X wisconsinense. The scatter-diagram (fig. 4), illustrates the intermediate shape of the hybrid leaf, as well as the unique short length of its capsules. Epilobium glandulosum var. adenocaulon is frequently as- sociated with E. coloratum, and not uncommonly plants of both taxa have been collected together and mounted on the same herbarium sheet. When comparing the phenology of these plants, one can observe that when Epilobium coloratum is in flower, E. glandulosum var. adenocaulon is generally in fruit, var. adenocaulon flowering about two and one half weeks earlier (fig. 5). This seasonal isolation may be im- portant in keeping these taxa relatively distinct. It should be noted that our hybrids occur in an area where the ranges of both parent species overlap. Epilobium glandu- 278 Rhodora [Vol. 65 EPILOBIUM FIG. 4 = € SEEDS MOSTLY ABORTED Q f O SEEDS MOSTLY FERTILE c O SEEDS STRIATE n'a Z O- SEEDS STRIATE - PAPILLOSE O-SEEDS PAPILLOSE ii O COMA WHITE oO -O COMA TAWNY I COLORATUM O PUBESCENCE GENERAL «r 9 PUBESCENCE IN LINES ~~ n'a ~~ o -o— H < 9 * a "i e- = J -e- e- P. + oe Ux + -e æ oo ° -e- -e e WISCONSINENSE O - mL fe) O O Q O GLANDULOSUM © al var. ADENOCAULON 2 3 4 5 6 CAPSULE LENGTH CM. —*— COLORATUM —o— GLANDULOSUM var. ADENOCAULON 2 -2-- WISCONSINENS g Q p-s E = O a 0 & WwW Oo a ul Or e = 2 Z Or L l L L L L L l 1 1 l L JUNE JULY AUGUST SEPTEMBER OCTOBER F | G . 5 FLOWERING DATE 1963] Epilobium — Ugent 219 losum var. adenocaulon has a wide northern distribution, ranging from Alaska to Newfoundland, south to Delaware, northern Illinois, and, at higher elevations along the Rocky Mountains, to Colorado. Epilobium coloratum, a species with eastern and southeastern affinities, extends from Geor- gia to Kansas north to Minnesota, and across northern Wis- consin to southern Quebec (Fernald, 1950; Gleason, 1952). Epilobium coloratum and E. glandulosum var. adenocaulon are both known to hybridize with other species. The fol- lowing hybrids are reported in Index Kewensis and the Gray Herbarium Card Index: Epilobium coloratum X Epilobium lineare Muhl., Epilobium coloratum X Epilobium commuta- tum Haussk., and Epilobium adenocaulon Haussk., X Epi- lobium canadense Levl. SPECIMENS EXAMINED. WISCONSIN: sine loc. [ca. 1860?] Hale s. n. (wis). Dunn. Co.: railroad tracks, Menomonie, Bachman & Patrick 7-10 (Wis). Grant Co.: Potosi, Davis s. m. (WIS); along streams, Boscobel, Sylvester 13590 (MIL). Jefferson Co.: disturbed sedge- goldenrod peat marsh, Town of Sullivan, sect. 13, Burger 152 (wis). Lafayette Co.: Fayette, Cheney s. n. (Wis). Polk Co.: St. Croix Falls, Baird s. n. (WIS); river bank, 8 mi. north of St. Croix Falls, Benner 363 (MINN); edge of road in wet swamp, not common, West Sweden tp. sec. 36, Johnson s. n. (WIS). Walworth Co.: spring-saturated sedge meadow, Delavan, Wadmond 17439 (MINN, WIS). I wish to thank Albert M. Fuller and Emil P. Kruschke, Milwaukee Public Museum (MIL), Thomas Morley, University of Minnesota (MINN), and Peter J. Salamun, University of Wisconsin-Milwaukee, for the loan of specimens; Mrs. Janice Paynter for her excellent drawings; the Wisconsin Alumni Research Foundation and the “| J Davis Fund for financial support; and especially Hugh H. Iltis for his advice and assistance. REFERENCES FERNALD, M. L., 1918. Some American Epilobiums of the section Lysimachion, Rhodora 29: 29-39. .1950. Gray's Manual of Botany. ed. 8. Am. Book Co. New York. GLEASON, H. A., 1952. The New Britton and Brown Illustrated Flora. Lancaster Press, Lancaster, Pa. Vol. 2. UGENT, D. 1962. Preliminary reports on the flora of Wisconsin No. 47. The orders Thymelaeales, Myrtales, and Cactales. Trans. Wis. Acad. 51: 83-134. HERBARIUM, DEPARTMENT OF BOTANY UNIVERSITY OF WISCONSIN THE IDENTITY OF BUMELIA LACUUM SMALL' OLGA LAKELA The author's interest in the genus resulted from collecting and identifying representative floristic components in the white sand formation of central Florida. Bumelia was commonly encountered in the scrub association of Persea humilis, Garberia heterophylla, Sabal Etonia, Quercus Chap- manii and other scrub species with the overhead of Pinus clausa and Ceratiola ericoides. B. lacuum was described by Dr. Small from the white sands of Highlands County. In studies of North American species of Sapotaceae, Dr. A. Cronquist regarded the taxon conspecific with B. tenax (L.) Willd.,? described from South Carolina. It was discovered there by Dr. Alexander Garden who dispatched specimens to Linnaeus. Despite limited knowledge of the genus, the author feels justified, as a result of her observations, placing on record certain characters of the two species. Pubescence has been found to be a reliable character in separating specific lines, (Wood & Channel).* Individual hairs have a central attach- ment, T-beam fashion. When the hairs are thinly acicular, straight and sericeous, the indumentum is appressed with a few diverging ends, as in the leaves of B. tenax. The dense, pannose indumentum of B. lacuum, mostly obscuring the lateral veins and the apical area of the midrib, at least when young, is due to long, sinuous implexed hairs with a ten- dency to curl. The overall pubescence in B. lacuum seems slightly coarser and less sericeous than that of B. tenax. In contrast to the pubescence, the pedicel-length has been 1Contribution No. 3. Botanical Laboratories of University of South Florida. The author is grateful for the loan of specimens from the Univer- sity of Georgia Herbarium. 2SMALL, J. K. Man. S. E. Fl. 1034. 1953. 3CRONQUIST, A. Studies in the Sapotaceae III. Jour. Arnold Arb. 26: 453-471. 1945. :Woop, C. E. and R. B. CHANNEL. The Genera of the Ebenales in the Southeastern United States. Jour. Arnold Arb. 41: 1-10. 1960. 280 1963] Bumelia — Lakela 281 found to be highly variable and unreliable (Wood & Channel, Op. Cit.). In the given species the variations in the length of the pedicels are consistent in comparable stages of dev- elopment. Seemingly short and thick, they are 2-5 mm. long Plate 1288 Bumelia lacuum Small. a. Surface of sand Josephine Creek, Highlands County, Florida. 282 Rhodora [Vol. 65 in B. lacuum, when the corolla begins to show between the spreading sepals. (Lakela, 25096 & 25120). When mature, they are 8-9 mm. long with fruits 9-12 mm. in length. The pedicels of B. tenax are slenderly clavate, 8-11 mm. long. (A. H. Curtiss, 5678 U. GA.). A part of this collection is a short twig in winter condition with two pedicels, 12 mm. long with calyces denuded of fruit. Mature fruit has not been available for study. In flowers of the two species there is a difference in the texture of the corolla tube. In B. lacuum it is membranace- ous, semi-opaque, obscuring the finer vascular traces. The median lobe in the fully expanded corolla is clawed ; flanking the claw on each side, the crescentic margins of the lateral lobes affect a small opening. The margins of all the lobes are more or less denticulate and narrowly translucent. Thin texture and broad translucent margins of corolla lobes are characteristic of the flowers of B. tenax. The terminal lobe is sessile. In support of this observation made on a limited amount of material reference is made to Sargent, Silva of North America, Tab. CCXLVI, illustrating the B. tenax flower.^ Dr. Clark segregated B. lacuum from B. tenax on the basis of shrubby versus tree-like habit.* Suffice it to state that the shrub is stoloniferous. The photograph shows five young plants with stolons radiating from the underground base of the parent shrub. (James D. Ray, Jr. and O. Lakela, 11054, 25 June, 1962. White sand scrub, Josephine Cr., east of US 27, Highlands Co., Florida). The colonial habit of B. lacuwm was not observed in a plant of B. tenax, 2.5 m. high, in coastal sands of northern St. Lucie County. The large plant had a single stem and young plants beneath its crown were proved to have originated from seeds. These observations seem to validate the specific status of B. lacuum. UNIVERSITY OF SOUTH FLORIDA, TAMPA *SARGENT, C. S. Silva 5: 167. 1893. "CLARK, ROBERT B. A Revision of the Genus Bumelia in the United States. Ann. Missouri Bot. Gard. 29: 155-182. 1942. 1963] Book Review — Hodgdon 283 JoHN LYON, NURSERYMAN AND PLANT HUNTER' Not a little of the pleasure that most botanists and horti- culturists feel for their work derives from the richness of human association with plants. That a certain kind of plant was used medicinally 150 years ago, for example, surrounds it with an added interest. Here is an account of the travels, business transactions, lists of plants collected and botanical and horticultural con- tributions of the Scotchman John Lyon during the years 1799-1814 in eastern United States. It was Lyon presumably for whom the ericaceous genus Lyonia, was named. More than 50 species were introduced or re-introduced by Lyon into the nursery trade. Pieris floribunda, Rhododen- dron calendulaceum, Nyssa sylvatica and Iris fulva, to name but a few were initial introductions by him. The relations of Lyon to Pursh, Nuttall and other bota- nists of his time are amply documented and it becomes evident that Lyon was hardly less influential in the field of botany than in horticulture. Following an introduction and discussion under several headings the main body of the work, as might be expected from the full title, is devoted to the presentation of Lyon’s journal. For several of the years there are day by day re- ports covering months of activity. These might become tedious except that they contain frequent information of historical interest. Often too they are relieved by footnotes of a most informative kind that, for this reviewer at least, greatly enhanced the account. A map showing Lyon’s journeys is included and a repro- duction for the first time of his printed plant sales catalogue of 1812. Several indexes permit easy use as a reference work. The book is attractively bound in heavy paper. ALBION R. HODGDON, UNIVERSITY OF NEW HAMPSHIRE. 1John Lyon, Nurseryman and Plant Hunter and his Journal, 1799- 1814 by Joseph and Nesta Ewan. pp. 1-69, 2 figs., 3 maps, Trans. Am. Philos. Soe. vol. 53. pt. 2, Philadelphia, 1963, $2.00. NEW RECORDS OF GRASSES FROM THE CHICAGO REGION AND LOWER MICHIGAN S. F. GLASSMAN All specimens cited in this article are deposited in the herbarium of the University of Illinois, Navy Pier (CHI). Stipa comata Trin. & Rupr. Cook (Illinois): Morton Grove, along Milwaukee R. R., loamy soil, June 21, 1961, Glassman 5607. Not previously recorded from any of the counties in the Chicago region. Apparently, the only other station in Illinois is Winnebago Co. (Fuller, Fell & Fell, 1949), in the northwestern part of the state. Deam (1940) listed it from only one locality in northeastern Indiana ; Fassett (1951) cited it from southern Wisconsin ; and Chase (1951) mentioned it for Michigan. Stipa comata resembles S. spartea Trin., common in all seven counties of the Chicago region, but can easily be distinguished from that taxon by the shorter culms, narrower leaf blades, and shorter glumes, lemmas, calluses and awns. Elymus arenarius L. Berrien (Michigan): 14. mile south of Concession building, Weko Beach, Bridgeman, on fore dune, scattered clumps, associated with Ammophila brev- iligulata Fernald, Aug. 13, 1961, A. S. Rouffa 5665. In the Great Lakes region, this species is otherwise known only from Illinois, between Wilmette and Waukegan (Steyer- mark & Swink, 1952), and from one locality in Wisconsin (Iltis, Reed & Melchert, 1960). Elymus mollis Trin., a closely related species, is found along Lake Superior in Up- per Michigan (Bowden, 1957) ; but, apparently this is the first record of E. arenarius for the state of Michigan. There are no specimens of this taxon for Michigan in the Univer- sity of Michiean herbarium, and Dr. E. G. Voss, curator of vascular plants, knows of no authentic records for the state. UNIVERSITY OF ILLINOIS, NAVY PIER LITERATURE CITED BOWDEN, W. M. 1957. Cytotaxonomy of the section Psammelymus of the genus Elymus. Canad. Jour. Bot. 35: 951-992. 284 1963] Great Wass Island — Pike and Hodgdon 285 CHASE, A. 1951. Hitchcock’s Manual of Grasses of the United States, ed. 2, 1051 pp. Misc. Publ. 200, U. S. D. A. DEAM, C. 1940. Flora of Indiana, 1236 pp. Indianapolis. FaAssETT, N. 1951. Grasses of Wisconsin. 173 pp. University of Wis- consin Press, Madison. FULLER, G. D., FELL, E. W. & FELL, G. B. 1949. Check List of the Vascular Plants of Winnebago County, Illinois. Trans. Ill. Acad. 42: 68-79. Intis, H., REED, J. & MELCHERT, T. 1960. Elymus arenarius and Diarrhena americana in Wisconsin. Rhodora 62: 199-201. STEYERMARK, J. A. & SWINK, F. A. 1952. Plants New to Illinois and to the Chicago Region. Rhodora 54: 208-213. NOTES ON GREAT WASS ISLAND, MAINE. — Great Wass Island projects well out to sea south of Jonesport in Washington County. In company with Ralph Burns of Bailey's Mistake, Lubec, we set out on July 24, 1963, to follow a trail leading toward the seaward end of the island. In an open area near the island's center were several features of interest. An attractive open bog, had the usual bog-plants but expanses of granitic ledge nearby, largely devoid of vegetation, sup- ported scattered trees of Pinus Banksiana Lamb. and on closer examination disclosed small patches of Hypericum gentianoides (L.) BSP. The pine had already been reported from Great Wass Island but some features of its manner of reproduction deserve comment. Most of the trees seen were bearing cones, the majority of which were opening soon after ripening. No evidences of fire in recent decades were evident yet numerous seedlings and young pines from 2-7 years of age were observed growing in cracks and in shallow soil at the edges of the ledges. It would appear that fire is not necessary in this area to open the cones of Jack-Pine although it seems to be the belief of most persons who know the species that the cones remain closed until subjected to fire. The Hypericum here is at least 25 miles northeast of its previously reported northeastern limit in Hancock County, Maine. Specimens of our collections are to be found in the Herbarium of the University of New Hampshire. RADCLIFFE B. PIKE and A. R. HODGDON, UNIVERSITY OF NEW HAMPSHIRE. NOTE ON PRIMULA LAURENTIANA IN MAINE In 1906 Joseph A. Cushman accompanied by S. N. F. San- ford visited Libby Island, Machiasport, Washington County, Maine, and reported the occurrence of Primula laurentiana (Rhodora 9: 217, 1907). He noted that this Primula was only to be found around the light and the nearby oil-house. He suggested that seeds of this plant were probably intro- duced here by birds being killed by flying into the light and falling at its base which is, of course, a regular occurrance. He also remarked that the soil near the buildings had been disturbed by grading. A visit by the author to this same island in 1962 showed Primula laurentiana to be still present in the neatly mowed turf around the base of the light. It gave the appearance of a lawn weed so thickly did the rosettes of leaves dot the ground. From certain angles the intensely white undersides of the farinose leaves gave a most unusual appearance to the light-keepers lawn. The plants were abundant within ten or 15 feet of the base of the light tower and completely absent further away. Several hours careful search of the rest of the island failed to reveal a single plant growing in any other location. There were several spots of disturbed soil such as flower beds, paths and grading about buildings in addition to a considerable space maintained as lawn. In any of these areas this aggressive primrose could conceivably have estab- lished itself. An examination of rock-crevices and niches in cliff-faces where one would normally expect to find this plant failed to show any. The soil on the island was uniform being a deep layer of humus typical of Maine coast islands. The vegetational cover was largely of ericaceous and other acid soil plants including: Empetrum nigrum, Vaccinium vitis- idaea, Ledum groenlandicum, and Rubus Chamaemorus. Sedum Rosea, Iris Hookeri and Campanula rotundifolia were also present. The vigorous growth of Primula laurentiana, a well- known calciphile under such circumstances and in such a restricted area, can possibly be explained by the whitewash which is applied to the granite tower of the light semi- 286 1693] Primula Laurentiana — Pike 287 annually. Small flakes of the whitewash could be seen on the surface of the ground near the light, the chalky surface of the tower being continually eroded by wind and rain and largely deposited on the surrounding soil. As whitewash is almost entirely lime this would serve to make calcium avail- able in the immediate vicinity of the light and thus provide an essential element for the growth and continued survival of Primula laurentiana. The use of ordinary paints, which have no calcium in their pigments, on the other buildings on the island probably ex- plains why the plant does not grow near these other struc- tures. Cushman postulated that other outlying islands and head- lands along the Maine coast may have had seeds of Primula laurentiana introduced by migrating birds similar to the situation on Libby Island. He found the plant near the Moose Peak Light House on Mistake Island and on Black Head on Head Harbor Island. It has also been found on Crumple Island still further west and on Schoodic Point in Acadia National Park. A visit to Moose Peak Light on Mistake Island, in Septem- ber of this year showed P. laurentiana to be present in great vigor in rock crevices under and beside the elevated walk- way which leads from the keeper’s living quarters to the light itself some hundred yards away. The guard rails on this walk-way have been whitewashed and again this is a possible explanation of the long continued presence of this plant. Along with the primrose Lomatogonium rotatum was also growing — a great rarity on the Maine coast. Black Head on nearby Head Harbor Island has changed a great deal since Cushman’s visit more than a half century ago. According to Mr. Thurman Alley who was born and brought up on the island the growth of trees has changed Black Head so that he can now hardly find his way. This may account for the fact that the author was unable to find any Primulas. However, the quantities of broken shells of crabs, sea urchins, clams and mussels all over the cliffs indi- 288 Rhodora [Vol. 65 cates that there would be a source of calcium for Primula laurentiana when other conditions are right. It is interesting to think that the continued presence of the rare Primula laurentiana in at least two places on the Maine coast may be due to the whim of the U. S. Coast Guard as to the kind of paint it uses. RADCLIFFE B. PIKE, UNIVERSITY OF NEW HAMPSHIRE Volume 65, No. 763 including pages 193-288, was issued September 30, 1963. =< rII Ü:nnARY CARLOW REFERENCE LIBRARI JAN 7 1964 o Hovova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by ALBION REED HODGDON, Editor-in-Chief ALBERT FREDERICK HILL S STUART KIMBALL HARRIS RALPH CARLETON BEAN ROBERT CRICHTON FOSTER ROLLA MILTON TRYON RADCLIFFE BARNES PIKE Associate Editors Vol. 65 October-December, 1963 No. 764 CONTENTS: Contribution to the Fungus Flora of Northeastern North America. III. Howard E. Bigelow and Margaret E. Barr 289 A New Station for the Massachusetts Fern. su E POMA oiiaii n a nni. 309 Streptopus oreopolus Fern., A Hybrid Taxon. Doris Love and Hinrich Harries cicccccccccccscssccsseececsecseesscsceeeees 310 Tall Wheatgrass, A New Roadside Species in Utah. Earl M. Christensen sisisissseressisstersattaanettn nates este a a 318 A Checklist and Key to the Species of Campanula Native or Commonly Naturalized in North America. E a ER TA LLL ES EENEI EEE E ETE 319 (Continued on Inside Cover) The Nem England Botanical Club, Ine. Botanical Museum, Oxford St., Cambridge 38, Mass. CONTENTS: — continued Alpine Zone of the Presidential Range (Review). Frederic L. Steele ............. eere eee eere ettet enne tno ate sn oto to ese tao aoo Chromosome Numbers in some North American Species of the Genus Cirsium. Gerald B. Ownbey and Yu-Tseng Hst ............. eee New Species of Digitaria and Trichachne. Jason R. Swallen .............. eerte eee eene entente nennen tnnt A Strange Form of Hypericum canadense. A. R. Hodgdon .......... eene nennen dés Southeastern Limit of Chamaecyparis thyoides. Daniel B. Ward .cccccccccocsccsssssssscsssssessssccsccsccssenscsccsesnssssansssseecsoes Commelina communis in New Hampshire. A. R. Hodgdon ............. eee Pr esee Note on Lomatogonium. Radcliffe B. Pike ......................-. A Buttressed Elm from Ontario. F. R. Fosberg ose Index to Volume 65............ eren nnne nnne Rbodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 65 October-December, 1963 No. 764 CONTRIBUTION TO THE FUNGUS FLORA OF NORTHEASTERN NORTH AMERICA. IT’ HOWARD E. BIGELOW AND MARGARET E. BARR? Many of the fungi reported in this contribution were col- lected during the 1962 field season in Baxter State Park, Maine, or in adjacent areas of Piscataquis and Penobscot counties. This large wilderness region abounds in resources for the scientist as well as for the sportsman and vacationer. A variety of habitats exists over extremes of elevation in the Katahdin Range and these provided interesting addi- tions to our earlier studies in Maine. We are grateful to Mr. Austin H. Wilkins, Forest Commissioner, for the privi- lege of studying the fungi of this valuable preserve, and to Mr. Helon Taylor, Supervisor of Baxter Park, for his kind assistance during our work. Also included are notes on species and collections from elsewhere in the northeast. Some of the records extend the known range of several taxa, while other data provide a better understanding of species described by Peck as well as Ellis and Everhart. The senior author is responsible for the investigations on agarics, and the junior author for those on pyrenomycetes. The colors noted in the descriptions of agarics are from Ridgway, R. 1912. Color standards and color nomenclature. Washington, D. C. We acknowledge with appreciation the support of Na- 'l. Rhodora 62: 186-198. 1960. II. Rhodora 64: 126-137. 1962. Contri- bution from the Department of Botany, University of Massachusetts, Amherst. *Mrs. Howard E. Bigelow. 289 290 Rhodora [Vol. 65 tional Science Foundation Grant G 19534. Dr. Clark Roger- son, Curator of the Cryptogamic Herbarium, New York Botanical Garden and Stanley J. Smith, Senior Curator of Botany, New York State Museum kindly have extended the opportunity to study the type material of Peck and Ellis and Everhart. BASIDIOMYCETES Clitocybe eccentrica Peck, Bull. Torrey Club 25:321. 1898, Plate 1289. Pileus 1-6 cm. broad, plano-depressed at first, soon expanding to infundibuliform, margin thin and narrowly inrolled, even, arched, often wavy and lobed in age, surface sometimes with thin canescent coating at first but soon glabrous, hygrophanous and moist, whitish to pale watery buff at first (“tilleul buff", *avellaneous"), darker in age (nearly “wood brown", “tawny olive"), white or sordid white when faded; flesh thin and cartilaginous, concolorous with the pileus, odor fragrant at times or absent, taste bitter or absent. Lamellae decurrent, crowded, narrow, thin, forked, not intervenose, whitish to pale cream, edges even and straight. Stipe 3-5 em. long, 2-7 mm, thick at apex, equal or nearly so, base densely white strigose and with numerous rhizomorphs, glabrous or appressed-fibrillose above, compressed at times, often eccentric, soon hollow, pliant, concolorous with pileus or lamellae. Spores 4.5-6 X 2.5-3.5 x, elliptical to pyriform in face view, sub- lacrymoid in side view, smooth, not amyloid, white or faintly cream tinged in mass; basidia 14.5-21 X 4-5 4, 4-spored; cystidia not dif- ferentiated; pileus tissue: homogeneous, hyphae mostly cylindrical, 2-8.5 n in diameter, clamp connections present; gill trama interwoven, hyphae mostly cylindrical, 2-7 & in diameter. Usually cespitose, sometimes gregarious. On hardwood logs. Material examined: Bigelow 3900, Madawaska Lake, Aroostook Co., Maine, August 10, 1956; 10342, Abol Field, near Baxter State Park, Piscataquis Co., July 18, 1962; 10766, Ragged Mt., Penobscot Co., August 5, 1962; 10849, near Norcross, Penobscot Co., August 8, 1962; 10995, Abol Field, August 14, 1962; 11111, 11112, near Norcross, August 18, 1962; 11260, Cedar Lake, Penobscot Co., August 24, 1962; 7185, Mt. Toby, Sunderland, Massachusetts, July 28, 1958; 7647, Mt. Toby, August 30, 1958; 8969, Conway State Forest, Conway, August 1, 1960; 9044, Conway State Forest, August 9, 1960; E. A. Burt, S. Dunmore, Vermont (type of Clitocybe eccentrica). Clitocybe leptoloma (Peck) Peck, N. Y. State Mus. Bull. 157: 68. 1912. Plate 1290, Agaricus leptolomus Peck, N. Y. State Mus. Rep. 32: 26. 1879. Clitocybe subbulbipes Murrill, N. Am. Fl. 9: 404. 1916. 1963] Fungus Flora — Bigelow and Barr 291 Plate 1289 Clitocybe eccentrica Peck. X 1 Rhodora 292 t 0671 9?31*[d T X “eq (12d) pu.ojo}da) 24/12031]2) bh Ke ey 1963] Fungus Flora — Bigelow and Barr 293 Pileus 1-6 cm. broad, broadly convex at first, soon expanding to plane with dise usually shallowly depressed, margin narrowly in- rolled and edge slightly white canescent, finally becoming infundibuli- form, margin often arched and undulate, at times repand or sinuate, occasionally faintly pellucid-striate, thin, surface glabrous, hygro- phanous, watery buff moist (“warm buff", “pinkish buff", “cinnamon buff", “pinkish cinnamon"), paler in age but disc often remaining dark and tinged with pink, fading to whitish or pale buff (“pale pinkish buff"), disc remaining moist and colored for some time after margin faded; flesh thin, watery and concolorous with moist pileus, whitish when faded, odor and taste not distinctive. Lamellae adnate to short decurrent at first, becoming moderately to long decurrent, narrow (1-4 mm.), close or sometimes crowded, occa- sionally forked, intervenose at times, whitish to pale yellowish or pale buff (“cartridge buff", “light buff", “pale pinkish buff", “pale pinkish cinnamon"), not fading with pileus, edges usually straight and even. Stipe (1-)2-5.5 (-7) em. long, (1-)2-5(-10) mm. thick at apex, equal or either end enlarged, base usually with watery buff tomentum, a few white rhizomorphs present, often curved, compressed at times, stuffed becoming hollow, glabrous or fibrillose-striate, ground color watery buff (“pale pinkish buff", “pinkish buff", “cinnamon buff"). Spores 4-5.5(-6.5) X 2.5-3.5 (-4.5) mu, broadly elliptical to elliptical, smooth, not amyloid, pinkish tinted (“pale pinkish buff") in heavy deposit; basidia 16.5-25(-28) 4-6(-6.5) „u, usually 4-spored, at times also 2-spored; cystidia not differentiated; pileus tissue: cuticular hyphae cylindrical, 1.5-4(-7) & in diameter, tramal hyphae cylindrical to inflated, 3-10(-18) 4 in diameter, clamp connections present; gill trama regular, hyphae cylindrical, 2-6.5(-14) u in diameter. Solitary, scattered, gregarious, or subcespitose. On or near hard- wood logs, or attached to buried hardwood debris. Material examined: Bigelow 4326, Guerette, Aroostook Co., Maine, August 24, 1956; 10407, near Norcross, Penobscot Co., July 22, 1962; 10565, near Millinocket, July 27, 1962; 11000, Abol Field, near Baxter State Park, Piscataquis Co., August 14, 1962; 11047, Medway, Penob- scot Co., August 15, 1962; 11141, near Norcross, August 18, 1962; 11154, Katahdin Lake Trail, Baxter State Park, August 20, 1962; 11192, Trout Mt., Piscataquis Co., August 22, 1962; 11428, Bear Brook, Piscataquis Co., August 31, 1962; 7186, 7187, 7189, Mt. Toby, Sunderland, Massachusetts, July 28, 1958; 7413, 7414, Amherst, Au- gust 9, 1958; 8159, Hawley, Franklin Co., July 16, 1959; 8182, 8183, Mt. Toby, July 18, 1959; 9678, Alum Pond, Fiskdale, Worcester Co., August 2, 1961; 9702, Conway, Franklin Co., August 7, 1961; C. H. Peck, Indian Lake, New York (type of Agaricus leptolomus); W. A. & E. L. Murrill 199, Lake Placid, July 17-19, 1912 (type of Clitocybe subbulbipes) ; Bigelow 5625, 5632, 5633, 5640, Lac Munroe, Mt. Trem- T )») [Vol. ( Rhodora I6ZI 381d Z X 'Opieoo€eg (Yd) Dorun] 29fi2031]) 1963] Fungus Flora — Bigelow and Barr 295 blant Park, Quebec, July 23, 1957; 5768, 5786, Lac Munroe, July 25, 1957; 5910, 5911, 5912, Lac Munroe, July 27, 1957; 5978 - 5984, Lac Munroe, July 28, 1957; 9734, 9735, Newfane - Wardsboro, Vermont, August 26, 1961. Clitocybe truncicola (Peck) Saccardo, Syll. Fung. 5: 184. 1887. Plate 1291. Agaricus truncicola Peck, Buffalo Soc. Nat. Sci. Bull. 1: 46. 1873. Pileus 1-3.5(-5.5) cm. broad, broadly convex with margin inrolled and slightly incurved at first, becoming plane, margin remaining nar- rowly inrolled, even, broadly depressed in age and margin wavy, sur- face heavily canescent appearing finely matted-fibrillose in places under a lens, white, sometimes faintly watery-yellowish about disc in wet weather, canescence appressed in age then faintly buff overall, rarely rivulose; flesh white, thin, firm, no distinct odor and taste. Lamellae adnate to short decurrent, close to crowded, narrow (up to 3 mm,, rarely 5 mm. broad), not forked or intervenose, whitish (*cartridge buff"), edges even. Stipe 1-3(-4) em. long, 1.5-5(-15) mm. thick at apex, equal or base slightly enlarged, rhizomorphs few or absent, basal mycelium sparse, apex slightly pruinose at times, innately-fibrillose below, stuffed (in- terior white), often curved, eccentric at times, terete, white, finally pale buff in age (“cream buff”, “pale pinkish buff"). Spores 3.5-4.5(-5) X 2.5-3.5(-4) &, subglobose to broadly elliptical or elliptical, smooth, not amyloid, white in mass; basidia 12-23 (-26) X 3.5-5(-6) u, 4-spored; cystidia not differentiated; pileus tissue: cuticu- lar hyphae cylindrical, 2-4(-5) » in diameter, tramal hyphae usually cylindrical, 2-10 » in diameter, clamp connections present; gill trama regular, hyphae cylindrical, 2.5-6 » in diameter. Scattered or gregarious, rarely solitary or subcespitose. On hard- wood logs and stumps. Material examined: Bigelow 4082, Madawaska Lake, Aroostook Co., Maine, August 17, 1956; 4120, Madawaska Lake, August 18, 1956; 10637, near Norcross, Penobscot Co., July 30, 1962; 10705, Abol Field, near Baxter State Park, Piscataquis Co., August 2, 1962; 10998, Abol Field, August 14, 1962; 11227, Abol Field, August 23, 1962; 11364, near Norcross, August 29, 1962; 11427, Bear Brook, Piscataquis Co., August 31, 1962; 11539, Katahdin Lake Trail, Baxter State Park, September 6, 1962; C. H. Peck, Croghan, New York, September (type of Agaricus truncicola). The three preceding Clitocybes of section Candicantes fruited abundantly during 1962 and provided several oppor- tunities for a close comparison of fresh carpophores. Al- though very similar in microscopic features these species are quite distinct when gross features are compared. C. 296 Rhodora [Vol. 65 c eccentrica generally grows in cespitose fashion on logs and is distinctive from the other two lignicolous species by long strigose hairs at base of stipe and a conspicuously depressed cap. C. leptoloma is separated by the pinkish to buff cap when moist (recalling C. diatreta) and pinkish-buff spores in heavy deposit. Typically, C. truncicola is heavily canes- cent. Although fully expanded or old specimens may lose this coating, especially if water-soaked, the white spore deposit always will separate C. truncicola from C. leptoloma. Faded specimens of the two species are virtually impossible to distinguish if a spore deposit is lacking. Hygrophorus chrysodon (Fr.) Fries, Epicr. Myc. p. 320. 1838. Material examined: Bigelow 11577, Conway, Massachusetts, Octo- ber 20, 1962; 11585, Conway State Forest, Conway, October 21, 1962. Hygrophorus fuligineus Frost apud Peck, N. Y. State Mus, Rep. 35: 134. 1884. Material examined: Bigelow 10027, Women's Club State Forest, New Salem, Massachusetts, November 7, 1961; 11586, Harvard Forest, Petersham, November 3, 1962. Marasmius strictipes (Peck) Singer, Lilloa 22:326. 1951. Collybia strictipes Peck, N. Y. State Mus. Rep. 41: 62. 1888. Gymnopus strictipes (Peck) Murrill, N. Am. Fl. 9: 357. 1916. Pileus 2-6.5 em. broad, convex at first with the margin narrowly in- curved and inrolled, not striate, expanding to broadly convex, plane in age, dise often rugulose, subumbonate at times, glabrous, moist, dise with irregular rusty-orange stains (“orange rufous”, “ochrace- ous orange", “mars orange"), margin watery buff (“warm buff" to “ochraceous buff"); flesh thin, watery pallid, odor and taste rather disagreeable. Lamellae adnexed to adnate, seceding at times, close or crowded, narrow (1-4 mm.), occasionally forked, whitish (“pale pinkish buff"), edges crenate or eroded under a lens, slightly undulate. Stipe 3-8.5 em. long, 3-9 mm. thick at apex, equal or the base slightly enlarged, base white tomentose or strigose, central, hollow, cortex brittle, compressed at times in age, surface white pruinose at first, becoming appressed, whitish to very pale yellowish. Spores 6-9(-10) X 3-4 u, narrowly elliptical to subcylindrical, base attenuated and somewhat curved in side view, smooth, not amyloid, white in mass; basidia 23-33 X 5.5-7.5 u, 2- and 4-spored; cheilocystidia present, 19-36 X 4-7.5 u, more or less basidioid in shape, forked at times, 2-celled at times; pileus tissue: cuticle pale yellow to pale orange in KOH, pigment intracellular or in wall, cuticle consisting of pilo- cystidia, cellular, clavate to globose or obtuse, mostly pedicellate, 1963] Fungus Flora — Bigelow and Barr 297 sometimes irregular in shape, 14-28 » in diameter, walls smcoth, thin or somewhat thickened, trama dextrinoid in Melzer’s reagent, hyphae cylindrical to inflated or irregular, 6-13 4 in diameter, clamp con- nections small and inconspicuous, lactifers present occasionally; gill trama regular to subparallel, dextrinoid, hyphae cylindrical to some- what inflated, 3-12 » in diameter. : E Often solitary, sometimes gregarious or subcespitose. Usually be- neath hardwoods, rarely, conifers. Material examined: Bigelow 10376, Katahdin Lake Trail, Baxter State Park, Maine, July 19, 1962; 10438, near Abol Campground, Baxter State Park, July 23, 1962; 10540, Foster Field, Baxter State Park, July 26, 1962; 10631, Norcross, Penobscot Co., July 30, 1962; 11099, Katahdin Stream Campground, Baxter State Park, August 17, 1962; 11411, Katahdin Lake Trail, Baxter State Park, August 30, 1962; 7493, Leverett, Massachusetts, August 14, 1958; 8145, Hawley, Franklin Co., July 16, 1959; 8781, Sunderland, July 18, 1959; 9784, Conway, September 5, 1961; C..H. Peck, „Catskill: Mts. New York, September (type of Collybia strictipes); Bigelow 9743, Newfane, Vermont, August 26, 1961; 9817, Whitingham, September 10, 1961. This species belongs to section Globulares of Marasmius as Singer (1951) has indicated. Although the general aspect certainly seems typical of a Collybia, the nature of the pileus cuticle and the dextrinoid reaction of pileus and gill trama clearly indicate the proper position within Marasmius. In New England, rarely have I found more than a single carpophore at a time, yet it is not an uncommon species under hardwoods during the summer. Field characters for easy recognition are the rusty-stained pileus with rugulose surface, and the straight, pale stipe. The distribution of M. strictipes is unusual on present records. It is known with certainty to occur from Michigan to Maine in the northeast, and in Mexico from the work of Singer (1958). Phaeocollybia christinae (Fr.) Heim, Encyc. Mycol. 1:71. 1931. Material examined: Bigelow 6704, North Sunderland, Massachusetts, July 7, 1958; 7154, Mt. Toby, Sunderland, July 28, 1958. Species in Phaeocollybia which have small spores and lack clamp connections on the hyphae of pileus are only two, P. christinae and P. jennyae. These are separated by differences in the taste of flesh, sur- face of pileus, and color of spores when revived in KOH. The Massa- chusetts specimens fit the interpretation of Smith (1957). Phaeocollybia rufipes Bigelow, sp. nov. Plate 1292. Pileus 1-4 em. latitudine, conicus demum convexo-umbonatus, vis- E [Vol. 65 Rhodora 298 666I 9?39[d D X aou *ds ‘moig sodiina vighjjooanyg 1963] Fungus Flora — Bigelow and Barr 299 cidus, glaber, hygrophanus, ferruginus demum pallido-alutaceus; odor raphanaceus, sapor farinaceus; lamellae emarginatae, confertae, an- gustae vel latae, primum pallide alutaceae tum argillaceae; stipes 8-15 cm. longitudine, apice 3-5 mm. crassitudine, pallido-alutaceus dein argillaceus, glaber, deorsum attenuatus, radicatus, rufescens; sporae 7-8.5 X 4-5 a; cheilocystidia filamenteo, 27-31 X 3-5 4; hyphae defibulatae. Typum legit H. E. Bigelow, n. 11478, Katahdin Lake Trail, Baxter State Park, Maine, September 1, 1962; in Herb. Univ. Mass. conserva- tum. Pileus 1-4 em. broad, acutely conic or cuspidate at first with margin slightly incurved and narrowly inrolled, even, becoming convex with a large acute umbo, in age broadly convex to nearly plane, umbo re- maining acute, margin becoming faintly striate, viscid, glabrous, hygrophanous, color brown (near "orange cinnamon") fading to buff (near “light ochraceous buff"); flesh thin, concolorous with pileus when moist, fading to whitish or pale buff, odor raphanoid, taste slowly farinaceous when crushed. Lamellae emarginate, close, narrow to moderately broad (2-5 mm.), “pinkish buff" when young, becoming brown with spores (“cinnamon buff" to *clay color"), edges fimbriate, wavy or crenate in age. Stipe 8-15 cm. long, radicate and tapering downward, apex 3-5 mm. thick, base deeply embedded in substrate, surface glabrous, cortex fibrous and easily splitting lengthwise, brittle, stuffed becoming hol- low, apex “pinkish buff" or “apricot buff", darkening somewhat to “cinnamon buff", rufous (dark “brick red") downward, darkest toward the tapered end, not blackening. Spores 7-8.5 X 4-5 n, broadly fusoid in face view, inequilateral in side view, apical pore distinct, sometimes with slight knob at apex, wall thickened, with short and rather inconspicuous spines, yellowish- brown in KOH, spore deposit “snuff brown”; basidia 19-31 X 4.5-8 n, 4-spored; cheilocystidia filamentous, 27-31 4 long, 3-5.5 u in diameter, smooth, hyaline, often subcapitate at apex; pileus tissue: cuticular hyphae with gelatinizing walls, hyaline, 3-4.5 1 in diameter, subcuticu- lar zone and trama brown in KOH, hyphae cylindrical to somewhat inflated, 4-12 » in diameter, walls finely encrusted with pigment or sinuous-thickened, clamp connections absent, yellowish-brown latici- ferous hyphae present; gill trama broad, subparallel, brown in KOH, hyphae cylindrical, 2.5-10 & in diameter, laticiferous hyphae present. Scattered to gregarious, under spruce and fir. Material examined: Bigelow 11478 (type), 11479, Katahdin Lake Trail, Baxter State Park, Maine, September 1, 1962. Numerous carpophores were found in each collection of this unusual agaric. They were deeply embedded in the needle beds and difficult to remove without breaking the 300 Rhodora [Vol. 65 stipes. Frequently, only the caps were visible above the ground level. In comparison with the North American species known previously in Phaeocollybia (Smith, 1957), P. rufipes is most closely related to P. laterarius Smith of section Phaeocolly- bia. From this species, P. rufipes differs in color of pileus, non-blackening stipe, more hair-like cheilocystidia, habitat, and both spore length and shape. The odor and taste of P. rufipes probably are distinctive as well, as is the hygrophan- ous nature of the pileus. These features are not mentioned specifically in the original description of P. laterarius. ASCOMYCETES Barya parasitica Fuckel, Symb. Mycol. p. 93. 1870. Material examined: Barr 3359A, on Bertia moriformis, Katahdin Stream Campground, Baxter State Park, Maine, July 10, 1962; C. H. Peck, on Bertia moriformis and decayed wood of Fagus, Catskill Mts., New York, September, (type and isotype of B. parasitica var. caespi- tosa Peck). My collection is a small one, whereas those of Peck’s bear hundreds of perithecia grouped over the Bertia and nearby wood. Seaver (1910) described a collection from New York City with numerous perithecia. Outside of these three localities, I have found no records of B. parasitica in North America. It is known in Europe, and has been reported to be rare. Munk (1957) has a good description of the species from Danish material. B. parasitica is a non-stromatic member of the Clavici- pitaceae. The yellowish perithecia are seated on a dingy whitish subiculum, and eventually darken to grayish yellow- brown. The asci are characteristic of the family, cylindrical with an enlarged and refractive apex, through which a nar- row canal can be seen. The ascospores are filiform, hyaline to faintly yellowish, several-septate and guttulate. Herpotrichiella spinifera (Ell. & Ev.) Barr, comb. nov. Page 302 figures 1-3. Melanomma spiniferum Ellis & Everhart, North Amer. Pyreno. p. 184. 1892. Ascostromata 70-112 u in diameter, globose to conical, entirely im- mersed or upper portion erumpent, scattered to gregarious, lower wall thin, light grayish brown to nearly hyaline, of two to four layers of , 1963] Fungus Flora — Bigelow and Barr 301 polygonal cells, 8-11 4 wide, upper exposed portion of wall blackened, short setose around apex, setae at times inconspicuous, 7.5-22 u long, 5-5.5 u wide near base, blackish-brown, often irregular, simple, ends blunt or pointed, wall thick, apical pore up to 25 y wide. Asci 27-44 X 7.5-15.5 mu, saccate, narrowed to rounded apex and sessile base, wall double, thickened above, few in a fascicle, pseudo- paraphyses indistinct. Spores (10-)12-15.5 X 3.5-4.5 u, hyaline when young, light olive brown or grayish brown at maturity, clavate-elliptical, broadest in upper third, tapered to rounded ends, straight to slightly curved, (1-, 2-)3-(4-, 5-) septate, not or slightly constricted at septa, cell above primary septum broadest, contents guttulate or one globule in each cell, wall thin, smooth or very finely roughened, crowded in the ascus. Immersed in pore surface of old resupinate basidiomycetes. Material examined: A. Commons, on Corticium sp.(?), Wilmington, Delaware, September, 1890, Ell. & Ev. N.A.F. Second series 2610; Barr 3326, on old Poria sp., Abol Stream, Baxter State Park, Maine, July 6, 1962; R. H. Petersen and C. T. Rogerson, on Sabacina incrus- tans, Toxaway River junction with Bear Wallow Creek, Transylvania Co., North Carolina, July 29, 1961; on old Sterewm, same data as above; on old Stereum, along Corbin Creek, branch of Whitewater River, Transylvania Co., August 9, 1961; on old Stereum, 4 miles south of Upper Falls of Whitewater River, Oconee Co., South Carolina, August 14, 1961. Dr. Rogerson has kindly communicated his collections of Herpotrichiella spinifera to me, and has compared them with type material of Melanomma spiniferum. N.A.F. 2610 is part of the material from which the species was originally described. My collection from Maine is identical with it and with Rogerson's material. The three collections on Stereum bear both H. spinifera and a representative of the Trichosphaeriaceae, as yet un- identified. The latter fungus is the more conspicuous, with superficial, collapsing, setose perithecia, and 3-septate ascos- pores. The ascostromata of H. spinifera in these collections are very short setose, and were at first thought to be glab- rous. Further examination has shown that setae are present, and in all respects these specimens are identical with the others cited above. Rogerson (in litt.) noted that the “glab- rous" form was similar to what Ellis called Melanomma porothelia (Berk. & Curt.) Sacc. However, M. porothelia Rhodora Figs. 1-83. Herpotrichiella spinifera (Ell. & Ev.) Barr, comb. nov.: 1, ascostroma, 2, ascus, 3, ascospores. Figs. 4, 5. Tubeufia cerea (Berk. & Curt.) Barr, comb. nov.: 4, ascus, 5, ascospores. Figs. 6-9. Nectria atrofusca (Schw.) Ell. & Ev.: 6, habit of perithecia on stroma erum- pent through bark, 7, perithecia on stroma, 8, ascus, 9, ascopores. Figs. 10-12. Niesslia barbicincta (Ell. & Ev.) Barr, comb. mov.: 10, peri- thecium, 11, ascus, 12, ascospores. Fig. 13. Phylleutypa wittrockii (Erikss.) Petrak: ascospores. Fig. 1, X 400; figs. 2-5, 8, 9, 11-13, X 750; fig. 6, X 10; figs. 7, 10, X 80. 1963] Fungus Flora — Bigelow and Barr 303 is identical with Litschaueria corticiorum (Hoehnel) Petrak, a member of the Xylariaceae, according to Petrak (1923). What Ellis considered M. porothelia is obviously different from Berkeley and Curtis's fungus. In most of the species at present known to belong to the genus Herpotrichiella, considerable variation in length and septation of ascospores occurs. Mueller (in Mueller and von Arx, 1962) considered that the genus Didymotrichiella Munk (1953), erected for species with one-septate spores, could not be distinguished satisfactorily from Herpotrichi- ella, and reduced Didymotrichiella to synonymy with the latter. In comparing the described species, H. spinifera appears to be most closely related to H. setosa Barr and H. fusispora Barr (1959). H. spinifera bears setae only at the apex of the ascostroma, whereas the other two species are setose over most of the wall. H. spinifera also differs from these two species in shorter spores, and in having the ascostromata more deeply immersed in the substrate. Munk (1953) erected the family Herpotrichiellaceae to accommodate H erpotrichiella and four other genera of simi- lar aspect. Mueller and von Arx (1962) reduced the family to synonymy under the Pleosporaceae. Such disposition ap- pears logical to me, as familial distinctions can scarcely be made on the bases of small size and greenish or olive grayish ascospores. Nectria atrofusca (Schw.) Ellis & Everhart, Journ. Mycol. 1:140. 1885. Page 302, figures 6-9. Sphaeria atrofusca Schweinitz, Trans. Am. Philos. Soc. ser. 2, 4: 206. 1832. Creonectria atrofusca (Schw.) Seaver, Mycologia 1:186. 1909. Cucurbitaria seriata Peck, N. Y. St. Museum Rep. 28: 75. 1876. Otthia seriata (Peck) Sacc. Syll. Fung. 1: 739. 1882. Otthiella seriata (Peck) Sacc. & D. Sace. Syll. Fung. 17: 662. 1905. Plowrightia staphylina Ell. & Ev. Proc. Acad. Phila. 1890: 248. Otthia staphylina (Ell. & Ev.) Ell. & Ev. North Amer. Pyreno. p. 251. 1892. Otthiella staphylina (Ell. & Ev.) Dearness & House, N. Y. St. Museum Bull. 266: 71. 1925. Stroma immersed in wood tissue, yellowish brown, composed of thin- walled, compact cells, with yellowish brown hyphae penetrating deep into wood, elliptical, rounded, or elongate, up to 4 mm. long, 150-600 304 Rhodora [Vol. 65 u wide, (80-)115-400 4 thick; perithecia grouped on and bases im- mersed in stromatic tissue, forming rounded or elongate rows and erumpent through splits in bark, brown to blackish and shining when dry, brown or yellowish brown when fresh, glabrous, apex collapsing or pinched in at sides when dry, but not cupulate, short papillate or conie, 150-275 4 in diameter, 180-300 u high, wall of two distinct layers, outer dark to light brown and crust-like, 16-33 x thick, com- posed of 4-6 layers of polygonal thick-walled cells, inner layer yellow- ish to hyaline, 8-13 x thick, of compressed rows of thin-walled cells, pore 20-33 u wide, surrounded by elongate cell layers, brown externally, lined with hyaline periphyses. Asci 50-82 X 6.5-12 u, clavate or oblong, narrowed to stipe of vary- ing length, apex rounded-truncate, wall single, thin but slightly thickened and with refractive line at apex, non-amyloid, (4-)8-spored, paraphyses delicate and slender, eventually compressed. Spores 9-15(-17.5) X 4-6(-7.5) 4, hyaline, light dull yellowish in mass and in age, oblong, elliptical, or ovoid, ends rounded, straight to inequilateral, 1-septate in the middle, not constricted, wall thin, smooth, in age finely roughened, contents homogeneous or minutely guttulate, obliquely uniseriate to partially biseriate in the ascus. Erumpent through bark on dead or dying limbs of Staphylea tri- folia, less often on Evonymus sp. Material examined: Barr 3182, South Deerfield, Massachusetts, May 10, 1962; L. N. Johnson 594, Ann Arbor, Michigan, April 1, 1893; C. Devol, Albany, New York, October, 1874 (type of Cucurbitaria seriata on Evonymus sp.) ; J. Dearness 1560, London, Ontario, (type of Plowrightia staphylina Ell. & Ev.); 663, London, May 25, 1891; 1848, Parkhill, May 24, 1892; Parkhill, May, 1892; same data, Ell. & Ev. N.A.F. Second series 3320; London, December, 1903, Fungi Columbiani, Bartholomew 1943; London, 1912, Rehm Ascomycetes 2041; London, June, 1913, Sydow, Fungi exot. exs. 184; B. M. Ever- hart 527, West Chester, Pennsylvania, October 1, 1885; same data, Ell. & Ev. N.A.F. Second series 1547; J. Dearness, Montreal, Quebec, November 8, 1898. According to Rogerson (in litt.), the type of Sphaeria atrofusca Schweinitz is identical with Ellis specimens desig- nated Nectria atrofusca and with my collection from Massachusetts. Type and authentic material of Plowrightia staphylina Ell. & Ev., and the type collection of Cucurbitaria seriata Peck are also identical. Wall and centrum structure, as well as the delicate paraphyses, thin-walled asci, and variable nature of ascospores, all lead to disposition of the fungus as a Nectria. According to Munk's (1957) description of Melanopsam- 1963] Fungus Flora — Bigelow and Barr 305 ma pomiformis (Pers. ex Fr.) Sacc., many points of simi- larity exist between it and N. atrofusca. Several specimens of M. pomiformis from Europe and North America were studied to compare the two species. The major distinction is in the perithecial wall which is carbonaceous in M. pomi- formis, much softer and more fleshy in N. atrofusca. They cannot be considered identical. Mueller and von Arx (1962) have transferred M. pomiformis to Chaetosphaeria. 'This genus belongs in the Trichosphaeriaceae (Sphaeriaceae), and within the family is close to Eriosphaeria. The Hypo- creaceae, while related to the Sphaeriaceae in ascus struc- ture, is distinguished by the bright or light colored, more fleshy perithecia. In the genus Nectria, N. atrofusca appears to belong in the Ochroleuca group, as delimited by Booth (1959). N. pallidula, as described by Booth, is similar in many respects. It differs in the brighter color of perithecia, roughened wall, and constricted spores. I have not seen authentic material, nor a good description of N. ochroleuca (Schw.) Berk., to compare with N. atrofusca, but apparently it is also much lighter in color than the latter species. Niesslia barbicincta (Ell. & Ev.) Barr, comb. nov. Page 302, figures 10-12. Byssosphaeria barbicincta Ell. & Ev. Journ. Mycol. 4: 63. 1888. Trichosphaeria barbicincta (Ell. & Ev.) Sacc. Syll. Fung. 9: 603. 1891. Herpotrichia barbicincta (Ell. & Ev.) Ell. & Ev. North Amer. Pyreno. p. 158. 1892. Perithecia 120-185 & in diameter, globose to conical, collapsing cupu- late when dry, black, shining, scattered singly or grouped on scanty or abundant subiculum of brown, thick-walled hyphae, apical pore small, periphysate, wall brittle, thin, 6-7.5 thick, composed of few compressed layers of cells, blackish brown externally, light brown to hyaline within, setose around base, setae few, curved upward around perithecium, dark brown, septate, apex blunt and often pallid, (27-) 60-200 u long, 3.5-7.5 u wide near base, frequently similar setae scat- tered on hyphae of subiculum, not associated with perithecia. Asci 46-66 X 6-10 u, oblong, apex rounded-truncate, stipe short, wall single, thin, with minutely refractive area at apex, non-amyloid, para- physes filiform, hyaline, thin-walled. Spores 10-16.5 X 2-4 4, greenish hyaline, oblong or fusoid, ends rounded or pointed, straight, inequilateral, or slightly curved, 1-septate 306 Rhodora [Vol. 65 in the middle, not constricted, wall thin and smooth, contents minutely guttulate, overlapping bi- to uniseriate in the ascus. Superficial on old fungi and on adjacent leaf or wood surfaces. Material examined: Barr 3766, Ruggles Pond, Wendell State Forest, Massachusetts, September 22, 1962; 1292A, The Gorge, U.M.B.S., Cheboygan Co., Michigan, July 22, 1953; J. B. Ellis, Newfield, New Jersey, October, 1887, N.A.F. Second series 1958 (cotype of Bysso- sphaeria barbicincta Ell. & Ev.). The distinguishing feature of N. barbicincta is the posi- tion of setae. They arise from the lower wall of the perithe- cium and curve upward beyond its apex. In N. exilis (Alb. & Schw. ex Fr.) Winter, setae develop over the entire peri- thecial wall and are short and stiff. N. exosporioides (Desm.) Winter, is said (Mueller and von Arx, 1962) to be similar to N. exilis except for longer setae. The third species recog- nized by Mueller and von Arx, N. crucipila (Hoehnel) Mueller, has branched setae. The spores of all three species are smaller than are those of N. barbicincta. N. exilis is the correct name for the fungus I have reported previously as N. pusilla (Fr.) Schroet. from northern Quebec (Barr, 1959) and Gaspé Parc (Barr, 1961). The genus Niesslia Auerswald is a representative of the Trichosphaeriaceae (Sphaeriaceae ss. auct.) with small, scat- tered, setose perithecia on scanty subiculum, and with thin- walled asci and hyaline, two-celled spores. It is closely related to Eriosphaeria Sacc., but differs in smaller, thin- walled perithecia and in lacking a thin stroma. Phylleutypa wittrockii (Erikss.) Petrak, Ann. Mycol. 39: 280. 1941. Page 302, figure 13. Material examined: Barr 3211, Linnaea borealis var. americana, near Katahdin Stream Campground, Baxter State Park, Maine, June 25, 1962; 3757, September 5, 1962; Stuntz and Allen. 1591, Isle Royal, Michigan, September 13, 1901; W. G. Farlow, Shelburne, New Hamp- shire, September, 1886; H. D. House, Lake Placid, New York, Septem- ber 9, 1917; Newcomb, Essex Co., July 18, 1922. P. wittrockii is much less common than the prevalence of colonies of Linnaea in northern regions would imply. I have found only the two Maine collections to date. The re- maining four collections are all in the Peck Herbarium at Albany (NYS). That collected by Farlow was cited both by Ellis and Everhart (1892) and Theissen and Sydow 1963] Fungus Flora — Bigelow and Barr 307 (1915). Wehmeyer (1942) has reported a collection from Nova Scotia. In Europe the only record I have seen is from Sweden. The collection made in Maine in June contained mature asci and spores as well as early infection and young stro- mata on living plants. The latter were blackened and dis- torted from the uppermost pair of leaves to tip of branch. The September collection, in the same locality, showed well- developed stromata. The perithecia were immature and lacked asci and spores. Tubeufia cerea (Berk. & Curt.) Barr, comb. nov. Page 302, figures 4, 5. Sphaeria cerea Berk. & Curt. Grevillea 4:108. 1875. Calonectria cerea (Berk. & Curt.) Sacc. Syll. Fung. 2: 551. 1883. Dialonectria cerea (Berk & Curt.) Cooke, Grevillea 12:110. 1884. Ophionectria cerea, (Berk. & Curt.) Ell. & Ev. North Amer. Pyreno. p. 118. 1892. Nectria fulvida Ell. & Ev. Journ. Myc. 1:140. 1885. Calonectria fulvida (Ell. & Ev.) Berl. & Vogl. Add. Syll Fung. p. 212. 1886. Dialonectria fulvida (Ell. & Ev.) Ell. & Ev. Journ. Mycol. 2: 122. 1886. Ophionectria everhartii Ell. & Galw. Journ. Mycol. 6:32. 1890. Ascostromata 150-160 & in diameter, 120-130 u high, superficial, scattered or grouped on surface of old pyrenomycete stromata and adjacent wood, yellowish brown, pulverulent, wall thin at base and lower sides, 10-11 » wide, of several layers of polygonal cells, brownish to yellowish, thickened as a ring or collar at upper sides, the outermost layers of cells protruding from surface, inner layers of cells yellowish, compressed, apex minutely papillate, pore filled with yellowish tips of pseudoparaphyses. Asci 66-74 X 9-11 4, oblong, apex rounded, base foot-like, wall double and thickened above, pseudoparaphysate. Spores 33-44 X 3-3.5 u, greenish hyaline, fusoid to nearly cylindrical, ends tapered and narrowly rounded, straight to slightly curved, 7- to 9-septate, delicate, not constricted, each cell with a single globule, wall thin, smooth, in two overlapping fascicles in the ascus. On old stromata of Diatrype spp. and Hypoxylon Spp. and adjacent wood surfaces. Material examined: Barr 3337, Diatrype stigma on white birch, Abol Field, Piscataquis Co., near Baxter State Park, Maine, July 6, 1962; 3360A, Hypoxylon sp. on beech, Katahdin Stream Campground, Baxter State Park, July 10, 1962; 2160, Diatrype sp. on birch, Lac Munroe, Mt. Tremblant Park, Quebec, July 28, 1957. 308 Rhodora [Vol. 65 This species is a member of the Pleosporaceae with super- ficial ascostromata, bitunicate asci and pseudoparaphyses. Booth (1959) noted that O. cerea belonged in the genus Tubeufia, but did not make the combination. The shape of the ascostroma, with thickened ring-like upper wall, is most distinctive. I have not examined the types of Nectria fulvida Ell. & Ev. and Ophionectria everhartii Ell. & Galw., but these names were placed in synonymy with Ophionectria cerea by Ellis and Everhart (1892). According to the description, Calonectria belonospora Schroet. (Ophionectria belonospora (Schroet.) Sacc.) is synonymous probably with Tubeufia cerea. This fungus has been reported from Europe and North America. According to the literature, in North America it has been collected in South Carolina, New Jersey, New- foundland, and Ontario. Although not rare, it is probably overlooked. DEPARTMENT OF BOTANY, UNIVERSITY OF MASSACHUSETTS, AMHERST LITERATURE CITED Barr, M. E. 1959. Northern Pyrenomycetes I. Canadian Eastern Arctic. Contr. Inst, Bot. Univ. Montreal 73: 1-101. 1961. Northern Pyrenomycetes II. Gaspesian Park. Can. Journ. Bot. 39: 307-325. BoorH, C. 1959. Studies of Pyrenomycetes: IV. Nectria. C.M.1I. Mycol. Paper 73: 1-115. ELLIS, J. B. and B. M. EvERHART. 1892. The North American Pyrenomycetes. Newfield, N. J. MUELLER, E. and J. A. voN ARX. 1962. Die Gattungen der didymos- poren Pyrenomycetes. Beitr. Kryptogamenfl, Schweiz 11(2): 1-922. Munk, A. 1953. The system of the Pyrenomycetes. Dansk Bot. Arkiv. 15(2) : 1-161. 1957. Danish Pyrenomycetes. Dansk Bot. Arkiv 17(1): 1-491. PETRAK, F. 1923. Mykologische Notizen VI. Ann. Mycol. 21: 182- 335. SEAVER, F. J. 1909. The Hypocreales of North America — II. My- cologia 2: 48-92. 1963] Massachusetts Fern — Poole 309 SINGER, R. 1951. The Agaricales (mushrooms) in modern taxonomy. Lilloa 22: 1-832. 1958. Fungi Mexicani, Series secunda — Agaricales. Sydowia 12: 221-243. SMITH, A. H. 1957. A contribution toward a monograph of Phaeocollybia. Brittonia 9: 195-217. THEISSEN, F and H. Sypow. 1915. Die Dothideales. Kritisch- systematische Originaluntersuchungen. Ann. Mycol, 13: 149-7406. WEHMEYER, L. E. 1942. Contributions to a study of the fungus flora of Nova Scotia. VI. Pyrenomycetes. Can. Journ. Res. C, 20: 572-594. A NEW STATION FOR THE MASSACHUSETTS FERN. — In Sep- tember 1962 on a collecting trip with three of my faculty col- leagues we stopped to open our lunch boxes in the old spring house at the Bradford Mineral Spring a short dis- tance from the highway between Bradford Center and East Washington, New Hampshire. While exploring in the vicinity after lunch I came across a sizable, vigorous colony of Dryopteris simulata Davenport growing in damp mixed woods a short distance north of the spring. The only other station for this species in Merrimack County, repre- sented by a specimen in the New England Botanical Club Herbarium, is in Contoocook near the middle of the county. The Bradford station is less than a mile east of the Sullivan County line. A station near the highway on Route 9 in South Stoddard in Cheshire County represents a more west- ern extension for this species for this part of the state but the Bradford station is the most northerly to be reported except for an early collection in 1903 made by Timothy O. Fuller at Lake Wentworth in Carroll County. A specimen from the Bradford station has been deposited in the Jesup Herbarium at Dartmouth College, Hanover, N. H. Others have been sent to the University of New Hampshire Her- barium in Durham, N. H., and to the Gray Herbarium and the New England Botanical Club Herbarium in Cambridge, Mass. JAMES P. POOLE, JESUP HERBARIUM, DARTMOUTH COLLEGE. STREPTOPUS OREOPOLUS FERN., A HYBRID TAXON DORIS LOVE AND HINRICH HARRIES In 1905, M. L. Fernald and J. F. Collins visited Mt. Albert on the Gaspé Peninsula in eastern Quebec. Above timberline they discovered on this mountain a remarkable Streptopus with deep claret-colored flowers. This plant proved to be a new taxon and was named by Fernald (1906) Streptopus oreopolus, the mountain-dwelling twisted-stalk. Next summer, when Fernald and Collins botanized again in the mountains of the Gaspé, doubts entered Fernald’s mind and he suggested (Fernald, 1907) that the new species might be a hybrid between S. amplexifolius (L.) DC. and S. roseus Michx. This hypothesis appeared to be supported by Fernald's observation that the fruits of S. oreopolus did not develop to maturity, though he considered also the pos- sibility that the extremely dry weather during the early part of this particular summer might have been responsible for the sterile condition. There were also certain features of the plant and its occurrence which made Fernald hesitant in proposing his hypothesis (p. 107) : “This uniform sterility of S. oreopolus throughout its known range suggests that it may not be a self-perpetuating species; yet, if this should prove to be the case, the profusion of the plant throughout the area and the constancy with which it maintains its char- acteristic pubescence and the color of its perianth is indeed remarkable." In 1929, a review of the literature on S. oreopolus and a discussion of its taxonomic position and distribution were given by Marie-Victorin in his work on “les Liliiflores du Québec". By this time the taxon had been reported from northwestern Newfoundland (Fernald, 1926), the Mingan Archipelago (Marie-Victorin, 1929), the Shickshock Moun- tains (Fernald, 1906, 1907), Mt. Katahdin (Stebbins, 1929), and Mt. Washington (Fernald, 1927), and plants with apparently well-developed fruits had been found on two occasions (Fernald, 1926; Marie-Victorin, 1929). Marie- Victorin rejected the possibility of a hybrid origin for S. 310 1963] Streptopus oreopolus — Lóve and Harries 311 oreopolus especially because of the dark red color of its flow- ers which he thought could not be the result of a hybridiza- tion between two parent species with respectively white and rose-colored flowers. Instead, he considered the taxon to be a good species which he classified (p. 98) “dans la catégorie des endémiques de la région non récemment glaciée qui entoure le golfe Saint-Laurent". The taxonomic position of S. oreopolus was discussed again by Fassett (1935). From Fassett's description of the taxon, as from the previous paper by Marie-Victorin (1929), it is apparent that S. oreopolus, while intermediate between S. amplexifolius and S. roseus in certain vegetative features and in the character of the fruits, is much more similar to S. amplexifolius, especially with respect to the flowers. Fassett reached the conclusion that S. oreopolus is a mere variety of S. amplexifolius and closely related to the var. denticulatus which occurs in the Lake Superior area and in western North America. The latter variety, accord- ing to Fassett, approaches S. oreopolus in its denticulate leaf margins and in its occasionally pink or reddish flowers. Fassett agreed with Marie-Victorin in rejecting the possi- bility that S. oreopolus might be a hybrid. He reports as additional evidence against such an interpretation of the taxon his observation (p. 101), “that S. oreopolus has the perianth-segments conspicuously papillate within, while S. amplexifolius var. americanus rarely has well-developed papillae, and the representative of S. roseus occurring in that region lacks papillae of the type found in the purported hybrid". The distribution of S. oreopolus appeared to Fas- sett to be likewise in disagreement with the hybrid hypothe- sis because (pp. 101-102), “if S. amplexifolius and S. roseus can hybridize, why do they not do so in the many other places where their ranges overlap, instead of only in one limited region". Fassett (1.c.) reported also the occurrence of a plant which appeared to be an intermediate or hybrid between S. amplexifolius and S. oreopolus (p. 102): “On Mount Washington, New Hampshire, just above Tuckerman's 312 Rhodora [Vol. 65 Ravine, may be found a most interesting series of variations of S. amplexifolius. Var. oreopolus is abundant, and var. americanus can also be found. A third type of plant has denticulate leaves like those of var. denticulatus. Since its leaf-margin is intermediate between that of var. americanus and that of var. oreopolus, and, in addition, the lower leaf- surface is less glaucous than in the former, but more so than in the latter, it is considered as a hybrid of these two varieties. The flowers, which were just beginning to open when observed by the writer on June 27, 1934, were nearly white and conspicuously papillate within. .. . The expanded flowers should be observed, especially for correlation of perianth-color with degree of toothing on the leaf-margin, and the fruit-colors should be noted." We ourselves have observed for several years the three kinds of Streptopus in the subalpine region of Mt. Washing- ton. There, S. amplexifolius is represented by its var. americanus Schultes and S. roseus by var. perspectus Fas- sett. Streptopus oreopolus has been collected at the Lakes of the Clouds (No. 7512, July 25, 1958, in deep moss around the Lakes at 5100’ alt., coll. A. & D. Love), along the Oakes Gulf Trail (No. 346, August 5, 1961, in subalpine birch-fir forest at 4300’ alt., coll. H. Harries), in the upper part of the Tuckerman Ravine headwall (August 28, 1962, in the sub- alpine Alnus-shrubbery around the trail, fruits coll. by D. Lóve), and on the east slope of Mt. Clay (No. 347, July 18, 1961, at 5350' alt., coll. H. Harries). In the subalpine region of Mt. Washington, the three taxa occur mainly in sheltered locations on the northeast, east, and southeast slopes between 4500’ and 5500’ alt. where be- cause of the thick and long-lasting snow cover the fir krummholz is replaced by a subalpine snowbed vegetation. In these localities, the three Streptopus forms grow asso- ciated with such snow-tolerant species as Betula glandulosa, Vaccinium caespitosum, Calamagrostis canadensis s.l., Des- champsia flexuosa, Luzula parviflora var. melanocarpa, Veratrum viride, Clintonia borealis, Houstonia caerulea var. Faxonorum, Solidago macrophylla var. thyrsoidea, and 1963] Streptopus oreopolus — Lóve and Harries 313 Dryopteris spinulosa var. americana. On such sites, the snow cover was observed in spring 1961 to disappear be- tween May 20 and June 13. In a few spots where the snow cover lingered into late June or early July, Streptopus was found to be lacking. In their subalpine localities, S. amplexifolius and S. ore- opolus grow mostly in the furrows which are frequently met with in snow bed areas on steeper slopes, and generally along drainage channels. In the forests of the lower alti- tudes, S. amplexifolius was encountered occasionally, mostly in close vicinity to a stream. S. oreopolus was found in such a habitat only once, namely in Oakes Gulf at 4300' alt. where it was observed together with S. amplexifolius in a subalpine birch-fir forest along the margin of a small stream. Streptopus roseus does not show such a pronounced pref- erence for drainage channels but is of a more general dis- tribution and is found even in some of the more luxurious kinds of heath vegetation. The species was noted by us in the Mt. Washington area only from subalpine localities and has not been observed in forest habitats, but it is cited by Pease (1924) for many low-altitude localities in the area. Streptopus amplexifolius and S. roseus resemble each other in their wide altitudinal range which extends from the valley levels up into the subalpine region where both species reach their altitudinal boundary, according to our observa- tions, between 5400’ and 5500’ alt. Streptopus o reopolus, on the other hand, was observed only between 4300' and 5400" alt. and appears to be a strictly subalpine taxon. In the subalpine region of Mt. Washington, the three Streptopus taxa were found to flower between late June and early August. In 1961, the peak of the flowering fell for S. amplexifolius and S. oreopolus between July 10 and July 20; that of S. roseus was about a week earlier. We noted in the subalpine region of Mt. Washington the color of the flower to be a good characteristic with all three taxa. The flowers of S. amplexifolius are creamy white with a dark purple streak at the base of each perianth lobe. The flowers of S. roseus are rose-colored with many irregularly 314 Rhodora [Vol. 65 scattered darker red streaklets. The flowers of S. oreopolus lack any kind of markings. They are white in the bud and become dark wine red at maturity (appearing almost black in herbarium specimens). We now have chromosome counts of all three taxa from the Mt. Washington area. Streptopus roseus is diploid (2n— 16, counted by Dr. S. Kawano on material collected at the Lakes of the Clouds in July, 1960). Streptopus oreopolus is STREPTOPUS = UN 32 l6 24 amplexi- roseus x oreopolus -folius Fig. 1l. Chromosomes of Streptopus roseus, S. oreopolus, and S. amplexifolius. In each karyogram the longest and the shortest sets, respectively, are indicated. triploid (2n—24, counted by A. & D. Lóve on coll. No. 7512, Lakes of the Clouds, July, 1958), and S. amplexifolius is tetraploid (2n—32, counted by Dr. S. Kawano on material collected at the Lakes of the Clouds in July, 1960). The chromosomes are large and easily distinguishable. The tri- ploid has distinctly three of each type in its karyogram, as can be seen in Fig. 1, where the smallest and largest chromo- somes are indicated. The diploid and tetraploid numbers are in conformity with those previously counted by Therman (1956). The tetraploid number has been counted earlier by Matsuura and Sutó (1935), by Sató (1932), and by Mattick (in Tischler, 1950). 1963] Streptopus oreopolus — Lóve and Harries 315 A further study of the plants collected on Mt. Washington showed that the pollen of S. oreopolus is completely sterile, whereas that of S. amplexifolius and S. roseus has 95% or more good pollen grains, when stained with aceto-orcein. In spite of the fact that all fruits collected looked mature and well-filled, the seeds of the S. oreopolus fruits proved to be shrivelled and infertile. All available herbarium material of S. oreopolus from Newfoundland, the Mingan Archipelago, and the Gaspé Peninsula was checked for these features and it can be stated with certainty that the taxon is sterile every- where in its range. The conclusion seems therefore unavoidable that Fernald (1907) was right in his suggestion that S. oreopolus might be a hybrid between S. amplexifolius and S. roseus. It remains to discuss those facts emphasized by Marie-Victorin (1929) and Fassett (1935) which might be considered as negative evidence against such an interpretation of the taxon. Streptopus oreopolus differs from its two parent taxa by the dark perianth color and by the copiously papillate peri- anth segments. No explanation for the appearance of these features in the hybrid can be given but the possibility should be considered that the hereditary material of the tetraploid S. amplexifolius could contain recessive characters which might become apparent in its hybrid with S. roseus. Neither in the field nor among herbarium material have we seen an intermediate or transitional type between S. oreopolus and either S. amplexifolius or S. roseus. The in- terpretation of the form observed by Fassett (1935) in the Tuckerman Ravine as a hybrid between S. amplexifolius and S. oreopolus appears to be incompatible with the observed complete sterility of S. oreopolus, and it seems likely that this plant might have been only a somewhat atypical speci- men of either S. oreopolus or S. amplexifolius. For a sterile hybrid taxon which can spread only by means of its rhizomes, the distribution of S. oreopolus is indeed a strange one. Whereas the ranges of the two parent species overlap over a wide area, the distribution of the 316 Rhodora [Vol. 65 hybrid is restricted to a few localities which are distin- guished by their wealth of taxa with relict character and highly localized distribution. In these localities, S. oreopolus seems generally to be frequent or even common. According to Fernald (1927, p. 76), S. oreopolus is an “abundant species of subalpine woods and meadows of the Shickshock Mts. of Gaspé and . . . equally characteristic of subalpine slopes of northwest Newfoundland". On the wide expanse of Tabletop Mt. in the Gaspé, Fernald (1907, p. 106) found it to be “always more abundant than S. roseus and S. am- plexifolius". Stebbins (1929, p. 142) observed S. oreopolus on Mt. Katahdin to be: “Abundant on damp slopes above timber line in both the North and South Basins". We our- selves found it on Mt. Washington to be only local in distri- bution and less frequent than the other two taxa. It does not appear possible at present to give a well- founded explanation for the distribution of S. oreopolus. In the subalpine region of Mt. Washington, S. amplexifolius and S. roseus were observed by us in many localities growing in large numbers side by side. The flowering seasons of the two species were found also to be largely overlapping. Streptopus oreopolus was seen always in the vicinity of the two other taxa except for the occurrence along the Oakes Gulf Trail at 4300’ alt. where no S. roseus was observed. Similar conditions are indicated for the Shickshock Mts. by Fernald (1907). In agreement with the interpretation of S. oreopolus as a hybrid, the suggestion might therefore be made that the range of S. oreopolus represents simply areas of maximum possibilities for hybridization between S. am- plexifolius and S. roseus. As an additional factor determin- ing the distribution of S. oreopolus the possibility might be considered that the hybrid taxon represents a genotype which is especially well adapted for a subalpine or cool coastal environment. H4525 KENSINGTON, MONTRÉAL 28. DEPARTMENT OF BIOLOGY, MOUNT ALLISON UNIVERSITY, SACKVILLE, N. B. 1963] Streptopus oreopolus — Lóve and Harries 317 Acknowledgements: Both authors are indebted to the Mt. Washing- ton Summit House Inc., its staff and its President, Col. Arthur Teague, for many courtesies and facilities extended to them during their stay on Mt. Washington. The senior author expresses her gratitude to the National Research Council of Canada for financial assistance. The junior author extends his thanks for financial support of his field work to the Cramer Foundation of Dartmouth College, to Sigma Xi-RESA, and to Rutgers University. For help and encouragement throughout his work he is thankful to Prof. F. H. Bormann of Dartmouth College and to Prof. J. C. F. Tedrow of Rutgers University. For stimulating discussions and for the preparation of herbarium specimens he is in- debted to Prof. J. P. Poole, acting curator of the Jesup Herbarium of Dartmouth College. LITERATURE CITED FASSETT, N. C. 1935. A study of Streptopus. Rhodora 37:88-113. FERNALD, M. L. 1906. The genus Streptopus in Eastern America. Rhodora 8: 69-71. 1907. Streptopus oreopolus a possible hybrid. Rho- dora 9:106-107. 1926. Two summers of botanizing in Newfoundland. Rhodora 28:89-111, 115-129, 161-178. . 1927. Streptopus oreopolus in the White Mountains. Rhodora 29: 76. MARIE-VICTORIN, FRERE. 1929. Les Liliiflores du Québec. Contr. Lab. Bot. Univ. de Montréal 14: 1 - 202. MATSUURA, H. and SvuTÓ, T. 1935. Contributions to the idiogram study in phanerogamous plants. I. Journ. Fac. Sci. Hokkaido Imp. Univ. V, 5: 33-75. PEASE, A. S. 1924. Vascular flora of Coós County, New Hampshire. Proc. Boston Soc. of Nat. Hist. 37, 3: 39-388. SaATO, D. 1942. Karotype alteration and phylogeny in Liliaceae and allied families. I and II. Jap. Journ. Bot. 12:57-161. STEBBINS, G. L. JR. 1929. Some interesting plants from Mt. Katah- din. Rhodora 31: 142-143. THERMAN, E. 1956. Cytotaxonomy of the tribe Polygonatae. Amer. Journ. Bot. 43: 134-142. TISCHLER, G. 1950. Die Chromosomenzahlen der Gefasspflanzen Mitteleuropas. s'Gravenhage. TALL WHEATGRASS, A NEW ROADSIDE SPECIES IN UTAH During the 1940's the old world species, tall wheatgrass, Agropyron elongatum (Host.) Beauv., was considered pro- mising for reseeding pastures at low elevations in the Inter- mountain Region although it was not readily available for use at that time (Stoddart and Smith 1943, Stoddart 1946). During the 1950's, tall wheatgrass was found to be one of the highest producers of forage on good sites on lowland and mountain brush areas, and it was recommended highly for reseeding in Utah (Stewart et al. 1951, Plummer et al. 1955, Stoddart and Smith 1955). The use of tall wheatgrass for reseeding pastures in Utah County occurred as early as 1951, and tall wheatgrass has been seeded commonly during the last decade. In 1956 the author observed a few plants that had escaped cultivation along the roadside near Spanish Fork City, Utah County. Tall wheatgrass is now common along the roadsides in the vicinity of Spanish Fork and Springville, and it occurs occa- sionally along irrigation ditches. Only future observations wil answer the following questions. Will tall wheatgrass continue to spread throughout the state along the roadsides and irrigation streams, and will tall wheatgrass also become established in the native vegetation of the lower elevations in Utah? EARL M. CHRISTENSEN, BOTANY DEPARTMENT, BRIGHAM YOUNG UNIVERSITY, PROVO, UTAH. LITERATURE CITED PLUMMER, A. Perry, A. C. HULL, JR., GEORGE STEWART, and JOSEPH H. ROoBERTSON. 1955. Seeding rangelands in Utah, Nevada, southern Idaho, and western Wyoming. U. S. Dept. Agr., Agr. Handbook 71. 73 pp. STEWART, GEORGE, L. A. STODDART, HAROLD J. BURBACK, J. A. LIBBY, R. L. WRIGLEY, D. A. GAUFIN, and ROBERT A. RouNDY. 1951. Recommendations for range reseeding in Utah. Utah State Agr. Sta. Ext. Bul. 212. 12 pp. STODDART, L. A. 1946. Seeding arid ranges to grass. Utah Agr. Exp. Sta. Circ. 122. 29 pp. STODDART, LAURENCE A. and ARTHUR D. SMITH. 1943. Range Man- agement. First Ed. McGraw-Hill Book Co., New York. 547 pp. Illus. 1955. Range Man- agement. Second Ed. McGraw-Hill Book Co. 433 pp. Illus. 318 A CHECKLIST AND KEY TO THE SPECIES OF CAMPANULA NATIVE OR COMMONLY NATURALIZED IN NORTH AMERICA STANWYN G. SHETLER INTRODUCTION This short conspectus of Campanula in North America (including Greenland) was prepared in the course of bio- systematic studies on the C. rotundifolia complex, as a means of better understanding the relationships of that com- plex to the whole genus. It is compiled in large part from the literature but includes a substantial number of original observations and measurements. The last conspectus of North American campanulas was published by Asa Gray in 1886 (Syn. FEI, ed. 2, vol. 2 (1) : 11-14, 395-96), in which 14 species were recognized. Six new species have been de- scribed since then, but there is at present no single source to which one can turn for even a simple listing, much less an inclusive key. The present conspectus is being offered, despite its brevity and compiled nature, in the hope that it will satisfy a practical need for many until an original re- vision can be completed. I have made the key quite descriptive, more so than usual, because separate descriptions are not included in the check- list. Thus, while it is constructed primarily for flowering material, the condition in which campanulas are usually collected, one should be able to name most fruiting specimens by careful comparison of the extra descriptive material in the various leads, including geographical and ecological data. Additional information beyond what is supplied here should be sought in one or more of the appropriate regional floras and manuals, cited in my bibliography. In the key, size ranges are sometimes given statistically, eg.. 28-208 (118) cm., when new measurements have been made. Statistical ranges can easily be recognized by the fact that the parenthetic number lies between the other two values given. This number represents the arithmetic mean, i.e., the average, and the other two numbers represent a 319 320 Rhodora [Vol. 65 minimum-maximum range, derived by subtracting and add- ing two standard deviations from and to the mean. (All numbers have been rounded-off to the nearest whole num- ber, so as to appear in the actual units of measurement.) These estimated ranges should include approximately 95% of the plants one will encounter. Owing to the tendency for size-distribution curves to be truncated below (no negative values) and asymptotically skewed above, a measurement as small as the estimated minimum will rarely if ever be found, and the transgressions of the estimated range that one will occasionally encounter will almost always be trans- gressions of the upper limit. Whenever possible, the statisti- cal ranges were derived from samples of 25 or more, de- pending on the variability of the character, but limited ma- terial occasionally prevented samples this large. Sometimes the parenthetic number is larger than the upper extreme, which means, according to customary pro- cedure, that this value represents not an average but simply an uncommon upper extreme, and the min-max range repre- sents observed values rather than statistical estimates. Ob- served values are given for relatively invariable or non- critical characters. In some cases the observed ranges were taken from the literature, but only after being verified personally. Both the relative dimensions of the capsule and the rela- tive position of the pores or slits of dehiscence can be ex- tremely useful if not diagnostic in the determination of campanulas, and whenever possible collectors should take specimens with flowers and fruits. Not infrequently, how- ever, one has only flowering material but really needs to use capsule characters for certain discrimination. While the relative dimensions (or shape) of the ovary are not likely to change significantly in the course of maturation, the posi- tion of the pores is virtually impossible to ascertain prior to actual dehiscence. However, the place of dehiscence tends, on a species basis, to be correlated with the width of the capsule, and more often than not the pores or slits will ap- pear near the widest portion. This means that as a fairly 1963] Campanula — Shetler 321 safe rule-of-thumb one can assume that dehiscence will be near the top if the ovary is widest above the middle and will be near the middle or below if the ovary is widest here. This rule will work best after anthesis and the beginning of maturation, when the capsule will quickly assume its characteristic shape if it has not before, but sometimes the rule will mislead on a given plant or even throughout an en- tire species (e. g., C. divaricata). Although species are very constant in the position of dehiscence, they can be quite vari- able in their capsule dimensions; hence, it is advisable to ascertain the average shape from as many ovaries as pos- sible before attempting to use this rule. The inflorescence is basically cymose in most species of Campanula, but this fact is seldom obvious. Secondarily, it is usually spicate, racemose, or paniculate, and floristic writers usually describe bellflower inflorescences in these terms. Such practice is simpler and not seriously misleading, and I have therefore followed it in my key. Previous workers have described infraspecific taxa for at least one-third of the 20 native species recognized here, and doubtless some of these taxa are good. In virtually all cases, however, thorough taxonomic investigation is needed before valid judgment can be made, and I have refrained from formal recognition of any infraspecific taxa in this con- spectus. Chromosome numbers, when known, have been given in the Checklist, and unless otherwise indicated these num- bers have been taken from the Darlington and Wylie Chro- mosome Atlas. ARTIFICIAL KEY TO THE SPECIES 1. Plants low or dwarf, 3-20 (35) cm. tall but mostly less than 15 cm., erect or decumbent perennials of alpine’ or arctic situations, or erect, divaricately cymose-branched (pseudo-dichotomous), tem- perate annuals of low elevations. 2. 2. Annuals, divaricately cymose-branched, without rosette; hypan- thium glabrous. Temperate species of low elevations (usually below 1200 m.), on dry rocky or grassy ridges and slopes, some- times in scrub or chaparral. 3. 222 Rhodora [Vol. 65 3. Corolla rotate, cut nearly to the base into linear-lanceolate lobes; style conspicuously exserted; capsule subglobose, open- ing at or below the middle. Grassy slopes. Chinsegut Hill, Fla. (Rare endemic) ................... eese C. ROBINSIAE 3. Corolla campanulate or funnelform, never rotate nor cut below the middle, the lobes broadly lanceolate or deltoid-ovate; style included or barely exserted; capsule various. 4. 4. Corolla inconspicuous, 2-3.5 (2.5) mm.’ long, about equaling length of calyx-lobes; stamens about 2 mm. long; leaves ovate, coarsely dentate; capsule urceolate, opening near the middle. Burns and disturbed places, Chaparral Zone. Mts. of N. Calif. (Endemic) ...................... C. ANGUSTIFLORA 4. Corolla showy, 5-20 mm. long, conspicuously longer than calyx-lobes; stamens 4-6 mm. long; leaves oblanceolate or lanceolate to linear-lanceolate, remotely denticulate; capsule various. 5. 5. Calyx-lobes usually 1/2-2/3 length of corolla; corolla 5-12 (9) mm. long; plants several times branched, often starting from a few cm. above base of stem; capsule urceolate, opening near the middle. Rocky ridges and talus, Chaparral Zone. Mts. of N. Calif. (Endemic) ...... ————————————————"—"— C. EXIGUA 5. Calyx-lobes usually less than 1/2 length of corolla; corolla 5-17 (11) mm. long; plants profusely branched, usually starting at base of stem; capsule ellipsoid or obovoid, opening above the middle, from the top of the pore down- ward (unique; all other species open from bottom up- ward). Granite areas. Cent. Tex. (Endemic) .................. NENNEN C. REVERCHONI 2. Perennials, unbranched or with a few short lateral branches near the top, never appearing dichotomous, often with prominent rosette; hypanthium glabrous or pubescent. Temperate alpine (usually above 1500 m.) or arctic species, never in dry scrub or chaparral. 6. 6. Hypanthium villous, sometimes sparsely or only in lines; stems unbranched, one-flowered or rarely short-branched above and few-flowered. High alpine or arctic fell-fields, barrens, screes, rubbles, cliffs, talus, alluvial fans, sandy and gravelly shores, turfy tundra slopes and plains, dwarf shrub-lichen heaths, etc. Low to high arctic species, reaching southward only in seattered high alpine situations, if at all. 7. 7. Corolla large, showy, usually over 20 mm. long, tubular or deeply campanulate or funnelform, spreading or nodding’; anthers never under 4 mm. long; calyx-lobes entire or toothed, erect or spreading, but never connivent; rosette usually prominent; lower leaves broadly obovate, spatulate 1963] Campanula — Shetler 323 or oblanceolate, uniformly crenate or serrate; plant not appearing fleshy; capsule broadly urceolate or obconic. 8. 8. Calyx with reflexed auricular appendages between the lobes, the lobes entire (rarely crenate-serrate); lower leaves broadly spatulate or obovate, crenate; capsule pendant, opening near the base. Alpine zone. Aleut, Is. CAmbpltu-PROHIC ui isse a C. CHAMISSONIS 8. Calyx without appendages, the lobes with 1-several sharp teeth or laciniations; lower leaves broadly oblanceolate, remotely sharp-serrate or laciniate; capsule erect, open- ing near the top. Partial to rocky or gravelly sites. Wide- spread in Alaska from Aleut. Is. to Arctic Slope, east- ward in scattered localities to Great Bear Lake, Mack. (NWT), southward in mts. to N. Wash. (Amphi-Pacific) C. LASIOCARPA Meettthsshihhsreshsussoessssososossussassesososeosssssesesssessoecssosseseeso 7. Corolla small, inconspicuous, seldom over 12 mm. long, fun- nelform, erect to nodding; anthers 1-3 mm. long; calyx- lobes entire, erect, appressed to corolla, sometimes conni- vent; rosette absent or few-leaved, not prominent; lower leaves narrowly obovate or spatulate, entire or less often irregularly crenate; plant appearing fleshy; capsule dis- tinctively obovoid or clavate, usually broadest just above the middle, erect, opening near the top. Partial to calcare- ous, often turfy or alluvial, sites. Widespread but scattered, in Alaska from E. Aleut. Is. to Arctic Slope, eastward to Ellesmere Is. and Greenl, southward in Rocky Mts. to Utah and Colo. and to Shickshock Mts., Que. (Circumpolar, With large TaNnve Caps) oaaae opo EC C. UNIFLORA 6. Hypanthium glabrous, scabrellous or hirtellous; stems charac- teristieally unbranched, one-flowered, but in some species rather frequently few-branched above. Habitats various. Temperate sub- and low alpine or arctic species. 9. . Hypanthium sparsely to densely scabrellous or hirtellous; plants rarely over 10 em. tall; leaves mostly basal or nearly so; capsule urceolate, ellipsoid, or short-obconic, opening near the middle or above. Alpine fell-fields, screes and crevices. 10. 10. Leaves sharply and coarsely serrate or dentate; plants sparsely to densely scabrellous above, including hypan- thium, glabrous below; calyx-lobes usually with l- several sharp teeth. Olympic Mts., Wash. (Endemic) C. PIPERI Besbtsssosssseseciess ihres usur ime] et rrr 10. Leaves entire; plants uniformly scabrellous or hirtellous throughout, usually densely so; calyx-lobes entire. N. Calif. to Wash., eastward to Ida., W. Mont. (Endemic) C. SCABRELLA 324 Rhodora [Vol. 65 9. Hypanthium glabrous; plants often much over 10 cm. tall; basal leaves frequently absent, cauline ones sometimes many; capsule various. 11. 11. Corolla rotate or shallowly campanulate-spreading, cut well below the middle; style conspicuously exserted, often recurved; calyx-lobes usually with 1-several prom- inent teeth near base; leaves narrowly lanceolate or oblanceolate, remotely denticulate; capsule obconic, erect, opening near the top. Dry limestone talus and cliffs, alpine crevices and rocky places, gravelly stream- banks, ete. South Slope of Brooks Range, Cent. Alaska, to Cent. Yuk. and S. W. Mack. (Endemic) .... C. AURITA 11. Corolla shallowly to deeply campanulate or funnelform, never truly rotate, seldom cut below the middle; style shorter than to barely exceeding corolla, not recurved; sepals toothed or entire; leaves and capsules various. 12. 12. Anthers 1-3 mm. long; corolla short-funnelform, seldom over 12 mm. long, erect to nodding; calyx-lobes erect, appressed to corolla, sometimes connivent, entire; plant appearing fleshy, one-flowered (very rarely few-flowered) ; lower cauline leaves entire or irregu- larly crenate; ovary and capsule distinctively obovoid to clavate, usually broadest just above the middle, erect, opening near the top. (Cf. No. 7 above for habitats and distrib.) ............... eee C. UNIFLORA 12. Anthers 3.5 mm. or mostly longer; corolla variously shaped, often much exceeding 12 mm. long, erect to nodding; calyx-lobes if erect and appressed never connivent, sometimes toothed; plant not appearing fleshy, frequently several-flowered (on short axillary branches above); lower cauline leaves usually sharp- toothed, at least remotely; ovary and capsule various- ly shaped but seldom broadest near the middie, erect or nodding, opening above or below. 13. 13. Plants 2-14 (8) cm. tall, completely glabrous; longest cauline leaves 7-17 (12) mm. long, ovate or lance- oblong with cuneate base, sharply denticulate or serrate at least at apex; corolla shallowly funnel- form, 9-15 (12) mm. long, cut nearly or quite to the middle, erect; capsule obconic, erect, opening near the top. Moist subalpine meadows and stream- banks. Mts. of N. Calif. (Endemic) .......................... 499993492 20 0479732303192 *59223 0259 83 ERAE E9244 Eos 14415 84983 F2 C. WILKINSIANA 13. Plants mostly over 10 em. tall and often to 20 em. or more, usually pubescent below at least in lines or 1963] Campanula — Shetler 325 with ciliate leaf bases; longest cauline leaves rarely under 20 mm, and usually much longer, linear- to ovate-lanceolate or oblanceolate, remotely serrulate occasionally close-serrate; corolla variously shaped, seldom as short as 15 mm. or cut to the middle; capsule various. 14. 14. Plants glabrous except for ciliate bases of lower cauline and rosette leaves, the cilia 0.5-0.7 (1.0) mm. long, typically one-flowered but not infre- quently with several axillary flowers above; corolla funnelform-spreading, erect; calyx-lobes erect, appressed to corolla, usually with several conspicuous denticulations near base; capsule obconic, erect, opening near the top. Subalpine meadows and open rocky or gravelly places. Rocky Mts.: Ariz. and N. Mex. to W. Mont, Cent. Ida., and Wenatchee Mts., Cent. Wash. (Endemic) ..... a eit iesnceeoscsetnese C. PARRYI 14. Plants glabrous or more often hirtellous below at least in lines on stem and leaf-bases, but hairs never longer than 0.2 mm., typically with 2-many flewers but sometimes reduced to 1 in extreme alpine situations; corolla deeply campanulate or funnelform to tubular, divergent or nodding; calyx-lobes if erect seldom appressed to corolla, always entire; capsule hemispherical or broadly urceolate, pendant (apparent in herbarium ma- terial by conspicuously arcuate pedicel), opening near the base. Subalpine to alpine and arctic meadows, fell-fields, screes, crevices and ledges, gravelly talus, alluvial fans, moist rocky shores, ete. Widespread, E. Aleut. Is. to S. Greenl., southward to Ore., Colo., Que., and Nfld. (Cir- cumpolar) .........555 4 9: EE a C. ROTUNDIFOLIA [Highly polymorphic species with extremely wide ecological amplitude. Only dwarf arctic-alpine specimens are intended to key out here; others should key out in next section.] 1. Plants tall, typically much over 20 em. but occasionally shorter (cf. especially C. scouleri and the marsh species, which all key out here), erect, decumbent, or weak and reclining biennials or per- ennials, or if annual then nct divaricately cvmose-branched (C. americana), only of temperate latitudes and low elevations (ex- cept C. rotundifolia, which keys out in both sections). 15. 15. Stem slender, weak and more or less reclining, sometimes ret- rorsely scabrous; inflorescence very lax, with terminal and 326 Rhodora [Vol. 65 axillary flowers on slender pedicels; corolla rarely to 20 mm., usually much shorter; capsule short-hemispherical to subglo- bose, opening at or below the middle. Wet meadows, marshes, swamps, and bogs. 16. 16. Corolla rotate, cut almost to the base into linear-lanceolate lobes; stem and leaves glabrous; leaves narrowly lanceolate or oblanceolate, crenulate or crenulate-serrulate. Marshes, swamps, and wet borders and waysides. Fla. Peninsula. (Endemic) .......... eee nnne C. FLORIDANA 6. Corolla campanulate or funnelform, rarely cut below the middle, the lobes broadly lanceolate or deltoid-ovate; stem and leaves (midribs and margins) usually retrorsely scab- rous; leaves various. 17. 17. Leaves linear or narrowly lanceolate or oblanceolate (rarely ovate), mostly 3 em. or longer, entire or remotely serru- late; corolla 6-12 mm. long; calyx-lobes entire or with a basal pair of denticulations. Wet meadows, marshes, swamps, and bogs. E. and Cent, U. S. and adjacent Can.: Sask. to Que., southward to Nebr., Mo., and Ga. (Endem- IC) BERNER C. APARINOIDES 17. Leaves short-ovate, elliptic or obovate, to 2.5 em. long, ob- scurely to prominently erenate-serrate at least above the middle; corolla 10-15 mm. long; calyx-lobes retrorsely scabrous on the margins near the base. Coastal fresh- water swamps and bogs. Cent. Calif. (Endemic) .............. LUETETERINSEPNTIPEATEENO SUP) EFAQURAERERARSARARARASARRASRATARSIUEERINSEDAER C. CALIFORNICA 15. Stem slender to robust, firm, erect or decumbent (sometimes reclining or pendant in C. rotundifolia, but then without retrorse pubescence or rotate corolla), never retrorsely scab- rous, though sometimes hispidulous with some reflexed hairs (C. prenanthoides) ; inflorescence lax or strict; corolla various but often longer than 20 mm.; capsule various, sometimes opening near the top. Habitat dry to moist but never a marsh, swamp or bog. 18. 18. Style conspicuously exserted, often recurved; corolla short- campanulate or cut to the middle or beyond and funnelform- spreading or rotate; hypanthium glabrous; inflorescence various but flowers never in terminal head. 19. 19. Corolla cut almost to the base, the lobes linear-lanceolate or broader, rotate or funnelform-spreading. 20. 20. Corolla rotate, the lobes broadly lanceolate to deltoid- ovate; style uniquely declined and upcurved; inflores- cence a spike of 1- 3(5)-flowered axillary cymes, usually very floriferous and wand-like; rank, more or less virgate annual (biennial), 28-208 (118) em. tall, but seldom under 50 cm.; stem very robust; leaves 1963] Campanula — Shetler 327 attenuate-acuminate, coarsely serrate; capsules slender- obconic, typically in triplets, stiffly erect on straight pedicels, opening near the top. Open woods, borders, shaded roadsides, and bottoms. Widespread, Long Is. through S. Ont. to Minn., southward to N. W. Fla., Miss., and Kan, (Endemic) ....................-— C. AMERICANA 20. Corolla funnelform-spreading, the lobes linear-lanceolate; style often variously recurved but not regularly de- clined and upcurved; inflorescence spicate, racemose, or paniculate but not with 3-flowered cymes, few-flowered ; slender, hardly virgate perennial (biennial), 20-80 cm. tall, but usually less than 50 cm.; leaves acute to short- acuminate, serrate; capsules hemispherical or urceolate, single or clustered but not in triplets, more or less pen- dant, on geniculate pedicels, opening at or below the middle. Dry open woods, Transition Zone below 1800 m. Southern B. C. to Cent. Calif., mostly west of Cas- cade Mts. (Endemic) .............. C. PRENANTHOIDES 19. Corolla cut shallowly, sometimes to the middle but rarely beyond, short-campanulate or funnelform-spreading, the lobes broadly lanceolate to deltoid-ovate, never approach- ing linear. 21. 21. Inflorescence a profusely branched compound panicle, the flowers numerous; corolla short-campanulate, seldom cut to the middle, 6-8 mm. long; leaves coarsely, almost laciniately toothed; style straight; capsule ovoid to short-obconie, erect, on straight pedicel, opening near the base. Cliffs, dry rocky woods and waysides in Ap- palachian and Blue Ridge Mts., chiefly at low elevations. S. E. U.S.: Md. to Ga. and Ala., rarely adventive far- ther north. (Endemic) ......................-..- C. DIVARICATA 21. Inflorescence spicate or racemose, simple or few-branched, the flowers few; corolla funnelform-spreading, cut to about the middle; leaves serrate or crenate-serrate; style often recurved; capsule urceolate or ovoid, more or less pendant, on geniculate pedicel, opening near the middle. Open to dense woods, talus and outcrops, Tran- sition Zone to about 1500 m. Alaska Panhandle to N. Calif., mostly west of Cascade Mts, (Endemic) ............. diasassastengegscostqsssesosbososadoeopaqva To FE UCIR IN OROERERRNE TORRE C. SCOULERI 18. Style included or barely exserted, never recurved; corolla deeply campanulate or funnelform to tubular, rarely cut as deep as the middle; hypanthium frequently pubescent; flowers sometimes in a terminal head. 22. 22. Flowers sessile, in involucrate terminal heads and axillary glomerules; leaves elliptic or oblong to ovate, obtuse or 328 Rhodora [Vol. 65 acute, crenate-serrate, petiolate below, often clasping above; capsule ovoid, erect, opening near the base. Waysides and waste places. Occasional escape from cultivation. (Eurasian) ............ ees C. GLOMERATA 22. Flowers with short to long filiform pedicels, solitary or clustered in the axils or in loose to strict racemes or panicles; leaves variously shaped, acuminate, entire to sharply serrate, sessile or petiolate, but never clasping; capsule broadly hemispherical, urceolate, ovoid, or near- ly globese, pendant, opening near the base. 23. 23. Hypanthium, calyx-lobes, and midveins outside unex- panded corolla bristly with white hairs, 0.5-1.5 (2.0) mm, long; lower cauline leaves deltoid or deltoid- ovate, distinctly petiolate, coarsely sharp-serrate or laciniate to crenate-serrate; flowers erect or spread- ing, in pedunculate axillary clusters or on slender pedicels, forming a loose raceme or panicle. Way- sides and waste places. Frequent escape from culti- vation. (Eurasian) ......... eee C. TRACHELIUM 23. Hypanthium and calyx-lobes glabrous or hispidulous, the hairs to 0.2 (0.5) mm. long; unexpanded corolla glabrous on the midveins; lower cauline leaves linear or narrowly to broadly lanceolate, sometimes sessile, entire or toothed; flowers divergent or nodding; in- florescence various. 24. 24. Hypanthium and calyx-lobes glabrous; flowers on slender pedicels, terminal and axillary, sometimes forming loose raceme or panicle, never secund; lower cauline leaves linear or narrowly lanceolate, sessile or acuminately tapered to winged petiole, entire or remotely serrulate (infrequently close- serrate); stems slender, more or less delicate, vir- gate or lax and decumbent, sometimes reclining or pendant from cliffs and crevices; corolla blue. Rocky or sandy shores, cliffs, mesophytic canyons and bottoms, open woods, and subalpine meadows. Widely distributed from E. Aleut. Is. to Greenl., southward to northern Mex., W. Tex., Mo., and W. Va. (Circumpolar, wide-ranging from temper- ate to arctic regions; cf. 14 above) vce —————— EH C. ROTUNDIFOLIA [Many Alaskan specimens and occasional others have broadly lanceolate or ovate, distinctly petio- late, and often closely serrate leaves; othcrwise the characters are the same.] 24. Hypanthium, calyx-lobes and pedicels usually his- 1963] Campanula — Shetler 329 pidulous; flowers on short pedicels, often nearly sessile, forming a strict, secund, spicate raceme; lower cauline leaves ovate-lanceolate, abruptly tapered to distinct, wingless petiole, finely to coarsely crenate-serrate; stems robust, the plants coarse and strictly erect; corolla violet. Waysides and waste places. Very common escape fron culti- vation, often becoming naturalized, (Eurasian) .... Cece E E E e C. RAPUNCULOIDES ‘Unless specifically modified, the terms “alpine” and “arctic” are used broadly here, as is customary, including those habitats or regions that more properly should be called “subalpine” and “subarctic.” "This is an observed range, and the parenthetic number is the mode, the most frequently observed value, not the mean. For characters having such narrow ranges of variability as this, the mode is perhaps more useful than the mean. "By “corolla... spreading or nodding” is meant, technically, “flower... spreading or nodding,” here and in all similar places in the key. CHECKLIST OF THE SPECIES CAMPANULA L. Bellflowers, Bluebells, Campanulas, Harebells. (Species prefixed with an asterisk (*) are endemic to North America.) *1, C. AMERICANA L. (C. acuminata Michx.; Campanulastrum ameri- canum (L.) Small. Tall Bellflower. N = 51. Perhaps the most characteristic American species. *2, C, ANGUSTIFLORA Eastw. Eastwood's Harebell. *9. C. APARINOIDES Pursh. (Incl. C. uliginosa Rydb.). Eastern Marsh Bellflower. Rydberg's C. uliginosa, the large-flowered form of the eastern marsh bellflower, is in its extreme quite distinct and, while hardly a good species, probably merits recognition as a separate race. C. AURITA Greene. Yukon Bellflower. C. CALIFORNICA (Kell.) Heller. (Wahlenbergia californica Kell.; C. linnaeifolia A. Gray). California Swamp Harebell. 6. c. cHAMISSONIS Fédorov. (C.dasyantha auct. pl., non M. á Bieb.; C. pilosa sensu A. Gray, non Pall. ex Roem. et Schult.). Aleu- tian Bellflower. N = 17. *7T. C. DIVARICATA Michx. (C. flexuosa Michx.). Appalachian Bell- ficwer. N = 20. *&. c. EXIGUA Rattan. Chaparral Campanula. *9. C. FLORIDANA S. Wats. ex A. Gray. (Rotantha floridana (S. Wats. ex A. Grav) Small). Florida Campanula. This species may prove to be merely a southern race of the widespread C. * cU A aparinoides. [c. GLOMERATA L. Clustered Bellflower. N = 17, 34. Intreduced. j 10. Cc. LASIOCARPA Cham. Alaska Bellflower. In the Aleutians this 330 Rhodora [Vol. 65 species frequently grows intermixed with C. chamissonis, and the two species are often confused by collectors and taken as one species, getting mounted on one herbarium sheet. Several good characters distinguish them, however, and only the most depauperate specimens should cause trouble (cf. key). *11. c. PARRYI A. Gray. (C. planiflora Engelm., non Lam.; C. langs- dorffiana sensu A. Gray, non Fischer ex A. DC.). Rocky Mountain Bellflower. N — 17 [Shetler, in manuscript]. Plants from Washington, Idaho, and Montana tend to differ from typical C. parryi by having entire calyx-lobes and leaves. The calyx-lobes and the corolla, which is borne on a shorter ped- uncle, are shorter on the average. In these characters and in the not infrequent puberulence of the hypanthium, the plants of this general region approach the otherwise quite distinet C. scabrella. McVaugh (1942) segregated these plants as var. idahoensis of C. parryi. Recent floristic workers in the Rocky Mountain region have generally followed him. While MeVaugh's taxon seems distinct enough on the whole, at least as he originally circumscribed it, further study is required to determine its true affinities. Some evidence sug- gests that it may represent a hybrid series between C. parryi and C. scabrella, and again other pieces of evidence hint that var. idahoensis has become a catch-basket epithet for several discordant elements and that certain plants hitherto referred here might in fact represent an as yet undescribed species. Until this question is resolved, occasional plants will continue to turn up from the Pacific Northwest that can be relegated to C. parryi, as var. idahoensis, only with doubt. This is par- ticularly true of plants from the Wenatchee Mountains of Washington and the mountains of southwestern Montana. Most of the plants from these areas presently available in collections are in the flowering condition, and what are ur- gently needed are mature capsules. *12. c. PIPERI Howell. Olympic Bellflower. N — 17. *18. C. PRENANTHOIDES Durand. (Asyneuma prenanthoides (Durand) McVaugh). California Harebell. By itself, this species strik- ingly resembles species of the Asiatic genus Asyneuma, to which MeVaugh (1945) has referred it. But the degree to which C. prenanthoides shares its characters in varying com- binations with other American campanulas is such that I find no compelling reasons at present for separating it from the other 19 species recognized here. Quite possibly intensive study will require a rather extensive realignment in the family Campanulaceae as a whole, but until more convincing and comprehensive data are available, isolation of this one American species seems premature. 1963] Campanula — Shetler 331 [C. RAPUNCULOIDES L. Rover Bellflower. N — 51. Introduced. By far the most common garden escape. Widely adventive or natur- alized along roadsides and borders, it is frequently confused with the native C. americana, which it only superficially resembles. Apart from the rank habit and sometimes similar habitat, these plants are wholly unlike. ] *14. C. REVERCHONI A. Gray. Texas Bellflower. *15. C. ROBINSIAE Small. (Rotantha robinsiae (Small) Small). Des- pite the fact that Small (1933, p. 1508) put this species in the segregate genus Rotantha, which he erected, along with C. floridana, because they share the rotate corolla, these spe- cies are not at all closely related, As Small (1926) himself so aptly pointed out in his original description, C. robinsiae is most closely related to the Texan C. reverchoni. According to him (p. 36), *It differs conspicuously, however, in the glab- rous leaves, the short hypanthium, the smaller calyx, the smaller corolla with a shorter tube, and the subglobose cap- sule." The distribution of this species is so restricted (Chin- segut Hill, Fla.) that one is led to raise certain questions about its origin. This part of Florida is not particularly note- worthy for local endemics, and in fact it is a bit hard to explain how this species or any other has come to survive only here. As one can readily see upon visiting Chinsegut Hill, the immediate vicinity has been disturbed by cultivation and plantings for a considerable period of time, and the possibility must not be ruled out that Campanula robinsiae represents a pre-1926 Eurasian introduction, perhaps accidental. 16. c. ROTUNDIFOLIA L. s.l. (Incl.: C. r. var. alaskana A. Gray; C. hete- rodoxa auct., non Vest in Roem. et Schult.; C. intercedems auct., non Witasek; C. latisepala Hult.; C. linifolia auct., non Scop.; C. macdougalii Rydb.; C. petiolata A. DC., C. rotundi- folia X latisepala sensu Hult.; C. sacajaweana Peck; C. scheu- chzeri auct., non Vill.). Harebell, Bluebells-of-Scotland. N = 17, 28, 34. This species is extremely polymorphic in North America, as throughout the Old World, and is comprised of several geographic races, which I have been studying. For the purposes of this synopsis, I have united the numerous biotypes under the single taxon, C. rotundifolia, which thereby becomes by far the most cosmopolitan North American campanula. It is the only species known from Mexico, where it has been col- lected in the states of Coahuila, Nuevo León, and Tamaulipas. Owing to its extreme variability, it is from time to time confused with almost every other American species. In the Rocky Mountain region, particularly, it is often confused with C. parryi, which however is amply distinct. The two species are distinguishable at a glance when mature capsules are 332 Rhodora [Vol. 65 present, but good floral and vegetative characters also exist (cf. key). *17. C. SCABRELLA Engelm. Downy Alpine Bellflower. *18. C. SCOULERI Hook. ex A. DC. Scouler’s Harebell. In addition to the means provided in the key, it can be distinguished from C. prenanthoides, to which it is most similar, as follows. While the latter is usually quite leafy below the inflorescence, having 12 or more leaves which are sessile, C. scouleri tends to have fewer than 10 leaves which are mostly petiolate. The iowest cauline leaves of C. scouleri are also more broadly ovate, often nearly rotund. [C. TRACHELIUM L. Nettle-leaved Bellflower, Throatwort. N — 17. Introduced. Of the species included here, it is more similar to C. rapunculoides, from which it can readily be distinguished by the pubescence. } 19. C. UNIFLORA L. Arctic Campanula, The only truly high alpine- high arctic campanula in North America. Single-flowered specimens of C. parryi and C. rotundifolia are mistaken for it frequently in the Rocky Mountain region, but there is little excuse for this. The much rarer C. uniflora is quite distinct and has no close relatives here. On the average it occurs at higher elevations than either of the other species. Typically, the hypanthium has long white trichomes, as C. chamissonis and C. lasiocarpa, but in the Colorado Rockies plants with a glabrous hypanthium are common. Perhaps they should be segregated as a separate geographic race, but further study is needed. *20. C. WILKINSIANA Greene. Wilkins’ Harebell. An endemic of Mt. Shasta and the Trinity Mountains in California, this cam- panula is surely quite close to the Olympic Mountain endemic of Washington, C. piperi. DOUBTFUL AND EXCLUDED SPECIES C. DASYANTHA M. á Bieb. (C. pallasiana Vest in Roem. et Schult.; C. pilosa Pall. ex Roem. et Schult.). An Asian species (not the C. dasyantha of American authors, cf. C. chamissonis) that does not to my knowledge occur in North America, despite the range statement of Fédorov in “Flora SSSR” (vol. 24: 278), including Canada. C. RENTONAE Senior. Described from plants grown in culture from seed originating in the Wenatchee Mountains of Washington. It seems to belong to the Parryi-Scabrella series usually re- ferred to C. parryi var. idahoensis, but needs more study. COMMENTS The single most striking fact deriving from this con- 1963] Campanula — Shetler 333 spectus is that of the 20 species recognized as native to North America 16 (80%) are endemic, sometimes to a very localized area of the continent. The specific differences a- mong the 20 species, though in some cases quite small, are nonetheless remarkably stable, and few taxonomists would dispute the discreteness of these species. If anything, the inclination might be to split several of them into two or more smaller species. Perhaps significantly, the endemics are confined largely to unglaciated parts of North America — areas south of maximum ice advance, refugial islands within the continental ice sheet, or areas where discontin- uous montane glaciers might have left sufficient nunataks for hardy species, like some of the Western endemic alpine bellflowers, to survive. Species illustrative of these three types of survival areas would be C. divaricata (Appala- chians), C. aurita (Yukon Tablelands), and C. parryi (Rockies), respectively. Although two widespread Eastern endemics, C. americana and C. aparinoides, presently occur inside the southern margin of the glaciated region, this dis- tribution probably represents a post-Pleistocene invasion from the south. Whether these and other endemic species occupied much greater areas of the glaciated region prior to Pleistocene times is hard to say, but it is very doubtful that any have done so in recent times. Greenland is without en- demic species. Only C. rotundifolia and C. uniflora, both cireumpolar, occur there as also in the Canadian Eastern Arctic. So highly endemic and localized is the North American campanula flora that one is tempted in constructing a key to to ignore the morphological characters and use geographical and ecological criteria for the primary dichotomies. The utility of geography is at once apparent from the following breakdown, in which I have divided North America into a number of arbitrary, though partially natural, regions and listed under each those species of Campanula native throughout or in some part of that region. For highly lo- calized species, the exact area is listed in parenthesis after the species. Any given species may appear in more than 384 Rhodora [Vol. 65 one regional list, but the endemics, which are starred (*), are generally confined to one of the regions. If not, they are starred only in the region of their principal range (cf. C. americana). Thus, by observing asterisks one can quickly grasp the relative distribution across North America of the 16 endemic species. ALASKA AND NORTHWEST CANADA *C. aurita C. rotundifolia C. chamissonis (Aleutians) C. scouleri (Alaska Panhandle) C. lasiocarpa C. uniflora CANADIAN EASTERN ARCTIC AND GREENLAND C. rotundifolia C. uniflora PACIFIC STATES AND ADJACENT BRITISH COLOMBIA *C. angustiflora (Calif.) *C. prenanthoides *C. californica (Calif.) C. rotundifolia *C. exigua (Calif.) *C. scabrella C. lasiocarpa (N. Wash., S. B.C.) *C. scouleri C. parryi (C. Wash.) *C. wilkinsiana (Calif.) *C. piperi (Wash.) ROCKY MOUNTAIN REGION (Northern Mex. to C. Alta.) C. lasiocarpa (Alta.) C. rotundifolia *C. parryi C. uniflora NORTHCENTRAL AND NORTHEASTERN U.S. AND ADJACENT CANADA *C. americana C. rotundifolia *C. aparinoides C. uniflora (Gaspé) SOUTHCENTRAL U.S. (All species infreq. and scattered or very localized) C. americana *C, reverchoni (C. Texas) C. aparinoides C. rotundifolia SOUTHEASTERN U.S. C. americana *C. floridana (Fla.) C. aparinoides *C. robinsiae (Fla.) *C. divaricata Any thoroughgoing consideration of relationships is not possible at present and speculation is premature; neverthe- 1963] Campanula — Shetler 335 less, the following species-groups are patent even to the casual student: I. DWARF ANNUAL ENDEMICS C. angustiflora C. exigua C. reverchoni C. robinsiae II. ARCTIC-ALPINE ENDEMICS C. aurita C. parryi C. piperi C. scabrella C. wilkinsiana III. MARSH-SWAMP-BOG ENDEMICS C. aparinoides C. californica C. floridana IV. PACIFIC COASTAL FOREST ENDEMICS C. prenanthoides C. scouleri V. AMPHI-PACIFIC ARCTIC SPECIES C. chamissonis C. lasiocarpa In conclusion, it should be emphasized again that this synopis is not in any way offered as a revision of Campanula in North America. At the same time there is little reason to expect that such a revision, when it does appear, will alter the present alignment of species greatly if at all. But while the species are reasonably well defined, the story at the lower levels is somewhat different. Insofar as the New World is concerned, the problems of Campanula have to do mainly with infraspecific variation and the question of geographic races within several of the more widespread and polymor- phic species (e. g., C. aparinoides, C. rotundifolia). ACKNOWLEDGMENTS To Drs. Olga Lakela, Univ. of South Florida, and Velva Rudd, Smithsonian Institution, I am grateful for informa- tion supplied on the exceedingly rare C. robinsiae, and to Dr. Daniel B. Ward, Univ. of Florida, I tender sincere 336 Rhodora [Vol. 65 thanks for taking me to Chinsegut Hill, the type and only locality for this species. SMITHSONIAN INSTITUTION LITERATURE CONSULTED ABRAMS, L., AND R. S. FERRIS. 1960. Illustrated Flora of the Pacific States. Vol. IV, pp. 72-77. Stanford Univ, Press, Stanford. BócHER, T. W., K. HOLMEN, AND K. JAKOBSEN. 1957. Grønlands Flora. pp. 195-196. P. Haase and Sons, Copenhagen. DARLINGTON, C. D., AND A. P. WYLIE. 1955. Chromosome Atlas of Flowering Plants. pp. 287-289. George Allen and Unwin Ltd., London. FERNALD, M. L. 1950. Gray's Manual of Botany. Ed. 8. pp. 1351-1353. American Book Co., New York. Féporov, A. 1957. “Campanulaceae.” In Flora SSSR, vol. 24, pp. 126-476, Acad. Sci. USSR, Leningrad. [In Russian] GLEASON, H. A. 1952 The New Britton and Brown Illustrated Flora. Vol. 3, pp. 314-316. Lancaster Press, Inc., Lancaster (Pa.). , AND A, CRONQUIST. 1963. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. pp. 662-663. D. Van Nostrand Co., Princeton. Gray, A. 1886. Synoptical Flora of North America. Vol. 2, pt. I, Gamopetalae after Compositae, ed. 2, pp. 11-14, 395-96. Ivison, Blakeman, Taylor and Co., New York. HARRINGTON, H. D. 1954. Manual of the Plants of Colorado. pp. 525- 526. Alan Swallow, Denver. HriTCHCOCK, C. L., et al. 1959. Vascular Plants of the Pacific North- west. Pt. 4, pp. 483-488, Univ. Wash. Press, Seattle. HULTÉN, E. 1960. Flora of the Aleutian Islands. Ed. 2. pp. 836-839. J. Cramer, Weinheim. 1949. Flora of Alaska and Yukon. Pt. 9. Lunds Univ. Arsskr. II Sect. 2, bd. 45: 1455-1464. McVavaH, R. 1942. A new Campanula from Idaho. Bull. Torrey Club 69: 241-243. . 1945. Notes on North American Campanulaceae. Bar- tonia 23: 36-40. 1951. “Campanulaceae.” I» Flora of Texas, by C. Longworth and collaborators. Vol. 3, pt. 5, pp. 331-366. Southern Methodist Univ. Press, Dallas. MuNz, PHILIP A. 1959. A California Flora. pp. 1062-1063. Univ. Calif. Press, Berkeley. RYDBERG, P. A. 1922. Flora cf the Rocky Mountains and Adjacent Plains. Ed. 2. pp. 823-824. Hafner Publishing Co., New York, (Re- print, 1954) 1963] Book Review — Steele 337 . 1932. Flora of the Prairies and Plains of Central North America. pp. 756-757. N. Y. Bot. Gard., New York. SMALL, J. K. 1903. Flora of the Southeastern United States. pp. 1140- 1141. Published by the author, New York. . 1926. A new bellflower from Florida. Torreya 26: 35- 36. . 19383. Manual of the Southeastern Flora. pp. 1289- 1290, Published by the author, New York. ALPINE ZONE OF THE PRESIDENTIAL RANGE! This attractive booklet describes certain aspects of the geo- logy, climate and organisms of the Presidential Range of New Hampshire. There is a brief section on mammals, birds and insects, but most of the space is devoted to plants. In- cluded is a checklist of plants of the alpine zone with photo- graphs, descriptions and discussions of a number of them. There is also an interesting discussion of the ecology of the region, a topic on which Dr. Bliss speaks with authority. The descriptions and photographs of the plants should be helpful to a beginner to the area. However, the ranges of many of the plants are much more extensive than is in- dicated. For example, Salix Uva-ursi is stated to be ‘‘com- mon in Alpine Garden near some of the streams and also on the summits of Mts. Franklin and Pleasant." Actually this plant, as noted by Pease in his "Vascular Flora of Coos County," is common throughout the alpine region, extending from Mt. Madison to Mt. Clinton and descending quite low on open ridges and ravines. The booklet contains a number of errors and omissions. Of Potentilla Robbinsiana, the most interesting plant in the Range, it is stated “that the only place in the world it occurs . is at the east end of Mt. Monroe cone." Actually it also occurs on the Franconia Range and is represented from there by a number of collections. He states of Geum Peckii "restricted to the Presidential and Franconia Ranges of the White Mountains." It occurs at several lowland stations between these ranges, on Cannon Mt., for example, and also 338 Rhodora [Vol. 65 on Brier Island, Nova Scotia, as is noted in Gray's Manual of Botany 8th Edition. The most notable omission in the checklist of plants found in the alpine zone is Rhinanthus borealis, but there are a number of others, examples of which are Claytonia carol- iniana, Ribes glandulosum and Viburnum edule. Plants typi- ical of the subalpine spruce-fir forest are noted with an asterisk, whereas alpine plants are not thus designated. However, there is some confusion in this procedure. Thus we find Carex atratiformis, known from New Hampshire in only one limited locality near Mt. Monroe, indicated as char- acteristic of the spruce-fir forest while Carex brunnescens is designated as limited to the alpine zone. (Carex brunnes- cens var. sphaerostachya is found at low elevations. It would have been better to list the Alpine plant as C. brunnescens var. brunnescens.) The booklet has a good index, and if the above limitations are kept in mind, should be most useful to anyone wishing an introduction to the flora of the Presidential Range. FREDERIC L. STEELE, ST. MARY'S-IN-THE-MOUNTAINS LITTLETON, N. H. ‘Alpine Zone of the Presidential Range, L. C. Bliss, Urbana, Il., 67 pages, 58 photographs, $1.25. CHROMOSOME NUMBERS IN SOME NORTH AMERICAN SPECIES OF THE GENUS CIRSIUM: GERALD B. OWNBEY AND YU-TSENG HSI As part of a comprehensive review of the systematics of Cirsium, the senior author, with the assistance and collabor- ation of the junior author, has accumulated data on chromo- some number in a number of species. In this genus, it appears that chromosome data can contribute significantly to an understanding of its taxonomy. The present report records those data accumulated during the past three years and it is hoped will represent only the first in a series of reports treating this subject. To date, chromosome numbers for eleven species of Amer- ican cirsiums have been published, viz., C. discolor and C. muticum (Ownbey, 1951; Frankton & Moore, 1963), C. cernuum (C. nivale) and C. subcoriaceum (Stoutamire & Beaman, 1960), C. flodmanii and C. undulatum (Frankton & Moore, 1961), C. horridulum (Turner, Ellison & King, 1961), C. edule and C. brevistylum (Moore & Frankton, 1962b), C. skutchii (Beaman & Turner, 1962), and C. altis- simum (Frankton & Moore, 1963). Twelve additional spe- cies are reported upon in the present paper. Counts for the total of twenty-three species are summarized in Table 1, and further details for the twelve newly reported and five pre- viously reported species are given in the body of the text. Chromosome numbers in sporophytic cells of American Cirsium species so far examined vary from eighteen to thirty-four. Evidently, the gametic number of chromosomes forms a continuous series from nine to seventeen. One or two extra chromosomes have been observed in somatic cells of several species. The data indicate that all American species of Cirsium are diploids, although Frankton & Moore (1963) report a possible instance of triploidy in C. muticum. Chromosome numbers for a large number of Eurasian and "This study was supported by National Science Foundation Grant G9071 to the senior author. 339 340 Rhodora [Vol. 65 Japanese species of Cirsium have been recorded. In all of these species the base or x number equals 17. All of the Eurasian species except two are diploids and these are tetraploids. In Japan, however, triploid, tetraploid, and hexaploid species have been reported by Aishima (1934) and Arano (1957). For a recent tabulation of chromosome numbers in Eurasian Cynareae the reader should refer to Moore & Frankton (1962a). It has been suggested by Frankton & Moore (1961) that many American species of Cirsium may have arisen through the same process of reduction in chromosome number known to have occurred in other Compositae. In brief, this is ac- complished by successive translocations from a chromosome to non-homologous chromosomes. When a centric segment has lost all genetic material essential to the viability of the race, the centric itself may be lost through meiotic accident without detriment to the survival of the race. The present writers see nothing inconsistent with this hypothesis in the data presented here and in fact the situation in the series Undulata, for which data are more nearly complete, affords abundant support for the hypothesis. Morphology of the individual chromosomes of Cirsium species leaves much to be desired if this is to contribute substantially to the taxonomy of the genus. Due to the small size of the chromosomes, it is very difficult to pinpoint the location of the centromere on all or even a majority of the chromosomes of a given cell, even in good preparations. In gross size, the chromosomes in all species we have examined form a closely graded series from smallest to largest. One pair of satellite chromosomes probably is present in all species examined and one or two additional satellite chromo- somes were clearly visible in some preparations. In all cases the satellite was part of a medium or large-sized chromo- some. A useful method of making chromosomal comparisons be- tween closely related species of Cirsiwm has been adopted by Frankton & Moore (l.c.). By carefully measuring the length of each chromosome of a diploid cell at metaphase, a total length of chromosomes for the cell ean be established. In 1963] Cirsium — Ownbey and Hsi 341 practice, several cells are studied and an average figure calculated. This figure can be compared with that of other related species and certain probable relationships inferred. Nomenclature of the subgeneric taxa above the species level adopted here follows exactly that of Petrak (1917). In this paper he treated a majority of the species which occur north of the Mexican border. The species found in Mexico and Central America were treated by the same author in earlier papers (Petrak, 1910, 1911). Petrak placed all of the American species of Cirsium in the subgenus Eucirsiwm. The subgenus for North America is divided into six sections. A very large majority of our species are assigned to a single section, the Onotrophe. The section Onotrophe is taxonomically complex. It is di- vided into eleven subsections, many of which are further broken ‘down into named assemblages of species which pre- sumably represent series. In Table 1 and elsewhere through the paper, twenty-one species of the Onotrophe are men- tioned. Chromosome counts have also been made on one species of the section Cirsiopsis and on one species of the section Erythrolaena as shown in the table. The remaining three sections of North American cirsiums are not mentioned in this paper. It is not our intent here to evaluate Petrak’s subgeneric classification of Cirsium, nor would it be possible to do so without more exhaustive knowledge of the genus. For the present Petrak’s system provides a useful and, un- doubtedly to a high degree, accurate framework upon which to base further studies of the genus. In the accompanying Table 1, it has been necessary to interpolate a few species which were unknown to Petrak or not fully treated by him. The species include C. brevistylum, C. foliosum, C. hillii, certain Undulata and C. skutchii. METHODS Almost all of our counts have been made from root-tip squashes. Seeds are scarified by chipping off part of the pericarp from the distal end and side, then laid on moist filter paper in covered petri dishes. The embryos begin to imbibe water immediately and within 24 hours the pericarp 342 Rhodora [Vol. 65 can be removed and discarded. The embryo is then placed in another sterile dish, allowed to grow for another 24 hours and then sterilized for three minutes in dilute potassium permanganate solution. It is then washed in distilled water and grown under sterile conditions as above for 4-7 days. Valuable seedlings may be lost if the cultures are allowed to become contaminated with fungi. When the primary root is of sufficient length (ca. 15 mm.) the tip is excised and placed immediately in .0025 mol./liter aqueous oxyquinoline solution where it remains for 3-4 hours. It is then trans- ferred to Warmke's solution for 4-5 minutes and preserved in Carnoy's solution No. 2. Root tip squashes stained with acetocarmine are prepared according to standard proce- dures. We have made most of our preparations permanent. The effect of pretreating with oxyquinoline is to cause maximum contraction of the chromosomes at metaphase and this makes for easier and more accurate counting. In our materials, however, a certain amount of morphological de- tail is lost in the maximally contracted chromosomes and this has some disadvantages for karyotypic analysis. Fur- thermore, as a consequence of unusual shortening of the chromosomes, it will be found that the measurements of length dimension are somewhat less than are reported by other workers. Removal of the tip of the primary root for cytological purposes in no way injures the seedlings. They are per- mitted to remain in the sterile dishes until the cotyledons turn green and the branch rootlets appear well started. They are then planted in sand-filled flats and fed with a balanced aqueous nutrient solution. The plants can be trans- ferred to a regular soil mix in individual pots or flats when they begin to show vigorous growth. If proper records are maintained, it is possible to grow the seedlings from which chromosome data has been obtained to maturity and to col- lect voucher specimens from them. In practice, we have often grown seedlings only to a stage sufficient to establish their specific identity for certain so as to eliminate unde- tected human errors in handling the materials. All of the species examined in the course of these studies belong to the section Onotrophe. 1963] Cirsium — Ownbey and Hsi 343 Subsection STENANTHA: Cirsium mexicanum DC. n = 10; 2n = 22. MEXICO. Curapas: About 1 mile east of Teopisca, along Route No. 190, R. M. King, No. 2855 (MIN). 2n = 22. Oaxaca: About 8 miles northwest of Nochistlán, along Route No. 190, R. M. King, No. 3532 (MIN, TEX). n — 10. Count made by B. L. Turner, W. L. Ellison and R. M. King, in 1960 and published by permission of Dr. Turner. TAMAULIPAS: Rancho del Cielo, 5 miles northwest of Gómez Fariás, alt. 3300 ft., B. E. Harrell, No. 258 (MIN). 2n — 22. One and probably two pairs of chromosomes with satel- lites were observed in No. 2855. If our understanding of this species is correct, it is the most common and widespread Cirsium in Mexico. Superficially it appears to be a member of the C. undulatum alliance, but the presence of crisped hispidity in varying degrees on the upper leaf surfaces miti- gates against its placement there. Petrak (1917) placed it in another subsection, the Stenantha, of the Onotrophe. An inconsistency exists between the chromosome number communicated to the senior author by Dr. Turner and our own observations on other collections identified as the same species. Possibly this variation in number actually exists. Subsection ACAULIA : Cirsium drummondii T. & G. 2n = 34. CANADA. MaNiTOBA: Riding Mountain National Park, about 5.5 miles north of Wasagaming, Highway No. 10, G. B. Ownbey, No. 2879 (MIN). SOUTH DAKOTA. Pennington Co.: 4.6 miles south of Hill City, G. B. & F. Ownbey, No. 2861. (C. coccinatum Osterh.) (MIN). Two pairs of chromosomes with satellites were observed with some certainty in the South Dakota material. Only one pair was observed in the Canadian material, but another pair could easily have been obscured in the preparation. C. drummondii has been confused with C. foliosum (Hook.) DC. by one of the present writers (GBO) and others in the past, but we are now convinced that the two are fully distinct although closely related species. The presence of C. drummondii in the Black Hills of South Dakota repre- sents an important disjunction from its primary Canadian range. Plants collected from the two localities cited above are, however, morphologically indistinguishable. The Black Hills plant was described as Cirsium coccinatum by George 344 Rhodora [Vol. 65 E. Osterhout in 1934. Evidently, Osterhout was unfamiliar with C. drummondii, but at the same time recognized the distinctness of the Black Hills specimens from the omnipres- ent C. foliosum of the western mountainous areas. Cirsium foliosum (Hook.) DC. 2n — 34, 36? COLORADO. Jackson Co.: 1.8 miles north of Walden, G. B. and F. Ownbey, No. 1497 (MIN). 2n — 34. LA PLATA Co.: 21.9 miles east of Durango, U. S. Highway No. 160, alt. ca. 7200 ft., G. B. Ownoey & Y. Hsi, No. 2642 (MIN). 2n = 34, 36? MONTANA. PowELL Co.: 5 miles southwest of Avon, U. S. Highway No. 12, alt. ca. 4800 ft., G. B. Ownbey & Y. Hsi, No. 2908 (MIN). 2n — 34, 36? One pair of chromosomes with satellites was observed in most cells at mitotic metaphase. This variable, wide- ranging species is in need of extensive cytotaxonomic anal- ysis. No. 2908 is the segregate species described as Carduus kelseyi Rydb. Subsection ODORATA: Series PUMILA: Cirsium hillii (Canby) Fern. 2n = 30. MINNESOTA. HENNEPIN Co.: Near Veteran’s Hospital, Fort Snelling, G. B. Ownbey, No. 2857. (MIN). One pair of chromosomes with satellites was observed. C. hillii is perennial by means of root and crown sprouts and this has been and still remains the most useful way to distinguish it from its closest relative, C. pumilum (Nutt.) Spreng., a biennial. Cirsium pumilum (Nutt.) Spreng. 2n — 30. CONNECTICUT. New Haven Co.: Vicinity of New Haven, J. R. Reeder, s.n. No chromosomal peculiarities were observed. Plants were grown in the garden from seeds provided by Dr. Reeder. Subsect. ACANTHOPHYLLA: Series MUTICA : Cirsium muticum Michx. 2n = 20. MINNESOTA. HENNEPIN Co.: 5 miles southwest of Edina, U. S. Highway No. 212, G. B. Ownbey & Y. Hsi, No. 2859. No chromosomes with satellites were observed in our preparations. They are, however, almost certainly present. Frankton & Moore (1963) have recently reported somatic numbers of 20, 22, 30 and 31, in C. muticum. The two higher 1963] Cirsium — Ownbey and Hsi 345 counts were obtained from materials originating in Florida. Earlier somatie counts by Ownbey (1951) and the number reported here have not deviated from the regular number of 20 for the species as it occurs in north-central United States. Further chromosome studies of this wide-ranging plant should prove of great interest. Series ALTISSIMA: Cirsium altissimum (L.) Spreng. 2n — 18, 20. IOWA. Boone Co.: 3 miles east of Boone, G. B. Ownbey, No. 2863 (MIN). 2n — 18, 20. WISCONSIN. Barron Co.: About 9 miles east of Cameron, 0.4 mile west of Rusk-Barron Co. line, U.S. Highway No. 8, G. B. € F. Ownbey, No. 2855 in part (MIN). 2n — 18. One pair of chromosomes definitely had satellites, and there was a suggestion of a second pair which could not be confirmed. The fully contracted chromosomes measured 0.3 to 1.5 microns in length. No. 2855 consisted of a mixed population of C. altissimum and C. discolor together with many putative hybrids. The counts were from seedlings of a “pure” C. altissimum plant. No. 2863 was the segregate called C. to10ense (Pammel) Fern. in some manuals. We consider C. iowense to be either synonymous with C. altissimum or a hybrid derivative of that species and C. discolor. The type locality of Cnicus iowensis Pammal is the Ledges, a short distance south of Boone. In 1959 one of us (GBO) visited the Ledges and searched carefully but without success for Pammel’s segre- gate but found only typical C. discolor. However, numerous plants of undoubted C. iowense, i.e., C. altissimum in the broad sense, were found along a railroad embankment east of Boone. A few viable seeds were obtained. Because of the lateness of the season it was impossible to form any opinion about the fertility of the naturally occurring plants. Four plants grown from seeds, however, proved to be fully fertile. The cytology of C. altissimum has been examined re- centlv by Frankton & Moore (1963). No deviation from a sematie number of 18 was found in the two collections studied. Cirsium discolor (Muhl.) Spreng. 2n = 20. WISCONSIN. Barron Co.: About 9 miles east of Cameron, 0.4 346 Rhodora [Vol. 65 miles west of Rusk-Barron Co. line, U. S. Highway No. 8, G. D. & F. Qwnbey, No. 2855 in part (MIN). Growing with C. altissimum and numerous putative hy- brids. The counts were made from seedlings of a “pure” C. discolor plant. At least one and possibly two pairs of chromosomes with satellites were observed. Length of the fully contracted chromosomes varied from 1.0-1.7 microns. An extra chromosome was reported in one seedling of C. discolor by Frankton & Moore (1963). Counts by the present authors and by the senior author in a previous re- port (1951) have not deviated from 20. Series UNDULATA: Cirsium brevifolium Nutt. 2n — 22. IDAHO. Laran Co.: Moscow, G. B. Ownbey, No. 2661 (MIN). One and possibly two pairs of chromosomes with satel- lites were observed. Length of the fully contracted chromo- somes varies from 1.0-1.5 microns. Cirsium canescens Nutt, 2n — 34, (36). COLORADO. JACKSON Co.: 1.3 miles north of Walden, G. B. & F. Ownbey, No. 1496 (MIN). 2n — 34. NEBRASKA. CHERRY Co.: 18.6 miles southwest of Merriam, G. B. Ownbey & Y. Hsi, No. 2601 (MIN). 2n = 34. SOUTH DAKOTA. BUTTE Co.: 17.0 miles northeast of Newell, U. S. Highway No. 212, G. B. Ownbey & Y. Hsi, No. 2507 (MIN). 2n = 34. SOUTH DAKOTA. FALL RIVER Co.: 9.0 miles north of Delrichs, S. Dak. Highway No. 79, G. B. Ownbey & Y. Hsi, No. 2600. 2n = 34. WYOMING. ALBANY Co.: 0.8 mile northwest of Cen- tennial, alt. 8100 ft., G. B. & F. Ownbey, No. 1807 (MIN). 2n — 34. WYOMING. ALBANY Co.: Veedavoo Glen, ca. 20 miles southeast of Laramie, a short distance east of U. S, Highway No. 30, G. B. Ownbey & Y. Hsi, No. 2611 (MIN). 2n — 34 (3 plants), 36 (1 plant). At least one and probably two pairs of chromosomes with satellites are present in this species. Fully contracted chro- mosomes measured from 0.5-1.2 microns in length. Two extra chromosomes were observed in one plant of No. 2611. Other less reliable figures not recorded here indicate an un- usual degree of instability in the normal diploid number for this species. Possibly this is due to unbalanced genomes resultant from hybridization with other species, an hypo- thesis for which we have considerable observational evi- dence. 1963] Cirsium — Ownbey and Hsi 347 The characteristic monocarpic form of this species native to the sand hill areas of Nebraska has been accorded species rank as Cirsium plattense (Rydb.) Cock. ex Daniels by some authors. This robust form intergrades imperceptibly to the west and north with a more branched and sometimes longer- lived form with smaller heads. It is to this latter phase of the species that Nuttall’s type specimen belongs. Plants of No. 2611 collected in the wild were not typical of the species and genetic contamination from some other species was suspected. No. 1496 came from a mixed popula- tion of C. canescens and C. foliosum (Hook.) DC. and puta- tive hybrids of these two. Cirsium canescens Nutt. X C. foliosum (Hook.) DC. 2n = 34. COLORADO. Jackson Co.: 1.3 miles north of Walden, G. B. & F. Qwnbey, No. 1497a (MIN). The count was made from a seedling grown from seed of a putative natural hybrid. Cirsium flodmanii (Rydb.) Arthur n — 11; 2n — 22 (24). NEBRASKA. Drxon Co.: 14.1 miles east of Laurel, U. S. Highway No. 20, G. B. Ownbey & Y. Hsi, No. 2487 (MIN). n — 11; 2n — 22. NORTH DAKOTA. WILLIAMS Co.: 7 miles east of Ray, M. Ownbey, No. 3228 (MIN). 2n = 22, 24. WYOMING. Crook Co.: 6 miles west of Alladin, M. Ownbey, No. 3215 (MIN). 2n — 22. One and probably two pairs of chromosomes of this spe- cies bear satellites. Length of the fully contracted chromo- somes of No. 2487 varied from 0.8-1.4 microns. The presence of two extra chromosomes in one plant of No. 3288 was fully verified. Haploid counts of No. 2487 were made from PMC squashes. Only one satellite-bearing chromosome was visible in this preparation. Frankton & Moore (1961) report a maximum number of four satellite chromosomes in C. flodmanii. They found no variation from the somatic number of 22 chromosomes in a large number of collections from southwestern Canada and Montana. Cirsium ochrocentrum Gray 2n — 32 (30, 31, 32) ARIZONA. CocoNnINo Co.: 16.7 miles south of Grand Canyon City, Ariz. Highway No. 64, G. B. € F. Ownbey, No. 1822 (MIN). 2n = 30, 31, 32. YAVAPAI Co.: 3.8 miles east of Seligman, M. & G. B. Ownbey, No. 3001 (MIN). 2n = 32. COLORADO. MowTROSE Co.: 7.1 miles south of Montrese, U. S. Highway No. 550, alt. ca. 6200 ft., G. B. 348 Rhodora [Vol. 65 Ownbey & Y. Hsi, No. 2635 (MIN). 2n — 32. SOUTH DAKOTA. FALL RIVER Co.: 14.3 miles north of Ardmore, G. B. Ownbey & Y. Hsi, No. 2504 (MIN). 2n — 32. One and almost certainly two pairs of chromosomes with satellites are present in this species. Fully contracted chro- mosomes in No. 2504 were found to vary from 0.5-1.0 mi- cron in length. Of the four collections studied, only No. 1822 showed any plant to plant variation from the usual 16 pairs of chromosomes characteristic of this species. Collection numbers 2504 and 2635 represent the typical phase of this species ranging from the Great Plains west- ward to Wyoming, eastern Utah and New Mexico. This phase normally has strongly decurrent leaves and pale red- dish-lavender corollas. Numbers 1822 and 3001 represent the phase found in Arizona southward and eastward to New Mexico, Sonora and Chihuahua. These plants have less strongly decurrent leaves with often broad and semi-clasping bases and bright scarlet-red corollas. Further study may indicate that the latter plants should be recognized nomen- clatorially as a distinct subspecies or species but for the present we are leaving them with the typical form. Cirsium pitcheri (Torr.) T. & G. 2n = 34. WISCONSIN. Door Co.: near White Fish Bay (northeast of Stur- geon Bay), G. B. Ownbey & Y. Hsi, No. 2655 (MIN). One pair of chromosomes with satellites was observed. Maximally contracted chromosomes varied from 0.5-1.2 mi- crons in length. C. pitcheri is one of the comparatively few flowering plant species having a Great Lakes distributional pattern. It is fully distinct from, but in its vegetative mor- phology similar to, C. canescens Nutt. They appear to have similar ecological requirements to the extent that C. pitcheri is confined to the areas of partially stabilized sand back of the shore lines whereas over much of its range C. canescens is a characteristic component of the Sand Hills flora. Both the Sand Hills C. canescens and C. pitcheri are monocarpic. We think that C. pitcheri may be a derived species and its very narrow ecological spectrum and morphological homo- geneity may indicate an advanced stage of depletion of its genetic variability. 1963] Cirsium — Ownbey and Hsi 349 Cirsium subniveum Rydb, 2n — 32. UTAH. UraH Co.: 22.4 miles southeast of Thistle, U. S, Highway No. 6-50, alt. ca. 7000 ft., G. B. Ownbey & Y. Hsi, No. 2660 (MIN). One pair of chromosomes with satellites was observed. Maximally contracted chromosomes measured 0.5-1.2 mi- crons in length. Although we have compared our collection with the type of C. subniveum and the similarity between the two is close, there is still a small residuum of doubt as to the correct identification of our specimens. We place C. sub- niveum in the series Undulata of the subsect. Acanthophylla with hesitation. Its closer affinities may lie elsewhere. Cirsium tracyi (Rydb.) Petrak 2n — 24. COLORADO, DELTA Co.: 1.8 miles southeast of Paonia, G. B. Own- bey & Y. Hsi, No. 2631 (MIN). (Carduus acuatus Osterh.) EAGLE Co.: Wolcott, alt. ca. 7000 ft., G. B. Ownbey & E. Hsi, No. 2623 (MIN). (Carduus floccosus Rydb.) MoNTEZUMA Co.: 2 miles northwest of Pleasant View, alt. ca. 6800 ft, G. B. Ownbey & Y. Hsi, No. 2647 (MIN). MONTROSE Co.: Cimarron, alt. ca. 7000 ft., G. B. Ownbey & Y. Hsi, No. 2683 (MIN). UTAH. SAN JUAN Co.: Elk Mountain road (Elk Ridge), 2.0 miles north of its junction with the Natural Bridge Naticnal Monument road west of Blanding, alt. ca. 8400 ft., G. B. Qwnbey & Y. Hsi, No. 2657 (MIN). One pair of chromosomes with satellites was observed. The maximally contracted chromosomes measured from 0.8- 2.4 microns long. The type locality of Carduus acuatus Osterhout is Paonia, the point of origin for our No. 2631. Comparison of the types establishes this species as a syno- nym of C. tracyi. Similarly, the type locality of Carduus floccosus Rydb. is Wolcott, the source of our No. 2622, and a comparison of the tvpe of C. floccosus with that of C. tracyi convinces us that they are synonymous. A search was made for specimens of C. tracyi at the type locality, Mancos, Colorado, but none was found there. We collected the species at its nearest occurrence to the westward, near Pleasant View, Colorado, our No. 2647. A study of the chro- mosomes of these three collections and two additional ones revealed no data inconsistent with combining the several proposed species under a single binomial, which in accord- dance with the International Rules must be Cirsium tracyi (Rydb.) Petrak. 350 Rhodora [Vol. 65 Cirsium undulatum (Nutt.) Spreng. n — 13; 2n — 26. MONTANA. LAKE Co.: 4.5 miles northwest of Polson, U. S. High- way No. 98, G. B. Ownbey & Y. Hsi, No. 2900 (MIN). MINERAL Co.: 7.6 miles west of Alberton, U.S. Highway No. 10, alt. ca. 5000 ft., G. B. & F. Ownbey, No. 2662 (MIN). NEBRASKA. Box BUTTE Co.: 21.8 miles south of Crawford, Nebr. Highway No. 2, G. B. Ownbey & Y. Hsi, No. 2503 (MIN). SHERIDAN Co,: 10.9 miles east of Alliance, Nebr. Highway No. 2, G. B. Ownbey & Y. Hsi, No. 2499 (MIN). SOUTH DAKOTA. CusTER Co.: 14.3 miles south of Custer, U.S. Highway No. A85, G. B. & F. Ownbey, No. 1297 (MIN). WASHING- TON. ApAMs Co.: 16 miles east of Washtuena, Wash. Highway No. 118, G. B. € F. Ownbey, No. 2662 (MIN). n — 13; 2n —26. WYOM- ING. Crook Co.: 0.5 mile east of summit of Bearlodge Mountains, on road between Alva, Wyo. and Belle Fourche, S. Dak., G. B. Ownbey & Y. Hsi, No. 2917 (MIN). Crook Co.: 7 miles northwest of Hulett, M. Ownbey, No. 3213 (MIN). One pair of chromosomes with satellites was seen in nearly every preparation. When fully contracted the chro- mosomes measured from 0.7-1.5 microns in length. Frankton & Moore (1961) report a maximum of four satellite chromosomes in C. undulatum. A constant number of 26 chromosomes was found in a large number of collec- tions from southwestern Canada and from Montana, South Dakota and Idaho in the United States. This species is the most common and widespread Cirsium in western United States. It is found in the Great Plains from southern Canada to northern Mexico and westward a- cross the mountains to the Pacific Northwest. Some varia- tion in the species from region to region can be detected and it was thought that some variation in chromosome number or morphology might also occur. The species hes, however, proved to be surprisingly constant to date, but more studies are indicated. Cirsium wheeleri (Grav) Petrak 2n — 28. ARIZONA. NavaJo Co.: 0.2 mile southeast of Showlow, G. B. & F. Ownbey, No. 1806 (MIN). At least one pair and possibly two pairs of chromosomes bear satellites. The fully contracted chromosomes measured frem 0.5-1.2 microns in length. DISCUSSION Tt is probably premature to make any sweeping general- 1963] Cirsium — Ownbey and Hsi 351 izations about the chromosomal situation in American cir- siums. It has been, however, of some interest to bring to- gether all of the literature reports available to us at this time and to combine them in tabular form with those pre- sented for the first time in this paper. Of the twenty-three species listed in Table 1, several have been examined repeat- edly from widely separated parts of their ranges. The need for subjecting all of the wide ranging species to this form of analysis should be emphasized. In contrast to the stability in chromosome numbers of Old World species of Cirsiwm, exclusive of the polyploid series in Japan, the American species exhibit an astonishing prevalence of reduction in numbers. As previously noted, diploid numbers at all levels of 18 through 34 have been found in one or more species. This fact will inevitably weigh heavily in the future development of any hypothetical phyletic sequences within groups of related species. We find in the series Undulata an excellent example of a reduction series that may find its counterpart in other alliances of related species within the genus. If we hypothesize that in the Undulata the more primitive species from the standpoint of chromosome numbers are those with 34 somatic chromo- somes, then the other species in the series are derived. It does not follow, however, that the 34-chromosome species, namelv, C. canescens and C. pitcheri are, in a morphological or physiological sense, necessarily more primitive than the other members of their group. The junior author has ex- amined the correlations of chromosome numbers and mor- phological features of the Undulata to some extent (Hsi, mss., 1960). It is expected that a revised version of these observations will appear in a separate report, however, and they will not be detailed here. Probably changes in chromo- some number are no more than a secondary pathway by which genetic diversitv and subsequent morphological di- versity can be achieved. Such changes are significant, par- ticularly if numerous simple and reciprocal translocations of small chromosome segments are involved, in providing ge- netic recombinations not attainable by any other means. These rearrangements comprise a part, although probably 352 Rhodora [Vol. 65 not the most important part in any case, of the raw mate- rials upon which selection operates. In the case of American cirsiums, chromosomal rearrangements concomitant with reduction in numbers may have played an unusually impor- tant role in speciation. SUMMARY Chromosome counts for seventeen species of American cirsiums are reported. Twelve of these species are recorded for the first time; five species for which published records exist are verified. American species examined by all authors to date total twenty-three. All species so far examined ap- pear to be diploids and all possible genomes from nine through seventeen occur in at least one species. One or two extra chromosomes are sometimes found in individual plants. One or two pairs of satellite chromosomes were observed by the present authors in almost all species examined and it is suspected that they are universally present in the American species of the genus. A trend toward reduction of chromo- some numbers in related groups of species is noted, the series Undulata being the best documented example of this trend. Chromosome morphology by itself has not provided any criteria of taxonomic value in distinguishing species or species groups to the present writers, but the work of others indicates that total length of the somatic chromosomes at metaphase, i.e., the sum of the lengths of the individual chromosomes, may ultimately provide useful comparative data. DEPT. OF BOTANY, UNIVERSITY OF MINNESOTA, MINNEAPOLIS AND DEPT. OF BIOLOGY, NORTHLAND COLLEGE, ASHLAND, WISC. LITERATURE CITED AISHIMA, T. 1934. Chromosome numbers in the genus Cirsium I. Bot. Mag. Tokyo 48: 150-151. ARANO, H. 1957. The karyotype analysis and its karyo-taxonomic considerations in some genera of subtribe Carduinae. Jap. Jour. Gen. 32:323-332. BEAMAN, J. H. & B. L. TURNER. 1962. Chromosome numbers in Mexi- can and Guatemalan Compositae. Rhodora 64:271-276. 1963] FRANKTON, C. & R. J. Moore. 1961. Cirsium — Ownbey and Hsi 353 Cytotaxonomy, phylogeny, and Canadian distribution of Cirsium undulatum and Cirsium flod- manii. Canad. Jour. Bot. 39:21-33. & . 1963. Cytotaxonomy of Cirsium muticum, Cirsium discolor, and Cirsium altissimum, Canad. Jour. Bot. 41:73-84. Moore, R. J. & C. FRANKTON. 1962a. Cytotaxonomic studies in the tribe Cynareae (Compositae). Canad. Jour. Bot. 40:281-293. & . 1962b. Cytotaxonomy and Canadian distribution of Cirsium edule and Cirsium brevistylum. Canad. Jour. Bot. 40:1187-1196. Natural hybridization in the genus Cirsium — OwNBEY, G. B. 1951. I. C. discolor (Muhl. ex Willd.) Spreng. X C. muticum Michx. Bull. Torrey Club 78:233-253. PETRAK, F. 1910. Die mexikanischen und zentral-amerikanischen Ar- ten der Gattung Cirsium. Beih. Bot. Centr. 27:207-255. 1911. Beitráge zur Kenntnis der mexikanischen und zentral-amerikanischen Cirsien. Bot. Tidsskr. 31:57-72. 1917. Die nordamerikanischen Arten der Gattung Cir- sium. Beih. Bot. Centr. 35:223-567. STOUTAMIRE, W. P. & J. H. BEAMAN. 1960. Chromosome studies of Mexican alpine plants. Brittonia 12:226-230. TURNER, B. L., W. L. ELLISON & R. M. KING. 1961. Chromosome num- bers in the Compositae. IV. North American species, with phyle- tic interpretations. Amer. Jour. Bot. 48:216-223. Table I. Summary of Counts of North American Cirsium Species. Nomenclature Author Gameto- Sporo- phytic phytic Subgenus EUCIRSIUM Sect. CIRSIOPSIS C. nivale (Kunth) Stoutamire & Beaman 17 Schz.-Bip. (1960) (C. cernuum Lag.) Sect. ONOTROPHE Subsect. MINUTIFLORA C. brevistylum Cronq. Moore & Frankton 34 (1962) C. edule Nutt. Moore & Frankton 34 (1962) Subsect. STENANTHA Series MEXICANA C. mexicanum DC. Ownbey & Hsi 22 Turner, Ellison & King 10 Subsect. ACAULIA C. drummondii T. & G. Ownbey & Hsi 34 354 C. foliosum (Hook.) DC. C. skutchii Blake Subsect. ODORATA Series PUMILA C. hillii (Canby) Fern. C. pumilum (Nutt.) Spreng. Series HORRIDULA C. horridulum Michx. Subsect. ACANTHOPHYLLA Series MUTICA C. muticum Michx. Series ALTISSIMA C. altissimum (L.) Spreng. C. discolor (Muhl.) Spreng. Series UNDULATA C. brevifolium Nutt. C. canescens Nutt. C. flodmanii (Rydb.) Arthur C. ochrocentrum Gray C. pitcheri (Torr.) T. & G. C. subniveum Rydb. C. tracyi (Rydb.) Petrak C. undulatum (Nutt.) Spreng. C. wheeleri (Gray) Petrak Sect. ERYTHROLAENA Subsect. CONSPICUA C. subcoriaceum (Less.) Schz.-Bip. Rhodora Ownbey & Hsi Beaman & Turner 17 (1962) Ownbey & Hsi Ownbey & Hsi Turner, Ellison & 16 King (1961) Ownbey (1951) ; Frankton & Moore (1963) Ownbey & Hsi Frankton & Moore (1963) Ownbey & Hsi Ownbey (1951) Frankton & Moore (1963) Ownbey & Hsi Ownbey & Hsi Ownbey & Hsi Frankton & Moore (1961) Ownbey & Hsi 11 Ownbey & Hsi Ownbey & Hsi Ownbey & Hsi Ownbey & Hsi Frankton & Moore (1961) Ownbey & Hsi 13 Ownbey & Hsi Stoutamire & Beaman 17 (1960) [Vol. 65 34 (36?) 30 30 22 34, 36 22 22, 24 32, 30, 31 34 32 24 26 26 28 NEW SPECIES OF DIGITARIA AND TRICHACHNE JASON R. SWALLEN Some time ago Dr. Richard A. Howard sent me two grasses from the West Indies for identification. They ap- peared to be common species of Digitaria and Trichachne, but on closer examination they proved to be undescribed. It is a coincidence that both species have a similar (geo- graphical) distribution, and differ from the related species in essentially the same characters. Trichachne affinis Swallen, sp. nov. Culmi graciles 60-70 cm. alti; laminae planae, laxae, usque ad 18 cm. longae, 7-9 mm. latae; racemi adscendentes, 8-10 cm. longi; spiculae 4 mm. longae, acuminatae; lemma sterile 5-nervium, glabrum, marginibus dense villosum. Perennial; culms 60-70 cm. tall, slender, erect, or decumbent at the base, the cataphylls densely villous; innovations extravaginal, some- times appearing like short rhizomes; sheaths mostly longer than the internodes, glabrous or very sparsely hispid; ligule a thin membrane about 2 mm. long; blades lax, flat, attenuate, as much as 18 cm. long, 7-9 mm. wide, glabrous, the margins scabrous; inflorescence 13-18 cm. long, composed of few to several erect or ascending racemes about 8-10 cm. long; spikelets in pairs, 4 mm. long, ovate, acuminate; first glume 0.5 mm. long, nerveless, obtuse; second glume narrow, 3-nerved, acuminate, as long as the fruit, glabrous between the nerves, long- villous on the margins; sterile lemma as long as the fruit, ovate, acuminate, 5-nerved, glabrous between the nerves, long-villous on the margins, the hairs extending about 1 mm. beyond the spikelet; fruit as long as the second glume and sterile lemma. Type in the U. S. National Herbarium No. 1556826, collected along roadsides at Quinigua, Valle del Cibao, Province de Santiago, Domini- can Republic, October 21, 1930 by E. L. Ekman (“Mus. Botan. Stock- holm" No. H-16090). This species is most closely related to T'richachne insularis (L.) Nees and T. sacchariflora (Raddi) Nees with which it has been confused. It is readily separated from both, how- ever, by the ovate, strongly 5-nerved sterile lemma which is glabrous between the nerves but densely villous on the margins with pale straw-colored or usually white hairs. The plants are also more slender with relatively narrower blades, 355 356 Rhodora [Vol. 65 although no clear differentiation can be made on these char- acters. The type was selected as an average specimen and the description drawn up from it. Depauperate specimens occur which are only 10-20 cm. tall with blades 2 mm. wide (Re- donda, Howard 15234), while others are as much as 15 dm. tall (Peru, Allard. 21150). Specimens from Mexico and South America usually have larger spikelets, 5 mm. long, and a longer more pointed fruit. West Indies and Mexico south to Bolivia and Brazil. Representative specimens: WEsT INDIES: PORTO RICO, Hess 428; ST. KITTS, Box 129; TRINIDAD, Freeman 7513; REDONDA (Leeward Islands), Howard 15234. MEXICO: QUINTANA ROO, Tancah, Swallen 2775; NUEVO LEON, Monterrey, Mueller 376, Harvey 1072; SAN LUIS POTOSI; Xititla, Sohns 1441. GUATEMALA: Escuintla, Hitchcock 9008. PANAMA: Point Chamé, Hitchcock 8156. COLOMBIA: VAUPÉS; Río Kuduyarí, Schultes 17867; TOLIMA; Armero, Cuatrecasas 10497. VENEZUELA: Cristobal Colon, Broadway 334; CARA- BOBO, Lake Valencia, Chase 12341. PERU: HUANUCO; Río Azul, Fer- reyra 12758; JUNIN, Colonia Perené, Hitchcock 22079; SAN MARTÍN, Tingo María, Allard 21150. BOLIVIA: Santa Cruz, Steinbach 5221, 6638; Mapiri, Rusby 246; Chulumani, Sur Yungas Hitchcock 22665. BRAZIL: RIO GRANDE DO SUL, Sáo Leopoldo, Dutra 566; Pelotas, Costa Sacco 86; SANTA CATARINA: Itajaí, Reitz & Klein 3365. Digitaria diversiflora Swallen, sp. nov. Culmi decumbentes 15-45 cm. longi; laminae 4-5 cm. longae, 3.5-5 mm. latae, sparse papilloso-hispidae; racemi divergentes, 4-8 cm. longi; spiculae 2.7-3 mm. longae, binae, biformes, spiculae inferioris lemma sterile glabrum, spiculae superioris lemma sterile inter nervos lateralis dense pilosum. Annual; culms erect or decumbent spreading and rooting at the lower nodes, 15-45 cm. long; sheaths shorter than the internodes, sparsely to rather densely papillose hispid, especially toward the base and along the margins; ligule membranaceous about 1.5 mm. long; blades 4-5 cm. long, 3.5-5 mm, wide, sparsely papillose hispid near the base with long coarse hairs; inflorescence composed of 2 to several digitate spreading racemes, 4-8 cm. long, the rachis nearly 1 mm. wide, scabrous along the margins; spikelets in pairs, 2.7-3 mm. long, the lower spikelet of each pair distinct from the upper; lower spikelet oval, acute, the first glume broad, obtuse, 0.2-0.3 mm. long, the second narrow, 3-nerved, 2/3 as long as the spikelet, the sterile lemma glabrous (rarely with a few short hairs on the margin), strongly 5-nerved, the nerves equidistant; upper spikelet usually a little longer 1963] Hypericum — Hodgdon 357 than the lower, lanceolate, the second glume pilose between the nerves, the sterile lemma 5-nerved, the lateral nerves approximate, densely fimbriate on the margins and between the lateral nerves; fruit acuminate, pale or straw-colored. Type in the U. S. National Herbarium No. 927914, collected in open ground along road, Black River, Jamaica, October 22, 1912, by A. S. Hitchcock (no. 9636). The type was filed under Digitaria sanguinalis (L.) Scop., which does not occur in the Tropics. The species is closely related to D. adscendens (H.B.K.) Henr., differing, how- ever, in the biform spikelets, the shorter obtuse first glume, and the strong equidistant nerves of the glabrous sterile lemma of the lower one of a pair of spikelets. Florida, Texas and the West Indies to Colombia and Venezuela. Representative collections: FLoRIDA: Key Largo, A. A. Eaton 446; Cocoanut Grove, Small & Carter 605. TEXAS: Corpus Christi, Hitchcock 5345; Sarita, Hitchcock 5440; Kingsville, Swallen 10264. MEXICO: BAJA CALIFORNIA; San Jose del Cabo, Brandegee 41; Ribera, Wiggins 5647; SONORA; Alamos, Rose et al. 12983; SINALOA; Rosario, Rose 1541: VERACRUZ; Veracruz, Hitch- cock 6566: YUCATAN; Gawmer 1029. GUATEMALA: IZABAL: Quirigua, Blake 7704. PANAMA: CANAL ZONE; Gamboa, Pittier 4440; Standley 28504; Balboa Heights, Killip 4255; San Jose Island, Johnston 991. CUBA: ISLE OF PINES; Herradura 43161; PINAR DEL RIO: Los Palacios, Shafer 11794. Porto Rico: Juana Diaz, Sargent 3210. DOMINICAN REPUBLIC: Citdad Trujillo, Allard 13,091, 15,010; Samana Peninsula, Abbott 499. HAITI: Port de Paix, Leonard 11,190. MONTSERRAT; Plymouth, Potter 5529. COLOMBIA: Cartagena, Hitchcock 9905; ANTI- OQUIA; Puerto Berrio, Archer 1413. VENEZUELA: FALCON; Caro, Tam- ayo 790. SMITHSONIAN INSTITUTION, WASHINGTON, D. C. A STRANGE FORM OF HYPERICUM CANADENSE — The recent report of Hypericum gentianoides L. from Great Wass Island, Maine (Rhodora 65: 285) is in error. What looked to me like that species actually was an odd assemblage of plants of H. canadense L. growing on or about a granite ledge. The herbarium material of these consists of 24 plants varying from 8-13 cm in height. They are stiffish, slender 358 Rhodora [Vol. 65 and entirely unbranched plants with very short and sub- appressed ascending leaves. From a short distance away, when first observed in the field, they seemed to be essentially leafless. The minimum leaf length in Gray's Manual, 8th edition is given as 1 em except for the var. galiiforme Fern- ald from southeastern Virginia with leaves down to 5 mm long but possessing other features that hardly match our material. The largest leaf on any of our 24 specimens is about 1 em long, the width being slightly more than 1 mm. Leaf length, in general varies directly with the height of plant, the shorter specimens having leaves not more than 5 mm long. Like many annuals H. canadense exhibits very different growth responses under different conditions. Many years ago in early September I collected a series of very un- usual specimens of this St. John's-wort at the water's edge in a nearly depleted reservoir in Somersworth, New Hamp- shire (Hodgdon No. 7984). These also showed a marked response, but of a somewhat different kind from the plants of Great Wass Island. It would seem desirable to revise the description of the species to include extremes such as these but it would not seem to be wise to give formal names to individual deviant populations in such a plastic polymorphic species. Preston Adams (Rhodora 64: 241, 1962), commenting on the Hypericum canadense complex, questions the status of the varieties magninsulare and galiiforme. The analyses of the Great Wass Island and Somersworth collections seem to support Dr. Adams' doubts. The sepal lengths of speci- mens from the Great Wass Island collection vary from 2.2- 3.2 mm, the mean being 2.68. The comparable data based on 18 plants of my Somersworth collection is 1.5-2.5 mm long the mean being 2.17. It is also of interest that flowers on the same plant may have very different lengths of sepals. When one considers the wide latitude of these sepal measure- ments and the great diversity in vegetative growth, it be- comes increasingly evident that these constitute very poor taxonomic characters in the species. A. R. HODGDON, UNIVERSITY OF NEW HAMPSHIRE. SOUTHEASTERN LIMIT OF CHAMAECYPARIS THYOIDES DANIEL B. WARD The Atlantic white-cedar, Chamaecyparis thyoides (L.) BSP., has long been understood to extend from southern Maine along the Coastal Plain through northern Florida and west to southern Mississippi (Rossbach, 1936; Little, 1953). At the southeastern limit of its range, Atlantic white-cedar was mapped by Korstian (1931), by Munns (1938), and by Brush (1947), as occurring throughout northern Florida and extending down the eastern half of the peninsula to just beyond Cape Canaveral. More recently, James (1961) has mapped Chamaecyparis in the Southeast and, although showing a gap in distribution between north- eastern and western Florida, indicates six stations for the species in the northeastern part of the state. Atlantic white-cedar does grow in western Florida, often forming dense stands in cool stream bottoms from Liberty County to the western edge of the state. Additional stations are known in Georgia ; James records four. But south of the Georgia line and east of Liberty County, Florida, Chamae- cyparis is one of the rarest of native trees. Herbarium records and confirmable published descriptions support the existence of only a single station of white-cedar in all of Florida east of Liberty County. A second station is now known, and the source of the erroneous published records has been traced. For many years it has been common knowledge among botanists in Florida that there was a small stand of Cham- aecyparis 4 miles south of Interlachen, Putnam County. West and Arnold (1946) apparently were the first to pub- lish a note on the presence of the species in this area. Speci- mens from this station have been widely distributed and form the basis for James’ southernmost record. This stand has recently been re-investigated and found to be much more extensive than previously thought. Scattered trees, and oc- casionally denser groupings, occur for several miles along a 359 360 Rhodora [Vol. 65 small clear stream known as Cabbage Creek almost to its mouth on the Oklawaha River. The number of trees is con- siderable, but few individuals exceed 15 to 18 inches in diameter at breast height, and all parts of the area show evidence of cutting and other disturbance. In January, 1962, Ranger Paul Bielling of the U.S.D.A. Forest Service called the attention of the author to a stand of Chamaecyparis thyoides in the Ocala National Forest, Marion County, Florida. Several trips have since been made to the area, and observations have been recorded both for the Chamaecyparis and for the associated flora. This newly discovered station is virgin timber and is believed to ter- minate the range of white-cedar in the southeastern United States. The Ocala National Forest is an extensive region covered in large part with a “scrub” characterized by sand pine, Pinus clausa, a species adapted to excessively drained and nutritionally deficient sands. Only in small “islands” of different soil type is there any appreciable growth of the more important commercial species such as longleaf pine, Pinus palustris. The region, therefore, was slow to become attractive to lumbering interests, and several small areas of great botanical interest have survived in an undisturbed state. The stand of Atlantic white-cedar extends along the stream bottom of a clear and cool spring-fed brook which enters Juniper Run, the outlet of Juniper Springs, one of the large ever-flowing springs in which north-central Florida abounds. The stand may best be reached by Forest Service road #71, an intractable “ball-bearing sand” fire trail off Florida Highway 19 at a point about 5 miles northeast of the Juniper Springs Recreation Area, itself 24 miles east of Ocala. The white-cedar is found along the stream for a dis- tance not to exceed one-half mile, in an area of less than 10 acres. The trees are common, but nowhere approach the dense single-species stands characteristic of white-cedar in the Atlantic coastal swamps. Thirty-nine of the larger trees have been measured, and averaged 18.9 inches in diameter. 1963] Chamaecyparis — Ward 361 Five of these were greater than 24 inches in diameter, with the largest attaining 34.2 inches. Since none of the larger trees has been cut, estimation of their age is difficult. Two small trees cut by campers, with trunk diameters between 4 and 5 inches, were 42 and 15 years old, suggesting a minimum age for the larger trees of perhaps 250 to 350 years. Recently Li (1962) has described white-cedar from Mis- sissippi, Alabama, and western Florida as a distinct species under the name Chamaecyparis henryae. Li (in correspon- dence with the author) has identified a specimen from the Ocala Forest stand as typical Chamaecyparis thyoides. Li has cited many differences in bark, in foliage, and in male and female cones between what he considers the two species and suggests that the Mississippi to west Florida variant be more closely related to Chamaecyparis nootkatensis of the West Coast than to the true Chamaecyparis thyoides. Such speculations merit further investigation. Associated with the Chamaecyparis in the Ocala Forest stand are several species of interest and of limited distribu- tion. Among the most curious is the anomalous Pieris phillyreifolia, an ericaceous vine whose stems burrow up- ward beneath the bark of the white-cedar and then, at heights up to 7 meters, burst through to form apparently angiospermous branches on the gymnospermous trunk. The local Illicium parviflorum and the needle palm, Rhapido- phyllum hystrix, are common in parts of the white-cedar swamp. There are several small trees of the very distinctive and very little known Salix floridana which, interestingly, has for another of its few stations the Chamaecyparis stand south of Interlachen. Among the more prominent of the herbaceous species associated with the white-cedar are numerous vigorous plants of Parnassia grandifolia at its only known station in Florida. This species is customarily assumed to extend into Florida and was specifically cited for the state by Wherry (1935), yet no prior supporting records exist. Wherry has suggested (correspondence) that the basis for the then- 362 Rhodora [Vol. 65 spurious records was confusion by himself and J. K. Small with Parnassia caroliniana which occurs locally in western Florida. Similar perpetuation of an original misinterpretation has been responsible for the published records of Chamaecyparis blanketing northern Florida and extending in quantity down the peninsula. The southernmost of the six northeast Florida stations reported by James was soundly based upon the Interlachen stand, but the remaining five were obtained from W. H. Duncan who, in turn, had received them from E. L. Little. Little (correspondence) had taken his informa- tion from an old map for white-cedar prepared by G. B. Sudworth, which had served as the source for the maps pub- lished by Korstian, by Munns, and by Brush. Little was able to trace the erroneous records to their source when he con- sulted Sudworth's card file of locality records, and there found them to be taken from R. M. Harper's Geography and Vegetation of Northern Florida (1914: 324, 334, 342). But no discredit falls to Harper; when he failed to observe a major forest tree in a particular area, and yet could not claim that no individuals were present, he customarily listed it under the conservative heading, “rare or absent.” Sud- worth chose to put the emphasis on the first possibility and thus began a belief in the Florida range of Chamaecyparis that is only now in part being zubstantiated. DEPARTMENT OF BOTANY, UNIVERSITY OF FLORIDA, GAINESVILLE LITERATURE CITED BRUSH, W. D. 1947. Knowing your trees — Atlantic white-cedar. Am. Forests 53: 218-219. HARPER, R. M. 1914. Geography and vegetation of Northern Florida. Fla. Geol. Surv. Rep. 6:165-451. JAMES, C. W. 1961. Endemism in Florida. Brittonia 13: 225-244. KorsTIAN, C. F. 1931. Southern white cedar. U. S. Dept. Agr. Tech. Bull. 251, map p. 2. , H. 1962. A new species of Chamaecyparis. Morris Arb. Bull. 13: 43-46. LITTLE, E. L. 1953. Check list of the native and naturalized trees of the United States. U. S. Dept. Agr. Handb. 41. Li 1963] Commelina — Hodgdon 363 Munns, E. N. 1988. Distribution of important forest trees of the United States. U. S. Dept. Agr. Misc. Publ. 287, map p. 59. RossBACH, G. P. 1936. Northeastward extensions in the Maine flora. Rhodora 38: 453-454. West, E. & L. E. ARNOLD. 1946. The native trees of Florida. Univ. of Fla. Press, 212 pp. Wuerry, E. T. 1935. The ranges of our eastern Parnassias and Sedums. Bartonia 17: 17-20. COMMELINA COMMUNIS IN NEW HAMPSHIRE — In early October 1963 a specimen of Dayflower, Commelina com- munis L. from Concord, New Hampshire, parasitized by a smut, was brought to Dr. Avery Rich, Plant Pathologist at the University of New Hampshire. Because the smut proved to be of particular interest, I checked through our herbari- um to determine if any of our specimens were similarly af- fected. To my surprise I found no collections from New Hampshire and on referring to our Manuals discovered that C. communis has not been reported as naturalizing north- east of Massachusetts. Certainly I am quite familiar with this species in the field having run across it a number of times in New Hampshire in recent years and have always as- sumed in fact that it was common enough not to merit much attention. Apparently the only herbarium record from New Hamp- shire is my No. 6432 collected in 1949 as a garden weed at my former home in Durham. It is interesting to report that several vigorous plants were observed at this same place as recently as October 9, 1963, still not stricken by frost. Col- lections were made at this time (Hodgdon & Pike No. 12754). A recent conversation with Mr. Carroll Durfee who had found the diseased Concord specimens revealed the inter- esting fact that these plants were also entirely naturalized, behaving much as those in Durham, growing each year from seeds and giving every indication of continuing to make themselves at home. Specimens of these recent collections from Durham and Concord, New Hampshire, are deposited in the Herbarium of the University of New Hampshire. A. R. HODGDON, UNIVERSITY OF NEW HAMPSHIRE. NOTE ON LOMATOGONIUM — When Fernald (1919) dis- cussed the taxonomy of the genus Lomatogonium it had not been reported in eastern North America south of the Mag- dalen Islands in the Gulf of St. Lawrence although the species L. rotatum (L.) Fries was well known along the coasts further north to Labrador and Western Greenland. Ten years later G. L. Stebbins, Jr. (1929) reported collect- ing this plant on August 4, 1928 at Schoodic Peninsula, Win- ter Harbor, Hancock County, Maine. Since 1919 there has been no other mention of Lomatogonium rotatum in Rhodora even in the exhaustive lists of Newfoundland and Labrador plants until it was reported from the Wolf Islands (Pike and Hodgdon 1963). This collection was made by A. R. Hodgdon 15 August 1960 on the northeast side of South Wolf near the beach of a small cove, since named Lomatogo- nium Cove. One dense clump formed a single station for this plant and only a few specimens were taken for the record. These have been deposited in the herbarium of the University of New Hampshire. The most careful search of this cove for the past three seasons has failed to reveal a single additional specimen. The site where the collection was made was a turfy bank vulnerable to northeast storms not far above the tide mark. A northeast gale occurred in the Bay of Fundy during the winter of 1960-61 at a period of high tides causing severe erosion on exposed shores. Thus, the site may have been washed away. However, it hardly seems credible that all the seeds of this profusely seeding annual plant would have been destroyed or elimi- nated. As some members of the Gentian family have seeds with long delayed dormancy it is hoped that this attractive gentian-like plant may reappear on the Wolves. While hunting for Primula laurentiana along the Maine coast (Pike 1963) on September 15 of this past season I found Lomatogonium rotatum growing on two islands in Englishman's Bay off Jonesport in Washington County. These islands are nearly midway between the Stebbins sta- tion at Schoodic Point and that on the Wolf Islands in New Brunswick. My first collection was on Water Island which is hardly more than a large granite ledge some 15 or 20 feet 364 1963] Lomatogonium — Pike 365 above high water with humus deposits and pools in protected depressions. One of these depressions on the shoreward side of the island was covered with a dense mat of Potentilla anserina or P. Egedei var. groenlandica (positive identifi- cation of which has not yet been made) throughout which were scattered plants of Lomatogonium rotatum in full bloom. Specimens for the plant press and the greenhouse were both collected leaving a large proportion of the colony for reproduction. Later the same day this rare plant was found on nearby Mistake Island again growing in turf in crevices of the granite ledges close to Moose Peak Light. Associated with it in this case were Primula laurentiana, Sagina procumbens, and Plantago juncoides. The stature of the plants was considerably less here than on Water Island, some of them being less than an inch high yet in full bloom. Fernald (1929) chose Lomatogonium rotatum as one of four species to illustrate epibiotic flora that in his opinion escaped the last glaciation. Hultén, (1955) in discussing the isolation of the Scandinavian Mountain Flora, cites Lomatogonium rotatum as one of a group “. .. of mountain plants with very wide gaps in their area" which “... have long been recognized as being very isolated in the Fenno- scandian mountains." He also regards the distribution pat- tern of this plant as due to escape from glaciation. The occurrence of these four stations raises some inter- esting questions concerning the history of the flora of the Maine Coast and Bay of Fundy and indicates the desirability of close scrutiny of other headlands and outlying islands. Specimens of these collections are deposited in the Her- baria of the New England Botanical Club and of the Uni- versity of New Hampshire. RADCLIFFE B. PIKE, UNIVERSITY OF NEW HAMPSHIRE, DURHAM LITERATURE CITED FERNALD, M. L. (1919) Lomatogonium the Correct Name for Pleuro- gyne, Rhodora 21 193. (1929) Some Relationships of the Floras of the North- ern Hemisphere; Proceedings of the International Congress of Plant Science 2:1494-1497, (with map). 366 Rhodora [Vol. 65 HULTÉN, Eric (1955) The Isolation of the Scandinavian Mountain Flora: Acta Soc. pro Fauna et Flora Fenn. 72,N :08,1955. PIKE, RADCLIFFE B. (1963) Note on Primula laurentiana in Maine, Rhodora 65:286. AND ALBION R. HopGpoN (1963) The Flora of the Wolf Islands New Brunswick Part I, Rhodora 65:94, A BUTTRESSED ELM FROM ONTARIO — Buttressing is a fea- ture commonly associated with tropical trees, although Ulmus americana L. as a street tree sometimes shows a tendency toward buttress roots, occasionally (Washington, D. C., Amherst, Mass.) producing tiny buttresses 3-5 dm high. In 1958, during a brief visit to Rondeau Provincial Park, Ontario, on the north shore of Lake Erie, I was surprised to see a great elm, about 30-35 m tall and with a trunk 6-7 dm thick, which had wide buttresses about 2 m high. This tree was in the central part of the park, in deep beech-maple forest with a scattering of other trees, including elms and an enormous Populus deltoides, taller and much thicker than the buttressed elm. It was not pos- sible to be positive as to which species of Ulmus the but- tressed tree belonged, but its form suggested U. americana. Examination showed that most of the trees in the immediate area had some slight development of buttresses. The area is low and rather swampy, with very low parallel ridges of sand. Boehmeria cylindrica, Collinsonia canadensis, and Onoclea sensibilis were the common herbs, forming a dense ground cover. In the tropies buttressing is frequently associated with swampy ground, but is by no means confined to trees in such habitats. In Ceiba pentandra, at least, it has been shown to be genetic (according to H. G. Baker, in talk given August 26, 1963 at Amherst, Mass.), rather than directly ecological in origin. It would be interesting to know how general the tendency toward buttressing is in the elms, and if it is more pronounced in swamps. F. R. FOSBERG, FALLS CHURCH, VIRGINIA Volume 65, No. 764 including pages 289-382, was issued December 26, 1963. INDEX TO VOLUME 65 New scientific names and combinations are printed in bold face type Abies balsamea 84; var. phanero- lepis 88; lasiocarpa 156 Acer pensylvanicum 93; rubrum 93; spicatum 93 Achaetogeron fisheri 234, 235; forreri 226, 231, 233; griseus 226, 231, 233; wislizeni 233 Achillea borealis 95; lanulosa 95, 231 Aconitum delphinifolium, ssp. paradoxum 216 Acrostichum rufum 56 Adelobotrys barbata 18; macro- phylla 18; praetexta 18 Agaricus leptolomus 290, 293; truncicola 295 Agropyron elongatum 318; re- pens, f. aristatum 88; f. tri- chorrhachis 88; trachycaulum, var. glaucum 88; var. majus 88 Agrostis alba, var. palustris 88; perennans, var. aestivalis 88; scabra 88; tenuis 88 Alaska, An Annotated List of Vascular Plants from Cape Sabine 208 Alnus crispa, var. mollis 90; rugosa, var. americana 91 Alopecurus alpinus, var. alpinus 212; pratensis 89 Alpine Zone of the Presidential Range (Review) 337 Amelanchier Bartramiana 92 2 laevis 92; laevis X Bartrami- ana 92 Ammophila breviligulata 284 Anagallis arvense, f. caerulea 81 Anaphalis margaritacea, var. sub- alpina 95 Androsace chamaejasme, ssp. leh- manniana 220 367 Anemone narcissiflora 211, 2106; ssp. interior 216; parviflora 216 Anthoxanthum odoratum 89 Aphanostephus ramosissimus 226, 231; ramosus 226 Aralia nudicaulis 93 Arctagrostis arundinacea 212: latifolia 212; var. arundinacea 212 Arctium lappa 368; tomentosum 81 Arctophila fulva 213 Arenaria arctica 216; lateriflora 91; obtusiloba 216; rossii 216 Argithamnia mercurialina 197 Arikuryroba Ruschiana 261; Tos- tana 261 Arnica louiseana, ssp. frigida 222 Artemisia aleutica 222; arctica 222; borealis 222; norvegica, ssp. saxatilis 222; richardsoni- ana 222; trifurcata 222 Arundinaria gigantea 134, 135 Pl; japonica 134; nipponica 134; tecta 134, 135 PL, 136 Asyneuma prenanthoides 330 Asplenium montanum 78 Asplundia ponderosa 1, 2 PL. rhodea 3 Pl., 4 PI. Aster acuminatus 95; f. discoi- deus 95; f. virescens 95; exilis var. australis 226; foliaceus var. arcuans 95; hirtifolius 236; nemoralis 95; novi-belgii 95; pauciflorus 226, 231, 234; radula 95; richardsonii 222; sibirieus 221; subintegerrimus 221; umbellatus, f. intercedens 95; var. pubens 95 Astragalus alpinus 218; polaris 219; umbellatus 219 368 Astranthium integrifolium 234; mexicanum 226, 227, 235; ortho- podum 227, 231, 234; purpura- scens 227; xanthocomoides 227, 231, 235; xylopodum 227, 231, 234 Athyrium — Filix-femina, var. Michauxii 87; f. lavrentianum 87 Atriplex glabriuscula 91; patula, var. hastata 91; var. patula 91 Baccharis glutinosa 227 Bactris Burretii 259; dakamana 259; essequiboensis 259; Glea- sonii 259; microcarpa 259; Schultesii 259; Stahelii 259; trinitensis 259 Bamboo Rhizome Anatomy, A New Feature In 134 Bambusa nipponica 134 Banisteriopsis caapi 9, 10; inebri- ans 10; Rusbyana 11 Barbosa Getuliana 260 Barkley, T. M., The Integradation of Senecio plattensis and Sen- ecio pauperculus in Wisconsin 65 Barr, Margaret E. and Howard E. Bigelow, Contributions to the Fungus Flora of Northeast- ern North America. III 289 Barya parasitica 300; var. caespi- tosa 300 Bellucia umbellata 18 Berlandiera lyrata 229, 236 Bertia moriformis 300 Betula exilis 216; glandulosa 216, 312; var. sibirica 215; lutea 23, 90; nana 216; ssp. exilis 215; nigra 77; papyrifera 156; var. cordifolia 84, 90; var. papyri- fera 90; populifolia 90 Bidens frondosa 95 Bigelow, Howard E. and Margaret E. Barr, Contribution to the Rhodora [Vol. 65 Fungus Flora of Northeastern North America. III 289 Boehmeria cylindrica 366 Botrychium matricariaefolium 76; simplex, var. tenebrosum 87 Braya siliquosa 211, 217, 218 Britton and Brown Illustrated Flora, A Companion Volume to the New (Review) 243 Bromis inermis ssp. pumpellianus 211; var. arcticus 213; pumpel- lianus 213; var. villosissimus 213 Bulbostylis capillaris 26, 31; var. crebra 26, 30 Bumelia lacuum 280, 281, 282; tenax 280, 282 Bumelia lacuum, Small, The Iden- tity of 280 Bupleurum americanum 219; tri- radiatum 219; ssp. arcticum 219 Byssosphaeria barbicincta — 305, 306 Cacalia sinuata 231 Cakile edentula 91 Calamagrostis canadensis s.l. 312; var. robusta 88; var. scabra 88; epigejos, var. georgica 74 Calathea acuminata 9 Calea palmeri 229, 236, 237; scabra 229, 231, 236, 237; tri- chotoma 236 Callitriche palustris 77; verna 77 Colonectria ^ belonospora 308; cerea 307; fulvida 307 Calopogon pulchellus 273 Caltha palustris 216; var. arctica 216 Campanula acuminata 329; ameri- cana 325, 327, 329, 331, 333, 334; angustiflora 322, 329, 334, 335; aparinoides 326, 329, 333, 334, 335; aurita 324, 329, 333, 334, 335; californica 326, 329, 334, 1963] 335; chamissonis 323, 329, 330, 332, 334, 335; dasyantha 329, 332; divaricata 321, 327, 329, 333, 334; exigua 322, 329, 334, 335; flexuosa 329; floridana 326, 329, 331, 334, 335; glomerata 328, 329; heterodoxa 331; inter- cedens 331; langsdorffiana 330; lasiocarpa 323, 329, 332, 334, 335; latisepala 331; linifolia 331; linnaeifolia 329; mac- dougalii 331; pallasiana 332; parryi 325, 330, 331, 332, 333, 334, 335; var. idahoensis 330, 332; petiolata 331; pilosa 329, 332; piperi 323, 330, 332, 334, 335; prenanthoides 326, 327, 330, 332, 334, 335; rapuncu- loides 329, 332; rentonae 332; reverchoni 322, 331, 334, 335; robinsiae 322, 331, 334, 335; rotundifolia 94, 286, 319, 325, 326, 328, 331, 332, 333, 334, 335; var. alaskana 331; f. albiflora 94; X latisepala 331; sacaja- weana 331; scabrella 323, 330, 3832, 334, 335; scheuchzeri 331; scouleri 325, 3277, 332, 334, 335; trachelium 328, 332; uniflora 221, 323, 324, 332, 333, 334; uliginosa 329; wilkinsiana 324, 332, 334, 335 Campanula Native or Commonly Naturalized in North America, A Checklist and Key to Species of 319 Campanulastrum americanum 329 Capsella rubella 91 Cardamine digitata 217; richard- sonii 217 Carduus acuatus 349; 349; kelseyi 344 Carex abscondita 76; angustior 89; aquatilis 214; atratiformis 338; atrofusca 211, 214; brun- nescens 338; var. brunnescens floccosus Index to Volume 65 369 338; var. sphaerostachya 89, 338; canescens, var. canescens 89; var. disjuncta 89; var. sub- loliacea 89; cephalantha 89; crinita, var. crinita 89; debilis, var. Rudgei 89; Deweyana 76; disperma 89; Emmonsii 77, 89; hormathodes 89; intumescens 23; var. Fernaldii 89; krausei, ssp. krausei 214; leptalea 89; limosa 89; lugens 211, 214; Mackenziei 89; membranacea 214; nigra, var. nigra 89; var. strictiformis 89; paleacea 89; pauciflora 89; paupercula, var. irrigua 89; physocarpa 214; rariflora 214; rostrata var. rostrata 89; var. utriculata 89; scirpoidea 214; scoparia 89; stans 214; stipata 89; tri- sperma 89; viridula 89 Carum Carvi 93 Carya ovata 77 Caryocar gracile 14 Cassandra calyculata 93 Cassiope tetragona 220 Castilleja pallida 221; ssp. cau- data 221; ssp. pallida 221; ssp. typica 221 Catalpa bignonioides 75; speciosa 75 Ceiba pentandra 366 Cerastium beeringianum 216; vul- gatum 91 Ceratiola ericoides 280 Chamaecyparis henryae 361; nootkatensis 361; thyoides 359, 360, 361 Chamaecyparis thyoides, South- eastern Limits of 359 Chamaedaphne calyculata var. angustifolia 273 Chambers, Kenton L., Hedyotis Australis in Georgia 271 Chaptalia dentata 231, 233, 238; integrifolia 238; nutans, var. 370 nutans 238; var. texana 238 Charette, L. A. and H. W. Vogel- mann, A New Station for Rho- dodendron Maximum in North- ern Vermont 22 Chenopodium album 91 Christensen, Earl M., Tall Wheat- grass, A New Roadside Species in Utah 318 Chrysanthellum mexicanum 229, 233, 237 Chrysanthemum integrifolium 222; Leucanthemum 95 Chrysosplenium tetrandrum 218; wrightii 211 Cinna latifolia 89 Cireaea alpina 93 Cirsium altissimum 339, 345, 346, 354; arvense 95; brevifolium 346, 354; brevistylum 339, 341, 353; canescens 346, 347, 348, 351, 354; X foliosum 347; cer- nuum 339, 353; coccinatum 344; discolor 339, 345, 346, 354; drummondii 343, 344, 353; edule 339, 353; flodmanii 339, 347, 354; foliosum 341, 343, 344, 347, 354; hillii 341, 344, 354; horridulum 339, 354; iowense 345; mexicanum 343, 353; mu- ticum 95, 339, 344, 354; nivale 339, 353; ochrocentrum 347, 354; pitcheri 348, 351, 354; plattense 347; pumilum 344, 354; skutchii 339, 341, 354; subeoriaceum 339, 354; sub- niveum 349, 354; tracyi 349, 354; undulatum 339, 343, 350, 354; vulgare 95; wheeleri 350, 354 Cirsium, Chromosome Numbers in Some North American Species of the Genus 339 Cladium mariscoides 26, 30 Claytonia caroliniana 338 Clevenger, Sarah and Charles B. Rhodora [Vol. 65 Heiser, Jr., Helianthus laeti- florus and Helianthus rigidus — Hybrids or Species? 121 Clintonia borealis 23, 90, 312 Clitocybe diatreta 296; eccentrica 290, 291 P1., 295; leptoloma 290, 292 PL, 296; subbulbipes 290, 293; truncicola 294 Pl, 295, 296 Cochlearia officinalis ssp. arctica 217 Cocos campos-portrana 260; getu- liana 260; mataforme 260; nuci- fera 260; Ruschiana 261; tos- tana 261 Coelopleurum lucidum 93 Collinsonia canadensis 366 Collybia strictipes 296, 297 Colombiae, De Plantis Principal- iter ex Orientalibus Partibus Notulae 1 Commelina communis 363 Commelina communis in New Hampshire 363 Compositae, Chromosome Studies in Mexican 225 Conioselinum chinense 93; cnidii- folium 219; vaginatum 219, 220 Conomorpha citrifolia 20; litho- phyta 20; magnoliifolia 20 Convulvulus sepium, f. malacho- phyllus 94; var. sepium 94 Conyza canadensis 227; aff. con- fusa 227, 231 PI. Cooley, George R., Phoradendron rubrum in Florida 190 Coptis groenlandica 23, 91 Corallorhiza trifida, var. verna 90 Cornus canadensis 23, 84, 93; stolonifera 93 Correll, Donovan S. Habenaria canopsea in North America 241 Corydalis pauciflora 217 Cosmos palmeri, var. palmeri 229, 233 PL, 237 1963] Creonectria atrofusca 303 Crepis nana 223 Cucurbitaria seriata 303, 304 Cypripedium acaule 23, 90 Cyperus brevifolius 26, 32, 33; dentatus 26, 30; diandrus 26, 30; Engelmannii 26, 32, 33; erythrorhyzos 26, 31; esculen- tus 26, 30; ferruginescens 26, 33; filicinus 26, 32; filiculmis 26, 31; var. macilentus 26, 30; fuscus 26, 29; Grayii 26, 32; grayioides 81; Houghtonii 26, 33; inflexus 26, 31; odoratus 26, 32; ovularis 26, 29; polystach- yos, var. texensis 26, 32; rivu- laris 26, 30; strigosus 26, 30; var. robustior X Weatherbi- anus 26, 33 Dalibarda repens 23 Danthonia spicata 88 DeJong, D. C. D. and E. K. Long- pre, Chromosome Studies in Mexican Compositae 225 Delphinium brachycentrum 216 Dennstaedtia punctilobula 87 Deschampsia caespitosa, SSp. caes- pitosa, var. glauca 212; flexu- osa 312; var. flexuosa 88; var. montana 88 Dialonectria cerea 307; fulvida 307 Diapensia lapponica 21 Diatrype stigma 307 Dicranopygium omichlophilum 5 PL 6 TL. 7T PL. Diervilla Lonicera 94 Digitaria adscendens 357; diversi- flora 356; sanguinalis 357 Digitaria and Trichachne, New Species of 355 Disporum calcaratum 185, 186; cantoniensis 186, 187, 188; Hamiltonianum 184; lanugino- sum 187; Leschenaultianum Index to Volume 65 371 187; multiflorum 187; parvi- florum 187; Pitsutu 187; ses- sile 165, 187; smilacinum 188; Smithii 187; trachycarpum 187 Dodecatheon frigidum 220 Draba alpina 217; bellii 211, 217; lactea 217; longipes 217; mac- rocarpa 217; nivalis, var. nival- is 217; pilosa 217 Dracocephalum formosius 63; in- termedium 61; moldavica 58; nuttallii 62; speciosum 63; vir- ginianum 58, 63 Drosera filiformis 273; inter- media 91, 273; rotundifolia 91; var. comosa 91 Drosera filiformis in Connecticut 273 Dryas integrifolia 218; octopetala 218; f. canescens-argentea 218 Dryopteris cristata 87; disjuncta 87; noveboracensis 87; Phegop- teris 87; simulata 309; spinu- losa 23; var. americana 87, 313; var. intermedia 87; var, spinu- losa 87; Thelypteris, var. pube- scens 87 Dulichium arundinaceum 27, 29 Dupontia fischeri, ssp. psilosantha 213 Dyssodia cancellata 229, 233 PE, hartwegii 229, 233 PI. Echium vulgare 368 Eleocharis acicularis 27, 29; var. submersa 27, 33; ambigens 2T, 33; calva 27, 31; diandra 21, 31; elliptica 27, 29, 89; Engel- mannii 27, 31; equisetoides X fallax 27, 33; halophila ZG 32, 89; intermedia 27, 32; me- lancearpa 27, 32; microcarpa, var. filiculmis 27, 33; nitida 27, 33; obtusa 27, 29; var. ellin- soidalis 27, 31; var. jejuna at, 31; var. Peasei 27, 33; oli- 372 vacea 27, 30; ovata 27, 30; var. Heuseri 27, 30; palustris 27, 30, 89; var. major 27, 29; parvula 27, 32; pauciflora, var. Fern- aldii 27, 30; quadrangulata, var. crassior 27, 33; Robbinsii 27, 30; rostellata 27, 31; Smallii 27, 30; tenuis 27, 31; tricostata 27, 32; tuberculosa 27, 31 Elm From Ontario, A Buttressed 366 Elymus arenarius 284; var. vil- losus 88; mollis 284 Empetrum nigrum 92, 286 Epilobium adenocaulon X E. can- adense 279; angustifolium, var. angustifolium 93; coloratum 274, 275 Pl., 276 Pl., 277, 278 P1., 279; coloratum X E. com- mutatum 279; X E. lineare 279; glandulosum 277; var. adeno- caulon 93, 274, 276 P1., 277, 278 PL, 279; var. occidentale 93; latifolium 219; leptophyllum 93; wisconsinense 275 Pl., 276 Pl. 278 Pl. Epilobium x Hybr. Nov. 274 Equisetum arvense, var. boreale 87; sylvaticum, var. paucira- mosum $87; variegatum, ssp. variegatum 212 Eragrostis curvula 81; multicau- lis 74; pectinacea 74 Erigeron compositus 211; var. glabratus 222; coronarius 227; delphinifolius 227, 231 Pl., 234, 235; aff. ssp. neomexicanus 227; var. neomexicanus 227; var. oreophilus 228; divergens 228, 231 Pl., 235; var. cinereus 235; pubescens 228, 231 Pl., 235 Eriophorum angustifolium 27, 30, 89; ssp. angustifolium 213; var. majus 27, 33; brachyantherum 213; gracile 27, 31; scheu- wisconsinense, Rhodora [Vol. 65 chzeri 213; spissum 27, 29, 30, 89; tenellum 27, 29, 89; vagin- atum, ssp. spissum 213; ssp. vaginatum 213; virginicum 27, 29, 89; viridi-carinatum 27, 30 Eriosorus accrescens 57; aureo- nitens 57; Lechleri 56; Orbi- gnyanus 56; rufescens 56; Stuebelii 57 Eriosorus, Notes On The Fern Genus 56 Erodium cicutarium 75 Erysimum inconspicuum 78; pal- lasii 217 Erythronium carolinianum 173, 185 Eupatorium betonicum 233; greg- gii 226, 231 Pl., 233 Euphorbia carunculata 203 Euphrasia americana 94; cana- densis 94; Randii, var. Farlowii 94; var. Randii 94 Falearia sioides 81 Fern, A New Station for the Massachusetts 309 Festuca rubra, 213; var. commu- tata 88 Fimbristylis autumnalis 27, 30; var. mucronulata 28, 31; vahlii 81 Fluggea virosa 199 Fosberg, F. R, A Buttressed Elm From Ontario 366 Fragaria virginiana, var. terrae- novae 92; var. virginiana 92 Fritillaria camtschatcensis 187 Fuirena pumila 28, 31 Fungus Flora of Northeastern North America. III, Contribu- tions to 289 Gaillardia pinnatifida 231 Galeopsis tetrahit, var. bifida 94 Galium labradoricum 94; tincto- rium 94; var. subbiflorum 94; 1963] trifidum 94; triflorum 94 Garberia heterophylla 280 Gaultheria hispidula 23, 93; pro- cumbens 273 Geum Peckii 337; rivale, f. vire- scens 92 Glassman, S. F., New Records of Grasses From the Chicago Re- £ion and Lower Michigan 284; Nomenclatural Changes in the Family Palmae 259 Glaux maritima 94 Glyceria canadensis 88 Gnaphalium lavandulifclium 229, 231 PL, 236; uliginosum 95; vuleanicum 236 Goodyera repens, var. ophioides 90 Gorgonidium (Araceae), The Geo- graphic Location of 68 Gorgonidium mirabile 68, 69, 70, 70-3 3 eee p» Graffenrieda candelabrum 19 Grasses From the Chicago Region and Lower Michigan, New Rec- ords of 284 Great Wass Island, Maine, Notes on 285 Grindelia oxylepis 228; sublanu- ginosa 228, 231 Pl., 236 Gymnadenia conopsea 241 Gymnopus strictipes 296 Gymnogramma Lechleri 56; Orbi- gnyana 56; rufescens 57; rufum 56; Stuebelii 57 Habenaria conopsea 241; clavel- lata 84; var. ophioglossoides 90; obtusata 84, 90; odoratis- sima 241; viridis, var. brac- teata 242 Habenaria conopsea in North America 241 Haplopappus spinulosa, ssp. cotula 236; ssp. scabrellus 228, 236, 239 Harries, Hinrich and Doris Lóve, Index to Volume 65 373 Streptopus oreopolus Fern., A Hybrid Taxon 310 Hedyotis australis 271 Pl, 272; crassifolia 271 Pl, 272; var. micrantha 272 Hedyotis australis in Georgia 271 Hedysarum alpinum, ssp. ameri- canum 219; mackenzii 219 Heiser, Charles B., Jr. and Sarah Clevenger, Helianthus laeti- florus and Helianthus rigidus — Hybrids or Species ? 121 Helenium flexuosum 76; nudiflo- rum 76 Helianthus annuus 130; decape- talus 130; hirsutus 130; laeti- flora-tuberosus 132; laetiflorus 121, 122, 128 Pl, 124 Pl, 125 Pl., 126, 127, 128, 129, 130, 131, 132; laetiflorus X laetiflorus 128; X tuberosus 128; rigidus 121, 122, 123 Pl, 124 Pl. 125 Pl., 126, 127, 128, 129, 130, 131, 132; X laetiflorus 127; X tu- berosus 127, 129; severus 121; silphioides 132; strumosus 130; subrhomboideus 121, 122, 128, PL, 124 Pl, 125 Pl. 126, 127, 128, 129, 130, 131, 132; « laetiflorus 125, 131; X rigidus 127; X tuberosus 126, 129, 131; superbus 121; tuberosus 121, 122, 123 Pl, 124 Pl, 125 Pl, 126, 127, 128, 129, 130, 131, 132 Hemicarpha micrantha 28, 31 Henry, L. K., Scheuchzeria in Western Pennsylvania 34 Hepatica americana 77 Herpotrichia barbicineta 305 Herpotrichiella fusispora 303; setosa 303; spinifera 300, 301, 302 Pl., 303 Herrania albiflora 11; atrorubens 12; Cuatrecasana 11; kanuku- ensis 11, 12; Mariae 12, 14; nitida 12; f. sphenophylla 12; 374 pulcherrima var. pacifica 14 Hieracium aurantiacum 95; flori- bundum 95; Pilosella 76; prat- ense 76; vulgatum 76 Hierochloé odorata 89 Hippuris vulgaris 219 Hodgdon, A. R., Commelina com- munis in New Hampshire 363; John Lyon, Nurseryman and Plant Hunter (Review) 283; A Strange Form of Hypericum canadense 357; and Frederic Steele, Hybridization of Rubus hispidus and Rubus setosus 262; and Radcliffe B. Pike, The Flora of the Wolf Islands, New Brunswick, Part I 82; Notes on Great Wass Island, Maine 285 Houstonia caerulea, var. Faxono- rum 312; minima 271; patens 271; pusilla 271 Hsi, Yu-Tseng and Gerald B. Ownbey, Chromoscme Numbers in Some North American Species of the Genus Cirsium 339 Hygrophorus chrysodon 296; fuli- gineus 296 Hypericum canadense 93, 357, 358; var. galiiforme 358; var. magninsulare 358; gentianoides 285, 357; virginicum, var. Fraseri 93 Hypericum Canadense, A Strange Form of 357 Illicium parviflorum 361 Illinois Flora I. The Genus Phy- sostegia, Contributions to 58 Ilex verticillata 92 Impatiens capensis 93 Iris fulva 283; Hookeri 90, 286; prismatica 273; versicolor 90 Nurseryman and John Lyon, Rhodora [Vol. 65 Plant Hunter (Review) 283 Juncus albescens 211, 214; arcti- cus, ssp. alaskanus 214; balti- cus, var. alaskanus 214; var. littoralis 90; brevicaudatus 90; bufonius 90; var. halophilus 90; castaneus 214; effusus, var. Pylaei 90; tenuis 90; triglumis 214 Kalmia angustifolia 93, 273; poli- folia 93 Krigia oppositifolia 273 Kuhnia chlorolepis 226, 231 Pl., 233 Lagotis glauca 220; var. stelleri 220, 221; minor 221 Lakela, Olga, Annotation of North American Polycarpaea 35; The Identity of Bumelia lacuum Small 280 Larix laricina 88, 156 Lathyrus japonicus, var. pellitus 92; latifolius 75; palustris, var. linearifolius 92; var. pilosus 92 Leandra rhodopogon 19 Ledum groenlandicum 93, 286; palustre, ssp. decumbens 220 Lepidium sisymbroides 45 Leontodon autumnalis 95 Leucelene ericoides 228, 231 Pl., 236, 239 Ligusticum scothicum 93 Linnaea borealis, var. americana 23, 94, 306 Linke, William R., Jr., Drosera filiformis in Connecticut 273 Linum arenicola 140 P1., 141, 143; bahamense 140 Pl., 141, 143; var. bracii 139, 141; var. coral- licola 141; catharticum 140 P1., 141; floridanum 140 Pl., 144; var. chrysocarpum 141, 142, 144; var. floridanum 139 Pl., 141, 142; hudsonioides 50, 52 1963] PL, 54; imbricatum 50, 51, 52 Pl.; intercursum 140 Pl., 141, 142, 144; lewisii 139, 140 PL, 142, 143; medium, var. medium 140 Pl., 141, 142, 144; var. tex- anum 140 Pl, 141, 142, 144; multicaule 50, 51, 53; rigidum 51, 140 Pl; var. carteri 141, 143; var. rigidum 139 P1., 141, 142; rupestre 140 Pl., 141, 143; selaginoides 53; striatum 139 Fl. 140 PL, 142, 144; sulcatum 140 PI., 141, 142; var. harperi 142; usitatissimum 139 P1., 140 PL, 142, 143; virginianum 140 Pl, 141, 142, 144; westii 140 Pl. 141, 142, 144 Linum: L. imbricatum and L. hudsonioides, Studies in 50; Pollen Morphology as a Taxo- nomic Tool in 137 Lipocarpha maculata 81 Listera cordata 84, 90 Litschaueria corticiorum 303 Lloyd, Robert M., Tetraploid Pas- siflora incarnata in North Car- olina 79 Lloydia serotina 215 Lomatogonium, Note on 364 Lomatogonium rotatum 287, 364, 365; f. americanum 94 Longpre, E. K. and D. C. D. De- Jong, Chromosome Studies in Mexican Compositae 225 Lonicera canadensis 94; villosa, var. Solonis 94; var. tonsa 94 Lotus corniculatus 75 Lóve, Doris and Hinrich Harries, Streptopus oreopolus, Fern., A Hybrid Taxon 310 Lupinus arcticus 218 Luzula confusa 214; multiflora, var. acadiensis 90; nivalis, var. latifolia 215; parviflora, var. melanocarpa 312; tundricola 215; wahlenbergii 211, 215 Index to Volume 65 375 Lycopodium amnotinum, var. al- pestre 87; var. pungens 87; clavatum 82, 87; inundatum, var. inundatum 77; obscurum, var. dendroideum 87 Lycopus uniflorus 94 McClure, F. A., A New Feature in Bamboo Rhizome Anatomy 134 Machaeranthera gymnocephala 228, 231 Pl; tanacetifolia 228 Mackenzie, Northwest Territories, Life-Forms in the Plains Flora of Southern 149 Maianthemum canadense 28, 90 Marasmius strictipes 296, 297 Matricaria matricarioides 95 Mayna amazonica 14, 16; muri- cida 14, 15 Pl, 16; toxica 16, IARI Medical Botany, Panorama in 97 Melandrium apetalum, ssp. arcti- cum 216 Melanomma porothelia 301, 303; spiniferum 300, 301 Melanopsamma pemiformis 304, 305 Menyanthes trifoliata, var minor 94 Mertensia maritima 94 Miconia astrotricha 19; filamen- tosa 19; fissa 19; marginata 19, 20 Miller, Kim I. and Grady L. Web- ster, The Genus Reverchonia (Euphorbiaceae) 193 Minuartia arctica 216; 216 Miscanthus, Paspalum and Rott- boellia (Graminae), Phytogeo- graphic Notes on 146 Miscanthus sacchariflorus 147, 148 PI. Mitella nuda 91 The Widening elegans 146, 376 Mohlenbrock, Robert H., Contri- butions to an Illinois Flora, I. The Genus Physostegia 58; Further Considerations in Sty- losanthes (Leguminosae) 245 Moneses uniflora 93 Monotropa Hypopithys 93; uni- flora 93 Myosotis alpestris, ssp. asiatica 220 Myrica Gale 273; var. Gale 90 Nectria atrofusca 302 Pl., 303, 304, 305; fulvida 307, 308; ochroleuca 305; pallidula 305 Nemopanthus mucronata 23, 92 New Plants in Old Places 73 New York Botanical Garden Wild Flower Books, The (Review) 189 Nezera imbricata 51, 53 Nicolson, Dan H., The Geographic Location of Gorgonidium (Ar- aceae) 68 Niess!ia barbicincta 302 Pl., 305, 306; crucipila 306; exilis 306; exosperioides 306; pusilla 306 Nymphaea tuberosa 78 Nyssa sylvatica 283 Oakesia floridana 173; puberula 170; sessilifolia 175; var. (?) nitida 170 Oakesiella floridana 173; nitida 170; puberula 170; sessilifolia 169, 175 Oenothera perennis 93 Ombrophytum zamioides 9 Onoclea sensibilis 87, 366 Ophionectria belanospora 308; cerea 307, 308; everhartii 307, 308 Orobanche uniflora 76 Osmunda cinnamomea 23, 87; Claytoniana 87; regalis, var. spectabilis 23 Rhodora [Vol. 65 Ossaea araneifera 20 Otthia seriata 303; staphylina 303 Otthiella seriata 303; staphylina 303 Ownbey, Gerald B. and Yu-Tseng Hsi, Chromosome Numbers in Some North American Species of the Genus Cirsium 339 Oxalis montana 92 Oxytropis glutinosa 219; leucan- tha 219; nigrescens 219; ssp. bryophila 219; ssp. pygmaea 219; viscida 219; viscidula 219 Palmae, Nomenclatural Changes in the Family 259 Parnassia caroliniana 362; gran- difolia 361; kotzebuei 218 Parrya nudicaulis 218; ssp. in- terior 218 Parthenium hysterophorus 237 Paspalum fimbriatum 146 Paspalum, Rottboellia, and Mis- canthus (Graminae), Phyto- geographic Notes on 146 Papavar alaskanum 217; counii 217; radicatum 217 Passiflora incarnata 79; lutea 79 Passiflora incarnata in North Carolina, Tetraploid 79 Pedieularis arctica 221; capitata 221; flammea 221; langsdorfii 221; oederi 211, 221; pennellii 221; sudetica, ssp. albolabiata 221; ssp. pacifica 221; verticil- lata 221 Pennsylvania, Western 34 Persea humilis 280 Petasites frigidus 222 Phacelia dubia 273 Phaeocollybia christinae 297; jen- nyae 297; laterarius 300; rufipes 297, 298 P1., 300 Phleum pratense 89 229, ma- Scheuchzeria in 1963] Phlox borealis 220; sibirica, ssp. borealis 220 Phoradendron flavescens 190; rubrum 190, 191; serotinum 190 Phoradendron rubrum in Florida 190 Phragmites communis, var. Ber- landieri 77 Phyllanthus niruri 196; penta- phyllus 196; platylepis 198; polygonoides 198 Phylleutypa wittrockii 306 Physalis pubescens, var. grisea 78 Physostegia angustifolia 58, 59, 60, 61, 62; formosior 63; inter- media 58, 59, 60, 61, 62; parvi- flora 58, 59, 60, 61, 62; speciosa 58, 59, 60, 61, 62, 63, 64; vir- giniana 58, 59, 60, 61, 62, 63, 64; f. candida 63 Physostegia, Contributions to an Illinois Flora, I. The Genus 58 Picea abies 74; glauca 84, 88, 155; mariana 88, 155; rubens 88 Pieris floribunda 283; phillyrei- folia 361 Pike, Radcliffe B., Note on Lom- atogonium 364; Note on Pri- mula laurentiana in Maine 286; and A. R. Hodgdon, The Flora of the Wolf Islands, New Brunswick, Part I 82; Notes on Great Wass Island, Maine 285 Pinus Banksiana 156, 285; clausa 280, 360; palustris 360; resin- osa '/4 Pitcairnia mituensis 7, 8 Pl., 9 Plantago juncoides 365, var. decipiens 94; var. glauca 94; var. laurentiana 94; lanceo- lata, var. sphaerostachya 75; major 94; oliganthos, var. fal- lax 94 Plants From Cape Sabine, Alaska, An Annotated List of Vascular 302 Pl., Index to Volume 65 377 208 Plowrightia staphylina 303, 304 Poa alpina 213; annua 88; arctica 212; ssp. arctica 212; compressa 88; glauca 212; komarovii 212, 213; lanata 212, 213; palustris 88; pratensis 88 Pohl, Richard W., Phytogeo- graphic Notes on Rottboellia, Paspalum and Miscanthus (Graminae) 146 Polemonium acutiflorum 220 Pollen Morphology as A Taxono- mic Tool in Linum 137 Polycarpaea, Annotation of North American 35 Polycarpaea corymbosa 41, 48; var. brasiliensis 48; cuspidata 44; nebulosa 35, 36 Pl., 37 Pl., 38 Pl., 39 Pl., 40 PI., 43 Polygonum aviculare 91; var. lit- torale 91; var. vegetum 91; bistorta, ssp. plumosum 216; Convolvulus 91; cuspidatum 75; sachalinense 75; sagittatum 91; viviparum 216 Polypodium virginianum 87 Polytrichum commune 273 Poole, James P., A New Station for the Massachusetts Fern 309 Populus balsamifera 156; del- toides 366; tremuloides 90, 156 Potamogeton alpinus, var. subel- lipticus 76; epihydrus, var. Nuttalli 76; gemmiparus 76; gramineus 78; nodosus 78 Potentilla anserina 92, 365; bi- flora 211, 218; Egedei, var. groenlandica 92, 365; emargi- nata 218; hyparctica 218; sim- plex, var. calvescens 92; norve- gica 92; ledebouriana 211, 218; palustris 92, 218; Robbinsiana 337; uniflora 218 Prenanthes altissima, f. integra 95; trifoliolata, var. nana 95 378 Primula borealis, var. borealis 220; egaliksensis 220; laurenti- ana 286, 287, 288, 364, 365; sibirica 211, 220; tschuktscho- rum 211, 220 Primula laurentiana Notes on 286 Protium heptaphyllum 114 Prunus americana 77; pensylvan- ica 92, 156; virginiana 92, 156 Psilactis asteroides 228 Psilocarya nitens 28, 33; poides 28, 33 Psilostrophe gnaphalodes 231, 233 Pl., 238, 239; tagetina 231, 233 PL, 238 Pteridium aquilinum, var. latius- culum 87 Puccinellia laurentiana, var. lau- rentiana 88; paupercula, var. alaskana 88; var. paupercula 88 Pyrenoglyphis microcarpa 259 Pyrola grandiflora 220 Pyrus americana 23; decora 92; melanocarpa 92 in Maine, scir- Quercus Chapmanii 280; har- vardii 204 Ranunculus acris 91; Cymbalaria 91; nivalis 216; pallasii 217; reptans 78; sulphureus 217 Raphanus Raphanistrum 91 Reverchonia arenaria 193, 201, 206 Reverchonia (Euphorbiaceae), The Genus 193 Rhapidophyllum hystrix 361 Rheum Rhaponticum 91 Rhinanthus borealis 338; Crista- galli, var. fallax 94 Rhododendron calendulaceum 283; canadense 78, 93; maximum 22, 23, 24 Pl., 25 Rhynchospora alba 28, 29; capil- Rhodora [Vol. 65 lacea 28, 32, 78; capitellata 28, 29; fusca 28, 30; inundata 28, 32; macrostachya 28, 31; Tor- reyana 28, 32 Ribes glandulosum 92, 338; hir- tellum, var. calciola 91; var. hirtellum 91; lacustre 92 Rickett, H. W., The New York Botanical Garden Wild Flower Books (Review) 189 Rivea corymbosa 116, 119 Pl. Rogers, C. Marvin, Studies in Li- num: L. imbricatum and L. hudsonioides 50; and K. S. Xav- ier, Pollen Morphology as a Taxonomic Tool in Linum 137 Rollins, Reed C., A Companion volume to the New Britton and Brown Illustrated Flora (Re- view) 242; Protandry in Two Species of Streptanthus (Cru- ciferae) 45 Rosa gallica 75; nitida 92; virgin- iana 92 Rotantha floridana 329; robinsiae 331 Rottboellia exaltata 146 Rottboellia, Paspalum and Mis- canthus (Graminae), Phytoge- ographic Notes on 146 Rubus adjacens 269; alleghenien- sis 92, 262; canadensis 92; chamaemorus 218, 286; hispi- dus 265, 266, 267, 268 Pl., 269, 273; X setosus 268 Pl.; idaeus, var. aculeatissimus 92; var. canadensis 92; var. strigosus 92; f. tonsus 92; jacens 269; pubescens, var. pilosifolius 92; var. pubescens 92; setosus 265, 266, 267, 268 P1., 269; spiculosis 269; tholiformis 269; trifrons 269; vermontanus 92 Rubus hispidus and R. setosus, Hybridization of 262 Rumex Acetosella 91; arcticus 1963] 216; domesticus 91; pallidus 91 Ruppia maritima, var. longipes 88 Russel, Norman H., Viola palust- ris L. in Arizona 49 Sabacina incrustans 301 Sabal Etonia 280 Sabazia humilis 229, 233 pl., 237 Sagina nodosa 91; procumbens 91, 365 Salix alaxensis, var. alaxensis 215; arctica 215; babylonica 75; Bebbiana, var. capreifolia 90; brachycarpa, ssp. niphocla- da 215; discolor 90; farrae, ssp. walpolei 215; floridanus 361; glauca 215; X brachycarpa, ssp. niphoclada 215; humilis 90; niphoclada 215; pedicellaris, var. hypoglauca 90; phlebophyl- la 215; polaris, ssp. polaris 215; ssp. pseudopolaris 215; reticu- lata 215; richardsonii 215; rigida 90; stolonifera 211, 215; Uva-ursi 337 Sambucus pubens 94 Sanguisorba canadensis 80 Sanguisorba canadensis, A Re- cent Illinois Collection of 80 Sanvitalia procumbens 229, 233 PL Sarracenia purpurea 91 Saurauia pruinosa 13 Pl., 14 Saussurea angustifolia 222 Savia erythroxyloides 196 Saxifraga bronchialis, ssp. funsto- ni 218; cernua 218; davurica 211; ssp. grandipetala 218; eschscholtzii 211, 218; flagel- laris, ssp. flagellaris 218; ssp. platysepala 218; foliolosa 218; hieracifolia, var. rufopilosa 218; hirculus 218; oppositifolia 218; punctata, ssp. nelsoniana 218 Scheuchzeria in Western Penn- Index to Volume 65 379 sylvania 34 Scheuchzeria palustris, var. amer- icanus 34 Schoenobiblus peruvianus 116, 117 Pl. Schultes, Richard Evans, Plantae Austro-Americanae XI. De Plantis Principaliter ex Colom- biae Orientalibus Partibus Notulae 1; The Widening Panorama in Medical Botany 97 Scirpus acutus 28, 29; americanus 28, 30; ancistrochaetus 28, 33, 34; atrocinctus 28, 29, 89; atro- virens 28, 30; var. georgianus 28, 29, 30; cespitosus, var. cal- losus 28, 30; var. delicatulus 28 32; Clintonii 28, 30; cyperi- nus 28, 30; var. pelius 28, 29; expansus 28, 31; fluviatilis 28, 33, 34; Hallii 28, 33, 34; hetero- chaetus 28, 33, 34; hudsonianus 28, 30; lineatus 28, 33, 34; Longii 28, 33, 34; maritimus, var. Fernaldii 28, 32; Olneyi 28, 32; paludosus, var. atlanticus 28, 32, 76; Peckii 28, 32; pedi- cellatus 28, 30; var. pullus 28, 30; polyphyllus 28, 32; Pur- shianus 28, 30; robustus 28, 32, 76; rubricosus 28, 3883, 34; rubrotinctus 28, 29; Smithii 28, 30; Steinmetzii 29, 38, 34; sub- terminalis 29, 30; Torreyi 29, 31; validus, var. creber 29, 89; verecundus 29, 31 Scutellaria epilobiifolia 94 Schkuhria anthemoidea, var. wislizeni 231, 233 Pl., 238, 239 Scleria pauciflora, var. caro'iniana 29, 33, 34; reticularis 29, 31, 81; triglomerata 29, 31; verticillata 29, 33, 34 Securinega buxifolia 199 Sedum Rosea 91, 286 380 Senecio atropurpureus, ssp. atro- purpureus 222; ssp. frigidus 222; aureus 65; conterminus 222; fuscatus 222; hyperbore- alis 211, 222; lugens 222; obovatus 66, 67; pauperculus 65, 66, 67; var. pauperculus 65; plattensis 65, 66, 67; resedifoli- us 222; smallii 66, 67; strep- tanthifolius 66, 67; sylvaticus 95; tomentosus 66, 67; vulgaris 95 Senecio plattensis and Senecio pauperculus in Wisconsin, The Intergradation of 65 Seymour, Frank C., New Plants in Old Places 73 Shetler, Stanwyn G., An Annot- ated List of Vascular Plants From Cape Sabine, Alaska 208; A Checklist and Key to the Species of Campanula Native or Commonly Naturalized in North America 319 Silene acaulis, var. exscapa 216 Silphium perfoliatum 75 Sisyrinchium montanum, var. cre- brum 90 Sixteenth Report of the Commit- tee on Plant Distribution 26 Smelowskia calycina 211; ssp. in- tegrifolia 217 Smilacina stellata 90; var. crassa 90; trifolia 90 Solanum platyphyllum 20, 21 Solidago canadensis 95; gramini- folia, var. Nutallii 95; macro- phylla 94; var. thyrosoidea 312; multiradiata 221; rugosa, var. villosa 95; sempervirens 94; uliginosa, var. terrae-novae 95 Sonchus arvensis 95; asper 95 Spartina X caespitosa 76 Spathantheum orbignyanum 68, 69, 72 Spergularia marina, var. leio- Rhodora [Vol. 65 sperma 91 Sphaeria atrofusca 303, 304; cerea 307 Sphenopholis nitida 78 Spiraea latifolia 92 Staphylea trifolia 304 Steele, Frederic L., Alpine Zone of the Presidential Range (Re- view) 337; A Double-Flowered Form of Diapensia lapponica 21; and A. R. Hodgdon, Hybrid- ization of Rubus hispidus and Rubus setosus 262 Stellaria calycantha, var. isophyl- la 91; graminea 91; longipes 216; media 91 Stephanomeria pauciflora 231, 233 Pl., 238; tenuifolia 231, 233 Pl., 238 Stipa comata 284; spartea 284 Streptanthus carinatus 45, 46, 47 Pl., 48; cutleri 45, 46, 47 P1., 48 Streptanthus (Cruciferae), Pro- tandry in Two Species of 45 Streptopus amplexifolius 184, 310, 311, 312, 313, 314 Pl., 315, 316; var. americanus 90, 186, 311, 312; var. denticulatus 311, 312; var. oreopolus 312; oreopolus 310, 311, 312, 313, 314 PI., 315, 316; roseus 187, 310, 311, 312, 313, 314 PI., 315, 316; var. per- spectus 90, 312 Streptopus oreopolus, Fern. A Hybrid Taxon 310 Stylosanthes angustifolia — 249, 355; biflora 249, 254 Pl., brac- teata 246, 250 Pl.; calcicola 246, 250 Pl, capitata 246, 250 PIl.; cayennensis 249, 255 P1., erecta 246, 247, 251 Pl.; figueroae 257 Pl., 258; fruticosa 247, 252 PL; guyanensis 249, 254 Pl.; hamata 245, 246, 248, 254 Pl; hippo- campoides 249, 255 Pl.; humilis 245, 257 Pl, 258; ingrata 246, 1963] 256 P1., 258; leiocarpa 249, 255 Pl.; macrocarpa 247, 248, 253 Pl.; macrosoma 256 Pl. 258; mexicana 246, 251 Pl.; montevi- densis 257 Pl., 258; mucronata 246, 247, 248, 252 Pl.; nervosa 247, 248, 253 Pl.; procumbens 245; scabra 247, 251 Pl.; seri- ceiceps 246, 248, 249, 254 Pl; suborbiculata 246, 256, 258 PI.; subsericea 248, 253 Pl.; suffrut- icosa 246, 248, 253 Pl.; sundaica 245, 246, 247, 252 Pl.; sympodi- alis 247, 249, 252 Pl.; tubercu- lata 247, 251 Pl.; viscosa 245, 257 Pl., 258 Stylosanthes (Leguminosae), Further Considerations in 245 Suaeda sp. 91 Swallen, Jason R., New Species of Digitaria and Trichachne 355 Swietenia Mahogani 190 Syagrus Campos-Portoana 260; coronata 260, 261; Getuliana 260; metaforme 260; X meta- forme 260; Ruschiana 261; Tos- tana 261; vagans 260 Tagetes lucida 231, 233 PI. Taraxacum erythrospermum 95; officinale 95; lyratum 222; phy- matocarpum 211, 222 Taxus canadensis 88 Tetracoccus ilicifolius 196 Thalictrum polygamum, polygamum 91 Thieret, John W., Life Forms in the Plains Flora of Southern Mackenzie, Northwest Terri- tories 149 Thuja occidentalis 77, 88 Tradescantia virginiana 75 Trichachne affinis 355; insularis 355; sacchariflora 355 Trichachne, New Species of Digi- taria and 355 var. Index to Volume 65 381 Trichosphaeria barbicincta 305 Tricyrtis hirta 186 Trientalis borealis 23, 94 Trifolium pratense 92; repens 92 Triglochin maritima 88 Trillium erectum 77; undulatum 23 Trisetum melicoides, var. majus 78; spicatum 212 Tsuga canadensis 77 Tubeufia cerea 302 P1., 307, 308 Tryon, Alice F., Notes on the Fern Genus Eriosorus 56 Ugent, Donald, Epilobium X Wis- consinense, Hybr. Nov. 274 Ulmus americana 366 Urtica gracilis 91 Uvularia acutifolia 184; amplexi- caule 184, amplexifolia 158, 184; Betua 184; calcerata 185; caroliniana 185; chinensis 186, 187; cirrhosa 186; distorta 186; flava 178, 179, 180, 181; flori- dana 161, 163, 164, 166 P1., 167, 169, 173 Pl; grandiflora 160, 161, 162, 163, 164, 165, 166 PI., 167, 170, 176, 178 Pl., 179, 181, 182, 183, 186, 187; f. latifolia 176; f. variegata 176; grandi- folia 186; Hamiltoniana 186; hirta 186; lanceolata 177, 186, 187; lanuginosa 187; DB major 187; Leschenaultiana 187; multiflora 187; nitida 170; op- positifolia 187; parviflora 187; perfoliata 158, 160, 161, 162, 163, 164, 165, 166 pl., 167, 168, 170, 176; 178; 179; 4130 PL, 181, 182, 183, 186, 187; a major 176; B. major 18: -Piotr 187; puberula 160, 161, 162, 163, 164, 166 PL, 101, 109, te PL, 185, 186; 187; vär. nitida 102, 170} pudica 187; var. nitida 170; rosea 187; sessiliflora 187; ses- 3 1753 00341 3876 silifolia 158, 161, 163, 164, 165, 166 PL, 167, 169, 174, 175, 176 Pl. 183, 184, 187; sessilis 165, 187; Smithii 187; umbellata 188; ? viridescens 188 Uvularia (Liliaceae), A Revision of the North American Genus 158 Vaccinium angustifolium, var. laevifolium 93, 273; caespitos- um 312; Oxycoccos 93; var. ovalifolium 93; macrocarpon 93, 273; myrtilloides 23; Vitis- Idaea 286; var. minus 93 Valeriana capitata 221 Veratrum viride 312 Verbesina callilepis 229, 233 Pl., 237; tetraptera 229 Viburnum alnifolium 23; noides 23, 94; edule 338 cassi- Viola aurea 49; canadensis 49; charlestonensis 49; cucullata, var. microtitis 93; incognita, var. Forbesii 93; var. incognita 93; Mackloskii, ssp. pallens 93; nephrophylla 49; nuttallii 49; palustris 49; pedatifida 49; pubescens 78; Rafinesquii 49, 273; septentrionalis 93; um- braticola 49 Viola palustris in Arizona 49 Virola calophylla 115 Pl. Viscum rubrum 191 Vogelmann, H. W. and L. A. Cha- rette, A | New Station for Rhododendron maximum in Northern Vermont 22 Rhodora [Vol. 65 Wahlenbergia californica 329 Ward, Daniel B., Southeastern Limit of Chamaecyparis thy- oides 359 Webster, Grady L. and Kim I. Miller, The Genus Reverchonia (Euphorbiaceae) 193 Wheatgrass, Tall, A New Road- side Species in Utah 318 Wilbur, Robert L., A Revision of the North American Genus Uvularia (Liliaceae) 158 Winterringer, Glen S., A Recent Illinois Collection of Sanguisor- ba canadensis 80 Wolf Islands, New Brunswick, The Flora of, Part I 82 Xanthocephalum gymnosperm- oides 228, 231 Pl., 236; serico- carpum 228, 231 Pl., 236 Xavier, K. S. and C. Marvin Rogers, Pollen Morphology as a Taxonomic Tool in Linum 137 Xenophya branceaefolia 69 Ximenesia encelioides 229, 233 P1., 238 Yuyba dakamana 259; essequi- boensis 259; Gleasonii 259; Schultesii 259; Stahelii 259; trinitensis 259 Zexmenia palmeri 229, 238 Zizania aquatica, var. aquatica 76 Zostera marina, var. stenophylla 76