Hovova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS ` RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD CARROLL EMORY WOOD, JR, VOLUME 58 1956 The New England Botanical Club, Ine. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. FARLOW H EFERENCE Linnmseny AR: hi LE kya Dodora - JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL ` STUART KIMBALL HARRIS RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD CARROLL EMORY WOOD, JR. Vol. 58 January, 1956 No. 685 CONTENTS: Authorship and Nomenclature of Bur Clovers (Medicago) Found Wild in the United States. Lloyd H. Shinners............. The Variations of Lilium canadense Linnaeus. B. Boivin and W.J. CE eher EIER UOI cS HORA 14 Nuttall’s Quarrel with Pursh. Jeannette E. Graustein........... 20 Validity of Nuttall’s Names in Fraser’s Catalogue. Arthur Cronquist, David D. Keck and Bassett Maguire.............. 23 Ilex glabra forma leucocarpa: a White-fruited Holly. Frank W. Wadi ido oes whe ee ey Os sS A 25 Viability of Seed of the Black Locust. Clarence R. Hanes....... 26 Silene virginica in the Gulf States. John Adam Moore.......... 27 The New England Botanical Club, Ine. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $4.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 40 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $4.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scien- tific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of publica- tion. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U.S. A. Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 January, 1956 No. 685 AUTHORSHIP AND NOMENCLATURE OF BUR CLOVERS (MEDICAGO) FOUND WILD IN THE UNITED STATES Luoyp H. SHINNERS Frew genera have accumulated a literature so abounding in misquotations, misinterpretations, and misapplications as Medi- cago. The only monograph, the preliminary one by Urban, was published more than eighty years ago. Only a small fraction of the synonymy is included by that author, and his taxonomic treatment is more often a cataloguing of morphological types than a delimitation of biological entities. The same can be said of most of the Old World floras. Lowe’s Manual Flora of Madeira is almost unique in supplying field observations, and synonymy in critical detail. Recent European floras (e. g., Willi Christiansen, Neue kritische Flora von Schleswig-Holstein, 1953) repeat the treatment of Urban with nomenclatural altera- tions made by Burnat and Briquet, or (P. Fournier, Les Quatre Flores de la France, 1946) revert to an assortment of invalid and quite unjustifiable names. Hylander’s Nomenklatorische und systematische Studien (1945) unfortunately mention only one species, M. minima, with an erroneous synonymy and a conclusion which I consider incorrect. What started as a seemingly minor routine task, checking the nomenclature of the species in my Flora of North Central Texas (which includes 6 of the 9 generally accepted as part of the North American flora), proved to be the most laborious and intricate biblio- graphic problem encountered in preparing the entire work. M. sativa L., M. falcata L., and M. lupulina L. present no special difficulties. They are therefore not discussed here, 2 Rhodora [Vor. 58 beyond mentioning that M. lupulina var. glandulosa Neilreich is in my opinion only a minor variation not worthy of varietal rank. The remaining species are a very different matter. The following annotated list gives the names and authorities correct under present rules of nomenclature, so far as I have been able to work them out. Synonymy is not complete, but is believed to include all names of importance in arriving at the accepted nomenclature, and those current in the United States. M. ARABICA (L.) Hudson, Fl. Ang. 288. 1762. The same combination published later by Allioni, Fl. Pedem. 1: 315. 1785.—M. polymorpha var. arabica L., Sp. Pl. 2: 780. 1753.— M. cordata Desr. in Lam., Encycl. Meth. Bot. 3: 636. Late 1791 or 1792. (Based on the preceding. Desrousseaux states that he does not believe 1t necessary to retain the Linnaean epithet for the plant ‘‘car elle est trés-vraisemblablement aussi étrangére à l'Arabie que le Vicia Benghalensis, Lin., Vest au Bengale; que la Scilla Peruviana, Lin., l'est à l'Amerique; . . .") —M. maculata Sibth., Fl. Oxon. 232. 1794. The same name, with M. polymorpha var. arabica and M. cordata listed as syn- onyms, published later by Willdenow, Sp. Pl. (ed. 4) 3 pt. 2: 1412. 1802. Early English writers treated M. arabica and M. polymorpha (M. hispida, M. denticulata) as one species, under the former name. Bartalini (Cat. Piante ... di Siena, 1776) listed Medicago polymorpha arabica as “nomen triviale” under the diagnostie phrase-name given by Linnaeus for the species poly- morpha—that is, he used arabica in the sense of var. typica, or var. polymorpha. Nearly all later authors have consistently understood M. arabica in the sense that 1t 1s today, as delimited in Urban's monograph. According to Urban (1883), the speci- men in the Linnaean Herbarium under this name belongs to an entirely different species, “M. ciliaris Willd." (considered by Urban as a species newly described in 1802, because of doubt as to its identity with M. polymorpha var. ciliaris L.; the older M. ciliaris (L.) Krocker, 1790, was overlooked). The Lin- naean citations are clear, agreeing with nearly all later usage, contradicting the specimen and the exceptional usage of Barta- lini. There is no reason to depart from present usage in this case. 1956] Shinners,—Bur Clovers in the United States 3 M. LACINIATA (L.) Miller, Gard. Dict. (ed. 8), Medicago 5. 1768. 'The same combination published later by Allioni, l. c. 316. 1785.—M. polymorpha var. laciniata L., 1. c. 781. 1753. According to Urban, both the Linnaean specimen and the citations belong to Allioni's (Miller's) species. M. MINIMA (L.) Bartalini, Catalogo delle Piante . . . alla Città di Siena, pp. 60-61. 1776. The same combination pub- lished later by Desrousseaux in Lam., |. c. 636. Late 1791 or 1792.—M. polymorpha var. minima L.,l. c. 780. 1753. Urban states that there is no Linnaean specimen, but that citations indicate Bartalini's species; this 1s in accord with present usage. As with M. arabica, some early post-Linnaean authors in- cluded M. polymorpha (M. hispida, M. denticulata) under this name. The combination is sometimes credited to Linnaeus’s student, Grufberg (dissertation, Flora Anglica, 1754; Amoen. Acad. 4: 105, 1759), or, in accordance with the early custom of crediting student publications to the supervising professor, to Linnaeus himself. Section Five of Grufberg's work is a list of species “secundum Systema sexuale.” In Section One, Grufberg explained that to allow easy comparison of the floras of Sweden and England, names of species found in both countries are put in italics. The list is in two columns per page, with bare name only, ‘‘brevitatis caussa,” and with numbers after each species referring to entries in Ray’s Synopsis Stirpium Britanniae, edition 3. Four medicagos are listed by name, each on a different line, followed by three lines of dashes, for species (or varleties) without Linnaean names listed by Ray: Medicago falcata 331-1. lupulina 331-2. polymor. arab. 331-1. minima 332-2. SHINE I knees 333-3. i et... soo 333-4. bL onic ae vm oo 333-5 It is possible that the bare name minima, on a line by itself, was intended to stand for a species. It is just as possible that a minor error in printing, or failure to indent because of lack of space, was responsible for the implied new combination. The second and third editions of Species Plantarum continue to list minima as a variety. There does not seem to be con- 4 Rhodora (Vor. 58 vincing evidence of an unequivocal new combination prior to Bartalini’s. Hylander (1945) writes “Medicago minima L. (Grufberg) 1754 (non M. minima Bartalini 1776)." This is certainly an error, perhaps due to the fact that Bartalini’s "N. Tr." (nomen triviale) occupies the first line on page 61, followed without break by the phrase-names of species no. 4, including that of Linnaeus's M. polymorpha. But Bartalini placed his binomials after the phrase-names, not before. Those phrases belonging with M. minima, species no. 3, are at the bottom of the preceding page, and include unequivocally a reference to M. polymorpha var. minima L., with page citation from the second edition of Species Plantarum. According to Urban, M. polymorpha var. hirsuta L., Sp. Pl. 2: 780, 1753, is the same as M. minima; M. hirsuta (L.) Bartal., l. e. 61, is an additional synonym, published at the same time as M. minima. Also to be regarded as synonyms, representing minor forms, are var. compacta Neyraut (pathological; cf. Reynier, 1906 and 1908) and var. longiseta DC., recognized in Fernald's 8th edition of Gray's Manual. Urban accepted var. longiseta, short-spined var. brachyodon Reichenb., and spine- less or nearly spineless var. pulchella Lowe, in addition to var. vulgaris Urban. Howell (1949) has reported var. pubescens Webb from California and var. viscida Koch from Arizona. These correspond to f. pubescens (Webb) Urban and f. viscida (W. Koch) Urban; f. mollissima (Roth) Urban is a third form described in the monograph. Unlike his style under M. hispida, and contrary to his statement that Greek letters indicated forms rather than varieties, Urban listed the three as if they were varieties, crediting them to Webb, Koch, and (Roth) Koch, respectively. All are listed together immediately follow- ing the description of the species; they are not put under var. vulgaris nor any of the three other varieties, which are all listed together after the three forms. On grounds of uniformity with other parts of the monograph, and because Greek letters pre- ceded them, Urban’s names for the three forms must be regarded as new combinations, though he failed to mark them as such. M. orpicuLaris (L.) Bartalini, Catalogo delle Piante . alla Città di Siena, p. 60. 1776. The same combination pub- lished later by Allioni, Fl. Pedem. 1: 324-325. 1785.—M. 1956] Shinners,—Bur Clovers in the United States 5 polymorpha var. orbicularis L., 1. e. 779. 1753.—M. orbiculata Gaertn., De Fruct. et Semin. Pl. 2: 349. 1791. (An evident error for orbicularis.)— Urban states that the Linnaean Her- barium contains under this name two specimens “of the usual form," with 3—4 compact spirals 13-15 mm. in diameter. M. POLYMORPHA L., Sp. Pl. 2: 779. 1753. This includes the plants commonly known as M. hispida, M. denticulata, and M. apiculata (detailed synonymy given below, under the varieties and forms). Few Linnaean species have undergone such thorough disin- tegration as Medicago polymorpha. Al the thirteen original named varieties were soon raised to specifie rank by various authors, and by the early 19th Century, the name polymorpha itself went completely out of use. (Later references to, or combinations based on “M. polymorpha Willd." were errors for M. polycarpa Willd.) Under present rules, however, the name must be reinstated for some member of the original ag- gregate. Since it has not been in use for a century and a half, it has not become a permanent source of confusion and error. Unlike the names of the yellow-flowered species of Trifolium, it cannot be rejected on that account. Urban’s account of the Linnaean species reports herbarium specimens of 10 of the 13 varieties, but none marked as polymorpha only. The name therefore cannot be typified by a specimen. Linnaeus’s under- standing of the components of his aggregate was very poor, as revealed in Urban's notes on the specimens and citations. The pre-Linnaean Historia Plantarum of Morison was much superior to Linnaean works in the treatment of bur clovers, and gives illustrations. Unfortunately, the Linnaean citations for Medicago polymorpha alone (not the varieties) were to pub- lications which offer little help. The diagnostic phrase-name is “Medicago leguminibus cochleatis, stipulis dentatis, caule diffuso," quoted from Hortus Cliffortianus and Hortus Upsalien- sis, and from Royen and Dalibardus. Next is quoted ''Tri- folium cochleatum, fructu nigro hispido," from Bauhin's Pinaz. No illustrations are cited. Following this is the list of 13 named varieties, each with one or more phrase-names quoted from various authors, and for 7 of them, references to figures. The fact that the first two phrase-names stand separate from 6 Rhodora [Vor. 58 all the named varieties indicates that the type should not include any of the latter. The doctrine of residues leads to the same conclusion, since all the varieties were early recognized as species under their own names. The only possibly helpful item in the first name-phrase is ‘‘stipulis dentatis” though this description was applied rather loosely to stipules varying from toothed to rather deeply lobed. The quotation from Bauhin specifies a plant with prickly pods, black at maturity. It is to be assumed that the species polymorpha must be a fairly common and widespread plant, and one which was known to Linnaeus. The plant best qualified, as fitting the meagre deseription, widespread, known to Linnaeus in 1753, and not one of the varieties named at that time, is the later-described var. nigra L., which is nomenclaturally identical with M. hispida Gaertn. (since the phrase-name for nigra is quoted in the original description of the latter, though the epithet is not). According to Urban, Linnaeus had specimens of M. hispida var. pentacycla (a synonym of var. hispida) filed under his varieties coronata, ciliaris, and nigra. The species best known to Bauhin (judging from Binz & Thommen’s Flore de la Suisse, p. 217, 1941) were M. arabica, M. hispida, and M. minima, of which only the second was not named as a variety by Linnaeus in 1753. (Although much is made here of the suggestiveness of Bauhin’s phrase, it is not to be inferred that Medicago polymorpha can be positively typified merely on that basis; it is likely that Bauhin included more than one species under his phrase-name. ) Admittedly the above arguments are weak, but something has to be chosen. There are specimens only for some of the Linnaean varieties. The pre-Linnaean citations are incon- clusive, and in any case should not be permitted to disturb later usage any more than can be helped. The few possible approaches to redefining Medicago polymorpha all point to the complex M. hispida Gaertner, as treated by Urban. (Gaertner’s figure is, to me, unidentifiable, and suggests M. ciliaris more than M. hispida; Urban’s interpretation, which did not include designation of a type element among the infraspecific taxa, is accepted because it included var. nigra, which is technically the basis of the species hispida.) The selection of M. hispida var. pentacycla (DC.) Urban as type element is made on the 1956} Shinners,—Bur Clovers in the United States 7 grounds that this form was represented in Linnaeus’s herbarium, even though misidentified by him under several varietal names. There is flimsy support from the fact that var. nigra repeats a key word from Bauhin’s phrase-name, and is identical with var. pentacycla. Medicago hispida as delimited by Urban is still quite polymorphic (a total of 15 infraspecific taxa are named); it is a common and widespread species (though less so in Linnaeus’s or Bauhin’s time than now); a majority of its forms have spiny pods, black at full maturity; it was known to Linnaeus in 1753, though not well understood by him. Lastly, replacement of M. hispida by M. polymorpha involves no general displacement of names in the genus, and need not occasion any great confusion. Medicago polymorpha Linnaeus is therefore here typified by the element later named var. nigra, or M. hispida var. pentacycla f. longeaculeata of Urban’s monograph, p. 75. As with M. arabica and M. minima, early writers frequently included more than one species or variety (as now understood) under the same name. Fortunately, Urban clarified the Lin- naean and many post-Linnaean names so that the purely bibliographic phase of revising his nomenclature is less formidable than it might have been. The biological interpretation of the taxa is far more difficult. No adequate account can be made without extensive investigations in the Old World, including intensive field work, and cultural and cytogenetic studies. Geographic evidence—so far as it can be gleaned from published floras—suggests that Urban’s arrangement of infraspecific categories (disregarding named forms) is a reasonable one, and may justifiably be retained for the present. It should be mentioned here that on page 44 of his monograph, Urban states that single Latin letters indicate varieties, Greek letters, forms. The latter have been erroneously quoted by Burnat and Briquet as subvarieties. Urban’s primary groups (aa. microcarpa and bb. macrocarpa) are at the level of subspecies, but were given no designation of rank by him; they are designated subspecies in Christiansen’s Neue kritische Flora von Schleswig-Holstein (1953, p. 313). To avold nomenclatural complications, they are not transferred here (see discussion under var. ciliaris). Only the last two 8 Rhodora [Vor. 58 varieties occur in the United States (one of these as yet only under cultivation). For completeness, the two strictly Old World varieties are also included, and essential synonymy given for all four. Following is a translation and partial revision of Urban's key (omitting forms), with order reversed because of modified typification. As additional characters under leads la-lb, Urban has “‘veinlets of legume anastomosing freely and well before reaching lateral nerves” vs. ‘‘veinlets of legume anastomosing only close to lateral nerves, or not at all." These statements contradict descriptions given by some European writers, and do not fit the specimens which have been examined; they are therefore omitted. la. Diameter of middle spirals of legume 7-10 mm. 2a. Spirals 4-6. 00 ee var. polymorpha. 2b. Spirals 1146-4... cece var. tricycla. 1b. Diameter of middle spirals of legume 4—6 mm. 3a. Spirals 5-6... 000 eee var. polygyra. 3b. Spirals 14-81... 0... var. ciliaris. M. POLYMORPHA var. POLYMORPHA. M. polymorpha L., Sp. Pl. 2: 779. 1753. (Emended.)— M. polymorpha var. nigra L., Mant. Pl. 2: 454. 1771.—M. hispida Gaertn., De Fruct. et Semin. Pl. 2: 349. 1791. (Based on the preceding.)— M. lappacea Desr. in Lam., Encycl. Meth. Bot. 3: 637-638. Late 1791 or 1792. (M. polymorpha var. ciliaris, M. poly- morpha var. nigra, M. hispida, all cited as synonyms.)— M. nigra (L.) Willd., Sp. Pl. (ed. 4) 3 pt. 2: 1418-1419. 1802.— M. denticulata var. lappacea (Desr.) Benth., Cat. Pl. Ind. Pyr. p. 103. 1826.— M. lappacea var. macracantha (Webb & Berth.) Lowe, Man. Fl. Madeira p. 158. 1868. (Based on M. denticu- lata var. macracantha Webb & Berth., Phyt. Canar. 2: 64. 1836-1840. This reference not seen. Lowe cites as synonyms M. lappacea Desr., M. nigra *(Willd.) DC.) According to Urban the following also belong here. M. pentacycla DC., Cat. Hort. Monsp. p. 124. 1813. M. lap- pacea var. pentacycla (DC.) Gren. & Godr., Fl. Fr. 1: 390. 1848.— M. hispida (bb. macrocarpa) var. pentacycla (DC.) Urban, Verh. Bot. Ver. Prov. Brandenb. 15: 75. 1873. M. POLYMORPHA var. tricycla (Gren. & Godr.) Shinners, comb. nov. M. lappacea var. tricycla Gren. & Godr., Fl. Fr. 1: 390. 1848. “Varie à courtes et à longues épines.” The form 1956] Shinners,—Bur Clovers in the United States 9 with long spines is taken as type.—M. hispida (bb. macrocarpa) var. tricycla (Gren. & Godr.) Urban, Ind. Hort. Bot. Berol. 1872 App. p. 3. (Not seen. Repeated in Verh. Bot. Ver. Prov. Brandenb. 15: 75. 1873.)—M. hispida (aa. macrocarpa) var. tricycla f. longispina Urban, 11. ce. Including f. microdon (Ehrenb.) Urban, 11. cc., based on M. microdon Ehrenb., Cat. Sem. Hort. Berol. 1827 (not seen). Urban also lists “M. obscura var. macrodon (Ehrenb.) Visiani" as synonym of this form, but Visiani (Fl. Dalm. 3: 284, 1852) merely comments that Ehren- berg’s plant “est varietas tuberculata M. obscurae," without making any formal transfer. M. POLYMORPHA var. polygyra (Urban) Shinners, comb. nov. M. hispida (aa. microcarpa) var. polygyra Urban, Ind. Hort. Bot. Berol. 1872 App. p. 4 (not seen); Verh. Bot. Ver. Prov. Brandenb. 15: 74. 1873. Two forms are named (f. inermis Urban, f. aculeata Urban); the first is taken as the typical form.— M. reticulata Benth., Cat. Pl. Ind. Pyr. p. 101. 1826. Judging from published references, this is a fairly definite geographic variety of the western Mediterranean region, spineless or with very short spines. In recent years experimental plantings of it have been made in Texas, under the erroneous name M. sardoa, which is a synonym of the next variety. M. polymorpha var. polygyra has not so far been found as a wild plant. M. potymMorpHA var. ciliaris (Ser.) Shinners, comb. nov. M. denticulata var. ciliaris Ser. in DC., Prodr. 2: 176. 1825. “M. ciliaris Savi cent. p. 148, non Willd." It is very unfortunate that Seringe did not supply a different epithet for this plant, so as to avoid confusion with M. ciliaris (L.) Krocker, based on M. polymorpha var. ciliaris L. As published, M. denticulata var. ciliaris must be treated as a new name, and is the first to be applied in varietal rank to the most widespread and com- monly planted of the bur clovers. The identity of Savi’s plant does not affect the application of Seringe's name, since the latter also gave a very brief description (“aculeis elongatis"), thereby validly publishing a new variety. An awkward situation arises in connection with the nomencla- ture of this variety, due to confliet between Article 35 of the !In a footnote under Leguminosae in the Flore de France it is stated ‘‘auctore Godron." The established custom of attributing new names to both authors is fol- lowed here, although very precisely only Godron should be mentioned. 10 Rhodora [Vor. 58 International Code (“If any infraspecific taxon which includes the nomenclatural type of the epithet of the next higher taxon is to be mentioned by a subdivisional name, that name must repeat the epithet of the higher taxon unaltered") and paragraph 3 of Article 16 (“For any taxon below the rank of genus the correct name is the combination of the generic name with the earliest available legitimate epithet or epithets validly published with the same rank"). On grounds of priority, it is necessary to adopt the combination M. polymorpha var. ciliaris. But if subspecies are recognized, M. hispida ssp. microcarpa Urban (emend. W. Christiansen) must be transferred to M. polymorpha. Since it is identical as to type with var. ciliaris, and since var. polygyra would also be recognized under it, Article 35 would require the rejection of var. ciliaris, and adoption in its place of the new name, var. microcarpa, cited without author. The use of automatic epithets for typical varieties is a happy im- provement when applied only to the type of a species. When extended to subspecies and varieties additional to the type, it becomes a very mixed blessing. I personally would retain Article 35 only for a hierarchy of epithets indicating the type of a species, but would reverse it with regard to other subspecies or varieties than the typical ones, on the grounds that only thus could changes of names be kept at a minimum. In the present case, for example, it would be more reasonable to say that var. ciliaris, being the oldest varietal epithet, must be retained; that as the earliest one, 1t should supply the subspecifie epithet for any subsequent varieties associated with it. In other words, we would say ssp. ciliaris (no author) var. ciliaris (Ser.) Shinners, rather than ssp. microcarpa (new combination) var. microcarpa (no author). Whichever procedure is followed, one has the license, so to speak, to remove names at will, merely by creating à new one in another category. As Article 35 stands now, the number of changes which must be or could be made is appallingly large. It would be made much smaller by the reversed arrangement for non-typical subspecies here suggested, though not entirely eliminated. For the present, the whole problem is evaded by omitting subspecies entirely. An additional synonym of M. polymorpha var. ciliaris is M. hispida (aa. microcarpa) var. oligogyra Urban, 11. ec. Urban 1956] Shinners,—Bur Clovers in the United States 11 lists three forms, treated by many recent authors as varieties, and by some as species. They are assigned names here solely for purposes of discussion, and convenience in listing synonymy. M. POLYMORPHA var. CILIARIS f. CILIARIS. Length of spines equalling or exceeding radius of spirals. This is the plant which passes in the United States as typical M. hispida, which it is in the sense of Urban, but not as to type. True hispida (M. polymorpha as here delimited) is not found either wild or commonly cultivated in this country. There is a long list of actual and sensu synonyms for f. ciliaris; only the following need be listed here. M. denticulata Willd., Sp. Pl. (ed. 4) 3 pt. 2: 1414. 1802.—M. denticulata var. vulgaris Benth., Cat. Pl. Ind. Pyr. p. 103. 1826. (Fide Visiani.)— M. hispida (aa. microcarpa) var. oligogyra f. denticulata (Willd.) Urban, 11. cc. M. POLYMORPHA var. CILIARIS f. apiculata (Willd., emend. Urban) Shinners, comb. nov. Spines shorter than radius of spirals. M. apiculata Willd., |. c. 1414. 1802. (As delimited by Urban, 11. cc.) —M. hispida (aa. microcarpa) var. oligogyra f. apiculata (Willd.) Urban, 11. ee.—M. hispida var. apiculata (Willd. Burnat, Fl. Alpes Mar. 2: 106. 1896.—M. denticulata var. brevispina Benth., Cat. Pl. Ind. Pyr. p. 103. 1826. (Fide Visiani.)—A ecording to Grenier & Godron and others, the following belongs here. M. sardoa Moris, Stirp. Sard. Elench. p. 15, 1827 (nomen nudum); ex G. Don, Gen. Syst. Dichlam. Pl. 2: 170. 1832. As mentioned before, plants in experimental cultivation as M. sardoa are M. polymorpha var. polygyra, or M. reticulata Benth. M. POLYMORPHA var. CILIARIS f. tuberculata (Gren. & Godr.) Shinners, comb. nov. Spines absent; spirals smooth or merely tubercled. M. polycarpa var. tuberculata Gren. & Godr., FI. Fr. 1: 390. 1848.—M. hispida (aa. microcarpa) var. oligogyra f. tuberculata (Gren. & Godr.) Urban, 11. cc.—M. apiculata var. confinis W. Koch, Syn. Fl. Germ. Helv. p. 164. 1835.— M. hispida var. confinis (W. Koch) Burnat, Fl. Alpes Mar. 2: 106. 1896.— The name apiculata has been applied to this form, and Willdenow’s original description (margine muricatis," "margine tuberculis admodum parvis spinescentibus obsita") suggests that he had this chiefly in mind. There is no sharp break between the three forms. Urban, who examined the 12 Rhodora (Vor. 58 Willdenovian herbarium, was entitled to emphasize “spinescenti- bus" in delimiting the form as he did. It should be noted here that Grenier & Godron do not mention the earlier-named M. tuberculata Moris, Fl. Sard. 1: 447, 1837. They do list as synonym of their var. tuberculata “M. confinis Koch, olim," overlooking the fact that Koch had put his epithet into print, as a variety. The three above forms intergrade completely. In the United States, selection of seed for agronomic purposes tends to fix a limited number of forms. Since escapes from cultivation are a major source of our wild plants, there is in reality an artificial selection, giving the impression that the forms are better marked or more permanent than is really the case. In eastern Texas, f. ciliaris is a common weed, while f. tuberculata is rare. The two intergrade freely when growing together, and this is said to be true of these two and var. polygyra when grown together under cultivation. Comments of some European authors (especially Lowe and Visiani) indicate that this is also true where the plants are native, and justifies Urban’s treatment of the three so-called species as forms. Reasons for accepting var. polygyra in that rank (likewise Urban’s treatment) have already been mentioned. M. PRAECox DC., Cat. Hort. Bot. Monsp., pp. 123-124. 1813. Reported by J. T. Howell as found in 1954 in Butte and Tehama counties, California (Leafl. West. Bot. 7: 199, 1954). Only those works giving helpful taxonomic, ecological, or geographical information are listed below. I am indebted to Mrs. Lazella Schwarten, Librarian, Harvard University Her- barium, for transcriptions of the original descriptions of Medicago reticulata Bentham and M. sardoa Moris.—SOUTHERN METHODIST UNIVERSITY, DALLAS, TEXAS. LITERATURE CITED BRIQUET, Jonn. 1913. Prodrome de la Flore Corse, Tome II. (Medi- cago, pp. 256-274.)—This usually reliable and very helpful work falls down badly with regard to Medicago. Burnat, EMiLE. 1896. Flore des Alpes Maritimes, Vol. IT. (Medi- cago, pp. 93-110.) CHRISTIANSEN, WILLI. 1953. Neue kritische Flora von Schleswig- Holstein. (Medicago, pp. 292-293, 313-314.)—Disappointingly uncritical. 1956] Shinners,—Bur Clovers in the United States 13 Heat, Gustav. 1923. Illustrierte Flora von Mittel-Europa, Band IV, 3. Teil. (Medicago, pp. 1248-1275.) HowEnL, Jonn THomas. 1949. Medicago minima in California and Arizona. Leafl. West. Bot. 5: 187. HYLANDER, Niis. 1945. Nomenklatorische und systematische Studien über nordische Gefiisspflanzen. Uppsala Universitets Arsskrift 1945: 7. (Medicago minima, p. 223.) INTERNATIONAL CODE or BOTANICAL NOMENCLATURE. 1952. (Regnum Veg. vol. 3.) Lowe, RrcHARD THomas. 1868. A Manual Flora of Madeira. Vol. I. (Medicago, pp. 156-157.) REYNIER, ALFRED. 1906. Deux anomalies végétales analogues. Bull. Soc. Bot. France 53: 65-68. 1908. La prétendue espèce Medicago ononidea De Coincy n'est, qu'une forme pathologique du M. minima Lmk.—Demonstra- tion concluante. l.c. 55: 553-557. URBAN, I. 1873. Prodromus einer Monographie der Gattung Medicago L. Verh. Bot. Ver. Prov. Brandenb. 15: 1-85. (Classification of M. hispida was outlined in Ind. Hort. Bot. Berol. 1872 App. pp. 3-4; publication not seen.) . 1877. [Remarks on Medicago, with seed key to 10 species, in Sitzung vom 31. August 1877.] Verh. Bot. Ver. Prov. Brandenb. 19: 125-134. . 1883. Die Medicago-Arten Linne's. Ber. Deutsch. Bot. Ges. 1: 256-262. VISIANI, RoBERTUS DE. 1852. Flora Dalmatica, Tom. III. (Medicago, pp. 279-285.) WAGNER, Frep H. 1948. The bur clovers (Medicago) of Texas. Field & Lab. 16: 3-7. 14 Rhodora [VoL. 58 THE VARIATIONS OF LILIUM CANADENSE LINNAEUS! B. BorviN AND W. J. Copy In the New Britton and Brown Illustrated Flora 1: 418. 1952, H. A. Gleason, apparently following the opinion expressed by E. D. Hull, Roopora 44: 220-7. 1942, has placed Lilium michiganense Farwell in the synonymy of L. superbum L. In RHODORA 44: 453-6. 1942, Hull’s opinion was carefully discus- sed and refuted by E. T. Wherry, who pointed out numerous differences between the two entities in pubescence of the leaves, dorsal ribbing of the tepals, length and curvature of the stamens, and extent of green zone at base of tepals. To these differences, M. L. Fernald, in Gray's Manual of Botany, 8th ed., 1950, p. 435, has added another character derived from the inflores- cence. We agree with Wherry and Fernald that these two entities are distinct, but the degree of distinctiveness is not as sharp as Wherry would lead one to believe. While L. michigan- ense is normally minutely scabrous along the margin of leaves and on the back of the main veins, occasional smooth specimens, such as L. H. Wright, Yates Co., N. Y., do exist and papillose leaf-margins and veins occur in L. superbum. In the latter species the anthers are longer, narrower and curved, but they may be as short as 11 mm. as shown by specimens and described by Wherry. On the other hand the anthers of L. michiganense are generally shorter, thicker and straight, but may occasionally be curved or thin and, as described by Fernald, up to 17 mm. long. The dorsal ribbing of the perianth segments was quite clearly illustrated by Wherry, op. cit., p. 454, but an occasional specimen of L. michiganense will show the characteristic ribs of L. super- bum. Unfortunately, the extent of the green zone at the base of the tepals cannot be checked on dried specimens, but this difference does not appear to be very sharp. It is described by Wherry, loc. cit., as follows: “In .. . L. michiganense this green area is mostly less than 10 mm. long . . . . In L. super- bum it is 10 to 15 mm. long. JU Heavily flowered specimens of L. superbum most often have the flowers in racemes; but umbellate, few-flowered and one- 1 Contribution No. 1488, from the Botany and Plant Pathology Division, Science Service, Canada Department of Agriculture, Ottawa, Ontario. 1956] Boivin and Cody,—Lilium canadense Linnaeus 15 flowered specimens which have the same type of inflorescence as L. michiganense, also occur. However, L. michiganense never presents the racemose type of inflorescence. In none of the six characters discussed above are these two entities sharply distinct; and, as frequent intermediates occur, retention of these taxa at specific level does not seem to be warranted. The two populations, however, are fairly well marked and most specimens can be referred quite satisfactorily to one or the other taxon and retention at an infraspecific level seems amply justified on the basis of the characters described by Wherry. L. michiganense is more closely related to L. canadense than to L. superbum. This was brought out by Wherry, op. cit., p. 435, who stated that “the intermediate L. michiganense surely does not ‘belong to L. superbum,’ lying instead close to L. cana- dense." With the above statement we are in full agreement, although on general appearance alone L. michiganense could be readily confused with L. superbum because both entities have strongly recurved red-suffused tepals. The degree of distinctiveness of L. michiganense and L. cana- dense was submitted to close scrutiny. Of the four usually recognized diagnostie characters, one, the flower color, could not be used with any degree of accuracy because, upon drying, flowers often become more reddish or lose all traces of red color. This change in color is evidenced by the discrepancies between field notes and the present color of the dried flower. The three other characters are as follows: L. canadense L. michiganense 1—Tepals arched, not recurved; 1—Tepals strongly recurved, their stamens and style usually tips nearly reaching or surpass- included or nearly so. ing the summit of the tube; stamens and style long exserted. 2—Filaments straight or nearly 2—Filaments outwardly arched, so. frequently incurved at tip. 3—Stigma deeply lobed. 3—Stigma merely emarginate, subglobose to shallowly lobed. Of the 75 odd sheets at hand, only 49 have fully developed flowers showing all characters listed above. Sorting the speci- mens for all three characters gives us the following results: 16 Rhodora [VoL. 58 12 specimens are typical L. canadense 10 specimens are typical L. michiganense 27 specimens are intermediate This is a very high ratio of intermediates, but 17 of these intermediates are intermediate for style only. If we contrast this number of 17 intermediates for style against the 10 and 12 specimens of each entity that are typical for all three characters, it is obvious that the lobation of the style has no diagnostic value whatsoever. Re-sorting our specimens for the first two characters only gives us the following results and distributions: 17 specimens are typical L. canadense: 4 Nova Scotia, 2 New Brunswick, 7 Quebec, 1 New York, 2 Pennsylvania and 1 West Virginia. 22 specimens are typical L. michiganense: 14 Ontario, 1 Indiana, 1 Illinois, 3 Michigan, 1 Wisconsin, 1 Minnesota and 1 Kansas. 10 specimens are intermediate: 1 Nova Scotia, 1 New Brunswick, 2 Quebec, 4 Ontario, 1 Massachusetts and 1 Pennsylvania. With only two diagnostie characters and such a high propor- tion of intermediates, retention of specific status for these two entities seems unjustified. Yet the two populations appear to be sufficiently distinct both morphologically and geographi- cally to warrant retention at an infraspecific level. We therefore consider that the recognized taxa of the Liliwm canadense group should be brought together and subordinated as follows: Lilium canadense L. ssp. canadense var. canadense f. canadense f. rubrum Britton var. editorum Fern. ssp. michiganense (Farwell) Boivin & Cody f. michiganense f. uniflorum (Farwell) Boivin & Cody f. peramoenum (Farwell) Boivin & Cody ssp. superbum (L.) Boivin & Cody These taxa may be briefly characterized as follows: LILIUM CANADENSE L. ssp. CANADENSE. L. canadense L., Sp. Pl. 1: 303. 1753; L. penduliflorum DC. ex Redouté, Les Liliacées 2: 105. 1805 (non vidimus); L. pendulum Spae, Mém. Cour. Ac. Roy. Se. Brux. Vol. 19, Essai d'une monographie du genre Lis, p. 28. 1847 (non vidi- 1956] Boivin and Cody,—Lilium canadense Linnaeus 17 mus) nee L. pendulum Noronha 1790. Leaves usually scaberulous along the margin and along the nerves underneath; flowers solitary to umbellate (occasionally disposed in a group of umbels); tube with a green zone at base up to 10 mm. long; tepals recurved from near the middle, but not reflexed, yellow to orange, occasionally suffused with red, or completely red; midrib of petal more or less rounded on back; stamens and style included or only slightly exserted; filaments straight, rarely curved; anthers commonly oblong to lance-oblong, nearly always straight, usually less than 12 mm. long. LILIUM CANADENSE L. var. CANADENSE. Median cauline leaves lanceolate to linear-lanceolate, usually more or less acuminate at tip; flowers commonly yellow to orange, often suffused with red, rarely deep red; petals usually broad. LILIUM CANADENSE L. f. CANADENSE. L. canadense flavum Pursh, Fl. Am. Sept. 1: 229. 1814; L. canadense flavum Hort. ex Waugh, Bot. Gaz. 27: 356. 1899; L. canadense var. flavum (Hort. ex Waugh) Waugh ex Bailey, Cycl. Am. Hort. 2: 922. 1900; L. canadense f. flavum (Hort. ex Waugh) Vict., Contr. Lab. Bot. Un. Montr. 14: 15. 1929; L. cana- dense var. luteum Waugh ex Bailey, Cycl. Am. Hort. 2: 922. 1900 (nomen alternativum). Flowers yellow to orange, occasionally suffused with red; petals commonly broad. Specimens examined from Nova Scotia, New Brunswick, southern Quebec, New York, Pennsylvania and West Virginia. LILIUM CANADENSE L. f. RUBRUM Britton, Bull. Torr. Bot. Club 17: 125. 1890; L. canadense var. rubrum (Britton) Waugh ex Bailey, Cyel. Am. Hort. 2: 922. 1900; L. canadense B coccineum Pursh, Fl. Am. Sept. 1: 229. 1814. Flowers red; petals up to 20 mm. across, usually about 15 mm. broad. Specimen examined from southern Quebec (Huntingdon Co.). LILIUM CANADENSE L. var. EDITORUM Fernald, Rhodora 45: 393. 1943; L. canadense ssp. editorum (Fernald) Wherry, Bartonia 24: 7. 1947. Median cauline leaves elliptic or oblong to oblong-lanceolate, not acuminate at tip; flowers red; petals narrower, 8-13 mm. across, usually less strongly recurved than in the preceding variety. We have only two specimens at hand. This variety is reported to grow in moun- tains and upland dry woods from Pennsylvania and southern Indiana south to Alabama. LILIUM CANADENSE L. ssp. michiganense (Farwell) stat. n., L. michigan- ense Farwell, Bull. Torr. Bot. Club 42: 353. 1915; L. michiganense Farwell, var. umbelliferum Farwell, Bull. Torr. Bot. Club 42: 353. 1915; L. michi- ganense Farwell f. wmbelliferum (Farwell) Wherry, Bartonia 24: 8. 1947. L. pardalinum Kellogg var. Bourgaei Baker, Journ. Linn. Soc. London, Bot. 14: 242. 1875. Leaves usually scaberulous as in ssp. canadense; inflorescence also as in ssp. canadense; tepals strongly re- curved from below the middle or from near the base, the tips nearly reaching or surpassing the base of the tube as in ssp. superbum; flowers orange, commonly reddish; color at base of the tube as in ssp. canadense; 18 Rhodora [Vor. 58 midrib of the petal as in ssp. canadense, or rarely more or less sharply ridged as in ssp. superbum; stamens and pistil long-exserted, usually by about half their length; filaments usually outwardly curved, sometimes incurved again at tip, rarely straight or nearly so; anthers as in ssp. canadense. Specimens seen from southern Ontario, Indiana, Illinois, Michigan, Wisconsin, Minnesota and Kansas. The most reliable distinguishing character between this and the preceding subspecies 1s the degree of curvature of the tepals. Of all the specimens at hand there is only one which is inter- mediate in this character: Cameron at Niagara Falls, Ont., July 1892. It is a single-flowered plant with spreading tepals as in ssp. canadense, but with filaments as in ssp. michiganense. As this specimen comes from an area where ssp. canadense is not known to occur we have referred it to ssp. michiganense. For an interpretation of var. Bourgaei Baker, see Stoker, Roy. Hort. Soc. Lily Year-Book 4: 26. 1935. The type of this variety was reputedly collected in Manitoba “ad ripas lacus Winnipeg" by Bourgeau, but we seriously doubt the accuracy of this geographical statement as no specimen of this very showy plant has ever turned up since or before from any- where in Manitoba. Furthermore, Bourgeau's localities are notoriously incorrect. Four phases of this subspecies have been recognized variously as species, subspecies, varieties or forms. These phases may represent no more than ecological forms or more vigorous or depauperate individuals, yet the vigor of some individuals may be susceptible of being segregated genetically. Since this subspecies is definitely of horticultural value, it seems desirable to recognize the two extreme forms as follows: LILIUM CANADENSE L. ssp. MICHIGANENSE Í. uniflorum (Farwell) stat. n. L. michiganense Farwell var. uniflorum Farwell, Bull. Torr. Bot. Club 42: 353-4. 1915; L. michiganense Farwell f. uniflorum (Farwell) Wherry, Bartonia 24: 8. 1947. Plant usually smaller throughout; flower solitary, terminal. Throughout the range of the subspecies. LILIUM CANADENSE L. ssp. MICHIGANENSE f. peramoenum (Farwell) stat. n. L. peramoenum Farwell, Bull. Torr. Bot. Club 42: 354. 1915; L. michiganense Farwell f. peramoenum (Farwell) Stoker, Roy. Hort. Soc. Lily Year-book 4: 27. 1935. The whole plant more vigorous; leaves more numerous, eight to sixteen to a whorl; flowers large and very nu- erous. Throughout the range of the subspecies. 1956] Boivin and Cody,—Lilium canadense Linnaeus 19 The typical phase (forma michiganense), including var. um- belliferum Farwell, was originally described as having from 5 to 8 flowers. Yellow-flowered mutants have been reported by Wherry, Rhodora 44: 454. 1942, but this phase has not been deseribed and is not clearly represented among the material at hand. LILIUM CANADENSE L. ssp. superbum (L.) stat. n. L. superbum L.: Sp. Pl. ed. 2: vol. 1: 434. 1762. Leaves smooth or sometimes minutely papillose along the margin and the veins underneath; inflorescence umbel- late to racemose; flowers orange-red; green zone at base of tube 10-15 mm. long; tepals strongly reflexed as in ssp. michiganense; petals sharply ridged on back along the midrib; stamens and pistil long-exserted; fila- ments recurved; anthers arcuate, linear, 11-25 mm. long. This subspecies has been reported to range from Georgia and Alabama north to New Hampshire, New York and perhaps Indiana. This range should be extended to include Maine as we have at hand an excellent sheet of this subspecies (H. N. Moldenke 18994, Penobscot Co., South Lincoln, July 27, 1947). Color mutants have been reported for this subspecies, but none has been named and none is represented among the material at hand. In a recently published chromosome study, R. N. Stewart (The morphology of somatic chromosomes in Lilium, Am. Jour. Bot. 34: 9-26. 1947) has, on the basis of their karyotypes, brought together within the same subgroup L. superbum, L. philadelphicum, L. catesbaei, L. carolinianum (= L. michauxic), L. michiganense and L. canadense. He remarks: “Five plants of L. canadense, two of L. canadense var. rubrum and two of L. canadense var. flavum all possessed identical karyotypes and all were indistinguishable from L. michiganense." The identity of the karyotypes confirms the opinion previously expressed by Wherry that these two taxa are very closely related indeed. Another taxon of the same subgroup, L. superbum, is also closely related to L. canadense. A fourth taxon, L. michauxii (= L. carolinianum), clearly belongs with the L. canadense group, but the lack of adequate herbarium material does not allow us to express a considered opinion as to its relative taxonomic status. It is, however, noteworthy that it was at one time reduced to varietal rank as L. superbum L. var. carolinianum 20 Rhodora [Vor. 58 (Michx.) Chapman. The last two taxa, L. philadelphicum and L. catesbaei, have erect flowers, long-unguiculate tepals, ete. and are not particularly closely related to the L. canadense group. A key to the major taxa described above has been published by E. T. Wherry, A Key to the Eastern North American Lilies, Bartonia 24: 5-8. 1947. The present study is based primarily on the specimens pre- served in the Herbarium of the Botany and Plant Pathology Division, Science Service, Department of Agriculture, Ottawa (DAO), supplemented by those preserved in the Herbarium of the National Museum of Canada, Ottawa (CAN). NUTTALL'S QUARREL WITH PURSH JEANNETTE E. GRAUSTEIN UNTIL 1936 there was no certainty about Thomas Nuttall’s activities or location in 1812 and the years immediately following. Although we are now far better informed, little attention seems to be paid to the light which has been shed on this period. Thinking has not been brought into line with the facts that have been established.! Of first importance in Nuttall's situation was his contract with Professor Benjamin Smith Barton, signed on April 7, 1810, before starting on his long expedition through the northwestern Territories. By the contract his journals and observations were the exclusive property of Barton who financed the journey; he was, however, permitted to retain a part of all the specimens which he collected but must not dispose of them without Barton's consent less “they might otherwise fall into the hands of persons who would use them to my disadvantage." “Should I ever publish the journal, I hereby promise and bind myself to make a publie acknowledgement, that the journey was performed by you, and to give you full credit for what services you may have rendered to me." When Nuttall arrived at New Orleans in December, 1811, ! F. W. PENNELL, Travels and Scientific Collections of Thomas Nuttall,” Bartonia 18 (1936), 1—51. Jeannette E. Graustein, ‘‘Nuttall’s Travels into the Old Northwest," Chronica Botanica 14 (1950/51), 1-88. 1956] Graustein,—Nuttall’s Quarrel with Pursh 21 on his return from the upper Missouri, he found the COLUMBIA about ready to sail for Liverpool, the familiar ground of his apprenticeship and his home port which he had left almost four years before. Added to the natural desire of a young man (approaching his twenty-sixth birthday) to visit his family after so long a separation, was the serious threat of war between the United States and England which might postpone such a visit indefinitely. Before embarking for Liverpool he sent to Barton his share of dried plants and seeds with memoranda. There is evidence that in the notes he indicated his intention of naming the most flamboyant plant of his collection (which he believed belonged in an undescribed genus) in honor of Barton. When Nuttall reached England early in 1812 he had no intention of publishing anything but Bartonia; not only was he legally restrained but he knew that to his patron any further publication would be an unforgivable sin. In 1810 Barton described Nuttall as distinguished by “innocence of character," and that was his reputation with the Astorians. He was so lacking in guile that his one treasured genus was snatched from him by the unscrupulous Pursh. That his quarrel with Pursh was strictly limited to the theft of the genus Fartonia has not been widely understood. Nuttall met Pursh in London—doubtless at the Banksian Herbarium—and, since Barton had been the patron of both, eagerly showed him a specimen of the proposed Bartonia and the notes concerning it which he was planning to publish. Shortly thereafter, in August 1812, Bartonia decapetala appeared in the Botanical Magazine under Pursh’s name although the accom- panying colored plate was done from Nuttall’s specimen. This was the grievance Nuttall held against Pursh—that he filched from him the only plant that he was free to publish from among the abundant and interesting specimens collected during two arduous and hazardous years. It was a cruel disappointment: it would have been Nuttall’s first publication of any kind; it might have modified the disapproval of his somewhat estranged uncle whose support could facilitate his chosen work; it deprived him of the small unit of recognition in the botanical world for which he had hoped, leaving him with experience as his harvest from two years of labor. 22 Rhodora [Vor. 58 Nuttall evidently drew a line between scientific botany and florieulture and between scientific collections and horticultural so far as his contract was concerned for he sold to Fraser seeds and living plants which he had laboriously transferred to England. He needed any funds they might bring. There is no doubt that he drew up the Fraser Catalogue.2 It is equally certain that he did not consider it a valid publication of the species involved. That he intended to stand by his contract with Barton is indicated by the fact that he made no attempt to publish any of his new species in the Botanical Magazine—a device by which he could have anticipated a few items at least of Pursh’s Flora— and remained aloof while others became authors of his plants in that publication. When Pursh’s Flora Americae Septentrionalis appeared in mid-December 1813, Nuttall was disturbed by the treatment of many of the species, especially the new plants of the West. He found the descriptions scant, habitats lacking, range erroneous, and in some cases he disagreed with the classifications. These deficiencies were to an appreciable extent the result of Pursh's limited field knowledge for he had not been south of Virginia nor west of the Appalachians. It became Nuttall's aim to write a new edition of Pursh’s Flora in order to correct the errors. Although Shinners calls him “secretive” he announced this intention to William Baldwin in the autumn of 1815 not long after his return from England, and there can be no doubt that he told Zaccheus Collins and other friends of his plans. The outcome of this urgent wish was Genera of North American Plants and a Catalogue of the Species to the Year 1817, a work which grew from concern over the inadequacies of Pursh’s Flora and which naturally, therefore, comments freely on Pursh’s errors. This perhaps is the basis for Shinners’ opinion that Nuttall “was resentful of Pursh's having anticipated him in publishing.” Since Nuttall had known from the first that he could not publish the plants collected under Barton’s aegis he had nothing personally at stake in the matter.—501 W. 113, NEW YORK CITY. ? No convincing evidence to the contrary has been furnished by Lloyd H. Shinners in questioning Nuttall's authorship in ‘‘Non-Validity of Nuttallian Names in Fraser's Catalogue." RHODORA 57: 290—293, 1955. 1956] Cronquist et al.,—Nuttall’s Names in Fraser’s 23 VALIDITY OF NUTTALL’S NAMES IN FRASER’S CATALOGUE ARTHUR CRONQUIST, Davin D. Kreck AND BASSETT MAGUIRE Dr. Lloyd Shinners has maintained in a recent issue of Rhodora (57: 290-295. 1955) that the names usually attributed to Nut- tall in Fraser’s Catalogue, even those accompanied by descrip- tions, were not validly published. Fraser’s Catalogue, an English nursery-firm's pamphlet that appeared in 1813, was published without formal indication of an author, but it is universally acknowledged that many or all of the new names contained in it were those of Thomas Nuttall. Many of these names were later republished by Nuttall with full descriptions, some- times with a reference to Fraser's Catalogue, sometimes without. Dr. Shinners shows that botanists have not been entirely consistent in their approach to Fraser's Catalogue, with the result that its rejection would displace some well known names, even while preserving others. Penstemon grandiflorus Nutt., which has regularly been accepted in floras and monographie work dating back at least to the fifth (1829) edition of Amos Eaton's Manual of Botany, would for example be replaced by the unfamiliar P. bradburyi Pursh if Fraser's Catalogue is ban- ished to a nomenclatural limbo. We believe that the interests of nomenclatural stability would be better served by the ad- mission of Fraser’s Catalogue as a proper publication than by its rejection, but we do not rely on that argument to support our position. Dr. Shinners argues that Nuttall was not himself the author of Fraser’s Catalogue, and this may perhaps be true in a strictly literal sense, although it is obvious that the descriptions and comments accompanying some of the names must be Nuttall’s own. Dr. Shinners further states that since the names were “anonymous as published,” and “not avowedly accepted by any author,” they are not validly published. Presumably he is relying here on Article 43 of the International Code of Botanical Nomenclature (Utrecht, 1952), which says that “A name (1) which is not accepted by the author who published it . . . is not validly published." This rule, however, as shown both by the examples and by the published discussion (Proc. 6th Int. 24 Rhodora [Vor. 58 Bot. Congr. 1: 364—366. 1936. Leiden) which preceded its orig- inal adoption, clearly pertains only to the intent of the author as expressed in the published work itself. If a name is formally and intentionally used, the pertinent part of Article 43 1s com- plied with. It matters not that the author may change his mind, or never again use the name, so long as he has accepted it in the published work. Dr. Shinners’ contention that the names in Fraser's Catalogue were really anonymous, rather than being properly attributable to Nuttall, is immaterial. Anonymity of the author is no bar, under the Rules, to validity of publication of a name. Dr. Shinners suggests that his attempt to disqualify anonymous names in Fraser’s Catalogue is merely an application of a principle propounded by Rousseau (Taxon 4: 40-42. 1955) in the latter's attempt to dispose of the anonymous generic name Americus. We do not wish to comment on the status of Americus at this time, but we would point out that the question raised by Rous- seau as to the serious intent of the author of Americus does not apply to the names used in Fraser’s Catalogue. Dr. Shinners points out that some of Nuttall’s names published in Fraser’s Catalogue were later taken up by him, while others were not. From this situation he concludes that “His in- consistent treatment gives us sufficient legal grounds, if anonym- ity be not enough, to disregard all names in the Catalogue.” We disagree. The names in Fraser's Catalogue were accepted in that published work by whoever wrote it, and the subsequent action of Nuttall or any other possible author has no bearing on the validity of their publication. Without at this time trying to determine what constitutes a “description” under Article 42 of the Rules, we believe that those new names which appeared in Fraser’s Catalogue, ac- companied by descriptions, are validly published.— THE NEW YORK BOTANICAL GARDEN. N CN N Plate 1: Rhodora rpa, Ilex glabra f. leucoca 1956] Woods,—1lex glabra forma leucocarpa 25 ILEX GLABRA FORMA LEUCOCARPA: A WHITE-FRUITED HOLLY FRANK W. Woops! In November 1954, a white-fruited specimen of Jlex glabra (L.) Gray was observed about 4 miles south of Marianna, Florida. The characterization of this hitherto undescribed color form of the species follows: Ilex glabra (L.) Gray f. leucocarpa F. W. Woods, f. nov. A forma typica drupis albis, non nigris, differt. As in the typieal form but with white rather than black drupes. Both the drupe and the persistent calyx are entirely devoid of pigment. Albinism also extends to the leaves, which are a distinctly lighter green color than immediately adjacent and normally pigmented specimens of I. glabra. The specimen exhibits the typical growth form of Ilex glabra, being many-stemmed as a result of sprouting from rhizomes after being burned. Sprouts from rhizomes as far as 0.5 meter from the main clump also bear white fruits. There appears to be a single clone, covering an area of about 0.2 square meter. A more detailed deseription of this new form follows: Leaf blade obovate, glabrous, lustrous, coriaceous, 1-2 cm. wide, 2-3.5 em. long, margin remotely serrate at the distal end but otherwise entire; petiole 4-6 mm. long, round, glabrous. Fruit white, borne singly, 5-6 mm. long, 6-8 mm. wide; nutlets smooth; peduncle 4-6 mm. long; calyx glabrous, white. Twigs light green, angular toward the tip. A shrub 1.5 meters tall, similar to and growing in association with other specimens of dark- fruited Ilex glabra. Foliage is somewhat lighter than the neighboring specimens. The habitat of I. glabra f. leucocarpa is identical to that of the typical form. Soil is moist and sandy. Principal associates are the typical form of J. glabra, Baccharis, Rubus, Myrica, Pinus serotina, and P. palustris. Edge of woods about 4 miles south of Marianna, Jackson County, Florida, Jan. 20, 1955, Frank W. Woods, C.E.F. 2, holotype deposited in U. S. National Herbarium of U. S. National Museum, Washington, D. C. Isotypes have been presented to herbaria of U. S. Forest Service ! Stationed at East Gulfcoast Research Center, Marianna, Florida. 26 Rhodora [Vor. 58 at Washington, D. C. and New Orleans, La., Gray Herbarium of Harvard University, New York Botanical Garden, Missouri Botanical Garden, Chicago Natural History Museum, University of Florida, Florida State University, University of Georgia, University of Tennessee, and Univer- sity of Alabama.—SOUTHERN FOREST EXPERIMENT STATION, FOREST SER- VICE, U. 8. DEPT. OF AGRICULTURE. VIABILITY OF SEED OF THE BLACK LocusT.—In the early spring of 1930, on property now owned by Orland H. Soule, of Schoolcraft, Michigan, a mulberry hedge was set. In prepara- tion for the planting a trench about eight inches deep was made. This brought to the surface some of the deeper soil. Later in the same season along a five-rod length of this planting 57 seedlings of black locust, Robinia Pseudo-Acacia, two and three inches high were counted at one time.! I learned that locust trees that had been killed by borers were cut down in 1867 on property belonging to Dr. Nathan Thomas. The Thomas family kept diaries and I had first-hand the in- formation about the time and the condition of the trees when cut. The trees had been on the east side of a four-rod street; the mulberry hedge, set in 1930, was directly opposite on the west side. I have been familiar with these two homes for the past 60 years and no trees of this species have been there during that period. At the time the locust trees were removed wooden sidewalks were used and yards were enclosed by fences. The locust legumes naturally lodged against the fence and renewal of parts of side- walk and posts for the fence placed some of the seeds where germination would cease. Locust seed is very hard and sprouts only under the most favorable conditions. Some years ago to get locust seedlings for future fenceposts the writer soaked the seed for several days. Each morning it was stirred in boiling water. The mulberry hedge mentioned proved a nuisance and was cut in 1954. If it had been pulled out some of the deeper soil would have been brought to the surface and a favorable op- portunity afforded to discover whether any seeds were still viable after another 25 years. The only remaining project 1 HANES, CLARENCE R. AND FLORENCE N. Flora of Kalamazoo County, Michigan. 167: 1947. 1956] Moore,—Silene virginica in Gulf States 27 was to exchange soil. Mrs. Hanes and I did this May 7, 1955. The top four inches of soil, between sidewalk and the roots of the mulberries, were taken out at two separate spots, each one by two feet in area. Then two bushels of the deeper soil were removed, one bushel of which was taken to our home and the other returned to the place from which it had been dug, only it was topside now. At our home we placed the soil in a plot two feet square and about four inches deep. The plot was well watered before and after the soil had been deposited. We understood the chance for success in picking up soil with viable seeds was almost nil. Would the seed be present in the small amounts of soil under observation and 1f there were any, would they still be viable? We were pleasantly surprised for we have on July 12, 1955 in our home plot two vigorous locust seedlings, the one five inches, the other three inches high. The latest leaves of these have nine and seven leaflets respectively. 'The home plot was well watered; the one along the hedge on the Soule's property had only the natural rainfall, which was scarce in early June but much more abundant later. At the time this paper was written, July 12, 1955. two small seedlings large enough, however, to show their identity had appeared. The fact that four locust seedlings have grown in soil from the same location 25 years after the 57 that were discovered in 1930 appears to substantiate the conclusion that these seeds, dormant since the close of the Civil War, came from the trees cut in 1867. Some black locust seeds therefore stay viable for at least 88 years. We cannot say how long before 1867 the trees ceased to bear fruit.—CLARENCE R. HANES, SCHOOLCRAFT, MICHIGAN. SILENE VIRGINICA IN THE GULF STATES.— he occurrence of Silene virginica L. in north Louisiana was first reported to me by Mr. Roy Morgan, a professional forester working in this area. In addition to his collection from Union Parish, I have collected it along the south side of D'Arbonne Bayou basin in northeast Lincoln Parish. Hiteheock and Maguire! show the species as occurring in extreme northeast Arkansas only, and not in Louisiana or in 1 Hircucock, C. Leo AND Bassett MAGUIRE. A revision of the North American species of Silene. Univ. Washington Publ. Bot. 13: 1-71. 1947. 28 Rhodora [Vor. 58 the states immediately east of Louisiana. The only previous Louisiana reports are those of Riddell? and the statement of Miss Dormon “In rich soil from Georgia to Louisiana.” In a recent letter, Miss Dormon stated that she and her sister col- lected the species “in the vicinity of Shreveport.” The Union and Lincoln parish stations are about 150 airline miles from the nearest reported Arkansas collection. The col- ; 7— © RX N = Š J N SE il j ERU my N NS \\\ N Fig. 1. Distribution of Silene virginica L. in the Gulf States. lection from the Natchitoches area reported by Mr. George Ware of Northwestern State College (Fox, April 13, 1939) is about 135 miles south of the nearest known Arkansas collection, and about 70 miles from the Lincoln parish station. The report of Miss Dormon fills in some of the range gap that was previously apparent. Further interest in the distribution of the species led to a request to southern botanists* for distribution data for the gulf states as shown by their herbaria. The accompanying map is based on the literature cited and on the information furnished by coóperating botanists. ? RIDDELL, J. L. Catalogus florae ludovicianae. New Orleans Med. & Surg. Journ. 8: 743-764. 1852. 3 DonMoN, CanoLiNE. Wild Flowers of Louisiana. New York. 1934. 1956] Moore,—Silene virginica in the Gulf States 29 The county records reported to me are summarized together with their source as follows: Arkansas: Benton, Carroll, Clay, Garland, Hempstead, Little River, Marion, Newton, Polk, Searcy, Washington (Iltis); Fulton, Hot Springs, Independence (Goodman); Sevier, Stone (Tharp); Pope (Shinners); Baxter (Schmitt). Louisiana: Lincoln, Union (Moore); Natchitoches (Ware); Caddo (Dormon). Mississippi: Lafayette, Lee (Schmitt); Okitb- beha (Ray); Grenada, Kemper, Tippah, Union (Lowe); Alabama: Lee, Marshall (Davis); Calhoun (Sharp); Randolph, Tuscaloosa (Shin- ners); Clark, Cullman, Montgomery (Mohr)5 Georgia: Bartow, Burke, Clark, Cobb, Dade, DeKalb, Floyd, Forsythe, Fulton, Gwinnett, Heard, Lumpkin, Rabun, Walker, Whitfield (Duncan); Dawson (Shinners); Meriwether (Tharp).—JoHN ApAM MOORE, DEPARTMENT OF BOTANY, LOUISIANA POLYTECHNIC INSTITUTE, RUSTON, LOUISIANA. 4 Botanists cooperating were: Donald E. Davis, Alabama Polytechnic Institute; Wilbur H. Duncan, The University of Georgia, Miss Caroline Dormon, Saline, La, George J. Goodman, University of Oklahoma, Hugh H. Iltis, University of Arkansas; James D. Ray, Jr., Mississippi State College; John A. Schmitt Jr., University of Mississippi; A. J. Sharp, University of Tennessee; Lloyd H. Shinners, Southern Methodist University; B. C. Tharp, University of Texas; George Ware, Northwestern State College. 5 Lowe, E. N. Plants of Mississippi. Miss. Geol. Survey Bull. 17. 1921. 6 Mour, CHanLEs, Plant life of Alabama. Contr, U. S. Nat. Herb. 6. 1901. Volume 57, no. 684, including pages 325-851, was issued 11 January, 1956. TSF ERCNCE LIDNARY npn -7 1056 73 Dodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD CARROLL EMORY WOOD, JR. Vol. 58 February, 1956 No. 686 CONTENTS: Notes on Triphora trianthophora in Ontario. C. Harold Zaviiz and Lulu O. Gaiser... sl eet ee ER 31 Froelichia gracilis in Maryland. S. F. Blake................... 35 Shortia galacifolia in Gray's Manual Range. Dorothy L. Crandall 38 Notes on Utricularia biflora and U. fibrosa. Leonard J. Uttal.... 41 Yonkers Wool Mill Plant Records. Joseph Monachino.......... 43 Correa da Serra and American Botany (Review). Joseph Ewan.. 45 Nomenclatural Change in the Grass Genus Echinochloa. David E. FECI RMN eg esi «os sda hin tie bot cas dene oe wie ee SS 48 Two Plants New to the Illinois Flora. Robert A. Evers.......... 49 Juniperus communis var. saxatilis from Mount Washington. I 0c c S TURNER EE oak 50 The New England Botanical Club, Inc. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $4.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 40 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $4.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scien- tific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of publica- tion. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. me ice — CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientifie libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U. S. A. Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 February, 1956 No. 686 NOTES ON TRIPHORA TRIANTHOPHORA IN ONTARIO C. HAROLD Zavitz AND LULU O. GAISER EARLY in the summer oÍ 1950, while the senior author was inspecting a woodlot in south-eastern Essex County, his atten- tion was attracted by a small plant (approximately four inches tall) which was unlike any plant observed by him previously. The plant had a fragile semi-transparent appearance and was of a light green or yellowish green color. Stem leaves were few in number, and small, being less than one-half inch in length, oval in outline, rounded to a triangular tip, and with a sheathing or clasping base. Subsequent visits were made to the area and other similar plants, about a dozen in number, were found in various parts of the woodlot. On August 21st the first flowers were observed and it was then recognized as a member of the ORCHIDACEAE. By reference to 'Gray's New Manual of Botany (Seventh Edition)", the plant was determined as Pogonia trianthophora (Sw.) BSP., although it was not recorded as from Ontario. Britton and Brown's Illustrated Flora further substantiated the original identification. This plant is given in ‘Gray’s Manual of Botany (Eighth Edition)" and also in “Native Orchids of North America" by Correll as Triphora trianthophora (Sw. Rydb. Neither of these authoritative works records it from Ontario, but Correll does say “from Canada (old record) through New England," etc. The flowers arose from the axils of the leaves of the upper stem, the uppermost of the leaves being reduced in size so as to appear almost bract-like. The number of flowers on a plant varied from one to four. 32 Rhodora (Vor. 58 The flowers resembled, in general form, those of Pogonia ophioglossoides (L.) Ker, but were much smaller in size being just over half an inch long. The color was milk-white to trans- lucent. Under a low power lens, the papillose protuberances on the lip appeared almost crystalline. The crests of the lip showed a slight green coloration and a very slight suggestion of pink was observed on close inspections of some other portions. On specimens examined, one or two tubers were attached to the end of the main root. These tubers were ovoid with the distal ends somewhat pointed and the proximal ends quite blunt. One plant obtained on August 21st was preserved. On Septem- ber 12th, the same author again visited the area, and a specimen with a well-developed seed capsule was obtained. This specimen and the one obtained previously on August 21st are preserved in the author’s herbarium. In 1951, the area was inspected once, but no plants were observed. This peculiarity seems quite characteristic of this group of orchids, and may partially account for its lack of detec- tion in an area so well populated. In 1952, I was in England and therefore unable to check the locality. The next inspection of the area was on the 14th of August, 1953, when a specimen in flower was obtained. This was delivered in a fresh state that day to Dr. L. O. Gaiser who was summering at Crediton, Ontario, and she confirmed the original determination. The specimen was later verified by Mr. Charles Schweinfurth, of the Oakes Ames Orchid Herbarium and it has been placed in the Gray Herbarium, Harvard University. While obtaining this specimen a number of tubers were dis- covered which were not closely attached to foliaceous plants, but appeared to be evidence of the subterranean process of vegetative propagation which obtains in this and some related species. These tubers were usually in thick beds of hardwood leaves and humus, and somewhat remote from the mineral soil. Both authors visited the area in August, 1954, and saw at least twelve specimens in two neighbouring patches. Two inspections of the area in 1955 failed to reveal any plants. The plants occurred in a sandy soil rich in humus, in medium to heavy shade of the native hardwood forest. The chief species of trees in the area in which this plant was found were 1956] Zavitz and Gaiser,—Triphora trianthophora 33 beech, hard maple, red oak, white elm, hickory and red ash. The Carolinian flora was represented there by tulip, chestnut, and sassafras, and in the general locality by flowering dogwood, sour gum, hop tree, papaw, and red mulberry. Some authors class this orchid as a saprophyte. However, the amount of green in the plants observed would indicate that it is not entirely saprophytic and that under certain circum- stances, at least, this species has the ability to develop chlorophyll and thus manufacture some of its own plant food. The earliest reference to Triphora trianthophora (Sw.) Rydb. as having come from Canada, is in the Botanical Register (1825) t. 908. There, after the description of Pogonia pendula, John Lindley states; “Roots of this curious little plant were collected in Canada, by Mr. David Douglas, for the Horticultural Society, and flowered in an open border in the garden at Chiswick in August, 1824." A later paragraph includes the statement “Mr. Nuttall says it is parasitic round the roots of beech trees from New York to Kentucky, and that he has found it also in Canada." W. J. Hooker, in Flora Boreali-Americana 11, p. 202 (1840), merely lists the plant referring to the above reference, but adds the name of Goldie to the two previously given after “Hab. Canada." Thus concerning the same plant, J. Macoun, in the Catalogue of Canadian Plants IV p. 11 (1888), writes: “Damp woods, Canada (Goldie fide Hooker)." He adds: “We have never seen à Canadian specimen of this species. It should be looked for in south-western Ontario." While reports of this plant from Canada have been dropped from modern manuals, such as Gray's Manual (Seventh and Eighth Editions), Britton and Brown, and Gleason 1952, Correll (1950), referring to the “old record from Canada," gives credit to David Douglas for its introduction to England at Chiswick in 1824, as referred to by Lindley. In the Journal kept by David Douglas during his travels in North America 1823-27, Wm. Wesley & Son, London (1914), his journey can be followed: On October 1 (1823) he crossed the Niagara River at Queenston, thus passing into the U. S. A., for “Lewiston on my way to Lockport.” Then on page 17 begins the report: “Lockport, Friday, 3rd.—In company with 34 Rhodora [Vor. 58 Mr. Thomas who kindly offered me the ‘use of his person,’ as he said, I visited woods north of Lockport which were almost all beech; we found two specimens of Corallorrhiza—one like innata of Britain, the other a small one. I secured plants of them”; continuing on page 18 we find: “the soil was dead leaves, very dry. Along with them grew Triphora,! also a Cypripedium in partly shaded parts, soil dry sandy peat; V?ola sp. leaves round, large, flowers small white striated.” There is an appendix “Plants introduced by David Douglas during the years 1826-34" and on p. 334 is listed “Pogonia pendula, Lindl. in Bot. Reg. ¢ 908-North America." With these as the only references to Pogonia or Triphora, it would seem that if the belief that the species was collected in Canada is based on Douglas’ collection, it is ill-founded. A search has since been made for specimens or a report of a collection of this plant in Canada that would confirm the “old record.” Negative reports to inquiries for Canadian specimens have come from the curators of the following herbaria: National Museum and Dominion Experimental Farm in Ottawa, Botanical Garden in Montreal, the Universities in Toronto, Kingston, London, Hamilton and the Agricultural College at Guelph, Ontario. There is also no former specimen from Canada in the Gray Herbarium. When the inquiry was sent to the first of these herbaria, Mr. A. E. Porsild kindly offered to look for specimens of Triphora trianthophora at Kew and the British Museum when he visited those herbaria during the year 1954-55. We are deeply in- debted to him for his help. He has reported the following: At Kew, under the above name are seven sheets, each with 1-3 separate collections. One sheet, stamped ‘‘Herbarium Hookeranium 1867,” contains three col- lections, two of which are labelled T. pendula with given localities in the U. S. A. and a third labelled “Arethusa pendula," written on the original ticket, and in another hand, “Canada Gouldie.”’ On a second sheet are collections labelled T. pendula likewise “Herb. Hook. 1867,” each bearing a collector’s name but no other data. That in the lower right corner, in Hooker’s hand, bears the name “Douglas.” The rest are all specimens from the U.S. A. "1 Pogonia Benth & Hook" 1956] Blake,—Froelichia gracilis in Maryland 35 Mr. Porsild concluded, “Thus at Kew there is only one speci- men definitely labelled Canada. Possibly ‘Gouldie’ is meant for ‘Goldie’ ." He further reported that no Canadian specimen could be found at the British Museum. While there is neither evidence on the sheet bearing the Douglas specimen nor any record in his own Journal that he collected this plant in Canada, both Hooker’s Herbarium and his Flora confirm Goldie’s collection of it here. FROELICHIA GRACILIS IN MARYLAND S. F. BLAKE Froelichia gracilis (Hook.) Moq., of the family Amaranthaceae, is regarded as indigenous in the region west of the Mississippi River from Nebraska or Iowa to Colorado, south to Texas, Arizona and Chihuahua. Its extension of range eastward in recent years even to several of the coastal states indicates the possession of a roving disposition which might complicate the problem of determining precisely its original native habitat. Standley, in 1916, did not know it east of the Mississippi, for he assigned a range from Iowa to Colorado, southward to Arkan- sas, Arizona, and Chihuahua (North Amer. Flora 21: 127). The seventh edition of Gray’s Manual (1908) and both editions of Britton and Brown’s Illustrated Flora (the second in 1913) had given a similar but more restricted range. In the 8th edition of Gray’s Manual (1950) Fernald assigned essentially the same range as Standley had done, but added: ‘‘Adventive eastward to New York, New Jersey, and Virginia.”’ The first report from east of the Mississippi appears to have been that of H. D. House in 1924 (N. Y. State Mus. Bull. 254: 303) recording it from railroad tracks at Despatch, Monroe County, New York, 1920, where it was found by D. M. White. Apparently the species has been collected only once again in New York, on waste land near railroad tracks at Cold Spring, Philipstown, Putnam Co., 22 Aug. 1953, by K. L. Brooks, S. J. Smith, and J. J. Wurdack. In 1929 H. C. Benke (RHODORA 31: 146) reported it from Cairo, Illinois, where he had collected it in 1928. C. C. Deam in 1940 (Fl. Indiana 431. map 887) 36 Rhodora [Vor. 58 listed it from four counties in Indiana but indicated its presence in five on his map. He had first found it in 1930 along a railroad about four miles south of Vincennes, in almost pure sand, and three years later found that it had spread a quarter of a mile along the track. He predicted that it would “probably become a weed in the sandy area of this part of the country" (very likely referring to the sandy region around Lake Michigan in Indiana and Illinois), but there are only four additional counties recorded in the 14 annual supplements to his Flora of Indiana that have been published to date. The eastward spread of the species has taken place almost entirely along railroad tracks. Of 24 specimens in the her- barium of the University of Illinois coming from 19 different counties in all parts of that state, the data for which have kindly been supplied by Dr. G. Neville Jones, 20 were collected along railroads, 2 by roadsides, and for 2 the habitat was not specified, as is the case also with Benke's original collection in the Chicago Natural History Museum herbarium. Of 16 specimens from other states, cited below, all but the 2 from New Jersey were found along railroads; the New Jersey specimens were collected along a path and in sandy fields, respectively. All four of the Indiana localities given in Deam's Flora were along railroads. The occurrences of this species in the eastern states as repre- sented by specimens in the Gray Herbarium (GH), Department of Botany of the University of Illinois (ILL), New York Botanical Garden (NY), New York State Museum (NYS), Ohio State University (OS), Philadelphia Academy of Natural Sciences (PH), and United States National Herbarium (US) are sum- marized here. Dr. R. €. Rollins (GH), Dr. G. N. Jones (ILL), Dr. Arthur Cronquist (NY), Stanley J. Smith (NYS), Miss Clara Weishaupt (OS), and Dr. C. Earle Smith (PH) have kindly supplied the data for their respective institutions. In- dividual eolleetions have been cited from only a single her- barium, but some of them are also represented in other institu- tions. New York: Monroe Co., 1920 (au); Putnam Co., 1953 (Nys). New Jersey: Camden Co. 1944 (pH), 1946 (GH). PENNSYLVANIA: Berks Co., 1945 (pH); Delaware Co., 1942 (GH); Lackawanna Co., 1926 (pH); Mont- gomery Co., 1944 (pH); Northampton Co., 1946 (rH); Schuylkill Co., 1944 (GH). ViRGiNIA: Dinwiddie Co., 1939 (GH); Henrico Co., 1940 1956] Blake,—Froelichia gracilis in Maryland 37 (GH); Sussex Co., 1939 (GH). Onto: Scioto Co., 1951 (os). INDIANA: Knox Co., 1941 (ny); La Porte Co., 1942 (ny); reported from 7 other counties. ILuINos: First record, 1928, Alexander Co; represented (ILL) by specimens from this and 18 other counties in all sections of the state.— Not reported from West Virginia or Kentucky. On 29 Sept. 1954 and again a week later I had occasion to investigate an infestation of this plant on the farm of J. Thomas Simpson in Odenton, Anne Arundel County, Maryland, which apparently provides the first record for the species in that state. Thousands of specimens were growing in sandy soil on the edge of a cornfield and among the low weeds (Diodia, Digitaria san- guinalis, etc.) on the uncultivated border of the field, mostly in an area about 150 by 50 feet, in places very thickly, in others sparsely, and there were scattered plants in other parts of the field and in an adjoining pasture. Mrs. Simpson informed me that the plant had first appeared there about 5 or 6 years ago, but had not become abundant until 1953. Their cattle will not eat it. The specimens growing in loose open sand were very well developed, up to about 60 cm. high with several assurgent basal branches as long as or longer than the stem; those growing among dense weeds were mostly simple and shorter. The nearest railroad to Mr. Simpson's place is a spur track of the Pennsylvania RR. about a mile away, running in to the Bowie Race Track (in Bowie, Prince Georges County) and used in season for the transport of horses and their fodder and bedding. Because of the plant's close association with railroads, it seemed reasonable to suspect that this might have been the source of introduction to Mr. Simpson's farm, particularly since his daugh- ter, Miss Marion E. Simpson, informed me that her father had occasionally brought in straw from the race track. On 25 July 1955, when I visited the spot in company with Miss Simpson and Paul G. Russell, several score specimens were found on the railroad tracks between the stables and the Administration Building, where the horses and their supplies are unloaded. The weedy nature of the plant is emphasized by my own experience with it. I collected a number of specimens on my first visit to the farm, and must have unintentionally picked up à number of fruits on my person as well, for the next morning, in my home in Virginia, I found in my bed a number of rough 38 Rhodora [Vor. 58 little objects which felt like small pebbles but turned out to be the fruiting calyces of this plant. They must have been carried home attached to my stockings or trousers or in my shoes, but I am puzzled to explain their transfer to the bed. They are unpleasant to walk on barefooted, but not nearly as much so as the fruits of Tribulus or Cenchrus. However interesting the plant may be as an addition to the flora of Maryland, that state has no need of any more weeds, and Mr. Simpson is attempting to eradicate it by cultural treatment before it spreads further. The railroad tracks at Bowie are sprayed regularly each summer, but it is very improbable that this procedure would have much effect on the multitude of seeds that must have been scattered previously from the plants growing there.—PLANT INDUSTRY STATION, BELTSVILLE, MD. SHORTIA GALACIFOLIA IN GRAY’S MANUAL RANGE! DonoTHv L. CRANDALL Shortia galacifolia has been called by C. 8. Sargent (1888) “one of the rarest and most interesting plants of North America" and since the first pressed specimen of Shortia was rediscovered by Asa Gray among Michaux’s unknowns, search for this species has intrigued many botanists. In a recent article in RHopora, P. A. Davies (1955) lists seven counties in which Shortia has been found, one in Georgia, two in South Carolina, and four in North Carolina. No natural colonies have been reported from outside this rather restricted area and yet ecologically there are many similar habitats in other mountainous sections of the Southeast. Aware of its limited distribution, the author was somewhat surprised to discover a flourishing colony of Shortia along a creek in Amherst County, near Lynchburg, Virginia. On the relatively steep east-facing bank (altitude approxi- mately 650 feet) of this small creek, just a few feet above the water, was a patch of Shortia covering an area about three by six feet. On this date, April 9, 1955, only a few scattered blos- soms were evident, varying from white to pale pink. Most ! Contribution from the Botanical Laboratory, The University of Tennessee, N. Ser. No. 169. 1956] Crandall,—Shortia galacifolia 39 of the earlier flowers had withered without forming fruit. Photo- graphs were taken and specimens were collected. The latter are being deposited at the Gray Herbarium, the U. S. National Herbarium, and the herbaria of The University of Tennessee and Randolph-Macon Woman's College. Downstream a few feet away from the original group, a small clump of three plants was growing. A search upstream dis- closed three more plants about twenty feet away and another small group about four feet away. The stream was traced to its source, only a few hundred feet, but no more plants were seen. The author made two subsequent trips to the site, one in the company of Franklin Flint of Randolph-Macon Woman's College and the other with Mr. Flint and Ruskin Freer of Lynchburg College. About 300 feet downstream at a point just above where this creek joins a still larger creek that flows into the James River about a half mile away, another clump of Shortia ap- proximately two by three feet was located on a knoll over- hanging the creek. A search for several miles up the larger creek failed to disclose any additional stations. The first colony appeared to be well established and had spread both by stolons and by seeds. On bare and mossy soil and only seen when the overhanging leaves of SAortia were lifted, were dozens of seedlings of various sizes. In addition to these a few seedlings were observed several feet away from the parent plants, the seeds having washed down the steep bank. Bailey's Standard Cyclopedia of Horticulture (1933) states “It is difficult to procure seed as the flowering stem usually withers away before maturing, though shortia is readily propa- gated by division and runners." C. F. Jenkins (Arnoldia, 1947) adds “Apparently no plants were propagated in this way," referring to attempted propagation by seeds. On a visit to the area near the end of May a few flowers, perhaps a dozen or so in all, had formed capsules. The conspicuous canopy trees included Nyssa sylvatica, Fagus grandifolia and Liriodendron Tulipifera. In the under- story were Cornus florida, Hydrangea arborescens, Alnus serrulata and Kalmia latifolia. Intermingled with the colony were Poly- stichum acrostichoides, Dryopteris noveboracensis and Athyrium 40 Rhodora [VoL. 58 Filiz-femina. Important among the low flowering herbs were Asarum virginicum, Iris cristata and Medeola virginiana. Calax aphylla was abundant, the bank of the stream lower down 'arpeted with blankets of it. A search of the area disclosed no old gardens from which Shortia could have escaped and local inquiries have yielded nothing as to the origin of these plants. The creek is part of a small farm which has changed hands many times and until the present owner purchased the land about ten years ago, no one had taken any interest in improving the property. Any information as to the possible introduction of these plants would be appreciated. However this species may have first gained its foothold on this creek bank in Virginia, it is of interest that Shortia is flourishing and spreading both by seeds and stolons in an area within the range of Gray’s Manual and considerably north of its previously known distribution in the Carolina mountains.—RANDOLPH- MACON WOMAN’S COLLEGE, LYNCHBURG, VIRGINIA. LITERATURE CITED Battey, L. H. 1933. The Standard Cyclopedia of Horticulture. Vol. HI. Macmillan Co., New York. Davies, P. A. 1955. Distribution and Abundance of Shortia galaci- folia. Ruopona 57: 189-201. FERNALD, M. L. 1950. Gray’s Manual of Botany. (Sth ed.). Amer- ican Book Co., New York. Jenkins, C. F. 1942. Asa Gray and His Quest for Shortia galacifolia. Arnoldia 2: 13-28. SARGENT, C. S. 1888. The Story of Shortia. Garden and Forest. 1: 506-507. 1956] Uttal,—Utricularia biflora and U. fibrosa 4] NOTES ON UTRICULARIA BIFLORA AND U. FIBROSA LEONARD J. UTTAL Tue Bladderwort, Utricularia biflora Lam., grows in certain wet places in Coastal Plain pine barrens, from southeastern Massachusetts to Florida, Texas, and Oklahoma. Rare and local in the north, it nevertheless seems odd that but one station, herewith reported for the first time, seems to be known for this species in New Jersey, where likely habitats are so frequent. This station is on the Public Hunting and Fishing Grounds, at Archer Corner, Ocean County, New Jersey, where the writer collected plants in August, 1954. Less than a half-meter square, the colony grew in water just a few centimeters deep, the sub- mersed parts enmeshed in muck, amidst remnant eranberry and sphagnum of a receding bog, well invaded by upland flora, perhaps as a result of fishing management operations on the preserve. A colored photograph was made of the colony, later affixed to the voucher of this collection, deposited in the her- barium of the New York Botanical Garden. N Z š Pss E uunsogj[Htn«n€& — suvjdaJ4 ^ QO SNS le \ zx ESSE 5 1 J =\ED A Im i j fae S zx / nw ot 0 x Ie AN E, ye » .£ WZ : as = x = 2774 du ` ESEN 7 f e A C NU w s N C EN ess ` š Ni y 04012 N : —=—-— -Z UL | | upto = I NY SN ES | 4" + mn T--—- ---| ! p a NN “et «S i " y~-—-—.i SsOH ne —— 2 °! J > -— | P l H l 106 Rhodora IVoL. 58 Fig. 2. Calyx, pedicel, and stem pubescence of P. reptans var. reptans and P. reptans var. villosum: A. Pedicel and lower part of calyx of villosum; B. Pedicel and lower part of calyx of reptans; C. Stem of villosum, just above lowest cauline leaf (A, B, and C in optical section; 1 mm. line shows scale); D. Greatly magnified pedicel hairs of reptans and E, of villosum. All drawn from living material. are similar. Plants of villosum are so viscid that bits of humus and other foreign matter, and small insects are seen adhering to stems and pedicels. "These help to give herbarium specimens of villosum their characteristic aspect. An odor of skunk, similar to that of certain western species of Polemonium, is noticeable if living plants have been enclosed, as when collecting in plastic bags, and, at times, when handled in the open. In general, plants of villosum are lower than those of reptans, and the main stems diverge from one another at a wider angle. Basal leaves of villosum tend to have fewer leaflets than those of reptans. 'The ratio of length to width of leaflets differs in the two. Leaflets of reptans are more than twice as long as wide; those of villosum mostly less than twice as long as wide (Figs. 3, 4). Differences in leaves in the inflorescence may be 1956] Braun,—Variation in Polemonium reptans 107 less pronounced. Correlated with the broader and shorter leaflets of villosum, apical and basal angles are normally wider than those of reptans (Figs. 3, 4). Leaflets often are not “acute” as are those of reptans, the margins sometimes converging at angle of 105° or more, and often over 90° seldom less than 75°. Although the ranges of leaflet number and of length : width ratio overlap, populations of the two taxa are noticeably dif- ferent in leaf aspect. Flowers of villosum are smaller than those of reptans, 8-12 or rarely 13 mm. long, in contrast to 12-16 mm. long. The size differences of wide-open flowers (measured in the field) is more prominent: 11-17 mm. (av. 14 mm.) in diameter, as contrasted with 18-23 mm. (av. 19.5 mm.). Many pedicels of villosum are shorter than the calyx, both in flower and fruit. The calyx lobes are shorter and broader, not “lanceolate to lance-triangular” (Fernald, 1950) as are those of reptans, but more deltoid, often almost an equilateral triangle in shape. The calyx of villosum is glandular-villous, especially toward base, and its margin glandular-ciliate (Fig. 2A). In the field, reptans and villosum are so obviously different that they are readily distinguished. This has been emphasized more than once when the writer has been asked “what is the Polemonium in Adams County?" or told that one will not thrive under cultivation (as does reptans), or that “it is not worth growing, anyway." It is far less showy. To demonstrate differences in pure populations of reptans and villosum, pictorialized scatter diagrams of the type devised by Anderson (1949, 1952, 1953) have been used (Fig. 4). All parts of the plants of villosum are glandular-villous—pedi- cels and calyx, stems, and leaves. Such plants are represented on the pictorialized scatter diagram by solid dots, to contrast with the circles used for the essentially glabrous and non-glandular reptans (Fig. 4). Position of dots on this graph is based upon the ratio of length to width of median leaflets of cauline leaves, and upon pedicel-hair characters. For the length : width ratio, the median pair of leaflets (or leaflets just below the middle, if number of pairs was even) of the lowest good cauline leaf were used. Cauline leaves near to the base of the stem and resembling basal leaves were rejected. 108 Rhodora [Vor. 58 — NC Z < >. Fig. 3. Median leaflets (random sample) from basal leaves: A, of reptans, B, of villo- sum, and C, of a mongrel population. 1956] Braun,— Variation in Polemonium reptans 109 The length and greatest width of both leaflets were measured, and the ratios obtained from the two leaflets averaged. The pedicel score is based on length and character of hairs of the pedicel: 1. All hairs short and stubby, .05 mm. long; hairs short, but less blunt; or, occasional hairs to 2 mm. long scattered among short hairs. Many hairs .25+ mm., to some 3 mm. Many hairs 3+ mm. to some 4 mm. . Hairs 3, .5, and .7 mm., and also shorter hairs. . Many hairs .5 or .7 mm., some 8 or 1.0 mm., and also shorter hairs. ES S Hairs were measured by using an ocular scale ruled to half millimeters, using whole pedicels. Exact measurements on whole pedicels are not possible because hairs project at various angles. The use of smaller parts and higher magnification was impraeti- cal because of variation from place to place. Pedicels were scored in intermediate positions (as shown on the graph) in accordance with variations in length and dominance of particular lengths. Characters used on the glyphs are ones which cannot be measured or scored as closely as can the length : width ratio or length and nature of pedicel hairs. Angles were measured by superimposing on the form to be measured paper sectors differing from one another by 15°, i.e., sectors 30°, 45°, 60°, 75°, 90°, 105°. Apical and basal leaf-angles often do not express the immediate tip or base of the leaflet, but rather the angle at which the margins converge toward apex or base. As margins are often slightly convex, the exact angle could not be measured. Stem pubescence varies from base to inflorescence, and from internode to node. The three-score system used is: 1. Stems essentially glabrous, or with scattered hairs. 2. Hairs widely spaced on internodes, more closely placed near nodes or toward base of stem. 3. Hairs dense, or very dense, and long, some 1 mm. or sometimes 2 mm. long. A three-score system was also used for leaves: 1. Leaves glabrous except for a few hairs in rachis groove or on petiole margin toward base, and except for a very small area of short hairs in concavity of base of midnerve of leaflets. 2. Rachis thinly pubescent, more densely so toward base. 3. Rachis villous; both leaf surfaces pubescent. The complete separation on the diagram (Fig. 4) of glyphs representing flowering individuals of the reptans population from 110 Rhodora [Vor. 58 š. w y `< + wy wy ww w ] @=rillosum | w O= reptans l v * | s 4 Ë w ° % 4 Aly Wo 4 `w O | Ó |] 2] Oo 6 Ò | Ó o O 4 o 4 O | ° Ó Ô g / ò Ó o Ë ob ÒO ° nn T T T T nn u u u T 2lo T T T T alg T T T T sly T T T als Sepal Angle Apical Angle Basa/ Angle Stem Pubescence Leaf Pebescence O e 30° O6 &ov/ess OQ Z %/ess OS / oe / Ò è 252 0O.975- G @ 2o?» o% 2 or 2 Ó Ó soto O @ go vorer SO oshare “Oe 3 e 3 Fig. 4. Pictorialized scatter diagram of populations of villosum and reptans, based on characters of flowering plants collected in April. width of median leaflets of lowest good cauline leaf. as described in text. Horizontal axis: ratio of length to Vertical axis: pedicel hairs, scored Note complete separation of glyphs of the two populations. those of the villosum population is evident. Almost as complete separation was seen on a diagram representing summer basal leaves, on which the glyphs were placed in relation to the length: width ratio and number of leaflets. Pubescence of leaf, and apical and basal angles were scored as in Fig. 4. Two entities were illustrated by these diagrams. Some populations of reptans show slightly more variation than 1956] Braun,—Variation in Polemonium reptans 111 | Š 3: joo | by 2| wy ó [o T T T T T T T T T T T T T T T T pP: " | 2lo 2l5 3lo als Fic. 5. Pictorialized scatter diagram of mongrel population, based on characters of flowering plants collected in April. Axes and glyph appendages as in Fig. 4. the one illustrated by Fig. 4. Some herbarium specimens differ considerably from those of the population diagrammed, yet are obviously referable to reptans. What is the cause of this vari- ation? Discovery of a mongrel population in which some plants resembled reptans, some resembled villosum, afforded an oppor- tunity for analysis (Fig. 5). Open circles, with dot in center, are used in glyphs for this population, because such glands as may be present are inconspicuous and not evident under a 10 hand lens as are those of villosum. Although a few of the longer hairs are structurally similar to those of villosum, the gland appears to be non-functional and the conspicuous tip is absent. All glandular hairs deteriorate in summer, and it is not possible to distinguish functional and non-functional glands at this season. 112 Rhodora [Vor. 58 Comparison of Figs. 4 and 5 shows that the glyphs in Fig. 5 are not distributed like those of either reptans or villosum, that some are in the area of reptans glyphs, some in the area of villo- sum glyphs, and some in a position not occupied by either, but in the belt of low length : width ratio as well as in the position of puberulent rather than villous pedicels. Furthermore, there is a tendency for those of higher length : width ratio to be simpler —more like pure reptans—while those of lower length : width ratio tend to have more of the villosum characters. Such a tendency is indicative of hybridization. Although many plants resemble villosum in length : width ratio and in amount of pubes- cence, none has the characteristic glandular-villous inflorescence of that taxon. This character appears to be recessive, and to be controlled by different genes than is stem pubescence. Comparison of summer basal leaves of villosum and reptans populations with those of the mongrel population brings out similar although less marked differences. Number of leaflets is not as distinctive a character as is pedicel-hair character. Nevertheless, the tendency toward villosum-like plants in the area of villosum glyphs, and toward reptans-like plants in the area of reptans glyphs was present in scatter diagrams of the mongrel population. A high degree of sterility of some plants may be further evi- dence of hybridization. Failure of capsule development and accompanying excessive enlargement of calyx was noted in several specimens among the random samples. The *connection between hybridization and disturbed habitats" has been summarized by Anderson (1949). This mongrel popu- lation of Polemonium occupies an area where man has “hybrid- ized the habitat," an area on a wooded slope steepened many years ago when a road was built in the valley at its base. Typical reptans may have grown in this valley; and perhaps typical villosum occupied some of the ravines. The nearest known pure populations of the latter are in other valleys some four to six miles distant in an air-line and separated by high hills which interpose ecologic barriers. Typical reptans is much more dis- tant (Fig. 1). Anderson has stated “it is only where man or catastrophic natural forces have ‘hybridized the habitat’ that any appreciable number of segregates survives.” 1956] Braun,— Variation in Polemonium reptans 113 Hairiness of stem and leaf, and low length : width ratio are prominent in the mongrel population. Glandular-villous pedicels are almost absent. This is precisely what is seen when hundreds of herbarium specimens of P. reptans (inclusive) are examined. Some are pubescent, some have broader leaflets than others, some have the short, “squatty” form of villosum, but none has the elandular-villous pedicels. Hybridization, such as is clearly indicated by the mongrel population analyzed, may have taken place infrequently in the past. Some characters have been lost, some retained, in the en- suing back-crosses. The results of introgressive hybridization are evident over a wide geographic area, but pure or almost pure reptans populations prevail And large pure populations of villosum exist in a limited geographie area. Why have these not been swamped by the widespread reptans? This would seem likely if villosum had always been surrounded by reptans. The suggestion is here made that villosum is older than reptans, that the now restricted range of villoswm is a remnant of a former more extensive and perhaps more northern range; that at some time during the Pleistocene this range was decimated, leaving only those populations south of the immediate effects of ice sheets. The distribution of existing populations of villosum correlates well with the disjunct occurrences of a number of species generally interpreted as relic distribution. Biotype de- pletion is a usual accompaniment of restriction of range. The uniformity of populations of villosum and their failure to spread suggest such biotype depletion. The progenitors of villosum, as of other species of Polemonium, are still more ancient. In a chart “showing the postulated affini- ties between the recognized entities in Polemonium,” Davidson (1950) indicates that P. reptans is related to P. delicatum Rydb. and more distantly, to P. pulcherrimum Hook., both of which are western species. P. delicatwm of the Rocky Mountains of Colorado, Utah, Arizona, and New Mexico, is a glandular- pubescent species, which suggests closer relationship to villosum than to reptans. Another pair of species (or subspecies) displays a comparable east-west relationship—P. Van Bruntiae Britt. of circumscribed eastern range and P. occidentale Greene of the western mountains. Such relationships must be ancient. P. 114 Rhodora [Vor. 58 reptans, in the restricted sense, is a widespread species found almost throughout the Deciduous Forest (mostly north of the Coastal Plain), and northward into Minnesota, central Wisconsin, northern Vermont and New Hampshire. It is common in the glaciated territory. The ancestral stock, from which villosum and perhaps delicatum were early differentiated, probably gave rise to other variants, and later, perhaps in late Pleistocene, to the glabrous reptans. Other lines of development in Polemontum include both glabrous and pubescent forms. Rapid spread of the glabrous reptans in late Pleistocene and post-Pleistocene time brought it into proximity with some of the relie populations of villosum, resulting in occasional hybridization, wherever the two came into actual contact. Thus reptans acquired some of its variability. Analysis of the scatter diagram (Fig. 4) supports this contention. If the ten individuals most like each other and the ten least like each other on the basis of the measurements used on the hori- zontal and vertical axes of the graph be selected for determining average ray value for the other five characters (zero for no ray, one for a short ray, two for a long ray), it is found that perfect villosum would have a score of ten, and perfect reptans a score of zero. Actually, the ten individuals of villosum to the upper left in the diagram have an average ray value of 9.0; the ten individuals of reptans to the lower right have an average ray value of 0.4, both of which values are elose to the perfect score. The ten villosum closest to reptans have an average ray value of 7.8; the ten reptans closest to villosum, a value of 0.9. This means that all seven characters, the two on the axes and the five on the rays, are tending to go together. Comparable analysis of the diagram of the mongrel population (Fig. 5) shows that the ten most like villosum have an average ray value of 7.1 (lower than the lower value of the villosum population); the ten most like reptans have an average ray value of 3.5. "That is, the average values are intermediate, and the same tendeney of all characters to go together which is seen in Fig. 4 is illustrated here. Only further studies of P. reptans throughout its range will make possible the determination of centers of variability, of clines which could be the result of introgressive hybridization, of contamination of the glabrous reptans stock with the glandular- 1956] Braun,—Variation in Polemonium reptans 115 villous stock of villoswm. Such clines would give clues to time and direction of migration. The most prominent appears to extend in general westward of the area of the villosum popu- lations in southern Ohio; another, less marked, is evident in southeastern Pennsylvania, where pubescence of many plants suggests that a relic population of villosum once existed there- abouts. This situation in the Polemonium reptans complex is compar- able to that of the wild black cherries discussed by MeVaugh (1951, 52)—a common and widespread taxon (Prunus serotina subsp. serotina) ranging almost throughout the Deciduous Forest, in both glaciated and unglaciated areas, and a highly localized taxon (P. serotina subsp. hirsuta) occurring in areas where there are other disjuncts and rare local endemics. This latter is inter- preted as more ancient than the glabrous and widespread serotina, and more closely related to western subspecies. The entire range of this complex, as of many other phylads, is somewhat V-shaped, with an eastern and western arm. Polemoniwm has a range comparable to that of the Prunus serotina complex, with repre- sentatives in Mexico, the East, and the West. In view of the distinctness of reptans and villosum demon- strated by the pictorialized scatter diagram (Fig. 4), it is evident they should be maintained as distinct taxa. For the present, then, Polemonium reptans includes the two distinct entities, P. reptans L. var. reptans and P. reptans L. var villosum, E. L. Braun. Whether the status of villosum should be changed to sub- species (which now appears suitable), or whether it should be maintained as a variety of P. reptans as originally described (1940) awaits further investigation. If the zoologist's definition of subspecies be accepted—‘‘a geographically defined aggregate of local populations which differs taxonomically from other such subdivisions of the species"—then villosum is a subspecies, not a variety, which term is “limited to discontinuous variants within a single interbreeding population" (Mayr, Linsley, & Usinger, 1953). Certainly it is not a mere forma based on an “evanescent” character; it forms pure populations.—UNIVERSITY OF CINCINNATI. 116 Rhodora [Vor. 58 LITERATURE CITED ANDERSON, E. 1949. Introgressive Hybridization. John Wiley and Sons. —— —. 1952. Plants, Man and Life. Little, Brown and Com- pany. —— 1953. Introgressive hybridization. Biol. Reviews 28: 280-307. Braun, E. L. 1928. Glacial and post-glacial plant migrations indicated by relic colonies of southern Ohio. Ecol. 9: 284—302. 1940. New plants from Kentucky. RHoponma 42: 47-51. —— —————-. 1951. Plant distribution in relation to the glacial bound- ary. Ohio Jour. Sei. 51: 139-146. DavipsoN, J. F. 1950. The genus Polemonium [Tournefort| L. Univ. Calif. Publ. in Bot. 23: 209-282. FERNALD, M. L. 1950. Gray’s Manual of Botany. Ed. 8. American Book Co. GruEASON, H. A. 1952. Illustrated Flora of the Northeastern United States and Adjacent Canada. The New York Botanical Garden. Mayr, Ernst, E. G. Linsey, and R. L. UsivaER. 1953. Methods and Principles of Systematic Zoology. McGraw-Hill Co. McVauau, Rocers. 1951. A revision of the North American black cherries (Prunus serotina Ehrh., and relatives). Brittonia 7: 279- 315. 1952. Suggested phylogeny of Prunus serotina and other wide-ranging phylads in North America. Brittonia 7: 317-346. Tuomas, E. S. 1951. Distribution of Ohio animals. Ohio Jour. Sci. 51: 153-167. Wuerry, E. T. 1942. The genus Polemonium in America. Amer. Midl. Nat. 27: 741-760. Wo tres, J. N. 1951. The possible role of microclimates. Ohio Jour. Sci. 51: 134—138. 1956] Habeeb and Drouet,—List of Freshwater Algae 117 A SECOND LIST OF FRESHWATER ALGAE CHIEFLY FROM NEW BRUNSWICK HERBERT HABEEB AND FRANCIS DROUET This paper is a continuation of the work initiated in 1947, and reported upon in Ruopora Vol. 50, March, 1948. The material dealt with in this paper was collected during the years 1948-1951. In the winter of 1948-49 collections were made right through the cold months to show that the algal flora of our rivers and brooks is active the year round. In collections made from Little River and several brooks in the vicinity of Grand Falls during January and February of 1949, the following species are worth noting; Batrachospermum moniliforme, Calothrix parie- tina, Chlorotylium cataractarum, Draparnaldia plumosa, Ento- physalis rivularis, Leptosira Mediciana, Phormidium autumnale, P. purpurascens, P. Setchellianum, P. uncinatum, Rivularia haematites, R. rufescens, Rhizoclonium hieroglyphicum, Tetra- spora gelatinosa, Tolypothrix penicillata, and Spirogyra Grevil- leana.—This last being in a sterile state. The determinations were all made by Francis Drouet except that several numbers of Characeae were kindly determined for us by Mrs. Fay K. Daily, and that the naming of the Zygne- mataceae is the responsibility of Herbert Habeeb. In this con- nection, it may be well to mention that all the species reported in this list are based upon dried herbarium macrospecimens. We are grateful to Rev. Irénée-Marie for his careful examination of two of the numbers for desmids. This resulted in his description of the novelty, Staurastrum Habeebense, from small ledge pools in the canyon at Grand Falls, New Brunswick. The collection numbers given are those of Herbert Habeeb. Complete sets of the specimens are in the possession of the Cryptogamie Herbarium, Chicago Natural History Museum, and in the collections of Herbert Habeeb and Francis Drouet. Many of the numbers will be found duplicated in other herbaria. Unless otherwise indicated, the specimens are from Victoria County, New Brunswick. A number of specimens are included that were collected in adjacent Aroostook County, Maine; these are indicated by placing “Maine” after the collection number. Also listed are a number of specimens from Mt. Albert in the 118 Rhodora [Vor. 58 Gaspé region of Québec; these are indicated by placing “Gaspé” after the number. Some species from our first list of New Bruns- wick algae are again included to show recent changes in nomen- clature and present appropriate synonymy. AGMENELLUM THERMALE (Kiitz.) Dr. & Daily. synonyms: Merismo- poedia thermalis Kutz. COLLECTION NUMBERS: 10040. 10982. In a brooklet. ANABAENA OSCILLARIOIDES Bory. 10430. Wet swamp bottom. A. VARIABILIS Kütz. 11682, Albert County. 11345, Maine. 11468, Maine. In a ledge pool. In a ditch. On exposed mud of a pond. ANACYSTIS DIMIDIATA (Kütz.) Dr. & Daily. Gloeocapsa dimidiata (Kütz. Dr. & Daily. G. turgida (Kütz.) Hollerb. 10582. 11414 in part. 11620 in part, Madawaska County. Ledge pool. Shore of a lake. In a pond. A. MONTANA (Lightf.) Dr. & Daily. Gloeocapsa alpicola (Lyngb.) Born. 1792, Gaspé. 10456. 10476. 10556. 10558. 10586. 10646 10687. 10704. 11474, York County. 11581, York County. 11582 in part, York County. 11651. 11672 in part. 11673 in part. 11674 in part. 11685 in part, Albert County. On ledge and canyon wall. In ledge pools. On vegetation in a brooklet. Off a twig in a lake. On a boulder near a spring. On a concrete water-tub at a spring. On a rock in a lake, top Mt. Albert. A. MONTANA, f. MINOR (Wille) Dr. & Daily. 11491. In a shallow ledge pool. A. THERMALIS (Menegh.) Dr. & Daily. Gloeocapsa minuta (Kütz.) Hollerb. G. membranina (Menegh.) Dr. & Daily. 10671. 10966 in part. 11414 in part. In ledge pools. Shore of a lake. A. THERMALIS, f. MAJOR (Lagerh.) Dr. & Daily. Gloeocapsa limnetica (Lemm.) Hollerb. 11657 in part. In a lake. ANKISTRODESMUs FALCATUS (Corda) Ralfs. 10623. 10663 in part. On ledge above river surface. In a ledge pool. APHANOCHAETE REPENS A. Br. 10368. 10627. 10706. In ledge pools. Bog part of a small lake. AÁULOSIRA IMPLEXA B. & F. 10772. On wet ground. BULBOCHAETE MINOR A. Br. 10332. In a ledge pool. CALOTHRIX PARIETINA (Nig.) Thur. Calothrix parietina B. & F. 1778, Gaspé. On debris in a pool, top Mt. Albert. CAPSOSIPHON FULVESCENS (Ag.) Setch. & Gard. 13475, Sussex County. In a well at the salt springs, Sussex. CHAETOPHORA INCRASSATA (Huds.) Haz. 10347. 10613. 10839. 11206. 11208. 11538. 11625. On vegetation, rocks, ete. in lakes, brooks and ponds. CHARA GLOBULARIS Thuill. 11543. In a lake. Determined by Mrs. Fay K. Daily. CLADOPHORA CRISPATA (Roth) Kütz. 1966, Gaspé. In a falls, 1015 miles *North" of the hotel at Mt. Albert. 1956] Habeeb and Drouet,—List of Freshwater Algae 119 CoccoCcHLORIS ELABENS (Bréb.) Dr. & Daily. 10349. 11157. 11414 in part. 11437. 11522. 11650. 11655. 11657 in part. In lakes. C. Pentocystis (Kütz.) Dr. & Daily. 11087. In a lake. C. sTAGNINA Spreng, Anacystis rupestris (Lyngb.) Dr. & Daily. 10917. 10927. 11218. 11416. In a ledge pool. On ledge near falls. In lakes. CoLEOCHAETE SCUTATA Bréb. 11623. On clam shells in a lake. CYLINDROCYSTIS CYLINDROSPORA (Bréb.) Dr. & Daily. Cylindrocystis Brebissonii Menegh. 1779 in part, Gaspé. 10521. 10573. 10575. 11264. 11269. 11270. 11271. 13429 in part. In ledge pools. Ditches. Shore of a small lake, top Mt. Albert. Desmipium Swartz Ag. 10518. 10599, Maine. Wet bog. In a small pond. DICHOTHRIX ORSINIANA (Kütz.) B. & F. 11494. 13403 in part. In triekling water on a ledge. In a ledge pool. ENTOPHYSALIS LEMANIAE (Ag.) Dr. & Daily. Entophysalis Brebissonii (Menegh.) Dr. & Daily. 10407 in part. On rocks in a brooklet. E. RivuLARIS (Kütz.) Dr. Chamaesiphon polonicus (Rostaf.) Hansg. 1967, Gaspé. 10424 jn part. 10425, in part. 10447. 10453. 10500 in part. 10562. 10658. 10667. 10707 in part. 10957 in part. 10993. 11016. 11108. 13353 in part. 13372. 13373. 13375. 13405. 11480 in part, York County. On wet ledges. Ledge pools. On rocks, etc. in brooks. On ledge in a falls 10% miles “North” of the hotel at Mt. Albert. E. RIVULARIS, f. PAPILLOSA (Kütz. Dr. & Daily. 11446. 11681, Albert County. On wall of a garden pool. On rocks in a spring. EREMOSPHAERA VIRIDIS DeBary. 11171. In a pond. FISCHERELLA AMBIGUA (Nig.) Gom. Fischerella ambigua (B. & F.) Gom. 1783, Gaspé. Between wet stones, top Mt. Albert. ALOEOTRICHIA Pisum (Ag.) Thur. 11399. 11609. 11611 in part. 11624. In a pond. On moss and clam shells in a lake. GOMPHOSPHAERIA APONINA Kütz. 11463, Madawaska County. 11466, Madawaska County. 11620 in part, Madawaska County. In a pond. G. LACUSTRIS Chod. 10849. In a pool of a brooklet. GoNcGROsIRA DEBARYANA Raben. 10472. 10954. 11177. 13482, Sussex County. On rocks, ete. in springs and brooks. HAPALOSIPHON FONTINALIS (Ag.) Born. Hapalosiphon pumilus (Kirchn.) B. & F. 1790 in part, Gaspé. 1791, Gaspé. 1807 in part, Gaspé. 1899 in part, Gaspé. On wet ground, wet ledges, in tiny lakes, top Mt. Albert. HonMoriLA MUCIGENA Borzi. 10963. On rock in a brook. HYALOTHECA DISSILIENS (Sm. & Sow.) Bréb. 11150, Maine. 11210. 11220 in part. 11590, Maine. In a ditch. Ponds. Lake shore. H. mucosa (Dillw.) Ehrenb. 11836. In a river. PALMELLA Myosunvs (Ducluz. Lyngb. Hydrurus foetidus (Vill.) Trevis. 10420. 10454. 10661. 11680, Albert County. In a ledge pool. In brooklets and springs. 120 Rhodora [Vor. 58 LeprosiraA MEDICIANA Borzi. 11012. On a rock in a brooklet. LYNGBYA DicvETI Gom. 13471 in part, Sussex County. Well at the salt springs, Sussex. L. versicotor (Wartm.) Gom. 10742. 11626. 11673 in part. In a ledge pool. Off wood in a lake. On conerete water-tub at a spring. MESOTAENIUM RUBESCENS (Bréb.) Dr. & Daily. Mesotaenium macro- coccum (Kütz. Roy. 1787, Gaspé. 10555. 11546, Madawaska County. 11675. In a ledge pool. On concrete water-tub at a spring. On bare ground. On ground, top Mt. Albert. MicRocoLEUs RUPICOLA (Tild.) Dr. 10623 in part. 10633. 10910 in part. 11117. 13474 in part, Sussex County. On ground and ledges along rivers. On salty mud at the salt springs, Sussex. MicRosPORA AMOENA (Kiitz.) Rabenh. 11106. In a brook. MOUGEOTIA LAETEVIRENS (A. Br.) Wittr. 11536. 11615. In a lake. NETRIUM DIGITUS (Ehrenb.) Itzigs. & Rothe. 10527. 10971. 11338 in part. 13398. Bog end of à lake. Ledge pool. In a brooklet. In a river. N.iNTERRUPTUM (Bréb.) Lutkem. 11400 in part. 11401. In a pond. NITELLA FLEXILIS Ag. 10888, York County. In the St. John River, Fredericton. Determined by Mrs. Fay K. Daily. ` NosTOC COMMUNE Vauch.? 1744, Gaspé. On ledge, top Mt. Albert. N.PRUNIFORME (L.) Ag. 11158. 11159. 11436. 11457. In lakes. N. MicROoscOoPIUM Carm. Nostoc microscopicum B. & F. 1921, Gaspé. 1925, Gaspé. 1926, Gaspé. 1934, Gaspé. 1991, Gaspé. 1992, Gaspé. On mosses and hepaties in a ditch 2 miles “North” of the hotel at Mt. Albert. On mosses, swampy ground, 2 miles “North” of Cap-Chat. N. SPHAERICUM Vauch. 10583. 11438. 11458. 11520. 11654. In lakes. Ledge pool. N. vERRUCOSUM Vauch.? 1919, Gaspé. On a boulder, Mt. Albert. Oocystis Borcer Snow. 10381 in part. On wet ledge. O. crassa Wittr. 10656. In a ledge pool. OPHIOCYTIUM COCHLEARE (Eichw.) A. Br. 1788, Gaspé. 10363 in part. 11168. 11306, Maine. In ponds. Bog end of a lake. Bottom of a tiny pool, top Mt. Albert. O. PARVULUM (Perty) A. Br. 11069 in part, Maine. 11074, Maine. 11079, Maine. 11081, Maine. 11223. 11249 in part, Madawaska County. 11398. In ponds, ditches, etc. OSCILLATORIA AMOENA (Kütz.) Gom. 10426. 10625. 10828. 11103. 11211. 11289. 11424. 11515 in part. 11578 in part. 11601. 11631. 11638. 13380. 13381. 13414. On rocks, gravel, ledges, etc. in brooks and rivers. O. ANGUINA (Bory) Gom. 10364 in part. 11439 in part. 11440. 11459. Ina lake. Bog end of a lake. Pool on shore of a lake. O. ARTICULATA Gardn. 10850. 10673. In a ledge pool. Bottom of a pool on a brooklet. O. cHLonINA (Kütz. Gom. 10343 in part. 10355 in part. In lakes. 1956] Habeeb and Drouet,—List of Freshwater Algae 121 O. curviceps Ag. 11391. In a pond on a brook. O. GRANULATA Gardn. 10364 in part. 11434 in part. In a lake. Bog part of a lake. O. IRRIGUA (Kütz.) Gom. 10342. 10353 in part. 10477. 10595, Maine. 10942. 10969 in part. 11004. 11413. 11435. 11441. 11539. 11540. 11597? 11602? In lakes. In brooklets. Pond on a brooklet. O. LIMOsaA (Roth) Ag. 10777. 10779. On bottom and floating in an inlet to the St. John River. O. oRNATA (Kütz.) Gom. 11161. Ina lake. O. PRINCEPS Vauch. 10342 in part. 10343 in part. 11252, Mad- awaska County. 11411. 11439 in part. 11443 in part. 11451. 11452. In a pond. In lakes. O. srMPLICISSIMA Gom. 11207. 11214. In lake. Pool on shore of a lake. O. sPLENDIDA Grev. 10526. 10549. 11531, Maine. 11553. 11554. 11559. 11562. Ina bog. In ponds and lakes. O. TENUIS, vàr. TERGESTINA (Kütz.) Rabenh. 10341 in part. 10353 in part. 10355. 11180, Maine. 11198, Maine. In lakes. In a ditch. Brooklet from a swamp. PALMOGLOEA PROTUBERANS (Sm. & Sow.) Kütz. Gloeocystis Grevillei (Berk.) Dr. & Daily. G. confluens (Kütz.) Richt. 1743 in part, Gaspé. 1745, Gaspé. 1746, Gaspé. 1757, Gaspé. 1769, Gaspé. 1785 in part, Gaspé. 1786, Gaspé. 1807, Gaspé. 1946, Gaspé. 1964, Gaspé. 1984, Gaspé. 1987, Gaspé. 2034, Gaspé. On wet ledge, on moss, on peat, in a small lake, etc., top Mt. Albert. On a boulder in a ditch, 2 miles “North” of the hotel at Mt. Albert. On ledge at a falls, 1014 miles “North” of the hotel at Mt. Albert. In falls of a brook, 4 miles “North” of Cap-Chat. On moss on a stump, 2 miles “North” of Cap- Chat. On moss on a ledge near the St. Lawrence River, 3 miles “North” of Grosses Roches, Matane County. PENIUM MARGARITACEUM (Ehrenb.) Bréb. 11408 in part. In a pasture pond. PHORMIDIUM INUNDATUM Kütz. 10977. 13371 in part. 11307, Maine. On ledge in a brook. On a log in a brooklet. In a pond. P. PuRPURASCENS (Kütz.) Gom. 10907. 11007a. 11596. 11291. In a brooklet. P. Rerzu (Ag.) Gom. 11298, Maine. 11557. On rocks in a brooklet. In a lake outlet. P. VALDERIANUM Gom. 11093 in part. 11094 in part. 11097. 11098. 11102 in part. In water near shore of a lake, and in a nearby spring pool. PLECTONEMA NosrocoRuM Born. Plectonema Nostocorum Gom. 1927, Gaspé. Off moss in a ditch, 2 miles “North” of the hotel at Mt. Albert. P. ToMasriNIANUM (Kütz.) Born. 11037. On rocks in Little River. PLEUROTAENIUM TRABECULA (Ehrenb. Nag. 10411. Pond on a brooklet. RHIZOCLONIUM HIEROGLYPHICUM (Ag.) Kütz. 10720. 11008. 11023. 122 Rhodora [Vor. 58 11030. 11619, Madawaska County. In trickling water in canyon. In brooklets. In a pond. R. RIPARIUM (Roth) Harv. 13471, Sussex County. 13472, Sussex County. In wells at the salt springs, Sussex. RIvULARIA HAEMATITES (DC.) Ag. 10992. On rock in a brooklet. R. MINUTULA (Kütz.) B. & F. 10755. 11610. On ledges and rocks in Little River. On rocks in a pond. R. RUFESCENS (Nág.) B. & F. 11029. 11032. 11508. On rocks and ledge in springs and brooklets. SCENEDESMUS BIJUGA (Turp.) Lagerh. 10579. 10697. 10929. 11535, Maine. 11603, Maine. In ledge pools. Permanent puddle. Road- side pool. SCHIZOCHLAMYS AURANTIA (Ag.) Dr. & Daily. Schizochlamys gelatinosa A. Br. 11153, Maine. 11524. 11656. In a lake. On vegetation in swamp water. SCHIZOTHRIX ARENARIA (Berk.) Gom. 10770. 10920. On ledge near falls, and on rocks in the St. John River. S. CALCICOLA (Ag.) Gom. 10375. 10738. Ledge pools. S. Dairy: Dr. 10425 in part. In a brooklet. S. LACUSTRIS A. Br. Schizothrix lacustris Gom. 1794, Gaspé. Border of a tiny pool, top Mt. Albert. S. Lamy Gom. 1749, Gaspé. 1785 in part, Gaspé. On peat, shore of a small lake, top Mt. Albert. Š. LARDACEA (Ces.) Gom. 10967. 11561 in part. 11664 in part. On a boulder, shore of a lake. Side of a ledge pool. On wet wood in a lake outlet. S. LATERITIA (Kiitz.) Gom. 10928. In a ledge pool. 8. MELLEA Gardn. 1742 in part, Gaspé. 1776, Gaspé. In a water trickle on a ledge, top Mt. Albert. On peat, shore of a small lake, top Mt. Albert. S. MuELLERI Nüg. 11486 in part, York County. On ground in forest. S. PURPURASCENS (Kütz.) Gom. 11486 in part, York County. On ground in forest. SCYTONEMA DENSUM (A. Br. Born. 1752, Gaspé. On wet ledge, top Mt. Albert. S. MIRABILE (Dillw.) Born. Scytonema figuratum B. & F. 1742 in part, Gaspé. Ina water trickle on a ledge, top Mt. Albert. S. roLvpPorRICHOIDES Kütz. 11442. 11636. In a brooklet. Lake shore. SPHAERELLA LACUSTRIS (Girod.) Wittr. 1750, Gaspé. On rocks in a tiny pool, top Mt. Albert. SPIROGYRA CATENAEFORMIS (Hass. Kütz. (Also spelled CATENI- FORMIS). 10780. 10781. 10812. 10820. 10837. 10841. 10846 in part. 10939. 11068, Maine. 11073, Maine. 11176. 11287. 11384, Restigouche County. 11385, Restigouche County. 11447? 11594 in part. In brooklets, freshwater pools, springs, ponds, etc. 1956] Habeeb and Drouet,—List of Freshwater Algae 123 S. COMMUNIS (Hass.) Kütz. 11349, Maine. In a pond. S. CONDENSATA (Vauch.) Kütz. 11277. 11278. 11279. 11280. 11281 in part. 11282. 11341. Inlet of the Tobique River. Pond, bed of the Tobique River. In a spring, shore of the Tobique River. S. CRASSA Kütz. 11430 in part. 11444 in part. 11465, Madawaska County. In lakes and ponds. S. DUBIA Kütz. 10530. 10533. 10543. 10545. 10547. 10548. 10594, Maine. 11350, Maine. 11351 in part. 11429 in part. 11450, Maine. Ina ditch. Lakes and artificial ponds. S. FLUVIATILIS Hilse. 10915. 10925. 10926. 11110. 11472, Maine. 11495. In a ledge pool. On and between rocks in a pond outlet, Caribou. Š. GRACILIs (Hass.) Kütz. 11137 in part, Maine. 11235, Maine. 11309 in part, Maine. In a pond. Ditches. S. GREVILLEANA (Hass. Kütz. 10930. 10989. 11066 in part, Maine. 11067 in part, Maine. 11113, Maine. 11138 in part, Maine. 11142, Maine. 11304, Maine. In a ledge pool. Brooklet. In ditches. S. Hawrzscuu Rabenh. 11147, Maine. 11183, Maine. 11293, Maine. 11342, Maine. 11343, Maine. In a ditch. Our material is a perfect match with the type itself except for the ornamentation of the inner membrane of the spores. In all probability, our material was not ripe enough to show an ornamentation of the inner membrane. S. INFLATA (Vauch.) Kütz. 10490. 11066 in part, Maine. 11067 in part, Maine. 11138 in part, Maine. 11141, Maine. 11400 in part. In a pond. Freshwater pool. In ditches. S. JuERGENSH Kutz.? 10819. In a ledge pool. S. LONGATA (Vauch.) Kütz. 11136, Maine. 11139, Maine. 11348, Maine. 11351 in part, Maine. 11392 in part. In ponds. S. MAXIMA (Hass.) Wittr. 11430 in part. 11433. 11444 in part. 11445. Ina lake. S. NEGLECTA (Hass. Kütz. 10600, Maine. 10601, Maine. 10906. 11167. 11173. 11174. 11175. 11319. 11393. 11394. 11395. 11429 in part. 11616, Maine. Common in artificial ponds forming dense floating mats. S. MAJUSCULA Kitz. Spirogyra orthospira (Nag.) Kuütz. 11412. 11454. Ina lake. S. Weseri Kütz. 11118. 11228. 11257, Madawaska County. 11258, Madawaska County. 11259, Madawaska County. 11299, Maine. 11403. In ponds. Ledge pool. Slow part of a brooklet. STIGONEMA MAMILLOSUM (Lyngb.) Ag. 10515 in part. In a lake. S. PANNIFORME (Ag.) B. & F. 1742 in part, Gaspé. 1777, Gaspé. 1779 in part, Gaspé. 1790 in part, Gaspé. 1793, Gaspé. 1797, Gaspé. 1807 in part, Gaspé. On wet ledge, pool bottom, shore of a small lake, and in a tiny lake, top Mt. Albert. SYMPLOCA DUBIA (Nág.) Gom. 11489. On ledge in falls. S. MURALIS Kütz. 10692 in part. 11686, Albert County. 11517. In ledge pools. With Philonotis sp. on a damp ledge. 124 Rhodora [Vor. 58 TETMEMORUS GRANULATUS (Bréb.) Ralfs. 11322. In a ditch. TETRAEDRON MINIMUM (A. Br. Hansg. 10557. 10570. Ledge pools. TRIBONEMA BOMBYCINUM, f. TENUE Haz. 11146, Maine. In a ditch. TorvrorHRIX DISTORTA (Hofm.-Bang) Kütz. 10522. Bog shore of a lake. T. LANATA (Desv. Wartm. 10362. 11164. 11418. Lake and bog part of lake. In a swamp pool. T. PENICILLATA (Ag.) Thur. 10407 in part. 10471. 10489. 10788. 10956. 10974. 11005. 11035. 11038. 11044. 11048. 11488. On rocks and ledge in Little River and its tributaries. T. renuis Kütz. 10515 in part. 11163. 13477, Sussex County. 11498. 11611 in part. 11622. 11627. 11628. On clam shells, vege- tation, etc. in lakes. In a well at the salt springs, Sussex. ULOTHRIX IMPLEXA Kütz. 13478, Sussex County. Salt springs, Sussex. VAUCHERIA AVERSA Hass. 11085, Maine. 11302, Maine. 11311, Maine. In ditches. V. GEMINATA, var. RACEMOSA (Vauch.) Walz. 11111. 11308 in part, Maine. 11370, Restigouche County. In a brook. Small pond. In a ditch. ZYGNEMA CYANOSPORUM Cleve? 11337. 11634. In a spring. In a pool, bed of Little River. Z. DECUSSATUM (Vauch.) Ag. 11292 in part. In a brooklet. Z. INSIGNE (Hass) Kütz. 11570. 11588, Maine. 11589, Maine. In a ditch. Pasture pool. Z. PECTINATUM (Vauch.) Ag. 11281 in part. 11292 in part. 11340. In a brooklet. Inlet of the Tobique River. ZYGOGONIUM ERICETORUM Kütz. 1795, Gaspé. 10663 in part. 10664. 10665. 11420. Path surface through a bog. Ledge pools. In a flow- way between pools, top Mt. Albert.—GRAND FALLS, NEW BRUNSWICK, and the CHICAGO NATURAL HISTORY MUSEUM. 1956] Bean, Knowlton, Hill,—Plant Distribution 125 ELEVENTH REPORT OF THE COMMITTEE ON PLANT DISTRIBUTION The tenth report concluded the Monocotyledoneae. The pres- ent report begins with the Dicotyledoneae and includes the families through Moraceae, taken in the order of the eighth edi- tion of Gray's Manual. Unlike any of the previous reports the plants included here are all trees and shrubs with the exception of Saururus. The data for these reports have been compiled chiefly from the herbarium of the New England Botanical Club and the herbarium of the Arnold Arboretum and Gray Herbarium. PRELIMINARY Lists or NEW ENGLAND PLANTS — XXXVI 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 our flora. | Me N. Vt Mass. Conn. SAURURACEAE Saururus cernuus L. E "ur SALICACEAE Populus alba L. Populus balsamifera L. Populus balsamifera var. subcordata Hylander Populus canescens (Ait.) Sm. Populus deltoides Marsh. Populus deltoides var. missouriensis (Henry) Rehd. Populus X gileadensis Rouleau (balsamifera X deltoides) Populus grandidentata Michx. Populus heterophylla L. Populus nigra L. var. italica Muenchh. Populus tremuloides Michx. Populus tremuloides var. magnifica Vict. Populus tremuloides f. pendula Jaeger and Beisner Populus tremuloides f. reniformis 'l'idestr. Salix alba L. Salix alba var. calva G. F. W. Mey. Salix alba var. vitellina (L.) Stokes Salix alba X fragilis Salix amygdaloides Anderss. | — Salix arctophila Cockerell ET Salix argyrocarpa Anderss. | + + * + * * + + ++# + + +e ++ ++ *+++ + FRA > + + *ok K * + * * * * + + ++ * * dX * & * Gd & #* + * * * * * 126 Rhodora Salix babylonica L. Salix Bebbiana Sarg. Salix Bebbiana var. capreifolia Fern, Salix Bebbiana X discolor Salix Bebbiana X gracilis Salix Bebbiana X rigida Salix candida Fliigge Salix candida f. denudata (Anderss.) Rouleau Salix candida x Bebbiana Salix candida X gracilis Salix candida X rigida Salix Caprea L. Salix cinerea L. Salix cinerea X rigida Salix coactilis Fern. Salix cordata Michx. Salix cordata var. abrasa Fern. Salix discolor Muhl. Salix discolor var. latifolia Anderss. Salix discolor var. Overi C. R. Ball Salix discolor x humilis Salix fragilis L. Salix glaucophylloides Fern. Salix glaucophylloides f. lasioclada Fern. Salix gracilis Anderss. Salix gracilis var. textoris Fern. Salix gracilis x humilis var. microphylla Salix gracilis X sericea Salis X Grayi Schneid. (argyrocarpa x planifolia) Salix herbacea L. Salix humilis Marsh. Salix humilis var. hyporhysa Fern. Salix humilis var. keweenawensis Farw. Salix humilis var. microphylla (Anderss.) Fern. Salix humilis f. curtifolia Fern. Salix humilis f. festiva Fern. Salix humilis f. tortifolia Fern. Salix humilis X gracilis Salix humilis X sericea Salix humilis X humilis var. microphylla Salix incana Schrank Salix interior Rowlee Salix interior var. exterior Fern. Salix interior f. Wheeleri (Rowlee) Rouleau Salix X Jesupi Fern. (alba X lucida) Salix lucida Muhl. Salix lucida var. angustifolia Anderss. Salix lucida var. intonsa Fern. Salix X myricoides (Muhl.) Carey (rigida X sericea) Salix nigra Marsh. Salix X Peasei Fern. Me. ++++ +++ + + ++ +++ ++ ++ et ++ + + +++ ++ ++ — + * ++ ++++++-+-+ + * + ++ + +++++++ + + + + ++ + ++ 1956] Bean, Knowlton, Hill,—Plant Distribution Salix pedicellaris Pursh Salix pedicellaris var. hypoglauca Fern. Salix pedicellaris var. tenuescens Fern. Salix pellita Anderss. Salix pellita f. psila Schneid. Salix pentandra L. Salix planifolia Pursh Salix purpurea L. Salix pyrifolia Anderss. Salix pyrifolia, X discolor Salix rigida Muhl. Salix rigida var. angustata (Pursh) Fern. Salix rigida f. mollis (Palmer and Steyerm.) Fern. Salix rigida X discolor Salix rigida X gracilis Salix rigida x lucida Salix rigida X pedicellaris Salix rigida X subsericea Salix X rubens Schrank (alba X fragilis) Salix sericea Marsh. Salix serissima (Bailey) Fern. Salix X subsericea, (Anderss.) Schneid. (gracilis var. textoris X sericea) Salix x Smithiana Willd. (viminalis X caprea) Salix Uva-ursi Pursh Salix Uva-ursi f. lasiophylla Fern. Salix viminalis L. MYRICACEAE Comptonia peregrina (L.) Coult. Myrica Gale L. Myrica Gale var. subglabra (Chev.) Fern. Myrica pensylvanica Loisel. JUGLANDACEAE Carya cordiformis (Wang.) K. Koch Carya glabra (Mill.) Sweet Carya ovalis (Wang.) Sarg. Carya ovalis var. hirsuta (Ashe) Sarg. Carya ovalis var. obcordata (Muhl. and Willd.) Sarg. Carya ovalis var. obovalis Sarg. Carya ovata (Mill.) K. Koch Carya ovata var. Nuttallii Sarg. Carya ovata var. pubescens Sarg. Carya tomentosa Nutt. Juglans cinerea L. Juglans nigra L. CORYLACEAE Alnus crispa (Ait.) Pursh Alnus crispa var. mollis Fern. Alnus glutinosa, (L.) Gaert. Alnus rugosa (DuRoi) Spreng. Alnus rugosa f. Emersoniana Fern. Alnus rugosa var. americana (Regel) Fern. E + * ++ + *+ * +++ ++ N. H. Vt. Mass. +++ ++ + *+ * ++ +++ ++++ +++++ +++ ++ + ++ + ++ + *+ * +++ +++ ++ +++ ++++ ++++ > *+ ++ *++++++ ++ +++ +++ *+ ++++ ++++ +++ ++ + ++ * ++ +++ ++++ +++ ++ 128 Rhodora [Vor i|. $128 ed . = £ : ° z Z. = A e S Alnus rugosa var. americana f. hypomalaca Fern. + + + + Alnus rugosa var. americana f. tomophylla Fern. + Alnus serrulata (Ait.) Willd. + + + + + + Alnus serrulata f. noveboracensis (Britt.) Fern. + + Alnus serrulata var. subelliptica Fern. + + + + + Alnus serrulata var. subelliptica f. emarginata Fern. Alnus serrulata var. subelliptica f. mollescens Fern. + Betula alba L. * Betula borealis Spach + Betula X caerulea Blanch. (populifolia X caerulea-grandis) + + Betula caerulea-grandis Blanch. + + + Betula glandulosa Michx. + + Betula lenta L. + + + + + Betula lenta f. laciniata Rehd. + Betula lutea Michx. f. + + + + + + Betula lutea var. macrolepis Fern. + + + + + + Betula minor ('Tuckerm.) Fern. + + — Betula nigra L. + + + Betula papyrifera Marsh. + + + + + + Betula papyrifera var. commutata (Regel) Fern. + + + Betula papyrifera var. cordifolia (Regel) Fern. + + + + + + Betula papyrifera var. macrostachya Fern. + Betula papyrifera var. pensilis Fern. + Betula papyrifera x populifolia + Betula pendula Roth » * Betula populifolia Marsh. + + + + + + Betula populifolia f. incisifolia Fern. + Betula pumila L. + + + Carpinus caroliniana Walt. var. virginiana (Marsh.) Fern. + + + + + + Corylus americana Walt. + + + + + + Corylus americana f. missouriensis (A. DC.) Fern. +) +] + | + + Corylus cornuta Marsh. + + + + + + Corylus heterophylla Fisch. * Ostrya virginiana (Mill.) K. Koch + + + + + + Ostrya virginiana var. lasia Fern. + FAGACEAE Castanea dentata (Marsh.) Borkh. + + + + + Castanea pumila (L.) Mill. + Fagus grandifolia Ehrh. ES + + + + + Fagus grandifolia f. pubescens Fern. and Rehd. + Fagus grandifolia var. caroliniana (Loud.) Fern. and Rehd. + Fagus sylvatica L. * Fagus sylvatica var. atropurpurea West. * Quercus alba L. + + + Quercus alba f. latiloba (Sarg.) Palmer and Steyerm. + + + 1956] Bean, Knowlton, Hill,—Plant Distribution 129 w š £ S a al B Z Z. > Z en ° Quercus alba f. repanda (Michx.) Trel. + + Quercus X Bebbiana Schneid. (alba X macrocarpa) + Quercus bicolor Willd. + + + + + + Quercus coccinea Muenchh. + + + + + + Quercus X Faxoni Trel. (alba X prinoides) + Quercus X Fernowi Trel. (alba X stellata) + Quercus ilicifolia Wang. + + + + + + Quercus imbricaria Michx. * Quercus X Leana Nutt. (imbricaria x velutina) + Quercus macrocarpa Michx. + + + + Quercus macrocarpa f. olivaeformis (Michx. f.) Trel. + + Quercus Muehlenbergii Engelm. + + + Quercus Muehlenbergii f. Alexanderi (Britt.) Trel. + Quercus palustris Muenchh. — + + + Quercus X Porteri Trel. (rubra X velutina) + Quercus prinoides Willd. + + + + + Quercus prinoides var. rufescens Rehd. + Quercus Prinus L. + + ES 4 + + Quercus X Rehderi Trel. (ilicifolia X velutina) + + Quercus X Robbinsii Trel. (coccinea x ilicifolia) + + Quercus Robur L. * * * ` Quercus rubra L. -+ + + E + + Quercus rubra var. borealis (Michx. f.) Farw. + + + + Quercus X Saulii Schneid. (alba X Prinus) — + + Quercus stellata Wang. + + + Quercus stellata var. Margaretta (Ashe) Sarg. - Quercus X stelloides E. J. Palmer (stellata X prinoides) + + Quercus velutina Lam. + + + + + + Quercus velutina f. dilaniata Trel. + + + + + + Quercus velutina f. macrophylla (Dippel) Trel. + + + + + Quercus velutina f. pagodaeformis Trel. + + ULMACEAE Celtis occidentalis L. + + Celtis occidentalis var. canina (Raf.) Sarg. + Celtis occidentalis var. pumila (Pursh) Gray + + + + + Ulmus americana L. + + + + + + Ulmus americana f. alba (Ait.) Fern. + + + + + + Ulmus americana f. intercedens Fern. + + + + Ulmus americana f. laevior Fern. + + + + + + Ulmus americana f. pendula (Ait.) Fern. + + + + + + Ulmus glabra Huds. * * * Ulmus procera Salisb. * * * Ulmus rubra Muhl. + + + 4 + + Ulmus Thomasi Sarg. + + MORACEAE Broussonatia papyrifera (L.) Vent. * * Ficus Carica L. * Maclura pomifera (Raf.) Schneid. * ad Morus alba L. * * * * * * Morus rubra L. + + + + 130 Rhodora [Vor. 58 One tree that might have been included in the above list is Quercus falcata Michx. var. pagodaefolia Ell. as Mr. Eugene P. Bicknell in his “Ferns and Flowering Plants of Nantucket" records a single individual growing on Nantucket. He also refers to a similar tree growing at Fairhaven, Massachusetts. It seems doubtful if this variety is really a part of our New England flora. Castanea pumila is included as occurring in Massachusetts, based on a single specimen from Stony Brook Reservation. That this is a native to Massachusetts is questionable. Saururus cernuus, the only herbaceous species in this list, is a southern species occurring in the Naugatuck and Thames Valleys of Connecticut and at Little Compton, Rhode Island. Certain trees are included which very evidently have been planted. Ficus Carica is in the list but it could hardly be con- sidered as part of the flora. Other trees like Populus alba, Salix alba, S. fragilis and their hybrids are so firmly established that it is hard to realize that they are not native. Populus nigra var. italica is an interesting example of a widely distributed tree that depends entirely on propagation from cuttings. In the highway developments in recent years Salix pentandra has been exten- sively planted. Many hybrids have been found and probably more exist that have not been identified as such. The hybrids have been omitted from the following distributional categories as have also varieties and forms which seem to have no geographical significance. Geo- graphically the species here considered fall into groups which have already been used and defined in previous reports but in the present list there are no specimens to represent marine or estuarine habitats. I. GENERALLY DISTRIBUTED.—Populus grandidentata, P. tremu- loides, Salix Bebbiana, S. discolor, S. humilis, S. lucida, S. rigida, S. sericea, M yrica Gale, Betula lutea, B. papyrifera, D. papyrifera var. cordifolia, Ostrya virginiana, Corylus cornuta, Fagus grandi- folia. Wide ranging species are usually northern. Exceptions are Ostrya virginiana and Fagus grandifolia which represent species coming into New England from the Appalachian area in the southwest. M yrica Gale (typical) is wide spread but the var. subglabra 1956] Bean, Knowlton, Hill, —Plant Distribution 131 is rare, occurring only in Connecticut, Rhode Island and northern Maine. While the Canoe Birches, (Betula papyrifera and the var. cordifolia), are very generally distributed, the var. commu- tata is represented by too few specimens to point to a definite range and the var. pensilis has been reported only from Dacey Dam and Pembroke in Maine. II. RATHER GENERAL EXCEPT CAPE COD AND WASHINGTON COUNTY, MAINE.—Salix rigida, var. angustata, Betula lutea var. macro- lepis, Ulmus americana. The distribution of Salix rigida var. angustata seems to be much more restricted than the species. This may be more apparent than real as this variety has not been recognized until recently. Betula lutea var. macrolepis seems to be absent also from Aroostook County. Herbarium material of Ulmus ameri- cana may not be indicative of its actual range. Many collectors have failed to include it from an area in the belief that it might have escaped from cultivation. III. NORTHERN, WITH NUMEROUS STATIONS SOUTH OF 43° .— Populus balsamifera, Salix pedicellaris and S. pedicellaris var. hypoglauca, S. gracilis and S. gracilis var. textoris, Alnus rugosa, and its varieties. Salix pedicellaris is more northern and much more local than its var. hypoglauca. Although Salix gracilis and its var. textoris (formerly known as S. petiolaris) have approximately the same range, the variety is much more frequent. IV. NORTHERN, NOT MUCH SOUTH OF 43?.—Populus tremu- loides var. magnifica, Salix coactilis, S. lucida var. angustifolia and var. intonsa, S. pellita, S. pyrifolia, Alnus crispa var. mollis, Betula caerulea-grandis, Quercus rubra var. borealis. Salix coactilis occurs only in Maine where it is local, chiefly in the Penobscot River Valley. The varieties angustifolia and intonsa of Salix lucida are much more northern and restricted in their distribution than the typical form which is widely distributed in New England. V. aLPINnE.—Salix arctophila, S. argyrocarpa, S. herbacea, S. planifolia, S. Uva-ursi, Alnus crispa, Betula borealis, B. glandu- losa, B. minor. The exact distribution ean be better understood from the following table: 132 Rhodora [Vor. 58 Mt. Katahdin Mt. Katahdin and Mt. Katahdin, White Mt. Katahdin and White Mts. Mts. and Vermont northern Vermont Salir arctophila Š. argyrocarpa Š. planifolia Betula borealis S. herbacea S. Uva-ursi Betula glandulosa Alnus crispa HB. minor VI. CAPE COD, GENERAL IN MAINE SOUTH OF 45°, BUT NOT NORTHERN MAINE.—Salix nigra, Myrica pensylvanica, Comptonia peregrina, Alnus serrulata, Betula populifolia, Carpinus caro- liniana var. virginiana, Corylus americana, Castanea dentata, Quercus alba, Q. bicolor, Q. ilicifolia, Q. rubra, Q. velutina. Myrica pensylvanica is confined to southeastern New England and the Maine coast. Quercus velutina is largely in southern New England but extends into southern New Hampshire and southwestern Maine. Of more general distribution are Salix nigra, Comptonia peregrina, Betula populifolia, and Quercus rubra. Other species extend east only to the Penobscot Bay and River: Alnus serrulata (only 2 stations beyond the Kennebec), Carpinus caroliniana var. virginiana, Corylus americana, Cas- tanea dentata, Quercus alba, Q. bicolor and Q. ilicifolia (with only one station on Mt. Desert). VII. NEITHER CAPE COD NOR NORTHERN MAINE AND NOT IN WASHINGTON COUNTY, MAINE.—Ulmus rubra, Betula lenta, Juglans cinerea, Carya ovata. Ulmus rubra is general in New Hampshire, Vermont, Massa- chusetts (except Cape Cod) and Connecticut but rare in Maine and Rhode Island. Betula lenta has some stations in Vermont, southern New Hampshire and southeastern Maine. Carya ovata and Juglans cinerea do not extend beyond the Kennebec River. VIII. CHIEFLY THE THREE SOUTHERN STATES.—Quercus pri- noides, Q. coccinea, Carya ovalis, C. glabra, Alnus serrulata var. subelliptica. Quercus prinoides has a few stations in southern Vermont and southern New Hampshire while Q. coccinea with the same distri- bution extends to York and Cumberland Counties, Maine. Carya ovalis is represented by 2 Maine stations, 3 New Hampshire and one Vermont but is too poorly known for any discussion. Carya glabra extends into southwestern Vermont and southern New Hampshire. Alnus serrulata var. subelliptica is more southern than the typical form but has one station in New Hampshire and one in Vermont. 1956] Bean, Knowlton, Hill,—Plant Distribution 133 IX. CHIEFLY THE THREE SOUTHERN STATES BUT NOT IN WEST- ERN MASSACHUSETTS.—Saururus cernuus, Salix humilis var. micro- phylla, Carya tomentosa, Quercus stellata. Saururus cernuus comes in from the south and west and is con- fined to Connecticut with the exception of one station in Rhode Island. Salir humilis var. microphylla (S. tristis) reaches south- ern New Hampshire and has one station in Maine on Mt. Desert. Carya tomentosa has one New Hampshire station but none in Maine. Quercus stellata 1s limited to extreme southern Connecti- cut, Rhode Island, Cape Cod and the Islands. X. CHIEFLY THE THREE SOUTHERN STATES BUT NOT ON CAPE cop.—Carya cordifolia, Quercus Prinus, Celtis occidentalis and its vars. canina and pumila. Carya cordiformis extends up into western Vermont. Quercus Prinus occurs in western and southern Vermont, southern New Hampshire and southeastern Maine. Celtis occidentalis occurs in western Vermont and the var. pumila in the Connecticut Valley in Vermont and New Hampshire. XI. CHIEFLY THE THREE SOUTHERN STATES BUT NEITHER CAPE COD NOR WESTERN MASSACHUSETTS.—Quercus palustris. XII. COASTAL PLAIN.—Salir humilis var. hyporhysa, Ostrya virginiana var. lasia, Quercus prinoides var. rufescens, Quercus stellata var. Margaretta. The typical form of Quercus stellata although it also occurs on the trap ridges of the New Haven, Connecticut, area is usually considered to be a coastal plain species. XIII. cCALCICOLOUS, CHIEFLY WEST OF THE CONNECTICUT RIVER IN THE SOUTH BUT IN THE EAST MOSTLY NORTH OF 45°.—Salix candida, S. glaucophylloides, Betula pumila, Quercus macrocarpa. Salix candida has one station in Essex County, Massachusetts. S. glaucophylloides is found only in northern Maine. Betula pumila grows in the limestone areas of western Connecticut and western Massachusetts and in northern Maine with one station at Orono in central Maine. Quercus macrocarpa occurs in west- ern New England and central Maine. XIV. RIVER VALLEYS, CHIEFLY OF WESTERN NEW ENGLAND.— Salix serissima, Quercus Muehlenbergii, Ulmus Thomasi. XV. MISCELLANEOUS.—Populus deltoides, P. heterophylla, Salix amygdaloides S. interior, S. interior var. exterior, Juglans nigra, Betula nigra, Morus rubra. 134 Rhodora [Vor. 58 In some cases native species have been so widely planted be- yond their normal range that the plotting of their natural distri- bution is difficult. Populus deltoides, the Cottonwood, is a western species which reaches western New England as a native plant but has been extensively planted farther east. Juglans nigra is another western species which occurs as a native in western New England but is planted in the east. Morus rubra is native in western Connecticut, western Massachusetts and southwestern Vermont but is also planted in southern New England. Populus heterophylla, another western species, is only repre- sented in southwestern Connecticut. Salir amygdaloides is con- fined to western Vermont and western Massachusetts. Salix interior seems to be confined to the Connecticut River Valley and the Lake Champlain region, while the var. exterior occurs only in the Aroostock River Valley in Maine. Betula nigra is restricted to the Merrimac and Ipswich Rivers and their tribu- taries except for one station in southwestern Connecticut.—R. C. Bran, C. H. KNowrrow, and A. F. HILL. Volume 58, no. 688, including pages 79-102, was issued 4 May, 1956. 73.4 "3 Hovora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS l RALPH CARLETON BEAN RICHARD ALDEN HOWARD ( Associate Editors CARROLL EMORY WOOD, JR. Vol. 58 June, 1956 No. 690 CONTENTS: Additional Evidence for Five as the Basic Chromosome Number of the Andropogoneae. Robert P. Celarier................. 135 A New Elymus from Alaska. H.J. Hodgson.................. 144 Some New Primitive Mexican Cruciferae. Reed C. Rollins....... 148 Clarence Hinckley Knowlton. Ralph C. Bean.................. 157 Dentaria multifida in Central Ohio. Robert W. Long, Jr........ 160 A Correction in the Names of Two Species of Heterokontae. Hannah DEO ROREM RN EDIT Ec cl ee 161 Elodea Correct without Being Conserved. Lloyd H. Shinners... 162 Che New England Botanical Club, Ine. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $4.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 40 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $4.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scien- tific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of publica- tion. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U.S. A. Rbodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 June, 1956 No. 690 ADDITIONAL EVIDENCE FOR FIVE AS THE BASIC CHROMOSOME NUMBER OF THE ANDROPOGONEAE Rosert P. CELARIER BEcAUSE of the meager amount of information available on the cytology, genetics, and comparative morphology of the grass tribe Andropogoneae, it has not been possible to state with any great amount of assurance what the phylogenetic lines of development are for the tribe. From the cytotaxonomic point of view it is of considerable importance to know what the basic chromosome numbers are, how they originated, and their significance in establishing phylogenetic trends. Several basic numbers have been suggested for the tribe during the past twenty-five years. Avdulov (1931) found 2n = 36 from his studies of Miscanthus floridulus Warb (as M. japonicus) and suggested nine as the basic number for the tribe. Although there is some doubt about the number of this species, since Bremer (1934) has recorded 2n = 38, there is no question but that nine is the haploid number for several species in the tribe, but whether it is primitive or derived is not established. Seven has also been proposed (Moriya and Kondo, 1950), as well as six (Janaki-Ammal, 1940) but both are rare and there are no diploid species known with either of these numbers. There are also several higher numbers recorded but, for the most part, they are more easily explained as amphiploids involv- ing species with different chromosome numbers. By far the most common numbers recorded for the tribe are multiples of ten, or five (Darlington and Janaki-Ammal, 1945; Delay, 1951). Apparently the first known case of n — 5 was recorded by 136 Rhodora [VoL. 58 Karper (1930) for Sorghum versicolor. This was later confirmed by Longley (1932) and Karper and Chisholm (1936). The same number was found for S. purpureo-sericeum and S. dimi- diatum by Huskins and Smith (1932, 1934), S. nitidum by Ayyanger and Ponnaiya (1941), and S. intrans by Garber (1948). Garber (1950), in a detailed study of the genus, confirmed these findings for all of the species except S. nitidum which he found to be 2n = 20. Besides these he found two others, S. slipoidum and S. brevicallosum, with n = 5, and later S. matarankense was added to the list (Garber and Snyder 1951). Although no other diploid species (n — 5) are known there are several kinds of secondary information that may be brought to bear on the problem. A few cases of 2n = 30 have been recorded for the tribe. The first was Narenga porphyrocoma reported by Bremer (1925), Janaki-Ammal (1941) and Moriya (1950). In these studies, representing the species from most of its range of distribution, the meiotic behavior was regular and 15 bivalents were almost always seen at diakinesis. This was followed by a report of 2n = 30 for Hyparrhenia hirta by Garber (1944). He found the meiotic behavior to be essentially normal but with an oc- easional quadrivalent. This report was substantiated by Krupto (1953), who reported 2n = 30 types from three localities, and by de Wet (1954). Another example of 2n = 30 was found by Brown (1951) in Hrianthus strictus Baldw. (not Bluff. et Fingerh.) from Texas material. No meiotic studies were made. In the present study another genus is added to this group, as well as two additional species of Hyparrhenia and two more collections of H. hirta. Hyparrhenia hirta with 30 somatic chromosomes was found in two entries, one from Israel and the other from Assam, India. The latter was undoubtedly introduced since the species is not native to India. Neither of these entries was entirely regular in its meiotic behavior and both had approximately the same types of irregularities (Tables I and II). Some 40-55 per cent of the cells had 15 bivalents at diakinesis and metaphase I (Fig. 2), and although univalents and quadrivalents were not uncommon, no cells were found that had both univalents and quadrivalents. At anaphase and telophase I about fifty per 137 :hromosome Number Celarier,—The Basic € 1956] T ETE S £ 0'6 r-0 09 OF 0818-V 06 08 OI OI LT $£-0 £9 LE 149c-V m "E = E £ T 6-0 99 [ere 9cI€£-V GI 88 € € € 6 9-0 0€ 0€ PSTE-V 8 c6 q S I q—-0 g¢ cP 9T6E-V 1ajINUOLIVUL (puou “Bout sabprug Bp) 1122 tad "4yo *Bp) (pusou ON UU quan Lag 1u32 aq 199 Lag ‘ou "ay Bp] “ON qna quad Lad 1439 Lag quan Lag aiody pofiqq I aspydoja [, pun aspydvup MƏVLQ HHO4S AVA ANF JL ANV TV LV HOIAVHAG ANOSOWOUHD II W'ISV.L 0'00I IISI 0£ [Ze "v (q'$8'0 I9I£-V SooN sniofujmds tuniufibod fiy 0'6 0'0 rv 9 G6 ' SI F 98 AIC-O 'IICTI-II 'Ic-O OE vAuoy ‘PHM ULIS M Of£I£-V Jdvjs vsowof 'H est EEE 0'96 €$'60 OO AII-0 IIST-cE IF-O OE UAOUX JON 90€ Jdejs pins ^H eco 9G 0'0 6 88 S'S AIC-0 'IIST-OI IFO OE 1eosw2epv]N [e1jUo() HEWN ^W I496-V Jas ofn `H 0'c 0'0t 0'c O OF O'OF AIT-O‘IIST-E1 'IT-0 O£ eAuow ‘OVIM esuengs H OcI£-V Jdvjs pins ^H 0'c 0'0 O'STI 0 S€ 0OSGS AII-O IIST-ET ‘I1Z-0 OE *rpul 'uressy uosiopuy 'H 9I6£-V jdeys vuy CH. 0'0 0'0 0'0£ 0:08 0'OF AII-O IIST-eTI 'Ic-O OE PUISI I€€ A-Q9AON wuey "VY FSI£-V jdeig pity piuaysivdhyy S$49V1O II8I-IY IITI-IG AII-IISI IISI aDup uz t011D90”I 401991109 ‘ON sads YNN $1122 1u32 194 I 38VHdVL3]X ANV SISANINVIC] LV SNOLLVHADIJNO|) AWOSOWOUHQ I TIdV L 138 Rhodora [Vor. 58 cent of the cells had lagging chromosomes (Figs. 3, 4, 5) and these laggards were found both dividing and non-dividing at these stages. Very rarely bridges and fragments were seen (Figs. 3, 5) and these were very likely due to small inversions. At the dyad spore stage only a very small percentage of the cell pairs had micronuclei (Table II). In a few cells at late telophase I it was still possible to count the chromosomes in each of the daughter nuclei. In several of these cells more than 30 chromosomes were found, with both equal and unequal distributions, and it can be concluded that both divided and undivided univalents are often distributed to the daughter cells. "This will help to account for the much smaller percentage of dyads with micronuclei than anaphase or telophase cells with lagging chromosomes. Hyparrhenia rufa with 2n = 30 has also been found, and has been studied from three locations (Tables I and II). Two of the entries are essentially similar in their meiotic behavior but the entry 306 is quite different. However, this difference is due principally to the increased number of univalents found in 306, and the orientation of these univalents at metaphase I (Fig. 7) strongly suggests that this may be due principally to early ter- minalization of two bivalents. In A-2671 and A-3126 a high percentage (33-40) of the cells with 15 bivalents at diakinesis and metaphase I occur, and a similar percentage with one quadrivalent is found. Univalents are not uncommon in either case (Fig. 6). Also A-2671 was frequently found to have two quadrivalents. At anaphase and telophase I approximately 65 per cent of the PLATE 1226 Figures 1-11. Meiosis in the Andropogoneae. Fig. 1. Metaphase I in Sorghum intrans with five bivalents. Fig. 2. Diakinesis in Hyparrhenia hirta A-3916 from Assam, India showing 15 bivalents, five are noticeably larger than the others. Fig. 3. Anaphase I in H. hirta A-3184 from Israel with two lagging chromosomes that are dividing, and a bridge + fragment. Fig. 4. Telophase I in H. hirta A-3916 with three lagging chromosomes. Fig. 5. Telophase I in H. hirta A-3916 showing lagging chromosomes and a bridge + fragment. Fig. 6. Metaphase I in H. rufa A-3126 from Kenya with 2 univalents and 14 bivalents, five of the bivalents are larger than the others. (Chromosomes have been touched up with India ink fore ontrast.) Fig. 7. Metaphase I in H. rufa, 306 showing univalents whose association to one another suggests early terminalization—note the five large bivalents. Fig. 8. Meta- phase I in H. formosa A-3130 from Kenya with 15 bivalents, two are lying together. Fig. 9. Diakinesis in Hypogynium spathifolius A-3161 from Brazil with 15 bivalents, five of which are very small. Fig. 10. Metaphase I in H. spathifolius with 15 bivalents. Fig. 11. Metaphase I in Dichanthium sericeum with 10 bivalents, five being considerably smaller than the others. All fig. 1350 X. Rhodora Plate 1226 e. 1956] Celarier, —The Basic Chromosome Number 139 cells studied had lagging chromosomes, although never more than three were seen in any one cell These laggards were both dividing and non-dividing at these stages. Although only one entry was studied at the dyad stage (Table II) it was similar to H. hirta in that there were far fewer dyads with micronuclei than anaphase and telophase cells with lagging chromosomes. The reason is probably similar to that of H. hirta. Hyparrhenia formosa with 2n = 30 was seen in one entry from Kenya. It is similar to the other Hyparrhenias at diakinesis and metaphase I except that it had somewhat fewer instances of quadrivalents (Fig. 8) and more cells with univalents. It is also similar to the other species at anaphase and telophase I with 60 per cent of the cells having lagging chromosomes, and as many as four found in one cell. Although there are differences in frequencies in the various types of configurations found in the Hyparrhenia entries, the same general pattern of behavior is seen in all. It is noticeable that four chromosomes are frequently involved in non-bivalent associations, either as a quadrivalent or as univalents. The formation of quadrivalents is usually associated with reciprocal translocation, or homology of the four chromosomes, and it is virtually impossible to distinguish the two except by studies of pachytene chromosomes. However, translocations that produce quadrivalents in no more than 40 per cent of the cells, and with 35-50 per cent bivalents are rather special cases. Their con- stant occurrence throughout such a wide geographical area as represented by these materials and in three different species seems rather unlikely to the author. On the other hand partial homology between four or more chromosomes (two quadri- valents were found occasionally in H. rufa and H. formosa) might be expected in a group where speciation was rather active. It seems much more likely that the multivalent formations found in this study are the result of partial homology than of reciprocal translocations. Hypogynium spathiflorum was also found to have 30 somatic chromosomes. This species was studied from one entry col- lected in Brazil. It was found to have 15 bivalents at diakinesis and metaphase I (Figs. 9, 10), and to have completely regular meiotic divisions. Another type of secondary information concerning five as a 140 '. Rhodora [Vor. 58 basic number is seen in chromosome size. In the three species of Hyparrhenia included in this study it was noted that five of the chromosome pairs were considerably larger than the others (Figs. 2, 6, 7, 8). Also in Hypogynium there was con- siderable variation in chromosome size and five of the chromo- somes were very much smaller than the others (Figs. 9, 10). This difference in size is also often quite noticeable in Dich- anthium sericeum. This species has 20 somatic chromosomes with ten bivalents at metaphase I and completely regular meiotic divisions. At diakinesis and metaphase I it is usually easy to distinguish five large and five small bivalents (Fig. 11). Although no detailed study has been made of this, it has been noted in several collections of the species. DISCUSSION That the genus Sorghum has a basic number of five seems to be fairly certain, both from the numerous accounts of n = 5 known, and also from the occasional bivalents found in haploids of Sorghum vulgare (Brown, 1943; Kidd, 1952), and especially from the recovery of chromosome interchanges in the derivatives of crosses of haploids with diploids (Endrizzi and Morgan, 1955). Several other genera are now known that have species with the somatic chromosome number of 30, and this might be ex- plained in one of the following ways. 1. diploids with a basic number of 15. 2. triploids with a basic number of 10. 3. amphiploids resulting from a cross involving two other basic num- bers (6 x 9 or 7 x 8). 4. hexaploids with a basic number of five. 1. If they are diploid species they should be predominately» regular in meiotic behavior and any irregularities would be due mostly to struc- tural changes (i.e. quadrivalents would be due to reciprocal transloca- tions). The three Hyparrhenia species reported here all have quadri- valents but the meiotie behavior suggests partial homology rather than translocations. Also 15 as the basic number for the tribe is untenable in view of the large number of species with 2n = 10. 2. If they are triploid species one would expect at least one genome to be frequently involved in either the formation of trivalents or uni- valents. Not a single entry of any of the species studied has even ap- proached this order of irregularity and at least two species are apparently completely regular in meiotic behavior. It seems very unlikely that the 2n — 30 types could have had a triploid origin. 1956] Celarier,—The Basic Chromosome Number 141 3. If they are amphidiploids they must be of secondary origin involving crosses of species with n = 6 Xn =9 or n =7 X n = 8. Of these numbers six, seven, and eight are so seldom met with in the tribe, and none yet seen at the diploid level, that this possible origin seems an un- likely one. 4. If they are hexaploids it is possible that they may have arisen as trispecific hybrids in which there is no homology between the chromo- somes and as a consequence they have a completely regular meiotic behavior. However, it seems somewhat more likely that they were formed as the result of crosses between rather closely related species in which some homology existed between at least some of the chromosomes. Consequently one might occasionally expect to find species with some multivalent associations. Any selective advantage for fertility would favor chromosomal changes (in the partially homologous chromosomes) that would tend to diploidize them. If this event has occurred the meiotic behavior might indicate the relative time of origin of the hexa- ploids. This type of origin of the 2n = 30 types is not incompatible with the data at hand. Species are known with regular meiotic behavior (Narenga porphyrocoma and Hypogynium spathiflorum) as well as species that indicate partial homology between’ four or more chromosomes (Hyparrhenia species). Also Hyparrhenia is the most highly specialized genus of the ones considered, is probably rather recent in origin and very active in speciation at the present time. Thus if there is a selective advantage for diploidization it would be expected that it would be less well developed in the more recently formed species which in this study would be Hyparrhenia. Therefore the most logical conclusion as to the origin of the 2n = 30 types appears to be that they are hexaploids with a basic number of five. With this conclusion it is now possible to study the distribu- tion of x = 5 in the tribe as a whole. The tribe is generally divided into six subtribes: 1. Dimertinae, 2. Saccharinae, 3. Ischaeminae, 4. Rottboellinae, 5. Sorginae, 6. Andropogoninae. No cytological studies have been made in the Dimeriinae and only two genera in the Ischaeminae and five in the Rottboellinae have been studied. All other subtribes, from the most primitive Saccharinae, with Erianthus and Narenga, to the most advanced Andropogoninae, with Hypogynium and Hyparrhenia, have species with a basic number of five. Also the very closely related and highly specialized tribe Maydeae has n = 5 in 142 Rhodora [Vor. 58 Coix and five is generally regarded as the basic number for the tribe. There is little question but that n = 10 is the basic material from which most present day species of the Andropogoneae have been derived. Consequently any single basic number for the tribe, other than ten, must have developed n = 10 types several times during the evolution of the tribe. Of the numbers that have been proposed, the one that could most easily have made this change is five. It is not the feeling of the author that the problem of a basic number for the tribe has been completely solved, nor that it is entirely established that the tribe has a monophyletie origin. However, the available evidence strongly suggests that five may have been the basic number, that ten was derived from it principally by allopolyploidy, and that nine is of secondary origin from ten. This seems to offer a fairly reliable working hypothesis. SUMMARY Three types of evidence have been presented supporting five as the basic number for the Andropogoneae. First there are species with n — 5, secondly the species with n — 15, and third are the species that have five chromosomes of different mor- phology from the other chromosomes of the complement. Although the number of species that are shown, by these methods, to have a basic number of five are few, they are widely distributed in the tribe, being found in both the most primitive and the most advanced subtribes. It is pointed out that x = 5 is more compatible with the other known numbers in the tribe, and that it should be considered as a working hypothesis.—DEPARTMENT OF BOTANY AND PLANT PATHOLOGY, OKLAHOMA A & M COLLEGE, STILLWATER. LITERATURE CITED AvpULOV, N. P. 1931. Karyo-systematische Untersuchung der Familie Gramineen. Bull. Appl. Bot., Genet., and Pl. Breeding. Suppl. 43. Ayyancer, G. N. R., and B. W. X. Ponnatya. 1941. Studies in Para-Sorghum Snowden—the group with bearded nodes. Proc. Ind. Acad. Sci. 14: 17-24. Bremer, G. 1925. The cytology of the sugarcane III. Genetica 7: 203-322. 1956] Celarier,—The Basic Chromosome Number 143 ———————. 1934. De Cytologie van het Suikerriet VII. Arch. Suik. Ned. 42: 141-166. Brown, M. S. 1943. Haploid plants in Sorghum. Jour. Hered. 34: 163-166. Brown, W. V. 1951. Chromosome numbers of some Texas grasses. Bull. Torrey Bot. Club. 78: 292-299. DanuiNGTON, C. D. and E. K. JANAKI AMMar. 1945. Chromosome Atlas of Cultivated Plants. London. Detay, C. 1951. Nombres chromosomiques chez les Phanérogames. Rev. de cyt. et biol. végét. vol. 12. pg Wet, J. M. J. 1954. Chromosome numbers of a few South African grasses. Cytologia 19: 97-103. Enprizzi, J. E. and D. T. Moraan. 1955. Chromosomal interchanges and evidence for duplication in haploid Sorghum vulgare. Jour. Hered. 46: 201-208. GARBER, E. D. 1944. A cytological study of the genus Sorghum, sub-sections Para-Sorghum and Eu-Sorghum. Amer. Nat. 78: 89-93. 1947. The pachytene chromosomes of Sorghum intrans. Jour. Hered. 38: 251-252. . 1950. Cytotaxonomie studies in the genus Sorghum. Univ. of Calif. Publ. in Botany 23: 283-361. , and L. A. Snyper. 1951. Cytotaxonomic studies in the genus Sorghum II. Two new species from Australia. Madroiio 11: 6-10. Husxins, C. L. and S. G. Smiru. 1932. A cytological study of the genus Sorghum Pers., I. The somatic chromosomes. Jour. Genet. 25: 241-249. —. 1934. A cytological study of the genus Sorghum Pers., II: The meiotic chromosomes. Jour. Genet. 28: 387—395. JANAKI AMMAL. 1940. Chromosome numbers in Sclerostachya fusca. Nature 145: 465. . 1941. Intergenerie hybrids of Saccharum I, II, and III. Jour. of Genet. 41: 217-253. KanrER, R. E. 1929. Annual report of the Texas Agr. Exp. Station. ————————, and A. T. CunisHoLw. 1936. Chromosome numbers in Sorghum. Amer. Jour. Bot. 23: 369—374. Kipp, H. J. 1952. Haploid and triploid Sorghum. Jour. Hered. 43: 225. Krupto, S. 1953. Karyological studies and chromosome numbers in Hyparrhenia aucta and H. hirta. Jour. So. Afric. Bot. 19: 31-58. Lonetey, A. E. 1932. Chromosomes in the grass sorghums. Jour. Agr. Res. 44: 317-321. Moriya. A. 1950. Contributions to the cytology of the genus Sac- charum IV. Studies of chromosomes in wild Saccharum species in Formosa. Cytologia 15: 237-254. ,and A. Konpo. 1950. Cytological studies of forage plants I. grasses. Jap. Jour. Genet. 25: 126-131. 144 Rhodora [Vor. 58 A NEW ELYMUS FROM ALASKA! H. J. HopnasoN? A question concerning the identity of a species of Elymus considered in other studies arose during the winter of 1954. At that time herbarium specimens of representative material collected in the Matanuska Valley were examined independently by Jason R. Swallen of the Smithsonian Institution, J. L. Steb- bins, Jr. of the University of California, D. C. Smith and E. L. Nielsen of the Wisconsin Agricultural Experiment Station and R. W. Pohl of the Iowa Agricultural Experiment Station. All questioned the relegation of this material to E. canadensis L. as had been done by Anderson? and Lepage.‘ Certain features suggested the species indicated, but all agreed that it did not appear to be closely similar to //. canadensis with which they were familiar. The following quotation, taken from Mr. Swallen’s letter regarding the herbarium specimens in question, is relatively typical of those received: . . . I am not satisfied with calling it Elymus canadensis, and on the other hand there is no other species in Alaska or North America that it could be. "There is some resemblance to E. hirsutus, but none of the lemmas are hirsute toward the summit as in this species. It would seem, then, that this is either a hybrid in itself or a new species. Subsequently, Nielsen has had opportunity to examine numerous colonies of the plant in question as it occurs naturally in Alaska. It was his opinion that this appeared quite different from any Elymus canadensis he was familiar with in the States. The following description has been drawn to cover the plant for which the name Elymus pendulosus is proposed, calling atten- tion to the distinctive pendulose aspect of the spikes. Elymus pendulosus Hodgson, sp. nov. Planta perennis, caespitosa, rhizomatibus brevibus adscendentibus, viridis vel purpurea; culmi erecti vel basi decumbentes, 40-100 cm. 1 Contribution from Agronomy Department, Alaska Agricultural Experiment Station, Published with approval of the Director. ? Formerly Head, Agronomy Department, Alaska Agricultural Experiment Station. 3 Towa State College Jour. of Science 18:381—445. 1944 4 Nat. Canad. 79:241-260. 1952 1956] Hodeson, —A New Elymus from Alaska 145 alti; vaginae glabrae, apertae; laminae planae, supra seabrae, infra seaberulae, 5.0-12.0 mm. latae; spicae setosae, pendulae, 15-30 em. longae; spiculae remotae, saepe solitariae, interdum ternatae, paulum patulae; glumae 0.5-1.0 mm. latae, 4.0-8.0 mm. longae, 2-4 nervis, scabrae, summo nervis hispidulae, arista hispidula 0.5-7.0 mm. longa praedita; lemmata scabra, basi sparse pilosa, summo nervosa, arista divergente 1.0-3.0 em. longa praeditum; rachilla summo pubescens. Plant perennial, caespitose, with short ascending rhizomes, green or often highly pigmented with anthocyanin; eulms usually semi-decum- bent at base, 40-100 em. tall; sheaths glabrous, open; blades flat, scabrous above, scaberulous beneath, 5.0-12.0 mm. wide; spike bristly, pendant, 15-30 em. long; spikelets distant, often solitary, usually in pairs but occasionally three at a node, slightly spreading; glumes 0.5-1.0 mm. wide, 4.0-8.0 mm. long, 2-4-nerved, scabrous, slightly hispid on the distal portion of the nerves; awn 0.5-7.0 mm. long, hispidulous; lemmas scabrous, slightly pubescent at base, strongly nerved above; awn some- what divergent, 1.0-3.0 em. long; rachilla pubescent. Elymus pendulosus may be distinguished from E. canadensis by: (1) its longer and narrower spikes that are also less dense and more pendant (the latter character contrasts sharply with the nodding character in E. canadensis); (2) its spikelets are more distant, often solitary but usually in pairs, only rarely are there three or four per node; (3) its shorter glume awns; (4) the less flexuous lemma awns; and (5) plants are never glaucous. Plants are often highly pigmented with anthocyanin, particularly when encountered in localities away from the immediate coast. In numerous instances the intensity of pig- mentation approached a dark reddish purple. It is not possible to state at this time whether any significance can be attached to the fact that colonies observed at Fish Creek and in the Lowe River gorge immediately above Valdez exhibited no anthocyanin development. Additional study will be necessary to determine if distinct genic races occur naturally in the species. It was possible to define the Alaskan plant more clearly during the past summer. It appears to be indigenous to south central Alaska. It is generally distributed throughout the Matanuska Valley which lies northward and westward from the Knik Arm of Cook Inlet. It was found to be locally abundant along the Glenn Highway to about milepost 103. It becomes much less common in occurrence as one travels into the interior, and occurs only sporadically at Tok Junction beyond which it was 146 Rhodora [Vor. 58 Figure 1. Elymus pendulosus. Spike (X14), spikelet (52), lemma ( X5) and glume ( X6). 1956] Hodgson,—A New Elymus from Alaska 147 not observed. It also has been observed along the Alaska Railroad right-of-way northward from the Matanuska Valley though the northern limit of its range in this direction is un- known. It has not been observed in the Fairbanks area. Throughout its range it is confined generally to habitats such as borrow pits and highway shoulders and embankments. In these situations it appeared where competition from other species was not sufficient to limit its occurrence and spread. Several colonies were located near Fish Creek. Here it occurred in openings of birch, alder and poplar thickets along the banks of the Knik Arm. Numerous colonies occurred on ledges and escarpments along the Lowe River which flows into Prince William Sound at Valdez. Its common occurrence in the gorge a few miles east of Valdez, and the colonies at Fish Creek, were in situations that precluded further consideration of the plant as a recently introduced exotic as had been the writer’s earlier opinion. Here the species gave the appearance of being native to the flora. The table below gives measurements and descriptive data derived from a mass collection consisting of 47 specimens made throughout the range of the species. Figure 1 depicts a typical spike, spikelet, lemma, and glume of Elymus pendulosus. The description is based upon measurements and other in- Mean + S. D. Range Description Lemma ]lengthmm. 12.224 1.2 9.5 — 14.2 5-nerved, strongly nerved upper 2/8, short- width mm. 1.8+ 0.2 1.3 — 2.2 pubescent at base awn mm. 20.34 4.0 11.5 — 31.0 Palea length mm. 11.7 1.1 9.4 — 14.6 toothed, weakly scabrous on edges width mm. 1.12 0.1 0.7— 1.5 Glume length mm. 5.8 2 1. 1 — 8.0 very narrow scarious margin, linear- acuminate, width mm. 0.724 0.2 0.5 — 1.0 2-5 nerves awn mm. 39.024 1.8 0.5— 7.0 Anther length mm. 17+ 0.1 1.4— 1.9 Culm length cm. 66.0 + 11.7 42.2 — 98.4 Spike length cm. 22.6 + 3.2 14.8 — 29.0 pendant width mm. 10.0 — 20.0 Rachilla short-pubescent 148 Rhodora (Von. 58 formation derived from specimens taken throughout the entire range of the species, as well as casual field observations. The type specimen, H. J. Hodgson No. 234, was taken near Palmer, Alaska on July 29, 1955. This is deposited in the U. S. National Herbarium, Smithsonian Institution, Washington. Cotype specimens have been deposited in various herbaria as follows: No. 230, University of California (Berkley); No. 231, Missouri Botanieal Garden (St. Louis); No. 233, New York Botanical Garden; No. 235 Department of Agriculture, Ottawa, Canada; and No. 236, University of Wisconsin (Madison). Other specimens have been deposited in the Herbarium of the University of Alaska, Fairbanks, Alaska. - Elymus pendulosus is a 28-chromosome species as is generally true of E. canadensis. Meiosis is predominately regular with almost all diakinesis figures showing 14 bivalents. Bridges occurred in about 10 per cent of the Anaphase I figures examined. About 75 per cent viable pollen is produced by this species. The cytology of this species and its presumed hybrids with Agropyron sericeum Hitch. will be presented in a later paper. The author takes pleasure in expressing appreciation to Dr. Etlar L. Nielsen for reviewing the manuscript and for his con- structive criticisms, to Mr. Jason R. Swallen for reviewing the manuscript and assisting with the latin description and to Mrs. E. Herbold Froeschner for the line drawings in figure 1. SOME NEW PRIMITIVE MEXICAN CRUCIFERAE REED C. ROLLINS Present day phylogenists will readily admit the Cruciferae and the Capparidaceae to adjacent rungs on the phylogenetic ladder of the higher plants even though they may not agree in detail as to the point of origin of each family. For some time it was con- tended that the Cruciferae arose from some prototype plant somewhat similar to certain taxa of the Papavaraceae, but evi- dence has been piling up in favor of the view that the Cruciferae arose from some capparidaceous-like plant and that the two families probably had a common origin (von Hayek, 1911; Sehulz, 1936). Certainly there are genera of the two families 1956] Rollins,—Primitive Mexican Cruciferae 149 with marked similarities. Both Warea amplexifolia and Stanleya pinnata, for example, were originally described in the Cappari- daceae as species of Cleome. Those who stress the similarities of these families usually point to the presence in the Cruciferae of characters such as a gynophore and spreading equal stamens, which are regular features of the Capparidaceae (cf. Rollins, 1939). These characteristics are by no means common in the Cruciferae but among genera of the Western Hemisphere they may be found in Warea, Stanleya, Thelopodium, Romanschulzia and a few others. In the past, Stanleya has been thought of particularly as being a key to the phylogenetic link between these two families. Now as Romanschulzia is becoming better known, this genus may be viewed with increasing interest in this connection. Stanleya occurs only in the arid areas of western North America and ap- pears to be somewhat isolated morphologically from most other erucifers. It ties in with Thelypodium but appears not to have any other close connections. This is not the case with Roman- schulzia, because it ties in with Thelypodium in one direction and with Sisymbrium in another. Furthermore, it is not too distant from a genus like Macropodium of eastern Asia. The broad, flattened and reflexed siliques of Romanschulzia apetala, together with the long gynophore, the spreading exserted stamens and the greatly elongated inflorescence, particularly suggest such a rela- tionship. Because of the increasing evidence that Romanschulzia may occupy a key phylogenetic position among primitive cruci- fers, the new material coming from Mexico and Central America is of high interest. However, the paucity of collections of this genus still does not permit a proper assessment of it in terms of the phylogeny of the family as a whole. The three species described below are clearly distinct from the eight species covered in my former treatment of the genus (Rollins, 1942). It is hoped that the illustrations of them will aid collectors in procuring more material of this interesting genus. Romanschulzia Meyerif Rollins, sp. nov. Perennial from a deep root; stems up to 2 meters high, branched above, glabrous, pithy; basal leaves petiolate, runcinately pinnatifid to pinnately lobed, 1-2 dm. long including petiole, blade extremely thin and fragile, veins prominent, pubescent with sharp flat awl-shaped trichomes which 150 Rhodora [Vor. 58 Fic. 1-4, Romanschulzia Meyeri#. Fig. 1, basal portion of plant with basal leaves, X 1$. Fig. 2, central portion of stem with auriculate cauline leaves, X 1⁄4. Fig. 3, upper portion of an infructescence, X 4. Fig. 4, basal portion of a silique showing gynophore, X 3. Drawings by Dorothy H. Marsh. 1956] Rollins,—Primitive Mexican Cruciferae 151 occur mostly along the veins and blade margins, lobes irregular, terminal lobe making up most of the blade surface, oval to nearly orbicular, irregu- larly dentate, incised or somewhat erose, lateral lobes remote and small, usually oblong, petiole variable in length, 5-15 em. long; cauline leaves auriculate, clasping, thin, broadly oblong, obtuse or the upper pointed, prominently veined, glabrous or with an occasional trichome arising near a vein, reduced upwards, 1-2 dm. long, 3-8 em. wide, margin smooth or with minute callosities which project very slightly; infructescence greatly elongated, 3-7 dm. long; pedicels spreading at right angles to descending, expanded at apex, glabrous, 7-12 mm. long; siliques terete, somewhat moniliform, stipitate, widely spreading, usually curved upward, rarely nearly straight, 6-10 em. long, 1-1.2 mm. in diameter, glabrous, valves nerved from base to apex; gynophores 1.5-2.5 mm. long; styles slender, unexpanded at apex, 1-1.5 mm. long; seeds uniseriate, alternating on each side of the septum resulting in a marked displacement of the septum into the opposite valve, seeds plump, oblong, ca. 1.5 mm. long, < 1 mm. in diameter, with well defined but not high longitudinal ribs, circular scar at the funicular end prominent, radical shorter than the cotyledons; cotyle- dons incumbent, thick. Fig. 1-4. Herba perennis robusta; caulibus glabris superne ramosis teretibus ad 2 m. altis; foliis radicalibus petiolatis lyrato-pinnatifidis membranaceis pubescentibus 1-2 dm. longis; foliis eaulinis sessilibus integris auriculatis amplexicaulibus late oblongis obtusis glabris 1-2 dm. longis, 3-8 em. latis; infructescentiis 3-7 dm. longis; pedicellis fructiferis patentibus glabris 7-12 mm. longis; siliquis glabris teretibus moniliformibus stipitatis 6-10 em. longis; stylis tenuibus 1-1.5 mm. longis; seminibus oblongis ca. 1.5 mm. longis, cotyledonibus incumbentibus. Type in the Gray Herbarium, collected along a dry stream at an altitude of 1900 meters on the east side of Cerro Linadero, Dulees Nombres, Nuevo León, just east of border into Tamaulipas, 24? N., 99.5-100.5? W., Mexico, August 9, 1948, F. G. Meyer and D. J. Rogers 2892. Isotype at the Missouri Botanical Garden. This species possesses by far the longest siliques known in the genus and even though they are mostly above seven centimeters long, they show a marked tendency to curve upwards, being most frequently bow-shaped. The pedicels are widely spreading from a position nearly at right angles to somewhat reflexed, but remaining always at a wide angle of divergence from the rachis. There are no flowers present on the specimens. Romanschulzia Meyeri# appears to be most closely related to R. elata, but there is so little material available of these species that it is impossible to be sure where the relationship lies. K. Meyeri also has a number of characteristics in common with R. turritoides. The very long arcuate siliques of R. Myeri are 152 Rhodora [Vor. 58 distinctive and indicate that it is without doubt an independent species. Romanschulzia subclavata Rollins, sp. nov. Herbaceous perennial (?) basal portion unknown; stems up to 2 meters high, branched above, hirsute with short awl-shaped trichomes below, sparsely pubescent above; cauline leaves auriculate, clasping the stem, lanceolate, acute to acuminate, shallowly dentate to denticulate, some leaves with a pair or several larger teeth toward the base of the blade, greenish above, slate-colored below, often somewhat narrowed toward the base, 6-15 em. long, 1-2 cm. wide, sparsely hirsute to glabrous; infructes- cence up to 10 dm. or more long, narrow; pedicels glabrous, spreading at right angles to rachis to very slightly ascending, 5-15 mm. long; flowers numerous in a long raceme; sepals oblong, white to slightly purplish, adjacent sepals united below into pairs, 3-4 mm. long, 1.5-2 mm. wide; petals absent; stamens nearly equal, spreading at anthesis, filaments slightly flattened, white, expanded at base, anthers 1.8-2 mm. long; siliques linear to somewhat subclavate, erect, nearly terete, stipitate, glabrous, 1.5-2.5 em. long, ca. 1.5 mm. broad, valves with a single nerve from base to apex; gynophore 1.5-2 mm. long; pollen tricolpate, reticu- lately pitted, narrowly oblong, 24.14 X 13.13 (av. of 10 grains); styles slender, 0.5-1 mm. long, unexpanded at apex; immature seeds wingless, oblong, plump, ca. 2 mm. long; ca. 1 mm. wide, forming a single row in the siliques. Fig. 5A-D. Herba perennis (?) basi ignota; caulibus superne ramosis teretibus pubescentibus vel glabris 1-2 m. altis; foliis caulinis auriculatis amplexi- caulibus lanceolatis acutis vel acuminatis dentatis vel denticulatis sparse hirsutis vel glabris 6-15 em. longis, 1-2 em. latis; inflorescentiis 6-10 dm. longis; pedicellis patentibus glabris 5-15 mm. longis; sepalis oblongis, albis, 3-4 mm. longis, 1.5-2 mm. latis; petalis absentibus; siliquis sub- clavatis vel linearibus teretibus stipitatis glabris 1.5-2.5 em. longis; stylis tenuibus 0.5-1 mm. longis; seminibus immaturis oblongis ca. 2 mm. longis, ca. 1 mm. latis. Type in the Gray Herbarium, collected on Mt. Tancitaro in the Mu- nicipality of Tancitaro, Michoacán, Mexico, July 25, 1941, Wm. C. Leavenworth and H. Hoogstraal 1207. The site is noted as “just above cloud forest on exposed ridges" at an elevation of 9,500 ft. A second specimen from the same area has the following label data: “plant 5 ft. tall, straight, white flowers; damp cliffs in cloud forest, Mt. Tancitaro, 10,000 ft., July 9, 1941, Wm. C. Leavenworth and H. Hoogstraal 4034." (GH). Romanschulzia schistacea Rollins, sp. nov. Perennial or possibly biennial; stems relatively slender, 5-8 dm. high, glabrous, simple or branched; basal leaves not present; cauline leaves thin, over-lapping, auriculate, clasping the stem, green above, slate- colored beneath, glabrous or with a few scattered trichomes, 4-10 cm. long, 1.5-3.5 em. wide, narrowly oblong to nearly lanceolate, broadest 1956] Rollins,— Primitive Mexican Cruciferae 153 Fig. 5A-D, Romanschulzia subclavata. Fig. 5A, upper portion of plant showing inflorescences, X 14. Fig. 5B, flowers at full anthesis, semidiagrammatic, X 5. Fig. 5C, flower at the beginning of anthesis, semidiagrammatic, X 5. Fig. 5D, silique, pedicel and area of attachment to stem, X 1.5. Drawings by B. Tugendhat, 154 Rhodora [Vor. 58 above middle, only slightly to rather markedly narrowed toward base, acute to acuminate, entire, denticulate with minute teeth along margins, midvein prominent; infructescence relatively short, 1-2 dm. long; flowers not numerous, about 12-36 in each inflorescence; pedicels spreading, expanded into receptacle, 6-10 mm. long; sepals caducous, yellowish white, nonsaccate, glabrous, 3-4 mm. long; petals linear, widely spreading at anthesis, not differentiated into blade and claw, yellowish-white, 4—5 mm. long, 0.5 mm. or less wide, only slightly narrowed below; stamens spreading at anthesis, nearly equal, filaments ca. 4 mm. long, expanded at base, anthers 1.5-2 mm. long; pollen tricolpate, reticulately pitted, narrowly oblong, 29.18 x 12.54 (av. of 10 grains); glands well developed around the stamens and petals and forming a dise around the ovary; siliques stipitate to nearly sessile, glabrous, terete, somewhat constricted between the seeds, widely spreading, curved upward, 2.5-3.5 cm. long, < 1 mm. in diameter, lowest portion sterile and resembling a gynophore, valves without a prominent nerve; gynophore 0.5 mm. long to nearly wanting; style slender, unexpanded at apex, 1-1.5 mm. long; immature seeds oblong, in a single row in the silique. Fig. 6A-D. Herba perennis (?) caulibus tenuibus glabris simplicibus vel ramosis 5-8 dm. altis, foliis radicalibus ignotis; foliis caulinis sessilibus auriculatis amplexicaulibus imbricatis oblongis vel lanceolatis integris denticulatis acutis vel acuminatis 4-10 cm. longis, 1.5-3.5 em. latis; pedicellis fruc- tiferis patentibus glabris tenuibus 6-10 mm. longis; sepalis caducis glabris 3-4 mm. longis; petalis patentibus ochroleucis linearibus 4-5 mm. longis, 0.3-0.5 mm. latis; siliquis teretibus patentibus glabris 2.5-3.5 em. longis stipitatis, seminibus uniseriatis. Type in the Gray Herbarium, collected in moist woods, Barranca de las Verduras, north side of El Monte on trail from Zimapan to Mines of El Monte, August 11, 1948, H. E. Moore, Jr. and C. E. Wood, Jr. 1945. Isotype at the Bailey Hortorium. This taxon has a shorter inflorescence than most species of Romanschulzia, approaching some specimens which I have de- termined as R. arabiformis in that respect. However, the widely spreading, upcurved, relatively long siliques are distinctive. The seeds are definitely uniseriate in R. schistacea and the sepals lack the horn-like thickening at the apex, as found in R. arabi- formis. The slender yet prominent styles are not common char- acteristics of any closely related species in the genus. The leaves are exceedingly thin and fragile. R. schistacea lies about as close to Sisymbrium as any species of Romanschulzia and were it not for the spreading petals, the equal and spreading stamens, and the caducous calyx, I should be inclined to place it in that genus. 1956] Rollins,—Primitive Mexican Cruciferae 155 f Fie. 6A-D, Romanschulzia schistacea. Fig. 6A, plant with basal leaves missing, x 1$. Fig. 6B, silique, pedicel and area of attachment, X 2. Fig. 6C, flower at the beginning of anthesis, X 5. Fig. 6D, flower at full anthesis, X 5. Drawings by B. Tugendhat. Romanschulzia apetala Rollins The original description of this species was based on a single flowering specimen. Since then, a second collection made by Richard W. Holm and Hugh H. Iltis in 1949 in the general area of the type locality, has been available for study. Fortunately, 156 Rhodora [Vor. 58 part of an old raceme of fruits is included as part of the specimen preserved at the Missouri Botanical Garden. These are the first fruits I have seen of R. apetala and they are unique in the whole genus in being broad, flattened, and with a relatively long slender style. As predicted from the flowering material, the pedicels are somewhat reflexed although they are not strictly so. The flat siliques are tapered both above and below into a more or less lanceolate shape. However, the body of the silique has roughly parallel sides. The fertile portion of the silique is 3-4 cm. long and 3.5-4.5 mm. wide; the gynophore varies from 5-8 mm. long; and the slender style is 4-5 mm. long. The data from the label of this specimen are as follows: loose shrub up to 3 m. tall; flowers dark purple; disk after dehiscence of sepals prominently whitish green; young stamens purple. Virgin cloud-forest, 3000-3300 m., vicinity of Millsville, Pan- American Highway, about 3 km. above Nivel, Province of Cartago, Costa Rica, 22 July 1949, Holm and Iltis 536 (MO). Sisymbrium Standleyi Rollins, nom. nov., based on Roman- schulzia alpina Standley and Steyermark, Fieldiana 24: 377. 1946, not Sisymbrium alpinum Fourn., Recherch. Crucif. 131, 1865, which was based on Braya alpina Sternb. and Hoppe. At the time I was studying Romanschulzia previously (l. c.), I put aside an undetermined specimen collected by Alexander F. Skutch from the Department of Quezaltenango in Guatemala that represented an undescribed species of Sisymbrium. The material was considered for possible inclusion in Romanschulzia since the cauline leaves are auriculate, but the flowers do not have the equal spreading stamens, the caducous calyx, nor the gynophorate siliques of Romanschulzia. In the meantime, Standley and Steyermark have described R. alpina based on specimens collected by Steyermark in the Department of Hue- huetenango, Guatemala. Now a comparison of the type of R. alpina with the Skutch collection shows that they belong to the same species. The species is not closely related to any in Romanschulzia and does not possess the essential characters of the genus. On the other hand, it is related to species included in Sisymbrium by Payson and others, where it is best placed, pend- ing further studies of that genus. The known collections, all from Guatemala, referable to Sisymbrium Standleyi are as 1956] Bean,—Clarence Hinckley Knowlton 157 follows: dry, rocky, grassy slopes, between Tojquiá and Caxín Bluff, summit of Sierra de los Cuchumatanes, Dept. of Huehue- tenango, Aug. 6, 1942, Julian A. Steyermark 50144 (F, type); near Tunimá, Sierra de los Chuchumatanes, July 6, 1942, Steyermark 48923 (F); Volcán Santa María, Dept. Quezaltenango, July 27, 1934, A. F. Skutch 864 (GH).—GRAY HERBARIUM, HARVARD UNIVERSITY. LITERATURE CITED von HAYEK, AuGusr. 1911. Entwurf eines Cruciferen-Systems auf Phylo- genetischer Grundlage. Beiheft. Bot. Centralbl., Beib. 27 (127): 127-355. Payson, E. B. 1922. Species of Sisymbrium Native to America North of Mexico. Univ. of Wyo. Publ. Sci. 1: 1-27. Roruns, REED C. 1939. The Cruciferous Genus Stanleya. Lloydia 2: 109-127. 1942. A Tentative Revision of the Genus Romanschulzia. Contrib. Dudley Herb. 3: 217-226. ScnuuLz, O. E. 1936. Cruciferae. Die Natur. Pflanzenf. 17b: 227-658. CLARENCE HINCKLEY KNOWLTON RALPH C. BEAN The New England Botanical Club has had in its membership through the years many enthusiastic amateur botanists, men in business or professional life who have gained physical vigor and intellectual enjoyment in their botanical pursuits. Perhaps none of these has been so active in the Club for so long a period as Mr. Clarence Hinckley Knowlton who died May 10, 1956 at the South Shore Hospital in Weymouth, Massachusetts. Mr. Knowlton was born in Farmington, Maine September 9, 1876, the son of David Hunter Knowlton and Clara Armine (Hinckley) Knowlton. His father was the publisher of the local paper and had a job printing business in Farmington. There Mr. Knowlton began to study plants and for many years he collected in Farmington and the surrounding towns. Day Mountain, Mount Abraham and Mount Saddleback, all in Franklin County, stimulated his early interest in unusual plants. Even after he had established his home in Massachusetts he would return to Farmington to search for plants which might have escaped his attention in his earlier collecting. Mr. Knowlton’s preparation for college was at the Farmington 158 Rhodora (Vou. 58 State Normal School. He entered Harvard in 1895 and gradu- ated in the Class of 1899. After graduation he engaged in teach- ing for more than six years in Massachusetts. He was one year at Webster, for three years he was principal of the Chelmsford High School and from 1903 to 1905 he was submaster at the High School in Lexington. During these years he explored the local flora most thoroughly as the many sheets in the New England Botanical Club herbarium show. After these years in teaching he became interested in the book publishing business and early in 1906 he began work in the Boston office of D. C. Heath and Company, publishers. He remained with them for more than forty years until his retire- ment in 1949. For the first two and a half years he worked in the manufacturing department and thereby acquired a thorough knowledge of the books which the company published. Then he was sent out as an agent. From time to time he did editorial work as well. As a result of his business contacts he came to have a large acquaintance with the schoolmen of New England. He was one of the founders and the first secretary of the Educa- tional Salesmen’s Association of New England, an organization which brought men of like interests together and did much to ameliorate the asperities of competition in the school book and supply business. On September 1, 1907, Mr. Knowlton married Lydia Anne Hunter of Machias, Maine whose father was a physician in Machias for many years. From that time on he spent part of his vacation each year at his summer home at Roque Bluffs, near Machias, and collected extensively in Washington County in which both Machias and Roque Bluffs are located. In 1907, he published in Ruopora “Noteworthy Plants collected at Roque Bluffs, Maine" and in 1917 ‘Plants and Plant Societies at Roque Bluffs, Maine." Unfortunately the Flora of Washington County on which he had spent so much time was never pub- lished. However, his specimens are in the herbarium of the New England Botanical Club. After his marriage, he made his home at Hingham, Massa- chusetts where both Mr. and Mrs. Knowlton entered whole- heartedly into all phases of the life of the community. Mr. Knowlton was active in the First Church (founded in 1635) serving as the first president of the Unitarian Layman’s League 1956] Bean,—Clarence Hinckley Knowlton 159 Chapter and chairman of the Parish Committee. In Hingham he served as president of the Hingham Chamber of Commerce and had been treasurer of the Hingham Historical Society since its incorporation in 1914. He took a deep interest in the Historical Society building, the Old Ordinary, and in the col- lections which it housed. Another Hingham interest was the South Shore Nature Club of which he was one of the most active members and of which he was president for four years. Several publications of this organization were written by Mr. Knowlton such as “The Goldenrods of the South Shore," “ The Asters of the South Shore” and “The Trees of the South Shore.” Mr. Knowlton was elected one of the Trustees of Public Reser- vations and prepared a “Flora of the Rocky Woods Reservation" at Medfield, Massachusetts while serving in that capacity. He was an associate member of the American Association for the Advancement of Science and a member of the Boston City Club. It is clear that his greatest interest always remained in the field of botany. He became a member of the New England Botanical Club in 1900 and was closely connected with its various activities for over fifty years. He served as president for three years (1925 to 1928), as chairman of the Program Committee, a member of the Committee on the Local Flora, à member of the Committee on Plant Distribution, a member of the Committee on the Flora of Massachusetts, and a member of the Committee on the Flora of the Boston District. He was a member of the Vermont Bo- tanical Club and of the Josselyn Botanical Society of Maine. For many years he was treasurer of the latter society, attending the field meetings and contributing to the programs. In this connection it is interesting to note that the Josselyn Botanical Society held a meeting in Farmington in 1896 which Mr. Knowl- ton attended and again fifty seven years later in 1953 there was a second field meeting at Farmington where Mr. Knowlton was of great help because of his intimate knowledge of the area. This was his last field meeting with the Society. In the course of the years, Mr. Knowlton made some collecting trips outside of New England though the latter area held his lifeinterest. In 1930, with Mrs. Knowlton, he spent the summer in England which enabled him to attend the International Bo- tanical Congress at Cambridge. In 1911 in connection with his business, Vermont and western 160 Rhodora [Vor. 58 New Hampshire were added to his collecting territory. When he began to make his business trips by automobile he also ex- plored the areas he visited for botanical specimens. Many of the duplicates were given to the New England Botanical Club and finally in 1955 he presented his entire herbarium to the Gray Herbarium with the stipulation that the Club was to have any New England specimens needed. He was always alert in find- ing new plants and in learning about the plants themselves, their distribution and their relation to the soil and habitat. Many botanists have pleasant memories of days in the field with him. Mr. Knowlton through the years contributed many articles to Rhodora. Two such appear in 1899 in the first volume and his last contribution is appearing in the current volume in 1956.— WAKEFIELD, MASSACHUSETTS. DENTARIA MULTIFIDA IN CENTRAL Onto—In the spring of 1914 Dentaria multifida Muhl. was first collected in Delaware County, Ohio by Dr. C. E. O'Neal and his student, the late Dr. R. C. Friesner. "This station is an undisturbed, rich elm-oak-maple woodland locally noted for its abundance of spring wildflowers. The soil is slightly acidic clay loam. A specimen was later deposited by O'Neal at the Herbarium of Ohio State University.! This record was overlooked by later authors, however, and thus the species is represented as being restricted to an area bounded by Georgia and Alabama on the south and the extreme southern portions of Ohio and Indiana on the north.’ During the spring of 1955 a second station was discovered by the writer near the top of a ridge bordering Alum Creek approxi- mately seven miles north of the original collection. Here the plants were growing closely associated with D. laciniata Muhl. and D. diphylla Michx. The distribution of D. multzfida 1s an interesting problem in plant geography. Delaware County was covered by the Wisconsin ice sheet and plants must have come in after the final retreat of the glacier. But evidently northward migration 1ScHAFFNER, Jonn H. 1932. Revised Catalogue of Ohio Vascular Plants, Ohio Biological Survey 5: 89-215. 2 FERNALD M. L. 1950. Gray’s Manual of Botany, 8th Edition, American Book Co. 1956] Long,—Dentaria multifida in Central Ohio 161 of this species has been very limited, for the nearest recorded collection is from unglaciated Morgan County, in Southeastern Ohio. Montgomery? has shown that species of Dentaria belong to a sexually sterile, polyploid series. One can only speculate as to how this isolated, northern population became established. The data for the original collection are as follows:—wooded ravine along brook tributary to Alum Creek, 7.1 miles east of Delaware, near U. S. Route 36, Delaware Co., Ohio, May 12, 1931, C. E. O'Neal. Specimens from both areas are deposited in the Herbarium of Ohio Wesleyan.—Rosert W. Lona, JR., DEPT. OF BOTANY, OHIO WESLEYAN UNIV., DELAWARE, OHIO. A CORRECTION IN THE NAMES OF Two SPECIES oF HETERO- KONTAE—The genus Radiosphaera Pascher (1939, p. 549) of the order Heterococcales has recently been shown to be invalid by the careful work of Starr (1955, p. 48). Accordingly its two species R. sol Pascher and R. Nemiahi Croasdale must be placed in a validly published genus. Pascher, in 1932, consider- ing Radiosphaera a subgenus, classified his species R. sol in the genus Meringosphaera Lohmann (1903, p. 68), to which it can most naturally be returned. Likewise R. Nemiahi can best be assigned to this more inclusive genus, differing, with M. sol, from other species only in the more or less aequatorial distribu- tion of the spines. The following changes are therefore made: Meringosphaera sol Pascher 1932, p. 208, Fig. 14 (Radio- sphaera sol Pascher 1939, p. 550, Fig. 404). Meringosphaera Nemiaht comb. nov. (Radiosphaera Nemiahi Croasdale 1948, p. 279, Pl. 1118, Fig. 8). LITERATURE CITED CROASDALE, H. 1948. The flora of Penikese, seventy-four years after. II. Fresh and brackish water algae of Penikese Island. RHopoRA 50: 269-279. Loumann, H. 1908. Neue Untersuchungen über das Reichtum des Meeres an Plankton. Wiss. Meeresunters. Kiel, N. F. 7: 1-88. Pascuer, A. 1932. Zur Kenntnis mariner Planktonten 1. Merinosphaera und ihre Verwandten. Arch. Protisk. 77: 195-218. 1939. Heterokonten, in L. Rabenhorst, Kryptogamen-Flora Deutschlands, Österreich und der Schweiz 11: 1-1092. 3 MoNTGOMERY, F. H. 1955. Preliminary Studies in the Genus Dentaria in Eastern North America. RuHopora 57: 161—173. 162 Rhodora [Vor. 58 Srarr, R.C. 1955. A Comparative Study of Chlorococcum Meneghini and other Spherical, Zoospore-Producing Genera of the Chlorococcales. Indiana Univ. Publ. Sci. Ser. 20: 1-111. —HANNAH CROASDALE, DEPT. OF ZOOLOGY, DARTMOUTH COLLEGE, HANOVER, N. H. ELopEA CORRECT WITHOUT BEING CoNSERVED.—In the 8th edition of Gray’s Manual, the generic name Elodea Michx. (Fl. Bor.-Am. 1: 20, 1803) is retained with the statement ‘‘pro- posed for conservation," being supposedly a later homonym of Elodes Adans. (Fam. Pl. 2: 444, 1763), or Elodea Adans. ex Jussieu (Gen. Pl. p. 255, 1789), or Elodea Adans. ex Vent. (Tabl. Règne Vég. 3: 144, 1799). At the time the problem was dis- cussed by C. A. Weatherby (“On the nomenclature of Elodea,” Ruopora 34: 114-116, 1932), Elodes and Elodea were considered variant spellings of the same name, and Michaux’ genus had then to be called Anacharis. But as since modified, the rules treat the words as different names (International Code, 1952, Art. 82, paragraph 4 of examples after note 4). With support from this fact, we may consider Elodea as used by Jussieu and Ventenat merely an orthographic error (cf. paragraph 3 of ex- amples under Art. 82). By the former author it was mentioned only as a synonym, since he lists it under Hypericum with the comment that the genus “is perhaps to be divided into several, which are Ascyrum Tourn., Hypericum Tourn., Androsaemum Tourn., Elodea Adans., Knifa Adans.” Ventenat was merely elaborating Jussieu’s classification (the subtitle of his work is “Selon la Méthode de Jussieu”), and although Elodea appears as a distinct genus, it is not published as new, but is still credited to Adanson. Since this is clearly an orthographic error, the first valid publication of the generic name Elodea is that of Michaux, which therefore does not need to be conserved. It may be mentioned that Elodea is used in such recent critical European floras as Clapham, Tutin and Warburg’s Flora of the British Isles (1952) and Nils Hylander’s Nordisk Kärlväxtflora (1953).—Lrovp H. SHINNERS, SOUTHERN METHODIST UNIVERSITY, DALLAS, TEXAS. Volume 58, no. 689, including pages 103-134, was issued 28 May, 1956. AUG 20 1956 ido: did Farlow @rrenenee Limmanry Dodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS l RALPH CARLETON BEAN > Associate Editors RICHARD ALDEN HOWARD ( CARROLL EMORY WOOD, JR. Vol. 58 July, 1956 No. 691 CONTENTS: A Cytotaxonomic Study of the Genus Hymenopappus (Compositae). EHI T pr D p luris Seis ce tee 163 Neviusia alabamensis in Arkansas. Dwight Munson Moore...... 187 Humboldt and American Botany. Joseph Ewan............... 191 Asclepias syriaca var. kansana in New York State. Julian A. Steyermark and Floyd A. Suomk.......................... 197 The New England Botanical Club, Jne. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $4.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 40 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $4.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scien- tific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of publica- tion. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U. S. A. Rbhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 July, 1956 No. 691 A CYTOTAXONOMIC STUDY OF THE GENUS HYMENOPAPPUS (COMPOSITAE)! BILLIE L. TURNER Hymenopappus, an exclusively North American genus, has its greatest concentration of species in the United States. It has been conventionally placed in the tribe Helenieae of the family Compositae, a group of plants which is almost entirely North and South American. Of the 55 genera recognized for the tribe by Hoffmann (1897) in Engler and Prantl, only 3 occur on other continents. Thus the members of Helenieae should serve as excellent genera for monographic and evolu- tionary studies by American taxonomists, particularly since a considerable amount of the available herbarium material, including many types, may be obtained in this country. Equal- ly, if not more important, is the opportunity for extensive field and cytogenetic work with members of the group. "The genera in the tribe seem good subjects for chromosomal studies since both pollen mother-cell smears and root-tip squashes are easily made and excellent results are obtained. Since the naturalness of the tribe has been seriously questioned by Small (1919, p. 312) and more recently by Cronquist (1955), the accumulation of detailed knowledge of the included genera, particularly of evolutionary trends and chromosomal studies, is especially desirable. Comparatively few genera of the Helenieae have been studied 1 Part of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the State College of Washington. Post-doctoral field work was supported in part by the Plant Research Institute of the University of Texas. 164 Rhodora [Vor. 58 monographically, much less cytotaxonomically, in recent years.? The completion of a larger number of such studies, combined with information from other tribes, particularly the Heliantheae, should provide much evidence on the affinities of the included genera and on the naturalness of the tribe as a whole. In addition to this larger aspect of plant phylogeny and evolution, there are many problems at the specific level. Several genera of this tribe have occasioned considerable taxonomic confusion, particularly in the United States and Mexico. Hy- menopappus is such a genus. The natural groups of this taxon have been treated in various ways. Gray (1886) recognized 7 species; Rydberg (1914), 28 species (including three of Hy- menothrix); Johnston (1923), 14 species and one variety. The present account recognizes 25 taxa: 10 species, and 15 varieties (including one new species, and three new varieties). Most of the shuffling and reshuffling of names has been in the H. filifolius complex, a group of closely related taxa distributed principally in the Rocky Mountains, Great Basin, and adjacent plains and deserts. These infraspecific units (as treated in the present paper) occupy more or less distinct habitats and are, for the most part, allopatrie, although considerable overlap and intergradation often occurs peripherally with closely related taxa. This intergradation, combined with much spatial (but not altitudinal or ecological) overlap, has made the assembling of a great deal of herbarium material, seed for garden studies, and first hand field knowledge an absolute necessity for the understanding of the complex. The present treatment is an attempt to bring together this information so as to clarify many of the existing problems, as well as to show the relationships of the more easily recognized members of the genus. In general, the most valuable part of this information, aside from that obtained from herbarium specimens, has been obtained from field work and uniform garden cultures. Acquaintance with local and regional populations, their habitats, variability, and ranges, combined with later responses of representative indi- viduals in the garden, has been invaluable. Chromosomal 2 The following have been monographed or revised since 1930: Chaenactis (Stock- well, 1940); Eriophyllum (Constance, 1937); Gaillardia (Biddulph, 1944); Hymenoxys (Parker, unpubl.); Monolopia (Crum, 1940); Palaforia (Baltzer, 1944); Perityle (Everly, 1947); Schkuhria (Heiser, 1945). 1956] Turner,—A Study of the Genus Hymenopappus 165 information has proved to be significant in relatively few taxa, though much remains to be done along this line, particularly in the H. filifolius var. lugens complex. In some cases the problems raised by the author are more complex than those he attempted to solve, but remaining prob- lems are few, and it is believed that the present treatment to a large extent places the individuals in natural groups from the standpoint of morphology, geography, ecology, and probably phylogeny. HisTORICAL ACCOUNT The genus Hymenopappus was proposed by L'Heritier in 1788 to accommodate H. scabiosaeus, a species grown from seed that had been collected in Carolina by Michaux and sent to Europe. The original publication consisted of only three pages, one of these being a plate. Three years later Lamarck (1792) unaware of L'Heritier's description, redescribed the genus as Rothia, basing his name R. caroliniensis also on plants grown from seed collected by Michaux in Carolina.* From 1788 until Gray (1886) treated the group in his Synop- tical Flora of North America, seven of the ten species recognized in the present paper were described (one of these was included in the genus Leucampyx). Though Gray did not group the species of the genus into series, he clearly pointed out the dis- tinetions of these groups in his key to species. After Gray’s treatment a number of new species were pro- posed; nearly all of these fell into the immediate vicinity of H. filifolius, a highly variable species-complex of the western United States. This period was climaxed by Rydberg (1914) who described four new species for the genus and recognized, in all, 28 species and no varieties. In addition, he confused generic limits by transferring three species of Hymenothrix to Hymenopappus. 3 According to Britton and Woodward (1905) only a few of these sets (‘‘mono- graphs’’) were circulated, probably no more than five or twelve. In connection with the present study, two copies have been located, one in the library of the Linnaean Society in London, and the other in the library of the New York Botanical Garden. 4 Kuntze (1891) thought that L’Heritier’s genus did not appear until 1803, and argued that Rothia Lam. is the older name. Britton and Woodward (1905) give 1788 as the probable date for Hymenopappus. Since the generic name Rothia was already preoccupied at the time of Lamarck’s publication, according to the Inter- national Code of Botanical Nomenclature (1952), Hymenopappus is the name to be used for the genus whether the date 1788 is correct or not. 166 Rhodora [Vor. 58 Johnston (1923) gave a brief account of the genus in which he reduced most of the obvious relatives of H. filifolius to syn- onymy under that species. In many respects Johnston's treat- ment was like that of Gray, with the important exceptions of the reduction of Leucampyx to synonymy under Hymenopappus and the retention of Hymenothrix palmeri within the genus. 'The present treatment, at the specific level, is also like Gray's, but with several notable exceptions: Leucampyx is placed in synonymy under Hymenopappus; three species described as new since his time are considered to be valid; and H. filifolius has been treated as a species-complex composed of a number of recognizable geographic variants. In this latter respect this treatment is like that of Rydberg, most of his species being considered varieties in this paper. In addition, for the first time, a clear morphological distinction has been made between the genera Hymenothrix and Hymenopappus. CHROMOSOMAL STUDIES Because of the apparent variability and wide distribution of the members of the genus Hymenopappus, it was thought early in this study that chromosomal information might be of considerable help in unraveling taxonomic problems. It soon became apparent, however, that the number » — 17 was con- stant for nearly all members of the genus, the exceptions being two tetraploid varieties or races within the H. filifolius com- plex. Metuops.—Collections of buds and/or seeds for chromosomal studies were made in the field during the summer of 1951.5 Two different treatments were used for killing and fixing buds: either they were killed and fixed in Belling's modified Navashin's fluid directly (after removal of enclosing bracts), or they were fixed in Carnoy's fluid (9 parts absolute ethyl aleohol: 3 parts glacial acetic acid: 1 part chloroform) for 10 minutes then transferred to Navashin's fluid, a procedure suggested by Johan- sen (1940, p. 33). Buds collected in this manner were passed through a tertiary butyl alcohol series and embedded in paraffin. Sections were made at 15-20 microns, and the material stained with crystal violet. Prefixing of the material in Carnoy’s 5 Voucher specimens are deposited in the herbarium of the State College of Washing- ton; collections made during 1954 and 1955 are in the University of Texas Herbarium, 1956] Turner,—A Study of the Genus Hymenopappus 167 fluid before transferring to Navashin’s fluid gave best results, though, in general, neither treatment was very satisfactory for meiotic chromosomal studies. The bivalents tended to clump tightly together at metaphase obscuring individual chro- mosomes, although when meiotic material was found at late prophase good results were obtained. Seeds collected in the field were grown in a uniform garden and pollen mother cell smears were made from these plants. Though the species of Hymenopappus are perennial and biennial, it was found that by germinating the seeds in the greenhouse during December and January and transplanting to the garden in spring, flowering could be induced the same year. Seeds germinated after January usually produced only rosettes the first year. Freshly collected buds from garden plants were placed on slides and the anthers removed to a drop of aceto- carmine where they were macerated with a glass rod. Since the chromosomes tended to clump at metaphase of meiosis, considerable pressure on the cover slip was required to spread them sufficiently for counting. Less pressure is needed at late prophase, and most of the counts and drawings were made at this stage. In addition to smearing fresh material directly, buds were also fixed in a modified Carnoy’s fluid (4 parts chloro- form: 3 parts absolute alcohol: 1 part glacial acetic acid) and stored under refrigeration, a method suggested by Walters (1952). Material collected in this manner takes a better stain with acetocarmine than does similarly stained fresh material. Buds may be stored for at least six months in the fluid with good results, and probably much longer. Attempts to obtain counts and to determine chromosome morphology from root-tip material by standard embedding techniques was not rewarding. The root-tip chromosomes are extremely long, and because of their number (2n = 34), the metaphase plates appeared as a mass of interwoven chromo- somes. However, excellent results were obtained in the latter part of this study through the use of a method suggested by Speese and Baldwin (1952). Seeds were germinated on moist filter paper and the primary radicles excised after they had emerged about 6mm. These were placed in a saturated aqueous solution of paradichlorobenzene for two hours, washed in distilled 168 Rhodora [Vor. 58 e e⁄ e> 4, at D ST GB OS as & ve -— Š z 2 3 3€ | @ Fic. 1-12. Meiotic chromosomes of Hymenopappus species. Fig. 1. H. fili- folius var. parvulus (P. & T. 2802). Fig. 2. Same, first metaphase, crystal violet preparation. Fig. 3. MH. f. var. filifolius (P. & T. 2293). Fig. 4. H. f. var. nanus (P. & T. 2608). Fig. 5. H. f. var. lugens, white-flowered tetraploid race (P. & T. 2666). Fig.6. Same, yellow-flowered diploid race (P. & T. 2504). Fig. 7. H. new- berryi (T. 2894). Fig. 8. Same, second metaphase. Fig. 9. H. filifolius var. cine- reus (P. & T. 2830). Fig.10. H. flavescens var, canotomentosus (P. & T. 2670). Fig. 11. H. tenuifolius (T. 2895). Fig. 12. H. scabiosaeus var. corymbosus (S.N.F.). All X ca. 1000. 1956] Turner,—A Study of the Genus Hymenopappus 169 š " FR E pum S R f 275 4, >< 2%= 14 SNR gin’ "2j ty i rd à v *e x [ow Ue RA ise 2$ - x p 15 ch 13 EN 2 4 Be. ‘A of™ s, r x: ES > X xt , P A £ p f >< A: tf dues 559 i EF ae > ^f "Rp" IU! x N 2c ñ 2 f ot tyes <. f “ es Aka ay CS. ÉL 277 N l6 18 2 & E s ` az S E breed = ` 20 I9 Fic. 13-20. Mitotic chromosomes of Hymenopappus species. Fig. 13. H. fili- folius var. cinereus (P. & T. 2754). Fig. 14. H. flavescens var. flavescens (Warnock; Crane Co., Texas). Fig.15. H. filifolius var. nanus (P. & T. 2608). Fig.16. H.f. var. lugens, yellow-flowered diploid race (P. & T. 2645). Fig. 17. H. f. var. luteus (P. & T. 2872). Fig.18. H.f.var. polycephalus (S. V.F.; Golden, Colorado). Fig. 19. H. f. var, cinereus (P. & T. 2751). Fig. 20. Same, chromosome set aligned and traced from Fig. 21. All X ca. 1250. 170 Rhodora [Vor. 58 Chromosome Number TAXON SOURCE Root P.M.C. Tip H. artemisiaefolius TEXAS. Kleberg Co.: eastern Laureles Divi- 17 var. artemisiaefolius sion of King Ranch, M. C. Johnston 54411. (This plant somewhat intermediate to H. a. var. riograndensis). TEXAS. Colorado Co.: 5 mi. N.W. of Colum- 17 bus, Turner 3361. H. artemisiaefolius TEXAS. Kleberg Co.: S. end of Canelo Pas- 17 var. riograndensis ture, Santa Gertrudis Division of King Ranch, M. C. Johnston 54483. TEXAS. Hidalgo Co.: 5 mi. S. of Brooks 17 County line, U. S. Highway 281, M. C. Johnston 54528. JHI. filifolius COLORADO. Boulder Co.: 6 mi. N. of Boulder. 34 var. cinereus Q. Jones 835. *COLORADO. Chaffe Co.: 4 mi. W. of Salida, 17 Preece & Turner 2830. COLORADO. Gunnison Co.: Cimarron, Preece 34 & Turner 2793. COLORADO. La Plata Co.: 1 mi. N. of Dur- 34 ango, Preece & Turner 2754. NEW MEXICO. Sandoval Co.: 7 mi. N.W. of 34 Cuba, Preece & Turner 2751. *NEW MEXICO. Santa Fe Co.: 3 mi. S.E. of 17 Espanola, Preece & Turner 2739. *NEW MEXICO. Santa Fe Co.: 4 mi. S. of 17 Espanola, Preece & Turner 2741. H. filifolius OREGON. Morrow Co.: 2 mi. W. of Board- 17 var. filifolius man, Ownbey «& Preece 3366. OREGON. Wheeler Co.: 4 mi. E. of Mitchell, 17 Ownbey «& Preece 3362. *WASHINGTON. Adams Co.: 12 mi. W.N.W. 17 34 of Washtucna, Turner 2293. H. filifolius IDAHO. Lemhi Co.: 14 mi. S. of Salmon, 17 var. idahoensis Preece & Turner 2378. H. filifolius *ARIZONA. Coconino Co.: 9 mi, E. of Peach 17 var. lugens Springs, Preece & Turner 2610. *ARIZONA. Coconino Co.: 2 mi. E. of Ash 34 Fork, Preece & Turner 2618. ARIZONA. Coconino Co.: 7 mi. W. of Wil- 68 liams, Preece & Turner 2628. *ARIZONA. Coconino Co.: 10 mi. S. of Flag- 17 34 staff, Preece & Turner 2645. *ARIZONA. Coconino Co.: 5 mi. E. of Flag- 34 staff, Preece & Turner 2666. *ARIZONA. Yavapai Co.: 5 mi. W. of Selig- 34 man, Preece & Turner 2614. *NEw MEXICO. Catron Co.: 35 mi, N. of Alma, 34 Preece & Turner 2710. NEW MEXICO. Catron Co.: 15 mi, N. of Alma, 68 Preece & Turner 2711. vTAH. Kane Co.: 3 mi. W. of Long Valley 17 Junction, Preece & Turner 2466. *uTAH. Kane Co.: 1 mi. S. of Glendale, Preece 17 & Turner 2504. 1956 Turner,—A Study of the Genus Hymenopappus 171 Chromosome Number TAXON SOURCE Root P.M.C. Tip H. filifolius *WYOMING. Sweetwater Co.: 32 mi. S. of 17 34 var. luteus Green River, Preece & Turner 2872. H. filifolius *ARIZONA. Mohave Co.: 6 mi. E. of Peach 17 34 var. nanus Springs, Preece & Turner 2608. UTAH. Washington Co.: 7 mi. S.W. of Shiv- 17 wits, Preece & Turner 2558. H. filifolius uran. Kane Co.: 15 mi. N. of Orderville, 17 var. nudipes Preece & Turner 2499. H. filifolius *COLORADO. Gunnison Co.: 15.6 mi. W. of 17 var. parvulus Gunnison, Preece & Turner 2802. COLORADO. Saquache Co.: dry hills near 17 Elko, Preece & Turner 2816. H. filifolius ARIZONA. Coconino Co.: 20 mi. S.W. of 17 var. pauciflorus Marble Canyon, Turner 3066. H. filifolius COLORADO. Jefferson Co.: Golden, July 12, 68 var. polycephalus 1951, Rev. S. V. Fraser. COLORADO. Larimer Co.: 1.5 mi. E. of Liver- 34 more, Preece & Turner 2861. H. filifolius vTAH. Washington Co.: Zion National Park, 17 var. tomentosus near lower entrance to road tunnel, Turner 3065. H. flavescens *ARIZONA. Navajo Co.: 13 mi. S. of Holbrook, 17 var. cano-tomentosus Preece & Turner 2670. NEW MEXICO. Catron Co.: 8 mi. S.W. of 17 Horse Springs, Preece & Turner 2738. H. flavescens NEW MEXICO. Lea Co.: 2 mi. S. of Eunice, 34 var. flavescens Turner 2946. *NEW MEXICO. Lea Co.: 5 mi. N. of Eunice, 34 Turner 2947. *"rEXAS. Crane Co.: sand between Crane and 34 Odessa, June 15, 1952, B. H. Warnock. H. mexicanus ARIZONA. Coconino Co.: 10 mi. N. of Flag- 17 staff, Preece & Turner 2634. *ARIZONA. Coconino Co.: 4 mi. 8. of Flag- 34 staff, Preece & Turner 2641. H. newberryi COLORADO. Archuleta Co.: 1.5 mi. S.E. of ird Pagosa Springs, Turner 2894. *NEW MEXICO, Sandoval Co.: Jemez Mts., 17 5 mi. S.E. of La Cueva, Preece & Turner 2748. H. radiatus ARIZONA. Navajo Co.: 9.5 mi. N. of White IT River, Preece & Turner 2686. H. scabiosaeus *kansas. Cloud Co.: 2.5 mi. W. of Aurora, 17 var. corymbosus July 17, 1951, Rev. S. V. Fraser. TEXAS. Austin Co.: 2.5 mi. N. of New Ulm, 17 Turner & Johnston 54333. H. tenuifolius NEW MEXICO. Santa Fe Co.: 14 mi. S. of 17 Santa Fe, Turner 2895. TEXAS. Coryell Co.: 3 mi. N. of Cooperas 17 Cove, Turner 3807. *rExAs. Dallas Co.: Flagpole Hill, N. E. 34 part of Dallas, Beaman 519. * Collections grown in the garden are indicated by an asterisk (*). 172 Rhodora [Vor. 58 water, and stored overnight in a modified Carnoy’s fluid (3 absolute alcohol: 1 chloroform: 1 glacial acetic acid). Prior to squashing, the root-tips were hydrolyzed in acid alcohol (1 hydrochloric acid [sp. gr. 1.19]: 1 absolute alcohol) for 10-15 minutes and returned to the Carnoy's fluid. After 30 minutes to several hours the tips were removed and squashed in aceto- carmine. Staining of chromosomes was intensified when the slides were allowed to stand for one to three hours. Cover- slips were ringed with white Karo syrup for temporary mounts or made permanent by McClintock’s method (1929). Drawings were made with the aid of a camera lucida; in some instances mitotic chromosomes were spaced when these over- lapped. hESULTS.—No previous chromosomal counts have been reported for Hymenopappus. In the following list counts are given for 20 of the 25 taxa, including eight of the ten species in the genus. All the species of the genus investigated were found to have the base number of « = 17. Two varieties of H. filifolius are tetraploid or have tetraploid races: var. poly- cephalus (n = 34), and var. lugens (n = 17, 34). Meiotic chromosomes at late prophase are comparatively short with similar morphologies and nearly always with two terminalized chiasmata. In all the collections from which meiotic counts were made, pairing appeared normal except in some of the tetraploid races of H. filifolius var. lugens where two or three quadrivalents were found consistently (Preece & Turner 2614, 2618, 2666). Mitotic chromosomes are more distinctive, showing con- spicuous differences in both length and centromere position. One chromosome pair with extended satellites is particularly outstanding, and has been found in the somatic cells of all collections of Hymenopappus so far examined. No attempt has been made to make extensive chromosomal comparisons within the genus, and it is possible that such study might reveal significant karyologic differences among the taxa. Though chromosomal studies have not been of much help in showing relationships or affinities within Hymenopappus, they have shown that the members of the genus are held together naturally. This is particularly interesting with respect to 1956] Turner,—A Study of the Genus Hymenopappus 173 H. newberryi which was originally described as a different genus and placed in a separate tribe (Anthemideae). In the light of the chromosomal as well as morphologic evidence, there can be little doubt now that the species is properly placed in Hymeno- pappus. GENETIC TESTS Meruops.—Seeds collected during the summer of 1951 were germinated the following winter on moist filter paper in petri dishes and the seedlings grown in greenhouse pots. In spring these seedlings were transplanted to an outside garden. Ten plants of most of the collections were grown to maturity and the crosses indicated in Figure 21 attempted. The plants were first tested for self-sterility by bagging portions of the inflorescence before anthesis. If seed failed to form in these bagged heads, it was assumed that the species was self- sterile. Two methods were used in cross pollination: (1) the heads were bagged while in bud and after anthesis the pollen was blown or washed off and foreign pollen introduced by brushing an anther held in tweezers across the styles, and heads then re- bagged; (2) peripheral florets were emasculated, foreign pollen introduced, and unpollinated florets removed before rebagging. In either case, if seeds with embryos developed the cross was interpreted as a successful one. It was found, in rare instances, that what were thought to be carefully bagged heads of self- sterile plants set one or two seeds whether foreign pollen was introduced or not, and it is suspected that some of the taxa are not completely self-sterile. Hence, successfully formed embryos as indicated in Figure 21 may not be an entirely accurate indication of crossability. In most instances, however, the crosses represent formation of several mature embryos per head. Results.—All species tested were found to be self-sterile or nearly so except H. newberryi and H. scabiosaeus var. corym- bosus. In general, it seems that crossing between the several species and varieties tested, so far as embryo development is concerned, occurs readily. Unfortunately, time was not suf- ficient to grow the supposed hybrid seeds to verify these assump- tions, but the seeds have been saved for future study. 174 Rhodora [Vor. 58 SER PERENNES AL LU ^^ SER BIENNES H.F 4 Ç PARVULUS i " ] a ON \ fà HS GORYMBOSUS 17 SER PERENNES P4 NON Z “SER a BIENNES a EMBRYO FORMED — —— —— CROSSING FAILEO Fia.21. Diagram of specific and varietal crosses attempted within Hymenopappus. Solid lines indicate successful production of mature seeds with embryos; dashed lines failure. Double lines show that the cross was attempted reciprocally. Arrows point to the female parent. MORPHOLOGY AND TAXONOMIC CRITERIA It is difficult to assign values to particular characters, for a distinguishing character in one group may be a highly variable one in another. This is especially true at the varietal level where distinctions are usually made on a combination of fea- tures. The characters listed below were useful principally in the delimitation of specific categories; however, application in some instances could be made at the infraspecific level. Hanrr.—Species of Series Biennes are tall, mostly leafy- stemmed biennials; those of Series Perennes tend to be lower, less leafy perennials. 1956] Turner,—A Study of the Genus Hymenopappus 175 Roor.—Biennial species have a single obconical tap-root whieh normally bears but one crown at its apex. Perennial species have a woody tap-root which, in older plants, bears branches with several to many crowns. Leaves.—Most of the species have pinnately compound leaves, although two species commonly have simple, entire leaves. The degree of leaf dissection and size of ultimate segments have been particularly useful in delimiting varieties. INFLORESCENCE.—In all the species the inflorescence is a few- to many-headed cymose panicle. Biennial species usually have several leafy lateral branches which contribute to the many-headed inflorescence; perennial species are usually with- out leafy lateral branches, the inflorescence being composed almost entirely of peduncles produced from the main stem. Heaps.—Radiate and discoid heads appear in species of both the biennial and perennial groups, the discoid type being much more common. In the discoid species the texture and size of the involucral bracts are often distinguishing. ConoLL4.—' The shape of the disk corolla, particularly that of the throat, has proved significant in several instances, both at specific and infraspecific levels. Two principal throat types are found in the biennial species: funnelform throats in the species of the eastern United States, and campanulate throats in the central and western species. In general, funnel-form- throated taxa also have long narrow tubes; campanulate-throated taxa have shorter tubes. "The perennial taxa all have campan- ulate throats. The size of the throat varies considerably in the H. filifolius complex, but when correlated with other characters serves as a convenient marker. COROLLA coLom.— This character is an important one and should be included on collection labels when possible. A num- ber of taxa can be separated by this feature alone, thus making a comparison of several combined characters unnecessary. 'There are two common flower colors in the genus, yellow and white. The perennial species are nearly all yellow-flowered, but a few varieties of H. filifolius have white flowers or white- flowered races. The biennial species are principally white- flowered, but include three yellow-flowered taxa. In addition, one taxon (H. artemisiaefolius) may have a reddish or wine- colored corolla. 176 Rhodora [Vor. 58 PUBESCENCE.—With some exceptions, pubescence has not proved to be of much significance. In the perennial species a dense tomentum or woolly mat of hairs is usually found in the axils of the basal rosette leaves. The biennial species are without this dense tomentum or, if it is present, it is of a lesser amount. Within the species or variety the overall pubescence varies considerably, and it is suspected that this is a result of environmental influences in some instances. Seeds collected in the field from tomentose-leaved specimens of H. filifolius var. lugens (Preece & Turner 2710), when grown in the garden at Pullman, Washington, produced plants with nearly glabrate leaves. Glandular hairs are found on the corolla tube in nearly all species of Hymenopappus, but have been of practically no use in circumscription of species or varieties. On the other hand, achenial pubescence seems to be relatively constant, and has been used in combination with other characters to de- limit certain taxa. ACHENES.— The achenes are basically alike in the genus, differing principally in size, degree of pubescence, and length of pappus. In most species the achenes are obpyramidal with four sides, but in some there is a tendency for them to be in- curved and diagonally compressed, especially at the periphery of the heads. STYLE BRANCHES.— The styles are essentially alike in all the species, having obtuse flattened lobes with distinet marginal stigmatie lines on the upper surface and papillose tips. MEASUREMENTS Except in those taxa described as new, all flower measurements were made from dried specimens. Throat/lobe ratios were calculated by dividing the throat length by that of the lobe from the point of reflection (not the sinus). Since the number of florets in each head varies considerably, even on the same plant (earlier heads contain more florets than later ones), ter- minal heads and/or the first one or two successive ones were used for these counts. Obviously late-flowering specimens with abortive flowers and injured plants were excluded. Achene- pubescence was measured to the nearest tenth of a millimeter by use of an ocular micrometer. Measurements were taken 1956] Turner,—A Study of the Genus Hymenopappus 177 y. papp from those hairs at the middle of a corner of the achene. With some exceptions, achenial pubescence proved fairly constant in any one taxon. Since leaf segmentation has frequently been useful in drawing varietal lines, especially in the H. filifolius group, a particular part of the leaf was needed for measurement so as to compare the various taxa uniformly. The terminal segment at the apex of the leaf has therefore been selected for the linear measure- ments used in the keys (but not in descriptions). Leaf silhouettes were made in the following manner: Leaves were collected fresh and pressed flat. When dry, these were inserted between sheets of white paper and a soft lead pencil used to shade the top sheet so as to reveal the exact outline of the leaf below. This outline was then filled in with India ink and the pencil smudge erased. SPECIFIC CONCEPTS In the present paper the natural units of Hymenopappus have been delimited by the combined information obtained from morphology, cytology, ecology, and geographic distribution. An attempt has been made to treat the taxa consistently with respect to their designation as species, variety, or population. In some instances distinct populations have been pointed out as perhaps worthy of varietal rank, but because of the complexi- ties involved, and the further studies needed to delimit these groups properly, nomenclatural status has not been assigned. The use of the term ‘variety’ for infraspecific units follows the usage implied by the International Code of Botanical Nomen- clature (1952, Article 14 and Appendix 3, Article C.5). The author believes this is desirable since the term has become ingrained in botanical usage and classification. In addition, the use of “subspecies” for the taxa here designated as varieties would not leave available a recognized term for the grouping of these infraspecific taxa, a course which has not been followed in this study but which may become advisable with accumula- tion of more information. GENERIC RELATIONSHIPS AND TAXONOMIC POSITION The species of Hymenopappus, as here delimited, make up a 178 Rhodora [Vor. 58 closely knit, natural group whose members are more similar to each other than to any known member of another genus. The total similarity of corollas, achenes, leaf patterns, and general aspect in the included taxa is truly remarkable, especially when one considers that the genus includes everything from perennial plants with chaffy receptacles and rayed heads to biennials with neither chaff nor rays. Some workers (Hoffman, 1897; Rydberg, 1914; Johnston, 1923) have included various species of the genus Hymenothrix in Hymenopappus, mostly on the basis of pappus morphology, or other achenial characters. Hymenothrix was proposed by Gray (1849) to accommodate H. wislizenii, an annual species of southern Arizona, New Mexico, and adjacent Mexico. Sub- sequently four other species from the same general area were added to the genus by various authors. Heretofore, the chief character used to distinguish the two genera was the excurrent awn or midrib on the pappus scales of Hymenothrix as opposed to the included midrib on those of Hymenopappus. However, it was found that Hymenothrix palmeri may or may not have awns extending from its pappus scales and, consequently, it was thought that this situation destroyed the distinctions be- tween the genera. Accordingly, Rydberg (1914) distributed the species of Hymenothrix in three genera: He recognized H ymenothrix as monotypic; established a new genus Trichymenia for one of the species (H. wrightii); and put the remaining species in Hymenopappus. Johnston (1923) accepted Rydberg's treat- ment stating, “I am following the only natural alternative of recognizing Hymenothrix and Trichymenia. The generic lines as here taken are very weak and those that maintain them must find their arguments in precedence and convenience.” There are several morphological characters which distinguish Hymenopappus from Hymenothrix (including T'richymenia), and so far as is known, the two genera are separated by discon- tinuities which are not bridged. Chromosomal studies should offer additional evidence bearing on this problem, but it is felt that the morphologie differences alone are sufficient to distinguish the two as genera, and it is doubted that they are even closely related. Inasmuch as it destroys the coherence of both groups, Rydberg’s treatment is highly artificial. The 1956] Turner,—A Study of the Genus Hymenopappus 179 important distinctions between these genera are enumerated below: HYMENOPAPPUS HyMENOTHRIX . Disk corollas regular, the lobes all 1. Disk corollas zygomorphic, the alike. lobes unequal (inconspicuously so in H. wislizenii). . Leaves bipinnately compound to 2. Leaves biternately or triternately simple. compound. . Plants biennial or perennial. 3. Plants annual (?). . Corolla lobes (after anthesis) 4. Corolla lobes erect or merely strongly recurved. spreading, not recurved. . Pappus seales with medial nerve 5. Pappus scales with medial nerve completely included. extending to the apex or excur- rent into a distinet awn. The author has made no attempt to separate Hymenothrix from Florestina or other related genera; that is a problem for the future, but the exclusion from Hymenopappus of all species of Hymenothrix, and the retention of these species together in the same genus (including Trichymenia) seems sound on the evidence available at present. Hymenopappus apparently has no immediate relatives within the Helenieae. The majority of the species of the genus are clearly derived from an ancestor with rays and chaff, perhaps not too unlike H. newberryi. Since such a species would not key to or fall naturally within the circumscription of the Helen- ieae, it would seem logical to look to other tribes for its relatives. In this connection, it is interesting to note that Hoffman (1897) in Engler and Prantl’s world treatment of Compositae included H. newberryi, as the genus Leucampyx, in the tribe, Anthemideae, as did Gray (1877). Bentham and Hooker (1873)* also placed Leucampyx in the Anthemideae stating, evidenter Anthemidi quam maxime affine, different patria, receptaculo ut videtur parum convexo, achaeniorum magnitudine et forma, et styli ramis minus truncatis." Thus it seems likely that Hymenopappus as a genus has its closest relationship with members of the tribe Anthemideae or perhaps some as yet uninvestigated member of the tribe Heliantheae (many of the characters of H. newberryi might place its ancestors here). This does not mean that the tribe “Genus ê Leucampyx was first published and credited to Gray by these authors, but no specific name was proposed. 180 Rhodora [Vor. 58 Helenieae in general is not a natural group, but rather that Hymenopappus appears to be not as closely related to members of the Helenieae as to members of the tribes mentioned above. DISTRIBUTION, ECOLOGY, AND POSSIBLE ORIGIN Hymenopappus occupies an extensive area in the western and southwestern United States and central Mexico. The genus 'an be divided into two well marked geographical and mor- phological groups: Series Perennes, the principal members of which occupy the mountainous areas of the western United States and central Mexico, and Series Biennes, the principal members of which occupy the plains and lower areas to the east and southeast of the Rocky Mountains. The species of Series Perennes appear to show, for the most part, a progressive adaptational trend toward xerophytism. The more primitive members of this series (see discussion below), H. newberryi and H. radiatus, have restricted distributions, and occupy more mesic situations (Pinus ponderosa to Picea engelmannii zones) than do the obviously more advanced taxa of the H. filifolius complex, the members of which occupy more xeric habitat types (Pinus ponderosa zone and lower). Species of Series Biennes may, in turn, be grouped into two main morphologieal and geographical types: (1) a group of taxa having bipinnately dissected basal rosette leaves and corollas with campanulate throats, these taxa ranging from the central Great Plains to western Texas and Mexico along the eastern periphery of the range of Series Perennes; (2) a group of taxa having less dissected basal rosette leaves and corollas with funnelform throats, ranging along the Gulf Coastal Plain to Florida, and northward up the Mississippi Valley to Kansas, Illinois, and Indiana. As in Series Perennes, the most primitive member of this group, H. biennis, is a rayed species. It occurs in the higher mountains of southern New Mexico and adjacent Trans-Pecos Texas, and commonly is associated with Pinus ponderosa, Pseudotsuga taxifolia, or Picea engelmannii. H. biennis seems to represent a surviving species or close relative of the ancestral stock which gave rise to the Biennes. The campanulate- throated members seem not to have migrated far from their 1956] Turner,—A Study of the Genus Hymenopappus 181 ancestral source, but have become adapted to more xeric con- ditions, developing (or retaining) bipinnately dissected leaves. The funnelform-throated groups appear to have branched off early enough from this ancestral biennial stock to retain less dissected, often simple leaves, perhaps indicating rapid spread and adaptation to more mesic situations to the southeast of the present-day Rocky Mountains. Combining geographical, ecological, and morphological evi- dence with the apparent evolutionary trends of the major groups it is possible to suggest a plausible history of the genus as concerns speciation, migration, and present day distribution. In order to understand the evolutionary trends in the genus it is necessary to reconstruct the appearance of the probable ancestral stock. This has been done by pooling all the assumed primitive characters now found in the various taxa. Such a procedure is justifiable on the assumption that the group is monophyletic and that the species existing today have retained some primitive characters while losing or modifying others. The critical part of this reconstruction is distinguishing between those characters which are primitive for the genus, and those that are derived. The following is a list of what the author considers primitive and corresponding derived characters of the genus Hymenopappus. Primitive Derived 1. Receptacle chaffy 1. Receptacle non-chaffy 2. Ray flowers present 2. Ray flowers absent 3. Root perennial 3. Root biennial 4. Leaves simple 4. Leaves dissected 5. Stem leafy 5. Stem subscapose 6. Inflorescence few-headed 6. Inflorescence many-headed 7. Achenes pubescent 7. Achenes glabrate 8. Involucral bracts mostly 8. Involucral bracts mostly non-membranous membranous 9. Pappus scales well developed 9. Pappus scales much reduced or absent 10. Receptacle large, convex 10. Receptacle small, nearly flat 11. Self-fertile 11. Self-sterile From this tabulation, the reconstructed ancestral species of Hymenopappus should be, in general, a leafy-stemmed perennial with simple leaves, having a few-headed inflorescence, ray florets, and chaffy receptacle (see diagrammatic scheme of 182 Rhodora [Vor. 58 CORYMBOSUS H.S. SCABIOSAEUS NEWBERRY! < — | FILIFOLIUS COMPLEX /, / — PLIOCENE — — MIOCENE Ei o o 5 $ & ë CIT — — OLIGOCENE LEAFY, RADIATE, CHAFFY, ANCESTOR Fig, 22. Schematic representation of the suggested origin and evolution of present- day Hymenopappus species. phylogeny, Fig. 22). No single species of present day Hymeno- pappus has all the primitive characters listed above, but H. newberryi retains seven of them (1, 2, 3, 6, 8, 10, 11), and is probably the least advanced or most primitive member of the genus. The two most primitive members of the Perennes, H. new- berryi and H. radiatus, and the most primitive member of the 1956] Turner,—A Study of the Genus Hymenopappus 183 M pap} Biennes, H. biennis, are restricted in their ranges, occupying the forested mountains of southern Colorado, east-central Arizona, and southern New Mexico (Fig. 24). These species (and those derived from them) seem not to have their immediate affinities with the present-day Composite genera of western North America; indeed, if past students of the Compositae were correct in assigning H. newberryi to the tribe Anthemideae, its nearest relatives should be in Eurasia, where most present- day members of the Anthemideae are concentrated (Bentham, 1873). Whether this is true or not, it seems likely that these species have come from the north via the Rocky Mountains. This is attested by the following facts: (1) The most primitive members of the genus are associated with a number of temperate mesophytic species, perhaps repre- senting relic members of the once more widespread Arcto- tertiary Flora. (2) Clearly more reduced species have apparently diverged from a southern Roeky Mountain center westward, southward, and southeastward (it would be most difficult to read this specia- tion in a reverse direction, as this would call for progressive development of rays and chaff—characters which nearly all workers in the Compositae agree are primitive). | (3) Laek of closely related genera to the south (absence of such groups to the north can possibly be explained by the de- struction of related genera, if such existed, by progressive cli- matic change since Oligocene time and the Pleistocene ice ad- vances). f Thus it seems possible that the ancestral stock of present day Hymenopappus migrated southward down the uplifting Rocky Mountain chain in Oligocene time, possibly along with other members of the Arcto-tertiary Flora (Chaney, 1947). The Miocene epoch saw the break up of this ancestral stock into several well-developed lines. Biennial taxa developed and spread southward to central Mexico and eastward to southern Texas. The funnelform-throated species seem to represent the terminal members of this imagined expansion, and appear to have developed in Pliocene time, and spread secondarily from southern Texas, occupying the Gulf Coastal Plain and inland habitats of the upper Mississippi Valley (evidenced by 184 Rhodora [Vor. 58 the considerable variability and presence of assumed primitive members in southern Texas; cf. H. artemisiaefolius var. rio- grandensis). The campanulate-throated biennial species, as already noted, remained mostly peripheral to their supposed ancestral center, and have retained many of the characteristics of H. biennis. Specifically, it seems likely that H. tenuifolius became adapted to the arid plains east of the Rocky Mountains, possibly in Pliocene time, and has since remained associated with the grasslands of this region; while H. flavescens var. cano-tomentosus became adapted to the more southern intermontane xeric habitats of Arizona, New Mexico, and Trans-Pecos Texas. Hymenopappus filifolius of the Perennes series is the most complex and widespread species of the genus. It is likely that its range, infraspecific variation, and complexity is a reflection of the climatic vicissitudes and geologic events of the region in which it 1s now located. Pliocene time saw the development of extremely arid condi- tions in the Great Basin area west of the Rocky Mountains (Axelrod, 1948). Thus, one might visualize the spread of some ancestral Hymenopappus filifolius stock from a more mesic habitat type of southern Colorado and northern New Mexico to this newly formed desert basin. With the stimulus of aridity, rapid evolution may have occurred, as suggested by Stebbins (1952). It seems likely that once a trend toward xerophytism had become established, rapid evolution and diversification might occur, especially if the area occupied is one of much climatic change and geologic unrest. The situation in the present- day H. filifolius with its numerous, complex and intergrading infraspecific taxa was perhaps what Axelrod (1948; pp. 134, 135) had in mind when he made the following statement con- cerning Great Basin environments: Furthermore, under the impact of fluctuating Pleistocene climates these environments were in a state of continual flux. In such areas a wealth of new entities may have arisen in latest Cenozoic time, further dif- ferentiating from their mid-Pliocene ancestors. Many new forms may thus have evolved in latest Cenozoic time due to the selective influence of rapidly developing and fluctuating localized environments which became successively more narrowly limited in extent. This apparently accounts far the wide ecological diversification of these herbaceous types which show a high degree of adaptation in the utilization of geologically recent, highly localized habitats. Differentiation of the various cate- 1956] Turner,—A Study of the Genus Hymenopappus 185 gories has been limited largely by the genetic variability of the popula- tions and by the diversity of localized environments. Nonetheless, the relatively short time that these highly-localized environments have been in existence may well account for the low degree of differentiation of many of the categories, and for the rapid evolution that is con- tinuing today. It seems likely that H. filifolius owes much of its infraspecific variation to such a history, along with that variation super- imposed by subsequent hybridization and introgression. There remains to be considered Hymenopappus mexicanus of the Perennes series. It apparently represents an early off- shoot of the same reduced line. This species does not seem to have given rise to any other present-day taxon, though consid- erable variability is found in the many populations composing it. In review, the author suggests the following history of the genus Hymenopappus based on the evolutionary trends as gleaned from the evidence offered by present-day morphology, ecology, and distribution (cf. Figs. 22, 23). (1) Migration of an ancestral stock southward down the rising Rocky Mountains in the Oligocene Epoch. (2) Separation of this stock in Miocene time into two main evolu- tionary lines, Biennes and Perennes. (3) Rapid spread of reduced biennial species southeastward from the Rocky Mountains in Pliocene time with adaptation of the more distant members to coastal, somewhat mesic, habitats; populations closer to the parental source of the Biennes became adapted to relatively xeric grass- land habitats. (4) Ancestral reduced stock of H. filifolius became adapted to the newly formed desert basin west of the Rocky Mountains in Pliocene time, and, as a result of aridity, climatic, and geologic disturbance in Pleistocene time, the species produced many variable and intergrading infraspecific taxa which became widely distributed over the Rocky Mountains, Great Basin, and peripheral areas. (5) Funnelform-throated biennial species have dispersed secondarily from southern Texas to the north and northeast in Pleistocene time. (6) Present-day ranges show the confinement of primitive species to relatively high, more mesic habitats of the southern Rocky Mountains; advanced reduced species are widespread and occupy more xeric habitat types, except where secondary adaptation and speciation has occurred. In addition to the broad outline given above, more specific information concerning ranges, ecology and, in some instances, possible origin has been given in the discussion for each of the various taxa concerned. ACKNOWLEDGEMENTS 'The author is especially indebted to Professor Marion Ownbey 186 Rhodora [Vor. 58 m H. a. artemisiaefolius "a H. a. riograndensis H. s. scabiosaeus H. s. corymbosus r- H. f. flavescens [ Biennes H. f. cano-tomentosus — H. flavomarginatus H. biennis — H. tenuifolius — H. f. polycephalus Hymenopappus 4 H. f. cinereus H. f. luteus H. f. pauciflorus l . tomentosus m . megacephalus eriopodus lugens nanus . idahoensis . filifolius . nudipes . parvulus fllifolius Perennes —| — CEEEEEEE oh PR h — mnm -= . mexicanus . radiatus . newberryi mm m Fig. 23. Hypothetical relationships in Hymenopappus. Arrows point to taxa of possibly hybrid origin; tail of arrow refers to putative parents; width of shaft indicates the relative amount of characters contributed, who suggested the problem and under whose direction the work was carried out. Grateful acknowledgement is also made to Rev. S. V. Fraser, Dr. Q. Jones, and Professor B. H. Warnock for seed collections; to Miss Nell C. Horner, Librarian of the Missouri Botanical Garden, and Professor L. H. Shinners for bibliographie information; and to Professor A. Cronquist for obtaining a photocopy of the original publication of Hymeno- pappus scabiosaeus and for his general interest in the problem. (To be continued) 1956] Moore,—Neviusia alabamensis in Arkansas 187 NEVIUSIA ALABAMENSIS IN ARKANSAS Dwicut Munson MOORE In late April, 1955, the writer’s attention was called to some strange shrubs growing on a low sandstone ridge about six air- miles northwest of Conway, Arkansas, in the southeastern corner of Conway County, next adjacent to Faulkner County in which Conway lies. These were growing in some quantity along the sandstone ridge, extending some two hundred yards. Specimens of the plant were collected and, although past their blooming period, were easily identified as Neviusia alabamensis Gray. Search in the University of Arkansas Herbarium revealed four other sheets collected in 1925 by D. Demaree along the cliffs of Cove Creek, in the northern part of Faulkner County, and identified then as Physocarpus. These also proved to be Neviusia alabamensis Gray. For nearly one hundred years, Neviusia alabamensis, a mono- typic species of Rosaceae, has been generally considered a very definite endemic in the state of Alabama. It had been found in 1857 by Drs. R. D. Nevius and W. S. Wyman along the cliffs of Black Warrior River a few miles from Tuscaloosa, Alabama. It was sent to Dr. Asa Gray who described it (1858) as a new genus and species, naming it for one of the discoverers, Dr. Nevius. Within the next fifty years other stations for this same shrub were found in three other counties of northern Alabama, but it was still considered a very limited endemic for that state. It has thus been referred to by Chapman (1884), Mobr (1901), Harper (1906), Small (1913, 1933), Rydberg (1918), and Bailey (1935). In 1918, Dr. J. C. Th. Uphof, while making an extended ecological study in southern Missouri, found a single specimen of this shrub about eight miles west of Poplar Bluff, Missouri. He reported this in 1921 in a German publication, and in 1922, while writing up the ecology of southeastern Missouri, men- tioned finding this plant. Unfortunately, his specimens were destroyed before he got them back to U. 8. herbaria from Hol- land, and little attention was paid to his find. As Dr. Uphof said, it was apparently an adventive, probably having been started from seed dropped by birds from Alabama or some [Vor. 58 Rhodora 188 Neviusia alabamensis Gray. 1 ! J£g -*] —— Mery Me Fic. 1. Map of southeastern United States showing location of collections of Neviusia alabamensis Gray at 1. Tuscaloosa (1857) (type locality); 2. other northern Alabama counties; 3. Poplar Bluff, Mo. (1918) 4. Cove Creek, Ark. (1925); Conway Co., Ark. (1955). Continuous line indicates margin of Coastal Plain. hitherto unknown location in Tennessee. However, Harper (1928), in his Catalogue of the Trees, Shrubs, and Vines of Alabama, acknowledged the new discovery and mentioned it as the only place outside of Alabama where this plant had been found growing wild. In 1931 Fernald called attention to the relationship of this unique plant to its nearest relative, Kerria, in eastern Asia, and Bailey (1935) refers to it as very limited in Alabama. Palmer and Steyermark (1935) apparently did not consider the report of this one specimen sufficient for including it in their Catalogue of the Flowering Plants of Missouri, or they doubted Uphof's identification. Comparison of Uphof's drawing, here reproduced, with Gray's original drawing (1858) seems to leave no doubt as to his accuracy. Even as late as 1944, Cain refers to Neviusia as an example of extreme local endemism, showing it on a map as located only at Tuscaloosa. With two Arkansas stations for this unusual plant, it seemed 190 Rhodora [Vor. 58 logical to pursue further the reported find in southeastern Missouri. Search in July, 1955, in the region indicated west of Poplar Bluff, failed to reveal any evidence of the plant. However, this is not surprising, when consideration is given to all the changes of roads, fields, and other marks of 'civiliza- tion.’ Furthermore, it may still be possible to find this or more plants of the same kind if it is sought in the spring when it would be in flower. This now seems to make a very definite picture. By placing these locations on a map (fig. 1) showing the Mississippi embay- ment of the old Gulf Coastal Plain, it may be seen that the orig- inal find lay on the eastern side of the embayment, while the one from Missouri and the two from Arkansas are on the north- western side of this embayment. In Alabama they are reported on sandstone, shale, and limestone. In Missouri it was reported on sandy loam, while in Arkansas the finds were both on sand- stone outcrops. In all cases they appear to be on carboniferous or subearboniferous formations. Geological evidence points to the fact that this particular area has been exposed above the encroachments of the sea since the late Paleozoic or early Meso- zoic. This might lead us to believe that Neviusia alabamensis would be classified as an epibiotic, a relatively old relic species, according to Cain (1944), rather than a strict endemic which he would consider to be a relatively youthful species. At least, Neviusia alabamensis may no longer be considered to be strictly an Alabama endemic, but like a few other species it must be shared with other similarly located states. Credit should be given to Mr. Marvin Lawson of Conway who had seen these plants there and recognized them as something unusual and who called them to the attention of the writer. Thanks are also extended to Dr. Hugh Iltis for his encourage- ment in the pursuit of this unusual find, and to Drs. Uphof and Harper who have aided in the location of some old records which have helped materially in getting this more nearly com- plete picture of Neviusia alabamensis. At Poplar Bluff, Mo., Mr. Frank Hearne was very coóperative in the attempt to relocate Uphof’s station.— UNIVERSITY OF ARKANSAS, FAYETTE- VILLE, ARKANSAS. 1956] Ewan,—Humboldt and American Botany 191 BIBLIOGRAPHY Barney, L. H. 1935. Standard Cyclopedia of Horticulture. (Illustr.) Cain, SrANLEY A. 1944. Foundations of Plant Geography. Harper & Bros. (Map) CHAPMAN, A. W. 1884. Flora of the Southern United States. 2nd Ed. p. 121. FERNALD, M. L. 1931. Specific Segregations and Identities in some floras of Eastern North America and the Old World. RHODORA, 33: 29 (Map 3). Gray, Asa. 1858. Neviusia, a new genus of Rosaceae.—Mem. Am. Acad. Arts & Sci., II. 6: 373-376, pl. 30. Harper, R. M. 1906. Notes on the distribution of some Alabama Plants. Bull. Tor. Bot. Club, 33: 532. . 1928. Catalogue of the Trees, Shrubs & Vines of Alabama. Monogr. 9. Geol. Surv. of Ala. p. 194-196. (Illustr.) Monm, CHARLES 1901. Plant Life of Alabama. Vol. VI: Contr. U. 5. Nat. Herb. U. S. D. A. (Illustr.) PALMER, E. J. and J. A. SrEvERMARK 1935. Annotated Catalogue of the Flowering Plants of Missouri. Ann. Mo. Bot. Gard., 22: 375-758. RYDBERG, Per AXEL 1918. North American Flora. 22: 481. SMALL, J. K. 1913. Flora of Southeastern U. S., p. 524. 1933. Manual of the Southeastern Flora, p. 625. (Illustr.) Urnor, J. C. Ta. 1921. Das Vorkommen von Neviusia alabamensis Gray im Suden von Missouri. Mitteilungen der Deutschen Den- drologischen Gesellschaft. p. 282-3. (Illustr.) 1922. Ecological relations of plants in southeastern Mis- souri. Amer. Jour. Bot. 9: 7. HUMBOLDT AND AMERICAN Borany.'—‘‘Alexander, that is a beautiful name. I seem to recall an earth-conqueror by that name. Do you wish to be a conqueror?” “Yes, Sire,—but with my head." So replied Alexander von Humboldt, aged eight, to King Frederick the Great. At the end of a long life there arrived in Berlin in 1857 an American, Bayard Taylor. He remarked, “I came to Berlin not to visit its museums, and galleries, its operas, its theatres, not to mingle in the gay life— but for the sake of seeing and speaking with the world's greatest living man—Alexander von Humboldt." For Humboldt had seen both Bogotá and Baltimore, had conversed with Indian tribes and that student of Indian vo- 1 HUMBOLDT. THe Lire AND TIMES OF ALEXANDER VON HuMBOLDT, 1769-1859. Helmut de Terra. Alfred A. Knopf, N. Y. i-xvi, 1-386, i-ix. 8 plates, 3 maps. 1955, 192 Rhodora [Vor. 58 cabularies, Thomas Jefferson, had climbed many of the highest peaks in tropical America and studied their faunas and floras. The historian George Bancroft wrote, “Humboldt was always the friend of young America." The latest biography of “the greatest scientific traveller who ever lived," as Charles Darwin characterized Humboldt, has just been written by the geographer, Dr. Helmut de Terra, and beautifully produced by Alfred A. Knopf. The author has turned over materials that have long slept in Marburg and Paris, as well as the familiar sources in this country, but it is rather of some adjacent pastures still fallow that I should like to direct attention. In 1804 at the age of thirty-five Humboldt paused in this country for eight weeks en route from Cuba to Paris, that is, at the conclusion of his great hegira to tropical America. In Philadelphia Humboldt particularly visited Prof. Benjamin Smith Barton, Benjamin Rush, and Charles Willson Peale. Peale’s Philadelphia Museum, the largest of its kind in the New World, tremendously interested the young Humboldt: He spent three weeks at Jefferson’s Monticello, talking, walking, going over many matters of mutual concern. He spoke to Jef- ferson of having Congress purchase the museum amassed by Peale and thus establish a national collection. It was late May when the party of six, conducted by Peale, set out for Baltimore and Washington by coach. The other four were Aimé Bonpland—familiar to botanists for the abbreviation “H. B. K.” which trilogy records the part played by the nephew of Humboldt’s former tutor, Karl Sigismund Kunth, who did not accompany Humboldt and Bonpland—, Don Carlos Montu- far of Quito, Dr. Anthony Fothergill, and the Rev. Dr. Nicholas Collin. The pilgrimage was not without its speechless moments. Peale’s “dental bridge having broken, he rushed out to look 2 For the tragic fate of Peale’s historic specimens, another ‘short chapter in the history of human stupidity,’ see Thomas Barbour, A NaATURALIST'8 SCRAPBOOK, 91—97. 1946. 3 Anthony Fothergill (1733-1813), M. D., nephew of John Fothergill, came to America about 1803. Cf. R. Hingston Fox, Dn. Jonn ForHERGILL AND His FRIENDS, 131—132. 1919. 4 Nicholas (or Nils) Collin (1746—1831), student at Upsala, who arrived at Penns Neck, N. J. in 1778 and became rector of Swedish Churches in Pennsylvania. Aside from collecting seeds and plants and sending them to his native Sweden, he was interested in lead-glazing, philology, colonial history, climatology, etc. Cf. Th. Krok, Binuro. Bor. SuECANA, 131, 1925; G. G. Simpson, Amer. Puiros. Soc. 86: 130. 1942, and W. J. Bell Jr., ÉARLY AMERICAN SCIENCE, 4, 15, 1955. 1956] Ewan,—Humboldt and American Botany 193 for a gunsmith to borrow tools to repair it. For one who had mounted the skeleton of a mammoth, this was a fast and easy job. Luckily he carried a few pieces of gold in his pocket, and so managed to mend his artificial teeth in time to rejoin the company for the rest of the evening.” Dinner at the Executive Mansion, visits with the first architect of the new capitol in Washington, Dr. William Thornton,’ a conference with Secretary of the Treasury, Albert Gallatin, and a boat trip down the Potomac to Mt. Vernon (where now only Washington’s negro body servant lived to tell his master’s story first-hand), filled his days and evenings. Of Bonpland, then twenty-nine years of age, we should particularly like to know more, but he was eclipsed by Humboldt’s overwhelming importance and handicapped by his inability to converse in English. The two visitors called on Henry Muhlenberg at Lancaster and the “American Linnaeus” packaged 160 grass specimens for Humboldt to compare with Michaux’s collections at Paris. Humboldt’s letter, not mentioned by de Terra, is of special interest to the botanist, it reads: My dear and honored friend, I use these last moments before my departure tomorrow, to express once more my heartfelt thanks to you for the great kindness which you showed me and my friends in Lancaster. Your grasses and your kind letter have come to hand and I promise to let you know Michaux’s names for them. Bonpland commends himself to you with gratitude for your kind remembrance. Yesterday he had a delightful noon with your brother the worthy General. Remember me to the good Ellicott® and Mr. William Barton; We saw the Hamilton garden with aston- ished delight. Yours most gratefully Humboldt’ Philadelphia the 27th June 1804 5 William Thornton (1761-1828), M. D., a close friend of Jefferson, who evidently owned a set of plant drawings. He was a founder and one-time curator of the Co- lumbian Institute for the Promotion of Arts and Sciences. Cf. E. M. Betts, THOMAS JEFFERSON'S GARDEN Book, 1766-1824, 398 et passim, 1944. 6 Andrew Ellicott (1754—1820), civil engineer, who while laying out state boundaries collected plants unfamiliar to him. A few of these are preserved in the B. S. Barton Herbarium (ANSP). Cf. W. J. Bell Jr., op. cit., 55—56. 7 William P. O. Barton (1786-1856), who succeeded his uncle, B. S. Barton, as professor at the University of Pennsylvania. Cf. F. W. Pennell, BAnTONIA no. 21: 45. 1940. s Paul A. W. Wallace, MUHLENBERGS OF PENNSYLVANIA, 313. 1950. 194 Rhodora [Vor. 58 Thus we have one more distinguished visitor to the “Woodlands”! Naturally it is of particular interest to this reviewer to note that Frederick Pursh was head gardener at Hamilton’s estate at the time of Humboldt’s visit. When is some member of the Phila- delphia Botanical Club reconstructing William Hamilton's guest book? From the good beginnings made by Sarah P. Stetson? and others this is now a practicable project. Audubon's friend and associate, Rev. John Bachman of Charleston, was present at the dinner given in honor of Hum- boldt at Peale’s Museum. Bachman was barely fifteen at the time and was probably invited through the influence of Alex- der Wilson. the two Bartrams, Wilson, the ornithologist, Lawson, his engraver, George Ord and a few others . . . Few speeches were made and those were short—there was no formality . . . Humboldt was then, as he was after- wards, in every society, “the observed of all observers" . . . [Bachman] saw him every day during the few days he remained in Philadelphia. He inserted my name in his note-book, and for the last sixty years we cor- responded at long intervals.! But when Humboldt finally sailed in July aboard the frigate Favorita down the Delaware for Bordeaux it was of Jefferson and those delightful weeks at Monticello that he recalled with special pleasure. It is not easy thoroughly to assay the in- fluence of Humboldt's visit on Jefferson's great plan for the exploration of the West, so well advanced by the nearly com- pleted expedition of Lewis and Clark, but it is certain that Humboldt’s bold approval of Jefferson's aims gave encourage- ment to what was not a wholly popular program at the time. Soon Humboldt's familiar pictorial diagram of Chimborazo demonstrating the zonation of vegetation “corresponding with its progress in different latitudes" was used as the frontispiece of a botany text published in 1829 by the vice-principal of Troy Female Seminary, Mrs. Almira H. Lincoln. Her Familiar Lectures in Botany enjoyed a sale of more than 275,000 copies in the succeeding forty years in the higher schools and academies of this country. ° Penna. Mag. Hist. Biog. 73: 26-33. 1949. Cf. also Thompson Westcott, Historic MANSIONS AND BUILDINGS OF PHILADELPHIA, 415 et seq. 1877; John T. Faris, OLD GARDENS IN AND ABOUT PHILADELPHIA, 150 et seq., 1932, but use with caution; J. Ewan, Proc. Amer. Purros. Soc. 96: 603-605. 1952. 19 F, H. Herrick, AUDUBON THE NATURALIST, 2: 284-285. 1917. 1956] Ewan,—Humboldt and American Botany 195 Humboldt’s American contacts continued throughout his lifetime. When Balduin Móllhausen, topographer and artist to Whipple’s Pacific railway survey, visited Berlin with a con- signment of animals for the city’s zoo and paid a call on Hum- boldt, he met the daughter of Humboldt’s valet. Later as Mollhausen crossed New Mexico with Whipple’s party he wrote long love letters to Alexandra Caroline Seifert. It is generally recorded that Dr. John M. Bigelow was the botanical collector on Whipple’s survey, but we find Móllhausen, in de Terra's words, “laden with plant and rock specimens for Humboldt to study." This Móllhausen material, which may represent authentic duplicates of species described from Bigelow's col- lections, may yet come to light, The diary of Dr. C. B. R. Kennerly, physician and zoologist to the survey, preserved in the Division of Manuscripts, Library of Congress, may hold the answer to this riddle. This is the Móllhausen called the ‘Ger- man Cooper' for his immensely popular Indian love stories, travel lore, poetry, and some eighty novelettes!" Fremont spread Humboldt's fame in this country. Using extensively and taking inspiration from Humboldt’s writings, Fremont was warmed by the praise accorded his Report. Alas! the argonauts who drank the bitter waters of the ‘Humboldt Sink’ in the rush for gold had little praise for either Fremont or the man he honored. Latin America preserves many witnesses of its appreciation for Humboldt. Bolivar and Humboldt stand side by side not only as twin peaks in the high range above the beautiful city of Merida but in the hearts of her patriots. Bolivar, to whom Humboldt gave encouragement, said that the “learning” of his friend had “done America more good than all the conquerors.” Mutis, pupil of Linnaeus, and Humboldt and Bonpland worked in adjoining houses in Bogotá, and Mu- tis’s biographer, A. Federico Gredilla, quotes long letters that passed between them.? We are still in desperate need of a biography in English of this great figure, whose library and eolleetions of plant drawings were said to be second only to those of Joseph Banks. Sir Robert Schomburgk's explorations 1 Mdllhausen’s contacts with John Xantus are interesting in this connection. Cf. H. M. Madden's XANTUs: HUNGARIAN NATURALIST IN THE PIONEER WEST, 72, 238, 1949. 12 Supplementary to Sprague's paper of 1926, cited by de Terra, cf. A. Dugand, Rev. Acap. CoLoMsE. 9 (35): 210-213. 1954. 196 Rhodora [Vor. 58 in Guiana keenly interested Humboldt, and he was again de- lighted when Prince Adalbert of Prussia’s Travels in Brazil were translated by Schomburgk. In a preface Humboldt prepared for the English edition, he remarked, “if human civilization, which is making such giant strides in North America, should ever penetrate into these fastnesses (a hope which St. Basil calls ‘day-dreams of man’), the structure of this great net of rivers, from north to south, may open channels of communica- tion such as are unknown in any other part of the world." Humboldt was a prodigious worker: in his early sixties he was working simultaneously on the final volume of his histoire de la geographie of America, completing his Asiatic studies—he had travelled 9600 miles through Asiatic Russia by coach in twenty- five weeks—and was planning the first volume of his great synthesis—now seldom read—entitled Cosmos. His writings exceeded in bulk those of any contemporary naturalist, rivalling Buffon and Leibnitz. His later writings were ponderous and at times repetitious; Arago, the physicist, had, in fact, insisted that Humboldt really did not know how to write a book, but that he wrote endlessly, and the result was not a book “but a portrait without a frame." The Scot, William Macgillivray, composed a useful resume of his travels and researches which went into a second edition. Corresponding and writing, the "Prometheus of our time" worked on, meeting American visitors, as when Prof. Benjamin Silliman of Yale visited him in 1851, and parrying with some over-zealous investigators: in 1843 a Dresden anatomist requisitioned Humboldt's skull upon hear- ing of his illness! Of that skull-duggery I know no more. “T used to admire Humboldt, now I almost worship him," wrote Charles Darwin to his teacher Professor Henslow. Dar- win insisted that his entire career was a consequence of his reading Humboldt’s Narrative of Travels. The botanist Berthold Seemann said that “by his glowing descriptions of tropical scenery" Humboldt “did more than any one living to encourage that desire for the exploration of equinoctial regions, to which we owe . . . directly the elaborate works of Martius, Griffith, Blume, and other botanists of eminence.” Humboldt was the fillip of Louis Agassiz's trip to Brazil. Prescott wrote that he was guided “by the light of his researches" and added that he 1956] Steyermark and Swink,— Asclepias syriaca 197 missed Humboldt’s hand in his study of Peru which lay just beyond Humboldt’s fields. Dr. de Terra’s biography is a highly successful portrait (with a frame!) done in sepia with plunging strokes. This biography does not intend to be exhaustive as the good index will quickly demonstrate. There is still a genuine need for a full and docu- mented biography along natural history lines, dressed for its immense potential reference value but blind to Brentano’s show window. This unborn biography will integrate from the rich reliquiae of Humboldt letters around the world, identify their personalia, trace the intercourse across the Atlantic, from amateur to mentor and back again, all arrayed at finger-tip accessibility by a complete index. The botanical collector Benedict Roezl discovered a new lily in California in 1869 and as a centennial gesture it was named for Humboldt. How fitting that Humboldt’s second centennial should be marked by an Olympian biography. Shall it be called “Prometheus Un- bound"?—JoskePH EWAN, TULANE UNIVERSITY, NEW ORLEANS, LA. ASCLEPIAS SYRIACA VAR. KANSANA IN NEW YORK STATE.— At the time of the 1952 meetings of the Botanical Society of America in Ithaca, New York, one of the local field trips planned for the Ecological Society included a stop northeast of Geneva on the east side of Seneca Lake. In the waste ground bordering the road was an extensive stand of Asclepias syriaca L. All gradations were observed from the essentially plain- surfaced fruits of forma inermis to the spiny-fruited var. kansana (Vail) Palmer & Steyerm. The collection data for the plants referred to var. kansana are “open ground bordering ditch along east side of highway 96 A, northeast of Geneva, on east side of Seneca Lake, Seneca Co., New York, September 11, 1952, Steyermark & Swink 74625." The range given for this spiny-fruited variety in Gray's Manual, eighth edition, is “Ta., Neb., Mo. and Kans." While Woodson in his recent monograph of the genus (Ann. Mo. Bot. Gard. 41: 105-108. 1954) does not consider var. kansana to merit taxonomic rank, he does acknowledge the fact that the plants of the western states are predominantly spiny-fruited, while those of the eastern states 198 Rhodora [Vor. 58 are predominantly smooth-fruited (loc. cit. p. 108, text and footnote). He relegates their status, however, to races and clines. Dr. Woodson’s careful monograph notwithstanding and with our highest admiration for this work, the fact still remains that the field worker and local botanist, examining the two extremes of the fruit, one from the west with the spiny processes conspicuous, the other from the east nearly or quite plain-sur- faced, are impressed by the constancy and prevalence of the spines or lack of them, depending upon the geographical location. The student and local botanist in Missouri and Kansas encounters the spiny-fruited type practically all the time, whereas in the eastern states the excessively spiny-fruited type is practically unknown. Since this morphological character can be combined to a certain degree with a circumscribed range, it does not seem to the present authors inappropriate to maintain varietal status for the predominantly western spiny-fruited var. kansana. If the spiny-fruited plants were found to be everywhere within the range of the smooth-fruited or less spiny types, there would be no point in maintaining a geographic status for the spiny- fruited plants. One expects to find every kind of intergradation where the boundaries of the one type pass over into those of the other, but that should not mitigate against the recognition of the two varieties, at least from the standpoint of a convenient category or name for purposes of identification. Also from the geographical standpoint, it would appear proper to retain taxo- nomic recognition of var. kansana. In addition to the single station for var. kansana, cited above for New York state, a new eastern record, the authors observed the spiny-fruited type in various parts of northern Illinois and Indiana. If the var. kansana is maintained, its range should be extended to include the states of Illinois, Indiana, and New York.—JuLIAN A. STEYERMARK AND FLovyp A. SWINK, CHICAGO NATURAL HISTORY MUSEUM AND COLLEGE OF PHARMACY, UNIVERSITY OF ILLINOIS. Volume 58, no. 690, including pages 135—162, was issued 12 July, 1956. 73 USCA ''S5vW 3923? WVS 7/273" PARLON REFERENCE LIGRARY 20 DIVINITY AVENUE ; CAMORICGE, MASS... O. S. A. M4) 3023AY ALINIAIC O2 YY 32N7H3471 MOINS JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS l RALPH CARLETON BEAN RICHARD ALDEN HOWARD ( Associate Editors CARROLL EMORY WOOD, JR. Vol. 58 August, 1956 No. 692 CONTENTS: On the Identity of Lesquerella angustifolia. Reed C. Rollins Some Michigan Records for Sarracenia purpurea forma hetero- phylla, Frederick W. Case, Jr... ccc ccc ccc eee eens 203 A Cytotaxonomic Study of the Genus Hymenopappus (Com- positae). Billie L. Turner (continued from p. 186) .... 208 Che New England Botanical Club, Ine. 8 and 10 West King St.. Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $4.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 40 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $4.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scien- tific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of publica- tion. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U. S. A. Rbodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 August, 1956 No. 692 ON THE IDENTITY OF LESQUERELLA ANGUSTIFOLIA REED C. ROLLINS FERNALD’s revision of Gray's Manual,! Gleason’s revision of Britton and Brown's Illustrated Flora? and Palmer and Steyer- mark’s Catalogue of the Plants of Missouri’ include Lesquerella angustifolia on the basis of its presumed occurrence in the state of Missouri. In this, these authors undoubtedly followed Payson who, unfortunately, misapplied the name in his mono- graph of Lesquerella.* He applied the name to a species occur- ring in southwestern Missouri and previously regarded as L. gracilis, but which is actually distinct from both L. gracilis and L. angustifolia. As a result of this confusion, Cory? was led to re-describe L. angustifolia as a new species under the name L. longifolia. The original description of L. angustifolia in Torrey and Gray® and that of Watson’ are definite enough and were obvi- ously based on the original collection of Nuttall and a subsequent collection by Leavenworth, both obtained along the Red River more than a century ago. Evidently, the fact that L. angustz- folia was not recollected in the interim between these early collections and the time Payson studied the group had some- thing to do with his misinterpretation of the earlier work. 1 Gray's Man. Bot. (ed. 8) p. 700. 1950. ? Ill. Fl. N. E. U. S. adj. Can. 2: 226. 1952. 3 Ann. Mo. Bot. Gard. 22: 547. 1935. * Ann. Mo. Bot. Gard. 8: 182. 1922. s Field & Lab. 18: 96. 1950. $ Fl. N. Am. 1: 101. 1838. 1 Proc, Am, Acad. 23: 253. 1888. 200 Rhodora [Vor. 58 In order to be sure about the proper application of the name L. angustifolia, | have studied the type at the British Museum of Natural History. The type consists of two plants. The lowermost leaves are missing on both plants which are in the ripe fruiting stage. The cauline leaves are linear, thickish, and are covered with minute lepidote trichomes. The pedicels are rigid and divaricately ascending. There are two ovules per locule in the siliques. The type was compared in detail with a specimen of one of my collections (Rollins 5583) from the Red River Valley in Choctaw County, Oklahoma. The type and this specimen were found to be closely similar in every respect and unquestionably represent the same species. The type of L. longifolia Cory at the Herbarium of Southern Methodist Uni- versity is also the same, and this name becomes a synonym of L. angustifolia. L. angustifolia is not uncommon, usually on thin soil overlaying horizontally bedded limestone, in Choctaw County, Oklahoma. I saw and collected the species in a number of localities and it was frequently in the same location with L. gracilis, although the latter species was usually to be found on deeper soils. These two species are somewhat alike in habit, but they can be readily distinguished by the pubescence type and by the ovule number. The trichomes of L. angustifolia are disc-like from the fusion of the rays basally and appear as if they were a lepidote covering of peltate scales. At their densest, they make the leaves and stems silvery. The stems and leaves of L. gracilis are greener and have a sparser covering of small stellate trichomes with discrete branches. There are two ovules per locule in the ovary of L. angustifolia, whereas, in L. gracilis, there are from eight to twelve ovules per locule. Evidence of hybridization was not seen where these species were growing together even though I looked for it at several locations. In addition to the Nuttall type and the Leavenworth collection, both from the Red River, but without any further specific locality data, the following collections of L. angustifolia have been studied; Oklahoma, CHOCTAW COUNTY: 6 mi. east of Hugo, April 13, 1955, Rollins 5581 (an); 6 mi. west of Fort Towson, Rollins 5583 (an); 5 mi. west of Fort Towson, Rollins 5585 (Gu); 2-3 mi. west of Fort Towson, May 24, 1953, Moore & Iltis 405 (Gu); 5 mi. east of Hugo, April 15, 1950, Waterfall 9289 (Gu); 9 mi. east of Hugo, Waterfall 9290 (GH). MCCURTAIN COUNTY: 2 mi. 356] Rollins,—Identity of Lesquerella angustifolia 201 northwest of Idabel, April 19, 1954, Waterfall 11820 (Gu); near Idabel, May 20, 1916, Houghton 3680 (au). Texas: 1.5 mi. southeast of Clarks- ville, Red River County, May 12, 1949, V. L. Cory 56031 (smu, type of L. longifolia). From the above data, it would appear that L. angustifolia is rather restricted in its range. It may be confined to a limited limestone area in southeastern Oklahoma and adjacent Texas. Such a restricted area of distribution would account, in part at least, for the long gap between the early collections and the recent ones. This species occurs in the same locality as Leavenworthia aurea which appears to be equally restricted in its distribution in southeastern Oklahoma and has had a similar history so far as paucity and infrequency of collections are concerned. The Missouri species of Lesquerella mistakenly identified as L. angustifolia by Payson is unquestionably a distinct species. The next question is by what name should it be known? I have searched the literature and find no available name. Therefore, the species has to be described as new. Lesquerella filiformis Rollins, sp. nov. Annual; stems numerous from the base, slender, branched, erect or the outer somewhat decumbent, densely covered with stellate trichomes, 1-2 dm. high, branches often filiform; basal rosette of leaves not formed; basal leaves few, entire, broadly spatulate to orbicular, rounded at apex, abruptly narrowed to a slender petiole, 1-2 em. long including petiole, 5-8 mm. wide, stellate trichomes on both surfaces numerous but not overlapping; cauline leaves entire, densely pubescent, 1-3 em. long, 1-6 mm. wide, lower cauline leaves petiolate, often with a broad blade, gradu- ally becoming narrower and sessile upward; inflorescence not crowded; pedicels filiform, divaricately ascending, nearly straight, 7-10 mm. long, densely pubescent; sepals narrowly oblong, densely pubescent, 3-4 mm. long, ca. 1 mm. wide, outer pair slightly saccate; petals light yellow, spatulate, 7-9 mm. long, 2.5-3 mm. wide; filaments slender, dilated at base; anthers ca. 1 mm. long; siliques globose, glabrous both on the exterior and interior, nearly sessile, erect, 2.5-3 mm. in diameter; styles slender, 3-4 mm. long; stigma unexpanded; ovules 2 in each loculus; funiculi attached to thin septum toward base only; seeds brown, somewhat flattened, not margined. Herba annua; caulibus erectis ramosis tenuibus pubescentibus 1-2 dm. altis; foliis stellato-pubescentibus; foliis radicalibus petiolatis integris late spathulatis vel orbieularibus 1-2 em. longis, 5-8 mm. latis; foliis caulinis inferne petiolatis superne sessilibus anguste spathulatis vel linearibus 1-3 em. longis, 1-6 mm. latis; pedicellis filiformibus rectis divaricatis pubes- centibus 7-10 mm. longis; sepalis anguste oblongis pubescentibus 3-4 mm. 202 Rhodora [Vor. 58 longis, ca. 1 mm. latis; petalis flavis spathulatis 7-9 mm. longis, 2.5-3 mm. latis; siliquis glabris globosis subsessilibus erectis 2.5-3 mm. diametro; stylis glabris 3-4 mm. longis; loculis 2-ovulatis; seminibus immarginatis. Type in the Gray Herbarium collected on thin soil, limestone barrens, Turnback, Dade County, Missouri, May 5, 1929, E. J. Palmer 35604. An additional collection in the Gray Herbarium is J. W. Blankenship s. n. made at Willard, Greene County, Missouri, May 7, 1887. The illustration and description in Payson's monograph (l.c.), so far as I can interpret them, apply completely to this species. The illustration shows the numerous stems, the branching habit and the lack of a basal rosette of leaves. As pointed out by Payson, the species is amply distinct from L. gracilis because of the subsessile siliques with 2-ovuled loculi and the filaments with dilated bases. Actually, L. filiformis is more closely re- lated to L. angustifolia than it is to L. gracilis. However, it has stellae with distinct rays instead of stellae of disc-like peltate scales with basally fused rays as found in L. angustifolia. Other- wise, the plants of L. filiformis are much smaller in every respect than those of L. angustifolia. With respect to height, the upper limit of 2 dm. in L. filiformis is about the lower limit for L. angustifolia. The lower leaves of the latter are dentate to lyrately pinnatifid, whereas, they are entire in L. filiformis. The two species are sufficiently distinctive in aspect to prevent confusion of identity.—GRAY HERBARIUM OF HARVARD UNI- VERSITY. 1956] Case,—Sarracenia purpurea 203 SOME MICHIGAN RECORDS FOR SARRACENIA PURPUREA FORMA HETEROPHYLLA FREDERICK W. Case, Jn! In 1822, Eaton described a plant from the region of North- ampton, Mass. as a supposed new species, Sarracenia hetero- phylla, distinguished by elongate leaves and yellow flowers. Later investigations have shown that the plant is not mor- phologically different from S. purpurea except that it differs in the color of its leaves and flowers. The elongate type of leaf is merely a change caused by shade, similarly shaped leaves being found on shaded plants of the typical form. Consequently, recent authors have reduced the taxonomic status of the plant from species to subspecies, then variety, and finally, a form (Fernald, 1922, 1950, and Gleason, 1952). Bell in his recent eytotaxonomie study of the pitcher-plants (1949) follows the last treatment. The purpose of the present paper is to present new data on the occurrence of this interesting plant in Michigan, the localities reported being a little over 600 miles west of the previously known localities, all of which are along the eastern seaboard. (See map, Fig. 1.) Sarracenia purpurea forma heterophylla, as currently under- stood, is characterized by the complete absence of red coloring anywhere in the plant. The flower petals of living specimens are of a pale lemon-yellow to greenish-yellow color. The leaves, although dull green in shade grown specimens, are yellow-green to rich yellow in sunny situations. In dried material the colors may appear golden- or brown-yellow. If any red coloration is present at all, the plant cannot be regarded as forma hetero- phylla. Until the presently described locations came to light, all authenticated records for Sarracenia purpurea f. heterophylla have been confined to the northeast coast of North America, and even here authentic records are extremely rare. Records do exist from the following localities, as reported by Bell (l. c.): 1 I wish to thank Drs. C. Ritchie Bell, W. H. Wagner, Jr., and E. T. Wherry, for their suggestions. I am especially grateful to Drs. Bell and Wherry for visiting the areas with me on September 4, 1955. 2 Bell (l.c.) suggests, however, that it is possible that forma heterophylla ‘‘may deserve higher taxonomic rank.” 204 Rhodora [Vor. 58 Northampton, Mass. (where it was collected by Eaton); Main Arm, Bonne Bay, Newfoundland (Fernald and Long); Exploits Hiver and Badger Brook, Newfoundland (Robinson and Schrenk); Young's Lake, Belle Isle, Nova Scotia (Fernald, and also Bell); and Forked River, New Jersey (Britton). Schallert applied the name “heterophylla” to herbarium material from North Carolina, but Bell discounts his records, since most of his specimens show red veining, and therefore should be excluded from this taxon (Bell, 1949). It is likely that all of Schallert’s identifications represent deep shade ecads of typical S. purpurea. Figure 1. The distribution of Sarracenia purpurea in North America. The triangles represent authenticated localities for the forma heterophylla. General distribution adapted from F. T. Wherry, 1935. 1956] Case, Sarracenia purpurea 205 In June, 1948, while botanizing in Montmorency County, Michigan, I came upon yellow-flowered plants of Sarracenia purpurea in à large spruce-tamarack bog, but at that time no record of the extent of distribution or of the number of yellow plants was made. Subsequent exploration revealed that forma heterophylla occurred in a number of bogs in this vicinity. When it became apparent that the numbers and distribution of the form in this area might possibly show some evolutionary pattern, a more extensive study was initiated. All of the bogs that the author could locate and to which he could gain access were explored during the years 1954 and 1955. This is not to imply, of course, that all suitable habitats in this bog-rich area have : d ç ! q SOR E. | : 2*6 O; | | IND ease. ong | | J i | 18 17 | Io | IS l | | N (CE p | | x | Wel ee ae . o 19 20 ! 22 l 22 | | | Io SP "OWN i l — it 1 “ CORNEKS l | 30 24 | 22 | 2 | | | | pole dodo o l s M NL LL | | | | | l | lmile " FiGURE 2. Portion of Township 30 N.. Range 1 E., Montmorency County, Michigan. showing location of the explored bogs. 206 Rhodora |Vor. 58 been covered; there are possibly others in which forma hetero- phylla will be discovered. The locations of the explored bogs are shown on the accom- panying map (Fig. 2). Forma heterophylla was found to occur in the bogs numbered 1, 2, 3, 8, and 9. A few of the plants found in bogs 6 and 7 appeared to be forma heterophylla but, as they were deeply shaded, final confirmation should be reserved until flowering specimens are seen. All the bogs are in Town- ship 30 N., Range 1 E. Montmorency County, Michigan. Bogs 1, 2, 3, and 5 are in the south half of Section 8; bog 6 on the line between sections 7 and 8; bog 7 in the north half of Sect. 7; and bogs 8 and 9 are in Sect. 18 of the same township. All of the bogs are rather open Picea-Larix associations on Green- wood Peat soils (Veach et al. 1930). Living specimens of taxon heterophylla from these locations are in the author’s greenhouse. Herbarium material, Case 900 to 908, and Bell 1548, will be deposited in the University of Michigan Herbarium, Gray Herbarium, the U. S. National Herbarium, and the University of North Carolina Herbarium. To determine something of the ratio of occurrence of the red and the yellow forms, a series of counts was made. It must be stressed that any color comparisons must be taken on plants growing in the open. Only in complete sunlight does the coloring present develop fully, eliminating the confusion of greenish ecads which are genotypically red with the true antho- cyanin-free forms. Plots 9 yards square were selected more or less at random from the open areas of bogs 1, 3, and 9. A record was made of the number of red-leaved plants and of the number of yellow-leaved plants with mature leaves in each plot. The findings are shown in the table below. Bog Plot Red form Yellow form ] l 129 16 — 2 77 21 3 1 29 37 — 2 7 34 9 1 28 26 — 2 25 20 The data are too few to be considered as a final indication of the rate of occurrence of the two forms in Montmorency Co., 1956| Case,—Sarracenia purpurea 207 Michigan. Yet they show that in some of the bogs, at least, forma heterophylla is very well developed. Especially interesting is the observation that in each of bogs 1, 3, and 9 there is a considerable number of plants of an orange-red color which do not fit the concept of either the typical red or the yellow form. In the above counts, however, such plants were treated arbitrarily as red. These orange-red plants, not found in bogs unless the yellow form was present also, lead me to believe that forma heterophylla is not behaving as a simple Mendelian recessive, and that the possibility of genetic blending must be considered. In order to obtain more information on the genetic nature of forma heterophylla I have initiated various greenhouse breeding experiments which should provide some insight into the situation, and which will be duly reported upon. SUMMARY The history and nature of Sarracenia purpurea forma hetero- phylla is briefly reviewed and maps of its known occurrence are presented. A number of new localities have been discovered in Montmorency Co., Michigan, over 600 miles west of the nearest localities heretofore recorded. Results of sample counts in various bogs where the two color forms occur together are given. A number of individuals in which the dominant color is orange-red suggest that genetic intermediates between the two forms may exist. LITERATURE CITED BELL, C. R. 1949. A Cytotaxonomie Study of the Sarraceniaceae of North America. Jour. Elisha Mitchell Sci. Soc. 65: 137-166. FEnNALD, M. L. 1922. Notes on the Flora of Western Nova Scotia. Ruopona 4: 165-183. 1950. Gray's Manual of Botany, 8th Ed. American Book Co. New York. VEACH, SCHOERMANN, MILLAR, and SHEARIN. 1930. Soil Survey of Montmorency County, Mich. U. 8. Dept. of Agriculture Bull. No. 39 (series 1930). Wuerry, E. T. 1935. Distribution of the North American Pitcher Plants. In Walcott, M. V., Illustrations of North American Pitcher Plants. Smithsonian Inst., Washington, D. C. 208 Rhodora [Vor. 58 A CYTOTAXONOMIC STUDY OF THE GENUS HYMENOPAPPUS (COMPOSITAE) Biri L. TURNER (Comtamued from paqe 186) Following is a list of herbaria consulted during this study. The abbreviations used in citation of specimens are those recom- mended by Lanjouw and Stafleu (1952). Grateful acknowledge- ment is made to the curators concerned for the loan of material. cas California Academy of Sciences; ps Dudley Herbarium, Stanford University; GH Gray Herbarium of Harvard University; mo Missouri Botanical Garden; ny New York Botanical Garden: pom Herbarium of Pomona College; RM Rocky Mountain Herbarium, University of Wyo- ming; RSA Rancho Santa Ana Botanic Garden; smu Southern Methodist University; TEX University of Texas; uc University of California, Berke- ley; us United States National Herbarium, Smithsonian Institution; ws State College of Washington. TAXONOMY Hymenopappus L'Hér. Hymenop. 1. 1788. Rothia Lam., Jour. Hist. Nat. 1: 16. 1792, non Rothia Schreb. Biennial and perennial, subscapose to leafy-stemmed herbs from a single unbranched, obconical tap-root (in biennial species) or from a woody, much-branched tap-root bearing several crowns (in perennials); stems slender to stout, erect, angled and sulcate, often pithy; leaves alternate, forming a basal rosette, mostly bipinnately dissected to rarely simple or merely lobed, reduced up the stem, minutely impressed-punctate, with usually inrolled margins; inflorescence a few- to many-headed cymose panicle; heads discoid or radiate, subturbinate to broadly campanulate, on slender conspicuous peduncles; involucre of 6-14 subequal bracts in 2-3 series, these membranous at the apex or rarely throughout; receptacle dome-shaped to nearly flat, without chaff, except in H. newberryi; ray florets, when present, pistillate and fertile, with conspicuous white ligules; disk florets with regular corollas (at least the throat), yellow to white or less often reddish-purple, narrowed below into a slender, mostly glandular tube, and above this (after anthesis) an abruptly flaring, campanulate to funnelform throat, lobes equal, ovate, reflexed after anthesis; style branehes more or less strongly flattened with obtuse, papillose tips, stigmatic lines marginal on the upper surface; anthers partially to com- pletely exserted, shortly cordate-sagittate at base with ovate terminal ap- pendages; achenes with 4 faces, obpyramidal to rarely ineurved and diagonally compressed, the narrow base often shortly stipitate (in biennial species), faces 0-3-nerved, glabrous to densely pubescent with ascending or spreading hairs; pappus when present of 12-22 linear-oblong or broadly 1956] Turner, —A Study of the Genus Hymenopappus 209 ovate, obtuse, hyaline scales, usually with an included medial nerve; chromosome number, so far as known, with a base of x — 17. Type species: Hymenopappus scabiosaeus L'Hér. KEY TO THE SPECIES OF HYMENOPAPPUS 1. Plants perennial, roots bearing several to many crowns; stem-leaves few, 0-10(-12); heads per stem few, 1-20 (-50); corolla throat 2-8 times as long as the lobes... . Series PERENNES. 2. Heads with conspicuous white rays. 3. Receptacle chaffy; pappus 0.1 mm. long or obsolete. .1. H. newberryi. 3. Receptacle naked; pappus 0.4-0.6 mm. long.......... 2. H. radiatus. 2. Heads eradiate (or in extremely rare specimens of H. fili- folius var. cinereus with poorly developed rays). 4. Leaves bipinnately dissected with narrow, linear divi- sions; achenes densely pubescent.................. 3. H. filifolius. 4. Leaves simple or once-pinnate with broad divisions; iba glabrata nan irki A 5 4. H. mericanus. 1. Plants biennial, roots bearing a single crown (rarely 2 through injury); stem-leaves many, 10-100 (rarely less); corolla throat 1-2(-3) times as long as the lobes.............. Series BIENNES. 5. Heads with conspicuous white rays..................... 5. H. biennis. 5. Heads eradiate. 6. Achenes glabrous; pappus 0.2 mm. or less long..6. H. flavomarginatus. 6. Achenes pubescent; pappus 0.2-2 mm. long. 7. Corolla tube 1.5-2(-2.2) mm. long; basal rosette leaves bipinnately dissected, with mostly linear ultimate segments, 0.5-6 mm. wide; flowers yellow or white; corolla throat campanulate (very rarely funnelform). 8. Flowers white; ultimate leaf segments narrowly linear, 0.5-1.5 mm. wide................... 7. H. tenuifolius. 8. Flowers yellow; ultimate leaf segments short, nar- row to broad, 1-6 mm. wide. 9. Leaf segments 2-6 mm. wide; leaves usually glabrous or less pubescent above; pappus 0.5-1 1.2) mm. long... . 8a. H. flavescens var. flavescens. 9. Leaf segments 1-2 mm. wide; leaves usually evenly pubescent on both surfaces; pappus 1-1.5 mm. long (rarely less) 8b. H. flavescens var. cano-tomentosus. 7. Corolla tube 2-3 mm. long; basal rosette leaves simple to bipinnate with broad ultimate segments 5-20 mm. wide; flowers white or vinaceous, never yellow, corolla throat funnelform (rarely campanulate). 10. Basal rosette leaves once-pinnate to bipinnately parted; flowers white; pappus 0.2-0.8(-1) mm. long; plants of clayey soils in grasslands of Texas, Oklahoma, Kansas, and Nebraska, or sandy soils of the Upper Mississippi Valley and S.E. United States. 210 Rhodora [Vor. 58 11. Peduncles with conspicuous ovate, membranous, petaloid bracts 5-14 mm. long (rarely less); involucral bracts white-membranous for half their length or more, 7-15 mm. long 9a. H. scabiosaeus var. scabiosaeus. 11. Peduncles ebracteate or with non-membranous, short, subulate scales; involucral bracts yel- lowish to white membranous for 2-4 mm. at the tip, 5-9 mm. long...9b. H. scabiosaeus var. corymbosus. 10. Basal rosette leaves simple to once-pinnate; flowers vinaceous or reddish-tinged (rarely com- pletely white); pappus 0.5-2 mm. long; plants in sandy soils of Louisiana and eastern and southern Texas. 12. Involucra] bracts 6-12 mm. long, snowy-white membranous for half their length or more; pappus 0.5-1(-1.5) mm. long; stem-leaves (6—)8-16...... 10a. H.artemisiaefolius var. artemisiaefolius. 12. Involucral bracts 5-7 mm. long, membranous for 1-2 mm. at the tip; pappus 1.5-2 mm. long; stem-leaves 2-8..10b. H. artemisiaefolius var. riograndensis. SERIES PERENNES 1. Hymenopappus newberryi (Gray) Johnston Leucampyx newberryi Gray, in Porter and Coulter, Synopt. Flora Colorado, p. 77, 1874. H. newberryi Johnston, Contrib. Gray Herb. n.s. 68: 96. 1923. Type examined (GH): Colorado, “La Pagosa, July 29, 1860," MeComb's Exped., Dr. Newberry. Isotypes (Ny, vs) give the lo- cality as “La Pajosa" and the date as 1859. Plants perennial, 20-60 cm. high, sparsely tomentose to nearly gla- brate; rosette leaves 12-25 em. long, 3-5 em. wide, bipinnately dissected with narrow, linear, mostly flattened ultimate divisions, evenly and sparsely canescent on both surfaces to nearly glabrate, inconspicuously impressed-punctate; stem leaves 1-3(-5), much reduced; heads 3-8 per stem, broadly campanulate, 2-3 em. in diameter, 60—150-flowered, on stout ultimate peduncles 6-15 em. long; receptacle chaffy, dome-shaped, 5 mm. across, 3 mm. high; chaff deciduous, 5-9 mm. long, 2-5 mm. wide, partly to completely (near the periphery of heads) enclosing the achenes, minutely glandular to nearly glabrous, yellow-membranous, except near the base; principal involucral bracts 8-10 mm. long, 4-7 mm. wide, white or yellowish-white membranous for 2-4 mm. from the mostly ob- tuse tip; ray flowers 8, pistillate and fertile, tubular at the base for about 2 mm., with a conspicuous white (rarely pink) ligule, 14-20 mm. long, 8-15 mm. wide, 3-cleft at the apex, the sinus 2-3 mm. deep; disk flowers yellow, 3.5-4 mm. long, the tube sparsely glandular to nearly glabrate, 1.5-2 mm. long, throat campanulate 1.5-2 mm. long with lobes reflexed, 2-3 times longer than the lobes; style branches about 1 mm. long, short and flattened, with well defined stigmatic lines along the upper margins; achenes black, glabrous, narrowly obpyramidal, diagonally compressed 1956] Turner, —A Study of the Genus Hymenopappus 211 and strongly incurved, 3.5-4 mm. long; pappus minute (less than 0.1 mm. long) or absent; anthers not much exserted, 2.2-2.5 mm. long; n = 17. DisTRIBUTION.—South-central Colorado and north-central New Mexico on clayey, principally igneous soils of mountainous regions, 7,000 to 10,000 ft., usually in open areas of pine, spruce or aspen woods (Fig. 24). June-September. Gray described this species as being the sole member of a new genus Leucampyz, and placed it in the tribe Anthemideae of the Compositae. He recognized its resemblance to Hymenopappus, but apparently thought this a superficial one, assigning more significance to the technical characters of ray florets and chaff. It was not until 1891, when Rose described the new species, H. radiatus, that this assignment of Leucampyx to the tribe Anthemideae was seriously questioned. The presence of an obviously rayed species of Hymenopappus no longer justified the tribal separation of Leucampyz, or even its maintenance as a separate genus. It remained for Johnston (1923) finally to reduce Leucampyx to generic synonymy under Hymenopappus and make the appropriate combination for the species. The work of the present author, both cytological and morphological, completely confirms this latter treatment.’ Hymenopappus newberryi has probably retained more primitive characters than any other member of the genus. Accordingly, from this species one may get an idea as to the main evolutionary trends that have occurred within the genus, foremost of which has been the progressive loss of parts such as chaff and rays. It seems probable that the H. filifolius complex had its origin from the H. newberryi line, or its immediate ancestor, at some early time. H. radiatus is thought to represent a more recently derived species which has differentiated directly from H. new- berryt. REPRESENTATIVE SPECIMENS.— Colorado. ARCHULETA CO.: Pagosa ? It appears likely that even the natural genetic barriers may rarely break down between H. newberryi and related species. John Beaman, of Harvard University, collected in 1954 several rather remarkable rayed specimens of what otherwise ap- peared to be H. filifolius var. cinereus. "These collections were made 10.3 mi. east of Taos, New Mexico, and were taken from a population growing near that of typical H. newberryi (Beaman 726; au, Tex). The plants in question have relatively short, poorly developed rays and are without chaff; however, the achenes and pappus are somewhat intermediate between the two taxa. If hybridization has been responsible for this variation, then its occurrence must be quite rare since similar variability was not detected in the numerous sheets examined from this area—nor was it detected in population studies during preliminary field work on the genus. [Vor. 58 Rhodora 212 H. radiatus (tri- H. mexicanus (disks), Fig. 24. Fig. 25. Distribution of Himenopappus species. angles), H. newberryi (disks), and H. biennis (circles). and H. flavomarginatus (circles). 24-25. Fia. 1956] Turner,—A Study of the Genus Hymenopappus 213 Springs, Bethel, Willey & Clokey 4371 (CAS, DS, MO, NY, RM, UC, US, WS). CHAFEE CO.: Salida, (w/o date), A. Lansing (Mo, UC). COSTILLA CO.: West Indian Cr., Rydberg & Vreeland 5476 (Ny, RM). CUSTER co.: Hard- scrabble Canyon, Penland 1266 (cas, Us). DOLORES CO.: 5 mi. S. of Rico, Maguire & Piranian 12925 (an, uc). FREMONT CO.: near Canon City, 1871, Brandegee 294 (Mo, NY). GUNNISON CO.: 16 mi. E. of Lake City, Cebolla Cr., Rollins 1420 (Mo, NY). HINSDALE co.: Spring Creek Pass, Hitchcock, Rethke, & van Raadshooven 4101 (DS). MINERAL CO.: Rio Grande Nat’l. Forest, near Elliott Cr., June 18, 1911, J. Murdock, Jr. 4623 (GH, MO, NY, UC, US). PARK CO.: South Park, 1871, Brandegee 294 (Mo, UC). SAGUACHE CO.: W. side of summit, Cochetopa Pass, Weber 5775 (smu, ws). New Mexico. corrax co.: near Vermejo, H. M. Hall R335 (vc). SANDOVAL co.: La Cueva Springs, Preece & Turner 2745 (SMU, WS). SAN MIGUEL CO.: Pecos R. Nat’l. Forest, Winsor's Ranch, Standley 4003 (GH, MO, NY, RM, US). SANTA FE CO.: vicinity of Santa Fe, Lake Peak, Arséne & Benedict 16016 (us). TAOS co.: 10 mi. E. of Taos, Rollins & Chambers 2415 (ps, GH, us). 2. Hymenopappus radiatus Rose Hymenopappus radiatus Rose, Contrib. U. S. Nat. Herb. 1: 122. 1891. Type examined (us): “Willow Spring, Arizona,” July (5-6), 1890, Dr. E. Palmer 615. According to Rose (1891), “Willow Springs is in the White Mountains near the pass leading to Ft. Apache." Plants perennial, 30-45 cm. high, sparsely to densely tomentose; larger rosette leaves 8-14 cm. long, bipinnately dissected with linear, some- what flattened, ultimate segments, 1-2 mm. wide; stem leaves 2-3, much reduced; heads 6-8 per stem, campanulate, 40-60-flowered, on ultimate peduncles 3-8 em. long; principal involucral bracts 7-9 mm. long, 3-5 mm. wide, tomentose, yellow-membranous for 1-3 mm. from the acute to obtuse tip; receptacle dome-shaped, without chaff; ray flowers 8, tubular at the base for about 2 mm., extending into a conspicuous white ligule 14-16 mm. long, 5-10 mm. wide, usually with an emarginate or indis- tinctly 3-cleft apex; disk florets yellow, 3-4 mm. long, the tube 1.5-2 mm. long, sparsely glandular to nearly glabrous, the throat campanulate, 1.5 mm. long with lobes reflexed, 3 times longer than the acute lobes; achenes purplish-brown, 4-4.5 mm. long, obpyramidal and 4-sided, becoming di- agonally compressed and incurved toward the periphery of the head; pappus of 14-18 small, obtuse scales, 0.4-0.6 mm. long; anthers partially exserted about 2.5 mm. long; n = 17. DisrriputTion.—Endemie to Arizona; sandy tuffaceous or igneous soils on south-facing slopes in pine woods of east-central Arizona (Mo- gollon Rim and White Mountains), 7000-9000 ft. (Fig. 24). May-July. Hymenopappus radiatus probably has its closest relative in H. newberryi. It differs from this latter species primarily by its non-chaffy receptacle, and less diagonally compressed, incurved achenes. The similarity of the two species is remarkable, 214 Rhodora [Vor. 58 and for this reason it is believed that H. radiatus has been derived directly from or through H. newberryi, this derivation requiring only the loss of chaff from the receptacle. REPRESENTATIVE SPECIMENS.—Arizona. APACHE CO.: Willow Springs, E. Palmer 615 (Gu, us). Coconino co.: Mogollon Mts., Collom 204 (au, MO, NY, US). NAVAJO CO.: Cooley's Ranch (semi-open flats), Goodding 1106 (NY, RM, vs): Holbrook to Showlow, J. Goodwin 112 (cas): Forest- dale, 66 mi. 8. of Holbrook, W. Hough 69 (us); Fort Apache, 1893, R. W. Hoyt (Ny); 9.5 mi. N. of White River, highway 73, Preece & Turner 2686 (SMU, ws); Fort Apache, 1892, E. A. Shuttleworth (us). 3. Hymenopappus filifolius Hook. Rothia filifolia (Hook.) O. Ktze., Rev. Gen. 1: 361. 1891. Plants perennial, supscapose to rarely leafy-stemmed, 5-100 em. high, densely tomentose to nearly glabrate; rosette leaves 3-20 cm. long, bipinnately dissected with linear, flattened or nearly filiform ultimate divisions, 2-50 mm. long, minutely impressed-punctate; stem leaves 0-12, usually much reduced upward; heads 1-60 per stem, subturbinate to broadly campanulate, 10-80-flowered, on ultimate slender peduncles 0.5-16 em. long; receptacle naked (very rarely with 6-10 well developed chaffy scales); principal involucral bracts 3-14 mm. long, 2-5 mm. wide, white or yellowish-membranous for 1-4 mm. from the acute to obtuse tip; corollas yellow or less often white, 2-7 mm. long, tube densely glandu- lar to nearly glabrous, 1-2.5 mm. long, throat campanulate, 1-3 mm. long with lobes reflexed, 2-8 times longer than the lobes; achenes ob- pyramidal, 4-sided, 3-7 mm. long, densely pubescent with hairs 0.2-3 mm. long; pappus of 12-22 linear oblong scales, 0.1-3 mm. long; anthers partially exserted, 1.6-4 mm. long; n = 17, 34. DisrRIBUTION.—Principally western United States in the Rocky Mts., Great Basin, Columbia Basin, northern Great Plains, and peripheral areas. KEY ro THE VARIETIES OF H. FILIFOLIUSS 1. Basal leaf axils without a dense tomentum, usually glabrous or sparsely tomentose; stem leaves 0 or 1(2); plants 5-30 (-45) em. high. 2. Ultimate peduncles 0.5-1.5 cm. long; corolla 2-3 mm. long; achenes 3-4mm.long........................... 3a. var. parvulus. 2. Ultimate peduncles 2-10 cm. long; corolla 3-5 mm. long, achenes 4-6 mm. long........................... 3b. var. nudipes. 1. Basal leaf axils with a dense conspicuous tomentum; stem leaves 0-12; plants 5-100 cm. high. 3. Corollas 2-3 mm. long; florets 10-30 (avg. 20) per head; involucral bracts 3-7 mm. long; leaf tips of basal rosette 1-6 mm. long. 8 Table 1 lists most of the important contrasts and should be used as a supplement to this key, which lists only the more conspicuous differences. 1956] Turner,—A Study of the Genus Hymenopappus 215 4. Stem leaves 0-3; pappus 0.1-0.8 mm. long; plants 5-20 cm. MEn rm rur m TN o 3c. var. luteus. 4. Stem leaves 4-9; pappus 0.5-1.5 mm. long; plants 20—35 cm high Ond. ed CM 3d. var. pauciflorus. 3. Corollas 3-7 mm. long; florets (15-)20-70 per head; in- volucral bracts (5—)6-14 mm. long; leaf tips of basal rosette 3-50 mm. long. 5. Anthers 3-4 mm. long; corolla (3.5-)4-7 mm. long; pappus (1.2-)1.5-3 mm. long; achenes 5-7 mm. long; corolla throat 3-8 times as long as the lobes. 6. Flowers white; peduncles 8-16 cm. long; stem leaves 2-7; achene hairs 0.5-1 mm. long............ 3e. var. eriopodus. 6. Flowers yellow or rarely white; peduncles 2-12 cm. long; stem leaves 0-6; achene hairs 1-2 mm. long. 7. Stem leaves (1)2-7; leaf tips of basal rosettes 5-30 mm. long; bracts 8-12(-14) mm. long; plants of low desert areas, 1000-5000 feet........ 3f. var. megacephalus. 7. Stem leaves 0-3; leaf tips of basal rosettes 3-15 mm. long; bracts 6-10 mm. long; plants of higher elevations associated with juniper, oak or pine, 5000-9000 feet. eel se eU ed s en 3g. var. lugens. 5. Anthers 2-3 mm. long; corolla 2.5-4.5 mm. long; pappus 0.2-3.0 mm. long; achenes 4-6 mm. long; corolla throat 2-5 times as long as the lobes. 8. Heads 5-60 per stem; stem leaves 3-8; leaf tips of basal rosettes 10-30 mm. long; throat 1.3-1.8 mm. long; plants of the northern Great Plains, E. of the cononentaldivide cre l So EU 3h. var. polycephalus. 8. Heads 1-40 per stem; stem leaves 0-12; leaf tips of basal rosettes 2-50 mm. long; throat 1.5-2.5 mm. long; plants of the Rocky Mts. or W. of continental divide (except for disjunct populations in the pan- handle of Texas). 9. Stem leaves 5-12; leaf tips of basal rosettes 3-6 mm. long; endemic to S. W. Utah.............. 3i. var. tomentosus. 9. Stem leaves 0-5; leaf tips of basal rosettes 3-50 mm. long. 10. Pappus 0.2-1.0 mm. long; leaf tips of basal rosettes 15-50 mm. long; plants 30-100 cm. high; heads (2-)6-40 per stem; plants of Oregon and Washington. ............... 3j. var. filifolius. 10. Pappus 0.6-3 mm. long; leaf tips of basal rosettes 2-30 mm. long; plants 5-50 cm. high; heads 1-6(-14) per stem; plants elsewhere. 11. Peduncles 3-15 cm. long; achene hairs 0.2- 1.0 mm. long; pappus 1.5-3 mm. long; plants of the Great Basin............... 3k. var. nanus. 11. Peduncles 1-6 cm. long; achene hairs 1-3 mm. long; pappus 0.6-2 mm. long; plants else- where. 12. Pappus 0.6-1.3 mm. long; stem leaves 0-2; plants Of [Idaho e. s Pes 3l. var. idahoensis. 12. Pappus 1.5-2 mm. long; stem leaves (0) 2-4; plants not of Idaho............ 3m. var. cinereus. 216 Rhodora [Vor. 58 3a. Hymenopappus filifolius var. parvulus (Greene), comb. nov. Hymenopappus parvulus Greene, Pl. Baker. 3: 30. 1901. Isotype examined (pom): “Gunnison,” Colorado, 7680 ft., July 17, 1901. C. F. Baker 449. Isotypie material of Baker 449 is not widely distributed. Only the one sheet was seen by the author. Cotype material, Baker 840, is typical of the variety and is found in several herbaria. Plants perennial, 10-25 cm. high; stems unbranched, very sparsely grayish-green tomentose to reddish-glabrate; larger rosette leaves sparsely Fic. 26-27. Floral morphology of Hymenopappus species. Fig. 26. H. fili- folius var. filifolius (T. 2293); (a) head, X 1.5; (b) floret, X 8; (c) pappus scales, X ca. 40; (d) style branches, X ca. 45; (e) upper surface of style branch, X ca. 45. Fig. 27. H. mericanus (Rusby 179); (a) peripheral floret. X 8; (b) corolla opened,- showing equal lobes, X 8. 1956] Turner, —A Study of the Genus Hymenopappus 217 pubescent to glabrate, without a dense tomentum in the axils, 3-10 em. long, 2-5 em. wide, typically once-pinnate with a few scattered, linear, secondary divisions, 0.5-1 mm. wide, 5-20 mm. long, petiole conspicuous, composing 14-24 of the leaf; stem leaves 0-2, much reduced; heads 3-10 (rarely-15) per stem, subturbinate, 10-18-flowered, on very short pe- duneles 5-15 mm. long; principal involucral bracts glabrate to pubescent below, 4-8 mm. long, 2-3 mm. wide, yellow-membranous for 1-2 mm. from the acute tip (very rarely red-tinged); corollas yellow, 2-3 mm. long, tube densely glandular, 1-1.5 mm. long, throat campanulate- tubular, 1.5-1.8 mm. long with lobes reflexed; achenes 3.5-4 mm. long, pubescent with hairs about 1 mm. long; pappus of 14-16 linear oblong scales, 1-1.8 mm. long; anthers partly exserted, about 2 mm. long; n — 17. DisTRIBUTION.—Sandy or rocky, gravelly slopes on foot-hills (5500- 8600 ft.) in sunny, exposed situations of central and southwestern Colorado (Fig. 28). Late June-August. Hymenopappus filifolius var. parvulus is one of the most distinet varieties of the H. filifolius group, being distinguished by its congested, small, few-flowered heads and by its almost complete laek of wool in the axils of basal rosette leaves (this latter character is shared only by var. nudipes in this complex). In the garden at Pullman, Washington, all these distinguish- ing characters appeared again on ten plants grown from seed collected in the field (Preece & Turner 2802), except that there was a tendency toward a more leafy rosette and generally in- creased size of vegetative organs. Johnston (1923), included this variety within his H. filifolius var. cinereus with some reservation, stating, ‘“Hymenopappus parvulus is referred here with great doubt, and I am inclined to believe that it probably is specifically distinct as it differs from the variety filifolius [sic] in its subscapose habit and small heads." This variety seems closest, morphologically, to the variety nudipes from which it differs in the fewer number, and smaller size of the florets, and its shorter-peduncled, more congested inflorescence. Though spatially the ranges of var. parvulus and var. cinereus overlap, the two are kept separate for the most part by altitu- dinal and topographic differences, H. filifolius var. parvulus being found usually at lower elevations on bare, sandy or gravelly slopes; var. cinereus occurs at somewhat higher elevations in clays and silts, mostly on vegetated valley bottoms and hill- sides. However, the two probably cross readily, and evidence [Vor. 58 Rhodora 218 0: 8-F SZ-O1 0c-£ (Z) 1—0 o'r 8'1-9'0 0 r-9'£ _ sn[nA.red £-S9'c 01-9 riami 0r—9 (Z) 1-0 Z'I-S9'0 S'Z-9'I 0°'9-0'F - sodrpnu £-9'C 01-9 001—0£ 068-61 ¢-@(—0) 0' I-F'O 0'I7c O0 0'9-9'F (+)+ + STI[OJIIU £-9'& 6-2 €t*—O0I 0£-9 6-0 OT €°1-9'0 0'9-S'F ++ sisuəou%ptr £-9'c 6-9 0e-c cI-€ (g)zZ—0 £-)1I-6'0 £-S'I(-D €'c-e't ++ snuvu c€'g-g 0r-9 09—0c SI-£ £-0 0°2-0'T S$'Z-6G'I 0°9-0'¢ ++ suəsn[ €'g£-g 01-4 OL-SF OF-9 L-g 0' I-9'0 0'Z-9'I 0:9-0'€ (+ +)+ snpodorio f€ — (FI-GI-s8 02-08 0g—ç 9-2(T) 0'Z-0'I 0'£-9'I 0°2-0°¢ + snjeydoorseur £-9'c OI-2 09-08 9-£ ZI-¢ 0 I—S:0 8'I-8'0 0'e-S'* ++ snsojueul0y Z-9'T 9-£ €£-0c 9-7 6- 0°Z-0'T S'I-9'0 €'t-0'£ + snaogroned 6 L-9 0c-€ 9-1 £-0 0'2-0'I 8'0-I'O 0°S-O'F ++ snojn| £-Z 6-9 OF-ST (SI-)8-@ t*-c(I-0 | 0O'£-O'I 0'Z-S'I 0'9-0' f ++ snoloUro £-e'c 8-6 09-07 . O£-O0I(—-P) 8-8(Z) 0'2-0' I 8'I-9'0 0°S-0'F + snjeydao4jod š> sh g = = SL SU SL X FEET SS B$ ` SS EIS 32 =š x s 3& ss Be Š$ S% ER EE: 2 = > 8 š sŠ sŠ => a= Ayana = <. 8 _ = s> * A = = E: 88 E 3 RES = X x €'I-9'0 OI-£ 0Z-O1 £-c S'I-9'I 0'£-0'Z — MOTION LI snjnased OI-c OI-I S$£-0I (€—)*-6 $'c-9'I 0'c-0'£ (+)- MOTOS LT sədrpnu SI-I Ot-9(-c) SSI €-9'€ g ¿2-9 I S'e (+)- MOT[OA LI SDI[OJIIU 9-1 FI-F 02-97 r-£ 0'z-€'I Sp- E (+)— Aollə£ n srsuooqept SI-£ 9-I 09—-0£ T-Z 0'Z-9'I 0 f-0'£ (+)- AOII9Á FSI LI snuvu eI-c 8-£ ZL-0Z 8-£ 0 £072 029-0 * (—)+ OUYM-MOTTPA FELT suəsnr 91-8 8-€ 09—0£ S-E 0 g-0 Z 0°S-0'F + '— EX EN — snpodorio oOl-z oI-€ 09—87 9—£ 0¢-0% O'L-0'F + '— AO[I94 — sn[eqdooesour 9-T FI-F 0f—0£ (9r-£ Sg'G-S'I S'O £ + *— AO[[94 Ll snsoyjuəuiro1 9-3 (0£—S1I-c 0£-0I £-6 9'I-0'I 0'£-0'G = AO[[94 ni sniogroned 4-1 06-6 SZ—SI +-g ` 0'Z-0'I 0'£-0'c + ‘- AO[[94 ni snojnj 9-I 9-1 0t-ec rec 0 4-9 '1 S'f-0'£ (+)- OUT M-MOTIOA LI snoJould 8-3 09-¢ 0S—-0z e-z 8'I-£'I €'£-0'£ _ AOI[[94 r£ snjeydao4 jod > < = Th > >= > = Q F ES ez $ 3 3o iE 89 29 Qa ze ^ < < < ° ° 32 Se a A °S = ° =° =Š 3° ° =s z “= 8 =g > S = S3 gÈ " š = 3 s 3 S 35 f]914DA a^ 3 š š 3 & =$ S E s 3 a SAITOJITIA SAdAVAONAWAH AO SHLILHINYA N3HONA LS SSONSUHOLAAICT MO SISA'IVNYV I WH'IHV.L 1956] Turner,—A Study of the Genus Hymenopappus 219 of introgression of one into the other may be seen at several places in the field, as well as among herbarium specimens. REPRESENTATIVE SPECIMENS.—Colorado. DOLORES CO.: boundary of Montezuma Nat’l. Forest, T39 N, R14 W, M. Ownbey 1475 (ps, GH, MO, NY, RM, UC, WS). GRAND CO.: Sulphur Springs, J. Wolf 476 (us). GUNNI- son co.: Gunnison, C. F. Baker 840 (ps, GH, MO, NY, POM, RM, UC, US, Ws). MINERAL CO.: 4.7 mi. S.W. of Creede, M. & G. B. Ownbey 3040 (NY, RSA, SMU, WS). PARK CO.: South Park, 1871, T. Meehan (NY). SAGUACHE CO.: hills about Elko, highway 50, Preece & Turner 2816 (SMU, Ws). 3b. Hymenopappus filifolius var. nudipes (Maguire), comb. nov. Hymenopappus nudipes Maguire, Am. Midl. Nat. 37: 143. 1947. Type examined (Ny): Utah, Kane Co., 15 mi. N. of Orderville, frequent under yellow pine and manzanita, sandy soil, 5800 ft., June 15, 1940, B. Maguire 18740. Hymenopappus nudipes var. alpestris Maguire, Am. Midl. Nat. 37: 144. 1947. Type examined (Ny): Utah. Iron Co. Cedar Breaks Rim, 10,400 ft., June 23, 1940, B. Maguire 19023. Plants perennial, 5-30(—45) em. high, sparsely tomentose to glabrate, up to 70 crowns from a single tap-root, larger rosette leaves 6-13 em. long, 3-5 em. wide, typically once-pinnate or with a few secondary divisions, these flattened and linear, 1-2 mm. wide, petiole frequently reddish- glabrate at base, comprising 1⁄—2⁄ of the leaf, without a dense tomentum in the axil; stem leaves 0 or 1(2), much reduced; heads campanulate, 1-12 per stem, 15-35-flowered, on ultimate peduncles 2-10 em. long; principal involucral bracts 6-10 mm. long, 2-5 mm. wide, yellow, or commonly reddish-membranous for 1-2 mm. from the obtuse tip; corollas yellow, 3-5 mm. long, tube glandular, 1.5-2.5 mm. long, throat campanulate, 1.5-2.5 mm. long with lobes reflexed; 2-4 times longer than the lobes; achenes 4.5-6 mm. long, pubescent with hairs 0.5-1.2 mm. long; pappus of 14-16 linear-oblong scales, 1.5-2.5 mm. long (rarely less); anthers about one-half exserted 2.5-3 mm. long; n = 17. DisrRIBUTION.—Barren, mostly exposed, gravelly, sandy, and sandy- clay soils, principally in Utah along the eastern slopes of the Wasatch Mountains southwestward to Kane Co., 6000-11,500 feet (Fig. 28). June-July. Hymenopappus filifolius var. nudipes is probably most closely related to var. nanus or var. parvulus. It differs from the former in the fewer florets, more glabrate leaves, and lack of a dense tomentum in the rosette leaf axils; it differs from the latter in having larger florets and heads, larger pappus, and a more open, longer-peduncled inflorescence. Thus var. nudipes seems to stand intermediate between these two varieties. The same situation exists as concerns their geographical distribution; 220 Rhodora [Vor. 58 var. parvulus occurs on the dry barren foothills of central and southwestern Colorado and var. nanus occurs to the west on the Great Basin ranges. The range of var. nudipes overlaps that of var. lugens in the southwestern part of Utah. Apparently, the two are kept separate, for the most part, by ecological factors; for example, var. nud?pes, at the type locality, occurs on sandy-clay, barren slopes, in what appear to be “pure” colonies. Not more than one-fourth of a mile distant and at somewhat higher elevations, iz] VL eset ENG Fra. 28. Distribution of some varieties of Hymenopappus filifolius. H. f. filifolius (large disks), H. f. nanus (circles), H. f. idahoensis (squares), H. f. tomentosus (small disks), H. f. nudipes (open triangles), H. f. pauciflorus (open squares), and H. f. parvulus (triangles), 1956] Turner, —A Study of the Genus Hymcnopappus 221 a “pure” colony of var. lugens may be found growing on a flat mesa top in sandy-gravelly soil. However, evidence of extra- varietal contamination or introgression 1s not completely lack- ing. Occasional plants, in the field as well as on herbarium sheets, are seen which have considerable amounts of tomentum in the leaf axils (but not so much as var. lugens), probably in- dicating gene-flow or introgression from that variety (Daly 7-BD). At lower elevations in the Bryce Canyon region where H ymeno- pappus filifolius var. nudipes comes close to the western-most range of var. cinereus, nearly complete intergradation of these two varieties occurs (Goodman «€ Hitchcock 1599; Preece & Turner 2476). Hymenopappus nudipes var. alpestris is considered to be a form of var. nudipes which occurs at somewhat higher elevations (10,000-11,500 ft.). It cannot be separated from other material of the variety by any constant, partieular characters, though there is a general tendency for a smaller, more pubescent, later- flowering plant to develop at these altitudes. Completely transi- tional types occur from higher to lower elevations, and com- parison of several collections from the type locality of Maguire's "ecotype" shows considerable variability in their characters. REPRESENTATIVE SPECIMENS.—Utah. DUCHESNE co.: about 18 mi. W. of Duchesne, highway 40, Ferris 11326 (CAs, DS). GARFIELD CO.: 6 mi. N. of entrance to Bryce Canyon, Goodman & Hitchcock 1599 (CAS, DS, GH, MO, NY, RM, UC). IRON co.: 1 mi. E. of Museum in narrow belt along Rim, 10-100 feet wide, Cedar Breaks, Preece & Turner 2436 (ws). KANE CO.: 15 mi. N. of Orderville, Maguire 18740 (GH, NY, MO, UC, US, WS). PIUTE co.: Brigham Peak, Marysvale, Aug. 29, 1894, M. E. Jones 5958k (POM, US). SEVIER CO.: Sawtooth, 15 mi. up Salina Canyon Keck 680 (CAs, GH, POM, UC). SUMMIT CO.: Uintah Mts., Mt. Elizabeth, Cow Hollow, Aug. 12, 1928, G. J. Goodman 552 (RM). uTAH co.: Soldier Summit, July 25, 1947, Holmgren, Boyle, & Ballenger 7055 (uc). Wyoming. LINCOLN CO.: Fossil, Aug. 9, 1895, G. W. Letterman (mo). (There are two specimens on this sheet representing two taxa, var. nudipes and H. filifolius var. poly- cephalus, neither of which has been collected near this locality. It seems more probable that the plant of var. nudipes goes with the label since the variety has been collected at Mt. Elizabeth, Utah, only a short distance across the border from southwestern Wyoming.) 3c. Hymenopappus filifolius var. luteus (Nutt.), comb. nov. Hymenopappus luteus Nutt., Trans. Am. Phil. Soc. 7: 374. 1840. Probable isotype examined (GH): “Rocky Mts., Nuttall!’ In the type 222 Rhodora [Vor. 58 3la Fic. 29-35. Leaf silhouettes of Hymenopappus filifolius varieties. Fig. 29a, b. H. f. lugens, yellow-flowered tetraploid race (P. & T. 2614). Fig. 30a, b. Same, yellow-flowered diploid race (P. & T. 2645). Fig. 3la, b, c. AH. f. luteus (P. & T. 2873). Fig. 32. H. f. nanus (P. & T. 2608). Fig. 33. H. f. parvulus (P. & T. 2802). Fig. 34. H. f. cinereus (P. & T. 2739). Fig. 35. H. f. idahoensis (P, & T. 3378). Al X ca. 1/3. 1956] Turner,—A Study of the Genus Hymenopappus 223 description Nuttall gave the following locality, “Rocky Mts., towards the Colorado of the west, particularly on Ham’s Fork.” The type was prob- ably collected in what is now Lincoln Co., Wyoming. Plants perennial, with slender stems, 10—-30(-45) cm. high, densely tomentose to nearly glabrate; rosette leaves 3-6 cm. long, 1-1.5 em. wide, bipinnately dissected with many very short divisions, 1-3 mm. long (re- sembling Achillea), the petiole composing less than 1⁄4 the leaf length; stem leaves 0-4, commonly 2, becoming reduced upward; heads 5-50 per stem, commonly 10 or more, campanulate, 15-30-flowered, on slender, often reddish, ultimate peduncles 1-7 cm. long; principal involucral bracts 5-7 mm. long, 2-3 mm. wide, yellow to reddish-membranous for 1-2 mm. from the acute to obtuse tip; corollas yellow, 2-3.5 mm. long, the tube densely glandular, 1-2 mm. long, the throat campanulate, 1.2 mm. long with lobes reflexed, (2-)3-4 times as long as the lobes; achenes 4-5 mm. long, evenly pubescent with hairs 1-2 mm. long; pappus of 14-18 linear oblong scales, 0.1-0.8 mm. long; anthers partially exserted, about 2 mm. long; n = 17. DistriBuTION.—Gravelly, calcareous or sandy foothills in the basin areas of central and southwestern Wyoming, extending into adjacent Utah (Uinta Basin) and Moffat Co., Colorado (Fig. 36). June-July. Hymenopappus filifolius var. luteus is a very distinct, easily recognized variety over a large part of its range; however, the characters of the variety intergrade completely with those of var. cinereus in Moffat Co., Colorado, and, to a lesser degree, with the characters of var. megacephalus in the Uinta Basin, Utah (E. H. Graham 8300, 9144; Ripley & Barneby 7793). Hymenopappus filifolius var. luteus is somewhat variable as to size of vegetative organs and number of heads produced, even within the same race, when habitats differ. Preece & Turner 2872, 2873, 2874 (all at ws) represent specimens that were collected from along a roadside in an old bladed area, from deeper soiled shallow diteh, and from an unbladed adjacent hillslope respec- tively. It was thought at first that there might be some con- spicuous genetic difference between the individuals comprising the three groups since there seemed to be considerable phenotypic variation, this appearing somewhat correlated with the habitats mentioned. Seed was collected from the extreme individuals of the groups (2872, 2873) within the race, and ten plants of each of these groups were grown from seed in a uniform garden. The resulting progeny (20 individuals) were essentially of the same type, being quite uniform in height, number of heads produced, and other features. Meiosis appeared to be normal. 224 Rhodora [Vor. 58 REPRESENTATIVE SPECIMENS.— Colorado. morrarr co.: 10 mi. W. of Maybell, Porter 3608 (ps, GH, RM, Uc, vs). Utah. DAGGETT CO.: 1.5 mi. S. of Manila, M. Ownbey 3255 (ws). viwTAH co.: N.W. of mouth of Split Mt. Canyon, above Island Park, Graham 9144 (au, wo). Wyoming. ALBANY CO.: Sheep Cr., Aug. 21, 1899, C. Schuckurt (?) (NY). BIG HORN co.: Trapper Cr., C. Finley 23 (nM). CARBON co.: Fort Steele, Nelson 7150 (GH, MO, NY, POM, RM, US). LINCOLN CO.: E. of Opal, Ripley & Barneby 7864 (CAS). NATRONA co.: Satanka Fm., near Alcova, Porter 4485 (GH, RM, RSA). SUBLETTE CO.: 6 mi. E. of Big Piney, E. B. & L. B. Payson 2594 (GH, MO, NY, POM, RM, UC). SWEETWATER CO.: Granger, Nelson 4695 (NY, RM, UC, ws). VINTA CO.: 16 mi. S.E. of Lyman, Rollins 1674 (DS. GH, MO, NY, UC, US). 3d. Hymenopappus filifolius var. pauciflorus (Johnston), comb. nov. Hymenopappus pauciflorus Johnston, Contrib. Gray Herb. n.s. 68: 97. 1923. Type examined (an): Utah. “along San Juan River, near Bluffs,” 1200-1500 m., Aug. 25-29, 1911, P. A. Rydberg & A. O. Garrett 9951 (Isotype Ny). Plants perennial, 20-35 em. high, tomentose to nearly glabrate, usually the much-branched stems from several crowns, giving a bushy aspect to the whole plant; rosette leaves 5-7 em. long, 1-2 em. wide, canescent to loosely tomentose, bipinnately dissected with short, linear divisions 1-5 mm. long, 0.5-1 mm. wide (resembling those of var. luteus); stem leaves 3-9, becoming reduced upward; heads 2-15(-30) per stem, sub- turbinate, 10—30-flowered, on ultimate slender peduncles 2-6 cm. long; principal involucral bracts 3-6 mm. long, 2-3 mm. wide, pubescent to nearly glabrate, yellow-membranous for about 1 mm. from the tip; corollas yellow, 2-3 mm. long, the tube densely glandular, 1-2 mm. long, the throat 1-1.6 mm. long with lobes reflexed, 2-3 times longer than the lobes; achenes 3-4.5 mm. long, evenly pubescent with hairs 1-2 mm. long; pappus of 14-18 linear oblong scales, 0.5-1.5 mm. long; anthers partially exserted, 1.6-2 mm. long; n = 17. DisrRIBUTION.—Low desert areas in sandy soils of southeastern Utah, adjacent Arizona and the southwestern-most corner of Colorado (San Juan R.), 3800-5500 feet (Fig. 28). May-September. Hymenopappus filifolius var. pauciflorus seems to have its closest relationship with varieties luteus and cinereus. In many respects it stands intermediate between the two. It differs from the former in having higher, more leafy stems with fewer heads, and from the latter in its more numerous heads which contain fewer, smaller florets. There seems to be considerable intergradation between H. filifolius var. pauciflorus and var. cinereus in the southeastern part of Utah at intermediate elevations where the two taxa 1956] Turner,—A Study of the Genus Hymenopappus 225 overlap in their ranges (Standley 7334); however, they are kept separate, for the most part, by altitudinal barriers. Var. cinereus occurs on the mountain ranges, in this area, from 5000-10,000 feet; var. pauciflorus, as noted above, occurs at lower altitudes from 3800-5500 feet. In addition, var. pauci- florus grades into var. tomentosus in southwestern Utah (S. B. Benson 146) and locally, in parts of its range, into var. mega- cephalus. The mechanism or method whereby this latter variety retains its distinctiveness in the range of var. pauci- florus is not known. Apparently they occur at similar eleva- tions, and evidence of introgression or contamination is seen in several collections (J. T. Howell 24714; M. E. Jones, May 19, 1914). In spite of the considerable intergradation with adjacent or overlapping varieties, var. pauciflorus retains a remarkable degree of uniformity throughout its range, especially in regard to its small florets and leafy stems. The *filifolius complex" in this area obviously needs extensive field study and detailed chromosomal investigation. REPRESENTATIVE SPECIMENS.—Arizona. APACHE CO.: N. end of Carrizo Mts., Standley 7334 (us). COCONINO co.: 1.5 mi. below Lee's Ferry, Cutler 3152 (ps, GH, MO, NY, SMU). NAVAJO CO.: Navajo Reserva- tion, Vorhies 130 (GH, Mo, Ny, UC). Colorado. MONTEZUMA CO.: San Juan Valley, Brandegee 1268 (Mo, Ny, vc). Utah. EMERY co.: San Rafael Swell, May 19, 1914, M. E. Jones (POM). KANE co.: Kaiparowitz Plateau, Apr., 1939, Tompkins (cas). SAN JUAN CO.: along San Juan R., near Bluffs, Rydberg & Garrett 9950 (NY, RM, UC, US). WAYNE CO.: 18 mi S. of Hanks- ville, June 10, 1947, Stoddart & Cook (vc). 3e. Hymenopappus filifolius var. eriopodus (Nelson), comb. nov. Hymenopappus eriopoda Nelson, Bot. Gaz. 37: 274-275. 1904. Type examined (nw): Utah. “Rocky volcanic draws. Diamond Valley," May 19, 1902, L. N. Goodding 880. Plants perennial, 40-80 em. high, sparsely to densely tomentose below, becoming glabrate above; principal rosette leaves sparsely tomentose to glabrate, 10-20 em. long, 3-7 em. wide, bipinnately dissected with mostly filiform, ultimate divisions, 10-20 mm. long, 0.4-1 mm. wide; stem leaves (2) 3-7, mostly glossy green and glabrate; heads 3-8 per stem, subturbinate to campanulate, 30—60-flowered, on slender ultimate pedun- cles 8-16 em. long (rarely less); principal involucral bracts 7-10 mm. long, 2-4 mm. wide, sparsely tomentose to glabrate, yellow-membranous for 1-2 mm. from the acute to obtuse tip (often tinged with red); corollas white (ochroleucous), 4-5 mm. long, the tube densely glandular to nearly glabrate, 1.8-2 mm. long, the throat campanulate 2-3 mm. long, with 226 Rhodora [Vor. 58 lobes reflexed, 3-5 times as long as the lobes; achenes 5.5-6 mm. long, evenly pubescent with hairs 0.5-1.5 mm. long; pappus of 12-16 linear- oblong scales, 1.5-2 mm. long; anthers partially exserted, 3-3.5 mm. long; chromosome number not known. DisrRiBUTION.—Limestone soils, associated with pines and junipers at high elevations (5,500-10,000 ft.) on isolated mountain ranges of H E l i T Lo. ` i Ja j PM Á ‘ TII ae M UWA Be. a | s es q 29 nay yet |! H ! lcd . p Pa A 4 oy "^A : K... cT d eee he ` ' | N H ` [ — ee S F i i e | i r i . Ü : K: : (i H 1 " `. oa dam o "- : `, , am, ; ' = i \ ` | i Í D a: + Ye... ` ' T t <4 I M H lea D M i " L n i Fic. 36. Distribution of Hymenopappus filifolius var. eriopodus (disks), H. f. megacephalus (triangles), and H. f. luteus (circles). southern Nevada, adjacent California (New York Mts.) and Washington Co., Utah (Fig. 36). Mid-May to mid-July. Hymenopappus filifolius var. eriopodus is a very distinct variety in its typical habitat, but at lower elevations it inter- grades completely with at least two yellow-flowered taxa of the H. filifolius group, var. megacephalus and var. nanus. Wilson's ranch in the Charleston Mts., Nevada, is outstanding in this 1956] Turner,—A Study of the Genus Hymenopappus 227 respect. At this locality var. eriopodus apparently extends down canyons to lower elevations where it comes in contact with var. megacephalus, a variety of the surrounding arid desert (1000-4500 ft.). Within this area of overlap clear Frs or their segregates have been collected (Clokey 8651; Maguire 17992), as well as both the putative parents. 'The introgression of var. eriopodus with var. nanus is less conspicuous, but its effect may be seen by the more leafy stem and pale yellow flowers in the latter variety. This apparent intergradation is conspicuous in the Beaver Dam Mts., Utah, where their ranges overlap (Preece & Turner 2558 cited under var. nanus). Hymenopappus filifolius var. eriopodus has the vegetative aspect of var. filifolius of the Columbia Basin, with which it has frequently been confused. Var. eriopodus can be distin- guished from var. filifolius by its white flowers, and larger size of pappus and throat. REPRESENTATIVE SPECIMENS.—California. SAN BERNARDINO CO.: Clark Mt., 1 mi. Š. of Coliseum Mine, Wolf 7069 (ps, RSA, UC, ws). Nevada. CLARK CO.: Charleston Mts., Kyle Canyon, Clokey 5618 (cas, MO, NY, POM, RM, RSA, UC, US, WS). LINCOLN co.: Deer Lodge, Galway 8265 (us). NYE co.: Nevada Nat'l Forest, Cherry Cr. Pass, Maguire & Holmgren 25610 (CAS, DS, GH, NY, UC, Us, ws). Utah. WASHINGTON co.: Diamond Valley, Goodding 880 (GH, MO, NY, POM, RM, US). 3f. Hymenopappus filifolius var. megacephalus, var. nov. Herbae perennes, caulibus plerumque tomentosis 1—6-foliatis, 30—60 em. altis; foliis inferioribus bipinnatis 8-20(-30) em. longis, segmentis 8-30 mm. longis, 1-2 mm. latis; involucris turbinatis campanulatisve, bracteis 8-12(-14) mm. longis; flosculis flavis 4-7 mm. longis, fauce 2-5 mm. longa; achaeniis 5-7 mm. longis; pappo conspicuo, squamellis 1.5-3 mm. longis. Plants perennial, 30-70 em. high, uniformly tomentose or (in the eastern part of its range) nearly glabrate; principal rosette leaves 8-20(—30) em. long, bipinnately dissected with coarse, flattened, linear divisions, mostly 8-30 mm. long, 1-2 mm. wide; stem leaves 2-6, becoming reduced up- wards; heads 3-14 per stem, subturbinate to campanulate, (20—)30—60- flowered, on ultimate peduncles 2-10 em. long; principal involucral bracts 8-12(-14) mm. long, 2-5 mm. wide, loosely tomentose, yellow to reddish- membranous for 1-2 mm. from the acute to obtuse tip; corollas yellow (apparently pale-yellow in some specimens) 4-6(-7) mm. long, the tube glandular, 2-3 mm. long, the throat campanulate-tubular, 2-4(-5) mm. long with lobes reflexed, 3-6 times longer than the lobes; achenes 5-7 228 Rhodora [Vor. 58 mm. long, pubescent with hairs 1-2 mm. long; pappus of 14-18 linear- oblong seales, 1-3 mm. long; anthers partially exserted, 3-4 mm. long; chromosome number not known. Type (row); Nevada. Clark Co.: Las Vegas, 1000 ft., April 29, 1905, M. E. Jones s.n., (POM, several sheets; DS, MO, US). DisTRIBUTION.—Sandy or gravelly soil in low desert areas of Clark Co., Nevada, adjacent California (Providence Mts.), eastern and southern Utah, adjacent northeastern Arizona, and western Colorado, 1000- 5500 ft. (Fig. 36). May-November. Hymenopappus filifolius var. megacephalus is a variety of peculiar distribution and wide variability. In Clark Co., Nevada, at lower elevations it is relatively constant in character, but at higher elevations of adjacent mountain ranges it inter- grades with var. eropodus (see discussion under that variety). In eastern Utah, western Colorado, and northern Arizona the taxon is quite variable, possibly as a result of long-time hybridi- zation and multiple introgression with at least five other varieties of the H. filifolius complex which surround or overlap its periph- ery: var. luteus in the north; var. cinereus in the east; var. pauci- florus in the southeast; var. lugens in the south; and perhaps locally with var. tomentosus in Washington Co., Utah. (For further information see discussion under the appropriate varie- ties.) In spite of these infringements, the variety maintains a line of continuity which is characteristic over most of its range and for a large part of the material examined. In general the variety has fewer leaves and is less tomentose in the eastern portion of its range as contrasted with its appearance near the type area. Var. megacephalus has apparently been responsible for much of the confusion existing in the H. filifolius complex of eastern Utah, northeastern Arizona, and western Colorado. Herbarium material of the variety from this region carried a hodge-podge of annotations: H. lugens, cinereus, pauciflorus, tomentosus, eriopodus, ete. In Colorado the name ZH. ochroleucous was given by Greene to a local (?) white flowered race of var. cinereus (as treated in this paper), which apparently has some of the charaeteristies of var. megacephalus (e.g., larger, coarser-pin- natifid leaves), but much more of the characteristics of var. cinereus proper. The type locality for Greene's species is near the eastern-most periphery of the range of var. megacephalus. 1956] Turner,—A Study of the Genus Hymenopappus 229 The possibility exists that the variety here typified includes only the individuals from Clark Co., Nevada, and vicinity, and that most of the remaining material to the east represents either a weakly defined separate variety or a common area of extensive hybridization and introgression among the several peripheral taxa mentioned above, such as that suggested by Gleason (1923) for Vernonia illinoensis. Gleason described this species from the Great Lakes region in 1906. Sixteen years later (1922), he decided that what he described as a species was, in reality, a hybrid complex of crosses and backcrosses resulting from the introgression of three species one into the other at a common area of overlap. Accordingly, he dropped the name. In the case of H. filifolius var. megacephalus the continuity of the combination of characters listed in the description and used in the key, combined with the correlation of supposed introgressant individuals with geographical overlap, argues against this latter hypothesis. In short, the peripheral variability is not random throughout the area of occupation, nor is it meaninglessly scattered. Further field study is needed to clarify the situation fully, but it is believed that the treatment used here is con- servative as well as practical. The variety is named for its large heads which are among the largest of the Hymenopappus filifolius complex, equaled or rivaled only by occasional specimens of the closely related var. lugens. REPRESENTATIVE SPECIMENS.—Arizona. APACHE CO.: National Petri- fied Forest (Natural Bridge), Osterhout 6855 (RM). COCONINO co.: Š. side of Painted Desert, Osterhout 6962 (POM, RM). NAVAJO co.: Betatakin, J. T. Howell 24518 and 24519 (cas). These last two cited specimens present good evidence for hybridization and introgression of var. lugens into var. megacephalus. California. SAN BERNARDINO co.: Providence Mts., Brandegee (uc, vs). Colorado. mesa co.: DeBegue, Osterhout 4281 (GH, RM). MONTROSE CO.: Naturita, Payson 365 (GH, MO, RM, WS). Nevada. cLaRK co.: Las Vegas, Goodding 2287 (GH, Mo, RM). Utah. CARBON CO.: Price, June, 1898, Stokes (uc). EMERY co.: 3.25 mi. N. of Emery, M. E. Jones 5456d (pom). GRAND co.: Moab, June 9, 1913, M. E. Jones (NY, POM). KANE CO.: 1 mi. S. of Zion Junction, Maguire 18888 (GH, NY, WS). SAN JUAN CO.: 3 mi. S. of White Mesa Dugway, Holmgren & Hansen 3406 (GH, MO, NY, UC, US, WS). UINTAH CO.: Uinta Basin, 4 mi. w. of Willow Cr., E. slope of Big Pack Mts., Rollins 1709 (GH, MO, NY, WS). WASHINGTON co.: Zion Nat'l. Park, Summit of Lady Mts., T. Craig 1433 (POM). 230 Rhodora [Vor. 58 3g. Hymenopappus filifolius var. lugens (Greene) Jepson Hymenopappus lugens Greene, Pittonia 4: 43. 1899. Hymenopappus filifolius var. lugens, Jepson, Man. Fl. Pl. Calif. 1128. 1926. Lectotype chosen (us): California. San Bernardino Co.: “Bear Valley,” San Ber- nardino Mts., on their eastern base, 6500 ft., June 24, 1894, S. B. Parish 3113. Greene did not designate a specific type, but based his name on two collections of Parish ‘‘whose earliest specimens were sent out as representing a variety of H. luteus, the later ones being labeled H. fili- folius." The collection cited above should serve as the type since it is a well-preserved specimen and is represented by isotypes in at least two other herbaria (Mo, NY). Hymenopappus gloriosus Heller, Bull. Torr. Bot. Club 26: 551. 1899. Type examined (ny): Arizona. Coconino Co.: “Collected about Mormon Lake,” 6000 ft., June 7, 1898, Dr. D. T. MacDougal 71. Hymenopappus scaposus Rydb., Bull. Torr. Bot. Club 27: 634. 1900. Type examined (Ny): Arizona. Coconino Co.: “vicinity of Flagstaff," : 7000 ft., June 18, 1898, Dr. D. T. MacDougal 1229. Hymenopappus macroglottis Rydb., Bull. Torr. Bot. Club 27: 636. 1900. Type examined (ny): Arizona, Coconino Co.: “Oak Creek, Aug. 5, 1883," N. H. Rusby s.n. The type does not have a collection number, but probable isotypes bear the collection number 692. Hymenopappus nudatus Wooton & Standley, Contrib. U. S. Nat. Herb. 16: 191. 1913. Type examined (us): New Mexico. Grant Co.: Burro Mts., 5500 ft., June 6, 1903, O. B. Metcalfe 107. Plants perennial, (20-)30-60 cm. high, densely grayish-tomentose to greenish-glabrate, except in the woolly basal leaf axils; rosette leaves (5-)6-14 em. long, bipinnately dissected with flattened, linear ultimate segments, mostly 3-10 mm. long, 1-3 mm. wide, petiole composing V6-24 of the leaf; stem leaves 0-2(-3), much reduced; heads 3-8 per stem, campanulate, 20-72-flowered, on ultimate peduncles 2-12 cm. long; principal involucral bracts 6-10 mm. long, 3-4 mm. wide, gradate or in two or three main series, glabrous to densely tomentose, the apex con- spicuously reddish to yellow-membranous for 1-2 mm. from the mostly obtuse to less often acute tip; corolla bright yellow or less often white, 4-6 mm. long, the tube sparsely glandular, 1.5-2 mm. long, the throat campanulate-tubular 2.2-3 mm. long, with lobes reflexed, (2.2—)4-8 times longer than the lobes; achenes 5-6 mm. long, pubescent with hairs 1-1.5(-2) mm. long; pappus of 16-18 linear-oblong scales, 1.2-2.5 mm. long; anthers partly exserted, 3-3.5 mm. long; n — 17, 34. DisTRIBUTION.— Mountain ranges of southern California and northern Baja California; northwestern to southeastern Arizona, adjacent south- western New Mexico, Nevada (Virgin Mts.), and southwestern Utah, mostly at 5000-9000 ft., commonly associated with junipers, pines, or oaks in gravelly, rocky, clay, volcanie, schistose, and limestone soils (Fig. 37). May-September. This is the most variable and perplexing variety of the Hy- 1956] 'Turner, —A Study of the Genus Hymenopappus 231 , y A pal Lm | a aS ve Fia. 37. Distribution of some varieties of Hymenopappus filifolius. H. f. lugens, yellow-flowered races (disks), white-flowered race (circles); H. f. cinereus (plus signs), and H. f. polycephalus (triangles). menopappus filifolius complex. The treatment here accorded this variety is admittedly inadequate and future studies will undoubtedly show that the variety, as delimited in the present treatment, is composed of at least two, and possibly three, wide- spread races that are perhaps worthy of varietal rank. "Though 232 Rhodora |Vor. 58 aware of these races, the author has been unable to distinguish between them with any degree of certainty from herbarium material, and it seems probable that proper delimitation will have to await detailed field and chromosomal study. For this reason, and on practical grounds, the group has been treated as a well marked variety composed of several widespread and intergrading races. In southern California and Baja California the variety is relatively constant in character. Almost without exception, specimens from this area have markedly gradate involucral bracts with red, almost crimson tips and edges; the florets are nearly always few in number (20-30); and the corolla throat is 5-8 times longer than the lobes. The variety has its greatest complexity in northwestern Arizona (Mohave, Yavapai, and Coconino counties). In this region two, and possibly three, races of var. lugens occur: a yellow-flowered diploid race with essentially the same mor- phology as the California race, except that it has somewhat more florets to a head (20-40); a tetraploid yellow-flowered race with about the same morphology as the above, except that the bracts are not so gradate and are mostly without the red or crimson margins and the florets are more numerous (30-72); a white-flowered tetraploid race which has heads with 20-30 florets, the corollas with smaller throats, and which possesses smaller leaves than the above two races. The diploid and tetraploid yellow-flowered races occur in the same general area (cf. Preece & Turner 2610, 2614, 2618). On herbarium sheets, as well as in the field, it is most difficult to distinguish between these groups, the characters mentioned above not being constant. When seed of both diploid and tetraploid plants were grown in a uniform garden, the tetraploids tended to be larger plants, with larger leaves (Fig. 29) and more numerous florets; diploid plants were smaller, with fewer florets. However, some of the tetraploid plants that were grown in the garden were not so well marked, being small and flowering late with fewer florets (perhaps due to injury during potting). Be- cause of this strong resemblance between diploid and tetra- ploid races and because the tetraploid race regularly showed 1 to 3 quadrivalents at meiosis, it was suspected early in the 1956] Turner,—A Study of the Genus Hymenopappus 233 study of this group that the tetraploid arose by autoploidy from typical var. lugens. However, subsequent investigation revealed the presence of a close relative, var. nanus in this region, which apparently introgresses with the diploid race mentioned above. (Var. nanus normally lacks the red-tinged involucral bracts, and has more numerous florets with shorter throats than does the diploid form of var. lugens.) Thus it now seems more likely that the tetraploid is a segmental amphiploid as defined by Stebbins (1950). Typically, diploid var. lugens occurs in this area at higher elevations in pine woods, whereas tetraploid var. lugens seems to occur more often in the transition zone of pine woodland and juniper belts, or exclusively in the latter region. The white-flowered race was not seen by the author west of Flagstaff in this area. It is possible that most, if not all, white- flowered var. lugens in Arizona is tetraploid. Unfortunately, most collectors have omitted flower color from their collection labels. The circles on the distribution dot map of var. lugens (Fig. 37) indicate those collections examined by the author that were obviously white-flowered (either the color was so indicated on collection labels, or the specimens were well enough preserved to determine this). Field observation indicates that these white-flowered plants are not just segregates in a popula- tion of primarily yellow-flowered individuals. In all instances the populations examined by the author were composed entirely of white-flowered individuals, though some populations were intermediate in this respect (i.e., all pale yellow). This white- flowered, tetraploid race loses most of its distinctiveness south- eastward from Coconino Co., Arizona, to Grant Co., New Mexico, as it seems to grade into a densely tomentose,? yellow-flowered race with long throats (perhaps the same tetraploid race found in Mohave and Coconino counties). Var. lugens intergrades to a small degree with var. nudipes and var. tomentosus in southwestern Utah, with var. cinereus and var. megacephalus in northeastern Arizona, and locally with var. nanus in Mohave Co., Arizona, and possibly in the Inyo Mts., California. ° The pubescence of these races is probably not a good character on which to dis- tinguish them, Plants grown in the garden from seeds taken from specimens that were quite tomentose in the field appeared nearly glabrate. 234 Rhodora [Vor. 58 In review, Hymenopappus filifolius var. lugens is a highly variable taxon of the southwestern United States which is composed of two or three closely related widespread races. These races are both diploid and tetraploid and their mor- phologies are probably complicated by introgression and inter- gradation from within and from without. In view of the com- plexities involved, it is believed that formal recognition should not be accorded these infraspecifie units until detailed study justifies such action. REPRESENTATIVE SPECIMENS.— Arizona. APACHE CO.: 3 mi. N. of Ft. Defiance, Cutler 2135 (cas, Mo, UC). CocHIsE co.: 5 Mi. E. of Burro Mts., Maguire 11570 (NY, us). COCONINO co.: Cape Royal, N. rim of Grand Canyon, Goodman & Hitchcock 1639 (ps, MO, NY, RM, UC): 9 mi. E. of Peach Springs, Preece & Turner 2610 (smu, ws); 2 mi. E. of Ashfork, Preece & Turner 2618 (smu, ws); Meteor Crater rim (on boulder near entrance), 38 mi. E. S. E. of Flagstaff, Preece & Turner 2667 (ws). GILA co.: Globe, Nelson 10341 (RM, UC). GRAHAM CO.: 8 mi. up Swift Trail Rd., Pinaleno Mts., B. & R. R. Maguire 11739 (GH, NY). GREENLEE co.: 27 mi. N. of Clifton, Maguire 11073 (NY). MOHAVE co.: Virgin Mts., pass above Whitney Ranch, Munz 16783 (Mo, POM, WS). NAVAJO CO.: 12 mi. N. of White River, White Mts., Goodman & Hitchcock 1310 (cas, DS, GH, MO, NY, RM, UC). PINAL CO.: Oracle, Aug. 27, 1903, M. E. Jones (POM). SANTA CRUZ CO.: 2 mi. N. of Vaughn, May 27, 1938, J. F. Arnold (CAS, GH, US). YAVAPAI CO.: 5 mi. west of Seligman, Preece & Turner 2614 (smu, ws). California. RIVERSIDE co.: El Toro Mts., Coyote Canyon, H. M. Hall 1178.1 (ps, UC, US). SAN BERNARDINO CO.: near Bear Valley, Abrams 2899 (CAS, GH, MO, NY, POM, UC, US): San Bernardino Mts., South Fork, upper Santa Ana Canyon, Hall 7507 (ps, GH, MO, NY, POM, RM, UC, US). SAN DIEGO CO.: between Julian & Cuyamaca (“Gumbo Hill opposite Talley's"), July 16, 1906, K. Brandegee (GH, NY, POM, UC, us). Nevada. cLARK co.: “Mica Mines" (Mica Springs), M. E. Jones 50729 (us). New Mexico. carron co.: 35 mi. N. of Alma, highway 260, June 26, 1951, Preece & Turner 2710 (smu, ws). GRANT CO.: Silver City, Eastwood 8402 (cas, GH, MO, us): Burro Mts., O. B. Metcalfe 107 (GH, MO, NY, POM, RM, UC, Us). Utah. BEAVER co.: 3 mi. E. of Beaver, May 30, 1913, H. Redeker (RM). GARFIELD CO.: 5 mi. S. of Panguitch, Maguire 18710 (NY, WS). KANE co.: 3 mi. W. of Long Valley Junction, Preece & Turner 2466 (SMU, ws). PIUTE co.: Marysvale, M. E. Jones 5355 (POM, US). SEVIER CO.: near Richfield, L. F. Ward 168 (GH, Mo, US). WASHINGTON CO.: near St. George, C. C. Parry 107 (GH, MO, NY). Mexico. Baja California. Tantillas Mts., “just over boundary (in) Lower California," Sept. 9, 1875, E. Palmer 183 (GH, MO, NY, UC, US). 1956] Turner,—A Study of the Genus Hymenopappus 235 3h. Hymenopappus filifolius var. polycephalus (Osterhout), comb. nov. Hymenopappus polycephalus Osterhout, Torreya 18: 90. 1918. Type examined (nw): “North of Livermore, Larimer Co., Colo., Aug. 11, 1917,” G. E. Osterhout 5680. Plants perennial, 30-60 em. high, one to several crowns from a single, woody tap-root; stems slender, single from each crown, much-branched, densely tomentose to nearly glabrate, often with conspicuous sessile glands; principal rosette leaves sparsely tomentose to nearly glabrate, 6-15 em. long, 4-6 em. wide, bipinnately dissected into linear, almost filiform, glandular-punctate segments 8-30 mm. long and about 0.5 mm. wide; stem leaves 3-8, becoming reduced upwards; heads 5-50 per stalk (often more on late-flowering specimens), 20—50-flowered, on ultimate peduncles 2-8 em. long; involucral bracts 5-8 mm. long, 2-4 mm. wide, yellow-membranous (very rarely red-tinged) for 1-2 mm. from the acute to obtuse tip; corollas yellow, 3-3.5 mm. long, the tube densely glandular, 1-1.8 mm. long, the throat campanulate, 1.3-1.8 mm. long with lobes reflexed, 2-3 times as long as the lobes; achenes 4—5 mm. long, densely pubescent with hairs 1-2 mm. long; pappus of 12-16 linear-oblong scales, 0.8-2 mm. (rarely less) long; anthers partially exserted, 2.5-2.8 mm. long; n = 34. DisrRiBUTION.—Holling, sandy plains and grasslands east of the Rocky Mountains from central Colorado to Canada (southern Alberta and Saskatchewan), extending on intermontane plains west to north- central Wyoming, western Montana, and east to western North Dakota, South Dakota, and Nebraska (Fig. 37). June-August. The variety has the aspect of H. tenuifolius, especially when poorly collected specimens are dried and placed on herbarium sheets. In the field and on well preserved specimens, the variety can be distinguished at once from this species by the yellow flowers and perennial root. H. tenuzfolius is a biennial with white flowers. Osterhout proposed the specific name polycephalus in 1918, but it was not taken up by later workers. He apparently was very familiar with the plant in the field, and his numerous col- lections preserved at the Rocky Mountain Herbarium, with their annotated comments, indicate that he was dissatisfied with placing it either under H. cinereus Rydb. or H. tenuifolius. Nevertheless, Osterhout thought polycephalus was more closely aligned with H. tenuifolius since, in describing the species, he stated, “The close relationship of Hymenopappus polycephalus, however, is not with H. cinereus but with H. tenuifolius Pursh. It is as high and as leafy as H. tenuifolius, but the flowers are 236 Rhodors [Vor. 58 yellow, not dull white, and it is a perennial, not a biennial.” Johnson treated the entity as a questionable synonym of H. cinereus stating, “Hymenoppapus polycephalus is a puzzling form which suggests a perennial phase of H. tenuifolius.” In the present treatment, polycephalus has been treated as a variety of JI. filifolius since its total characteristics indicate a much closer relationship to the members of this species. In spite of this closer affinity, var. polycephalus does have several characters which possibly link it to H. tenuifolius; indeed, there seems to be a considerable degree of intergradation be- tween the two entities with respect to leaf dissection, amount of tomentum on the stem, and throat/lobe ratio. It is interesting to note that var. polycephalus, so far as known, is tetraploid with n = 34, while both var. cinereus and H. tenuifolius are diploid with n = 17. Var. polycephalus, in Colorado, occupies a geographical position between these two taxa. From these facts it might be speculated that at some past time the Rocky Mountain var. cinereus came in contact with H. tenuifolius so that extensive hybridization occurred on the diploid level and concomitantly amphiploidy occurred. (Evidence of a once farther eastward extension of H. filifolius var. cinereus is found in the disjunct races of this variety in the panhandle region of Texas, a considerable distance from the normal range now occupied by the variety.) With recession of this peripheral contact at a later time, the amphiploids sur- vived in the habitats in which they were produced and have since spread to the area they now occupy. In the few meiotic slides prepared of var. polycephalus pairing appeared normal, there being 34 bivalents. Unfortunately, seed of this variety was germinated too late in the study of this group to provide material for genetical experiments to test this hypothesis. REPRESENTATIVE SPECIMENS.—CANADA, Alberta. Milk R., Macoun 10937 (GH). Saskatchewan. Coteau de Missouri (Base of Coteau), Macoun 98 (971) (GH, NY, US). Unirep STATES. Colorado. BOULDER co.: Lyons, Fritchey 50 (Mo). EL PASO CO.: edge of Black Forest, 15 mi. N. of Colorado Springs, God- dard 961 (UC). JEFFERSON CO.: | mi. w. of Golden, Beetle 2034 (GH, RM). LARIMER CO.: Fossil Cr., Crandall 3202 (NY, RM, Us, Ws). MORGAN CO.: Narrows, opposite mouth of Bijou Cr. Ward (Us). WELD CO.: near Keota, Osterhout 6035 (RM, vc). Montana. BEAVERHEAD Co.: Bannock City, S. Watson 220 (GH, US). BROADWATER CO.: 10 mi. Š. E. of Helena, Hitchcock & Muhlick 11782 (ws). CASCADE co.: Great Falls, Blankinship 1956] Turner,—A Study of the Genus Hymenopappus 237 721 (MO, POM, RM, US). GALLATIN co.: St. Joe Cr., 1899, W. W. Jones (uc). LEWIS AND CLARK CO.: about 12 mi. N. W. of Silver City, July 2, 1948, Hitchcock 17917 (RSA, UC, WS). MADISON co.: about 10 mi. S. W. of Crockett Lake, Gravelly Range, Hitchcock & Muhlick 12534 (ws). ROOSEVELT CO.: Culbertson, E. L. Larsen 123 (MO). SILVER BOW CO.: Melrose, Shear 5023 (Ny, Us). TETON co.: Deep Cr. Canyon, Aug. 14, 1899, Blankinship (RM). WHEATLAND co.: 20 mi. S. of Harlowton, Hitchcock 16314 (RSA, ws). YELLOWSTONE CO.: Custer, June 10, 1890, Blankinship 145 (Mo, uc, vs). Nebraska. BANNER co.: Wild Cat Mts., July 16, 1891, Rydberg (Ny, US). CHASE co.: S. E. of Enders in French- man Valley, Tolstead 411427 (uc). CHERRY co.: Merriman, July 11, 1899, J. M. Bates (RM). DAWES co.: Dunlap, June 16, 1939, H. Hapeman (mo). HOOKER CO.: Middle Loup R., near Mullen, Rydberg 1554 (Us). LINCOLN co.: (w/o locality) Barker 2600 (Mo). MORRILL co.: Broadwater, June 11, 1935, H. Hapeman (Mo, NY). SCOTTS BLUFF co.: N. of Mitchell, Osterhout 7208 (RM). stoux co.: Rocky outcrops, J. Kramer 7 (Ds, MO). THOMAS co.: Halsey, June 26, 1912, R. J. Pool & D. Folsom (wo). North Dakota. pivipr co.: Alkabo, E. L. Larsen 95 (GH, MO). BILLINGS co.: Medora, Bolley 402 (NY, RM). HETTINGER CO.: Mott, O. A. Stevens 143 (vc, vs). SLOPE co.: Marmarth, L. R. Moyer 480 (NY). STARK co.: Dickinson, O. A. Stevens 713 (Mo, UC, US). WILLIAMS CO.: Williston, W. B. Bell 246 (xm, uc). South Dakota. cusrER co.: Fairburn, Hayward 1498 (RM). FALL RIVER CO.: 3 mi. Š. of Hot Springs, P. Johnson 229 (GH, NY). HARDING CO.: Table Mt., Cave Hills, Visher 408 (RM). PENN- INGTON CO.: near Wall, E. J. Palmer 37252 (aH). Topp co.: Rose Bud Cr., E. J. Wallace 102 (NY). WASHABAUGH co.: Bear Cr., Visher 2085 (NY). Wyoming. ALBANY co.: Sybille Hills (Cr.), A. Nelson 328 (an, MO, NY, RM, US). BIG HORN CO.: Otto, Aug., 1878, N. Edwards (NY). CONVERSE CO.: E. of Douglas, C. L. Turner 92 (RM). cRook co.: Ewing, S. Seig 7 (RM). FREMONT co.: 13 mi. S. of Dubois, Porter & Rollins 5773 (RM). GOSHEN CO.: Ft. Laramie, A. Nelson 8310 (GH, MO, NY, POM, RM, US). JOHNSON co.: near Kaycee, A. & R. Nelson 1101 (GH, MO, NY, RM, UC, US, WS). LARAMIE CO.: Table Mt., A. Nelson 1369 (RM). NATRONA co.: Bates Cr., L. N. Goodding 199 (Ds, POM, RM, UC). NIOBRARA CO.: Lusk, F. H. Knowlton 106 (vus). PLATTE CO.: Guernsey Lake St. Park, C. L. Porter 4000 (GH, RM, SMU, UC). SHERIDAN co.: E. of Sheridan, S. S. Sharp 156 (nM). Weston Co.: near Newcastle, Degener & Peiler 16235 (MO, NY). 3i. Hymenopappus filifolius var. tomentosus (Rydb.), comb. nov. Hymenopappus tomentosus Rydb. Bull. Torr. Bot. Club 27: 633. 1900. Type examined (wv): Utah. “St. George," 1877, Dr. E. Palmer 270. Hymenopappus niveus Rydb. N. Am. Flora 34: 52. 1914. Type examined (Ny): Utah. “Springdale,” 4000 feet, May 14, 1894, M. E. Jones 5261. Plants perennial, 30-70 em. high, permanently densely white tomentose; larger rosette leaves 6-10 cm. long, 2-3 em. wide, bipinnately parted with many short, filiform divisions 3-6 mm. long; stem leaves 5-12, not much 238 Rhodora [Vor. 58 reduced; heads 4-14 per stem, subturbinate to campanulate, 30-40- flowered, on ultimate peduncles 1-6 em. long; principal involucral bracts 7-10 mm. long, 2-3 mm. wide, densely tomentose, yellow-membranous for 1-2 mm. from the acute tip (rarely red-tinged); corollas yellow, 3-4.5 mm. long, the tube densely glandular, 1.5-2 mm. long, the throat 1.5-2 mm. long with lobes reflexed, 3-4(-6) times longer than the lobes; achenes 4.5-5 mm. long, evenly pubescent with hairs 0.5-1 mm. long; pappus of 14-16 linear oblong scales 0.8-1.8 mm. long; anthers partially exserted, 2.5-3 mm. long; n = 17. DisTRIBUTION.—Sandy soil on rocky limestone hills and mesa tops, 3500-7000 ft.; known only from colleetions in Washington and western Kane counties, Utah (Fig. 28). May-June. Hymenopappus filifolius var. tomentosus occupies a position seemingly at the center of the H. filifolius complex as concerns maximum variability and total number of taxa present. In spite of its restricted range, the variety, as viewed from herbarium and field observation, shows evidence of intergrada- tion or introgression with at least three taxa of the H. filifolius complex: var. pauciflorus, var. megacephalus, and var. lugens. These varieties are kept separate for the most part by ecological and/or geographic barriers: var. lugens occurs at somewhat higher elevations on gravelly or sandy-clay soils; var. mega- cephalus occurs at lower elevations on sandy soils; var. pauci- florus occurs toward the southwestern periphery of the range of tomentosus on deep sandy soils. In this latter area almost complete intergradation between the two taxa is found. At one locality in Kane Co. (1 mi. S. of Glendale, Preece & Turner 2503) a few specimens, apparently intermediate between var. tomentosus and var. lugens, were found growing in what otherwise appeared to be a “pure” colony of this latter variety. These few intermediates were still in bud so that they lagged at least 1-2 weeks behind the others which were in full bloom. The intermediates collected were of good size (apparently 2-3 years plus in age). Ten plants grown in a uniform garden from seed of var. lugens from this locality maintained the char- acteristics of that variety, and no evidence of contamination from var. tomentosus was found. Seed (or open flowers) from the intermediates could not be obtained. Extensive inter- gradation of characters between varieties lugens and tomentosus apparently does not occur, though the varieties are probably fully compatible. Further field and genetical studies of the Hymenopappus filifolius complex in this area will undoubtedly 1956] Turner,—A Study of the Genus Hymenopappus 239 yield much information as to isolation mechanisms within several, spatially overlapping, closely related infraspecific taxa. Variety tomentosus has the flower size and shape of var. mega- cephalus, but the vegetative characteristics are those of var. pauciflorus. Its rather complete intergradation with these two taxa to the southeast and southwest makes the hypothesis that var. tomentosus had its origin from an early time of more exten- sive overlap and introgression attractive, although it is probable that the situation is much more complex than that visualized. Hymenopappus niveus Rydb. is a name based upon specimens which show some evidence of introgression from var. mega- cephalus (e.g., leaves somewhat more coarsely pinnatifid) ; other- wise it is typical of the variety tomentosus as here delimited. REPRESENTATIVE SPECIMENS.—Utah. KANE co.: 1 mi. S. of Glendale, highway 89, Preece & Turner 2503 (ws). WASHINGTON CO.: near Canaan Ranch, M. E. Jones 5262C (wo, pom, uc): St. George, 1877, E. Palmer 270 (GH, MO, NY, UC). 3j. Hymenopappus filifolius Hook. var. filifolius Hymenopappus filifolius Hook. Fl. Bor. Am. 1: 317. 1833. Probable isotype and fragment of type examined (Gu): “on the undulating grounds of the Columbia, near the Wallawallah" Douglas (quoted from informa- tion given in the original description). Hymenopappus tenuifolius Dougl. (non Pursh) in Fl. Bor. Am. 1: 317. 1833, as synonym. Hymenopappus columbianus Rydb. N. Amer. Fl. 34: 52. 1914. Type examined (us): Washington. “near Moses Lake, Douglas Co.” (now Grant Co.), 390 meters, July 4, 1893, J. H. Sandberg & J. B. Leiberg 975. Plants perennial, 30-100 em. tall, densely white tomentose to nearly glabrate; principal rosette leaves 10-20 em. long, bipinnately dissected with long filiform divisions, 10-50 mm. long, 0.5-1.5 mm. wide; stem leaves (0-)2-5, becoming reduced upward; heads (2-)6-40 per stem, campanulate, (15-)20-45-flowered, on ultimate peduncles 1-15 em. long; principal involucral bracts 6-10 mm. long, 3-4 mm. wide, densely tomen- tose to nearly glabrate, yellow-membranous for 1-2 mm. from the acute to obtuse tip; corollas yellow, 2.5-4.5 mm. long, the tube densely glandu- lar, 1.2-2 mm. long, the throat 1.6-2.5 mm. long, 3-5 times longer than the lobes, with lobes reflexed; achenes 4.5-6 mm. long, evenly pubescent with short hairs, 0.4-1 mm. long; pappus of 16-22 linear-oblong scales, 0.2-1 mm. long; anthers partially exserted, 2.5-3 mm. long; n = 17. DIsTRIBUTION.—Sand dunes and gravelly, sandy soils of the Columbia Basin of Washington and barren sandy-clay or clay soils of adjacent Oregon (John Day Valley, south to Harney Co., Fig. 28). May-Sep- tember. Hymenopappus columbianus Rydb. is a leafy, densely tomentose form 240 Rhodora [Vor. 58 of the variety. This is the common aspect of the plant in Washington; southward into Oregon both tomentose and glabrate forms occur, as well as leafy and non-leafy types. In the southern part of its range the taxon picks up characters of var. nanus, a variety of the Great Basin. However, var. filifolius is held together by a combination of characters, the most constant being the small pappus and shortly pubescent achene. Variety filifolius closely resembles var. polycephalus of the northern Great Plains, with which it has been confused in nearly all manuals. Variety polycephalus is a tetraploid with smaller throats to the corollas, less wool in the leaf axils, and shorter peduncles. REPRESENTATIVE SPECIMENS.—Oregon. CROOK co.: sandy, sterile hills on the Burns-Prineville road, Cusick 2631 (GH, MO, NY, POM, RM, UC, US). GILLIAM co.: hills S. of Arlington, Henderson 14472. (uc). GRANT CO.: Squaw Cr., Humphrey’s Ranch, John Day Valley, Henderson 5236 (CAS, DS, GH, MO, RM, US). HARNEY CO.: lower Willow Cr. Canyon, Peck 18990 (uc). MORROW co.: Paterson Ferry, Eastwood & Howell 3517 (cas). wasco co.: Muddy or Muddy Station (Muddy Ranch?), John Day Valley, May 12, 1885, T. Howell 447 (am, ny, US, Ws). WHEELER CO.: 5 mi. N. of Mitchell, Peck 10081 (NY, ws). Washington. ADAMS CO.: 12 mi. W. N. W. of Washtuena, Turner 2293 (smu, ws). BENTON CO.: 5 mi. N. W. of Richland, L. S. Rose 48154 (CAs, RSA, US). FRANKLIN CO.: 8 mi. N. of Kahlotus, Constance & McMurray 1139 (an, MO, US). GRANT CO.: near Moses Lake, Sandberg & Leiberg 375 (cas, GH, MO, NY, POM, RM, UC, US, WS). WALLA WALLA CO.: near Walla Walla, Brandegee 903 (GH, UC). YAKIMA co.: 3 mi. W. of Mabton, highway 3A, June 15, 1950, R. E. Fue 29 (ws). 3k. Hymenopappus filifolius var. nanus (Rydb.), comb. nov. Hymenopappus nanus Rydb. N. Amer. Fl. 34: 53. 1914. Type examined (NY): Nevada. Elko Co.: “ridge above Cave Creek Post Office,” East Humboldt or Ruby Mts., 9300 ft., Aug. 20, 1908, A. A. Heller 9494. Plants perennial, 5-50 em. high, evenly sparsely tomentose throughout or less so on the peduncles; principal rosette leaves 2-12 em. long, 14 em. wide, bipinnately dissected with linear, nearly filiform, ultimate divisions 5-15 mm. long, 0.5-1.0(-1.5) mm. wide, the secondary divisions few in number, the petiole comprising 14-24 of the leaf; stem leaves 0-2(3), much reduced; heads 1-5(-6) per stem, 30-60-flowered, mostly on long slender peduncles 3-15 em. long; principal involucral bracts 6-9 mm. long, 2-5 mm. wide, tomentose to nearly glabrate, yellow- membranous (often reddish-tinged) for 1-2 mm. from the acute to broadly ovate tip; corollas light yellow, 3-4 mm. long, the tube moderately glandular, 1.5-2 mm. long, the throat campanulate, 1.5-2 mm. long with lobes reflexed, 2-4 times longer than the lobes; achenes 4.5-5.5 mm. long, pubescent with hairs 0.2-1(-2) mm. long; pappus of 14-18 linear oblong scales, (1—)1.5-3 mm. long; anthers partly exserted, 2.5-3 mm. long; n =17. DisrRIBUTION.— Mountain ranges and foothills of the Great Basin, mostly in coarse limestone soils of juniper belts or at higher elevations 1956] Turner,—A Study of the Genus Hymenopappus 241 on rocky limestone and shaley outcrops, 4800-9800 ft., eastern Cali- fornia (Inyo Mts.) central and eastern Nevada to western Utah and Mohave Co., Arizona (Fig. 28). May-July. Rydberg assigned the specific name nanus to this variety because of its small habit. In the original description he stated, “stems 1 dm., rarely 1.5 dm. high," which is correct for the few specimens he examined; however, isotypic material (Mo) shows specimens up to 20 cm. high. An attempt was made to dis- tinguish a group of specimens that would “fit” Rydberg’s descrip- tion. Two groups of plants were sorted out of herbarium material: plants of smaller habit, smaller leaves and fewer heads, and plants of the opposite nature, otherwise essentially alike. When the completed distribution dot-maps were compared there was almost complete overlap of ranges. In addition, complete intergradation of habit, leaf size, and head number was encountered. In short, the variety appears to have a wide ecologic amplitude: where it grows at higher altitudes on rocky outerops there is a tendency to dwarfness; in more favorable habitats there is a tendency toward robustness. Variety nanus shows some of the characteristics of the typical var. lugens in the Inyo Mts., California (Hovanitz s.n.), the bracts, in particular, appearing much like this latter variety. The collections from Mohave Co., Arizona (cited below) appear to be established races that apparently have hybridized with var. lugens, since they show certain characteristics of that variety. Seed of var. nanus collected near Peach Springs (Preece & Turner 2608) were grown in a uniform garden along with collections of var. lugens from the same general region (P. & T. 2610).'!° The contrast between these taxa was main- tained (cf., Table 1). In addition, it was noted that the race of var. nanus did not produce flowers after mid-August, while var. lugens continued to produce flowers until frost set in (late September). A tabulation of the flowering dates of var. nanus taken from 37 herbarium sheets representing different collec- tions was made and the following data obtained :!! 10 Ten plants were grown of each of these varieties. u The method used in assigning the dates in this chart was as follows: The letter E was given to those specimens in early flower, M for those in mid-flower, and L for those in late flower. After this initial tabulation the L's and E's were moved back and up one week respectively, to approximate more closely equivalent flowering conditions. 242 Rhodora [Vor. 58 Specimens Specimens Month Week in flower Month Week in flower May 1 1 July 1 5 2 0 2 2 3 1 3 1 1 2 4 0 June 1 9 2 1 August 0 3 11 4 4 September 1 1 In addition to that with var. lugens, var. nanus apparently shows some intergradation with var. erzopodus at mid-elevations in the mountainous regions, where their ranges overlap (Preece & Turner 2558). The variety probably has its closest relationships with var. lugens and var. idahoensis. It can be distinguished from the former variety by its lighter yellow, more numerous florets with shorter throats, more linear, filiform leaf segments, and less crimson-tinged bracts. It may be distinguished from var. idahoensis, along with other characters, by its narrower, more pubescent leaf segments, and larger pappus scales. REPRESENTATIVE SPECIMENS.—Arizona,. MOHAVE co.: Hackberry, M. E. Jones 4556 (CAs, DS, NY, POM, RM, UC, US); 6 mi. E. of Peach Springs, Preece & Turner 2608 (smu, ws). California. INYo co.: Chris’s Cabin Flat, Inyo Mts., Alexander & Kellogg 2536 (aH, Mo, vc). Nevada. ELKO CO.: Ruby Mts., ridge above Cave Creek Post Office, Heller 9494 (MO, NY, UC, US). ESMERALDA CO.: 2 ml. W. of Lida, Silver Peak Range, Maguire & Holmgren 25616 (CAS, DS, GH, MO, NY, UC, US, WS). EUREKA co.: Monitor Valley, S. of Lone Mt., June 12, 1944, Ripley & Barneby 6221 (CAS). LINCOLN Co.: Pioche, Aug. 31, 1912, M. E. Jones (POM). NYE CO.: near Sunnyside, Ripley & Barneby 4007 (CAS). WHITE PINE co.: 30 mi. S. of Ely, Maguire & Holmgren 25627 (cas, DS, GH, NY, UC, US, ws). Utah. Beaver co.: Half-way station W. of Wa Wa, June 26, 1906, M. E. Jones (POM). IRON co.: 0.5 mi. S. E. of Cedar City, Degener 16618 (Mo, US). JUAB CO.: Fish Springs, June 4, 1891, M. E. Jones (POM). MILLARD CO.: 37 mi. from Delta, Maguire & Becraft 2864 (au, UC). SEVIER CO.: “Vermilion,” W. side of valley, June 4, 1901, M. E. Jones (POM). TOOELE CO.: Dugway, June 2, 1891, M. E. Jones (wo, UC, US). WASHINGTON CO.: 7 mi. Š. W. of Shivwits, Beaver Dam Mts., Preece & Turner 2558 (SMU, ws). (To be continued) Vol. 58, no. 691, including pages 163-198, was issued 6 August, 1956. 10-$- 56 73 PTT Lie AY FARLo® mart ^f" JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS. Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD CARROLL EMORY WOOD, JR. Vol. 58 September, 1956 No. 693 CONTENTS: Asa Gray and Andrew Jackson Downing: A Bibliographic Note. A. Hunter Dupree... n n cc cee 5 243 Additions to the Flora of St. Paul Island, Nova Scotia. J. S. BipmR ee i el ee c LT 245 A Cytotaxonomic Study of the Genus Hymenopappus (Com- positae). Billie L. Turner (continued from p. 242) 5... 045 250 Lygodium palmatum vs. Urbanization. Jesse F. Smih.......... 270 Clarence R. Hanes. Edward G. Voss and James S. Wilson....... 272 Flora of Winnebago County, Illinois [Review]. Julian A. Steyer- AY, el tl aed AEE ree eI DP UP DPI 273 The New England Botanical Club, Inc. 8 and 10 West King St.. Lancaster, Pa. Botanical Museum. Oxford St.. Cambridge 38. Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $6.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 60 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $5.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scien- tific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of publica- tion. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards, GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U. S. A. QTRbooora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 September, 1956 No. 693 ASA GRAY AND ANDREW JACKSON DOWNING: A BIBLIOGRAPHICAL NOTE A. Hunter DUPREE ALTHOUGH he labored for thirty years as director of Harvard's Botanic Garden, Asa Gray published little on horticulture, and his work in botany completely overshadowed the garden side of his career. An early publieation,! however, which does not ap- pear in the standard list of Gray’s published works,’ connects him both with horticulture and with Andrew Jackson Downing, its most important devotee in the pre-Civil War United States. In 1841, Wiley and Putnam of New York published, under the editorship of Downing and Gray, the first American edition of John Lindley, The Theory of Horticulture. Since Gray was in New York in 1841 he doubtless knew the young landscape gardener who lived on the Hudson River. And since the firm of Wiley and Putnam was the publisher both of Gray's Botanical Textbook and Downing’s Treatise om the Theory and Practice of Landscape Gardening, the origin of the collaboration is easy to visualize. Each editor signed the notes he added. Downing usually discussed practical problems of gardening in the United States, while Gray supplied data on American climate and also comments concerning plant physi- ology in terms similar to the 1842 edition of the Botanical Textbook, on which he was working at the time. 1 Joun LINDLEY, The Theory of Horticulture: or, an Attempt to Explain the Principal Operation of Gardening upon Physiological Principles (First American edition, with notes, etc., by A. J. Downing and A. Gray, N. Y., 1841). 2 [Sereno Watson and G. L. Goopark], “List of the Writings of Dr. Asa Gray, Chronologically Arranged, with an Index," American Journal of Science, 3rd ser., XXXVI (1888), Appendix, 3-07. 244 Rhodora [Vor. 58 The appearance of this American edition of the British bota- nist Lindley's work on horticulture illustrates two important tendencies in plant science in the United States about the year 1840. In the first place, it was a part of the campaign waged by Gray and John Torrey to bring American botany abreast of current European theory. In the second place, the book is a part of the first great effort to apply science to American agri- culture and horticulture. The translation of Justus Liebig’s epochal work on agricultural chemistry appeared in the United States in the same year of 1841.4 Gray explicitly recognized the importance of Liebig’s work in a footnote to Lindley’s chapter on soil and manure, where he said, “These remarks were perfectly applicable when this work was published (early in 1840); but the treatise on Organic Chemistry in its application to Agriculture and Physiology, by the distinguished Professor Liebig, which ap- peared a few months later, has greatly elucidated the whole subject.” The immediate popularity of Liebig’s work perhaps accounts for the obscurity into which the American edition of Lindley soon fell. That Gray and Andrew Jackson Downing worked together as young men has a certain importance to the later history of horti- culture and landscape gardening in the United States. Their collaboration seems to have ceased when Gray moved to Cam- bridge in 1842, but Downing in the next few years laid the basis for a brilliant and pioneering career in landscape architecture. He had just begun to lay out the Smithsonian Institution grounds in Washington when he died in 1852 at the age of thirty-seven while rescuing victims of a steamboat disaster on the Hudson.^ His influence nevertheless lived on and was definitely present in the planning of the Arnold Arboretum, established in 1872. Back in 1841 Downing had personally presented a copy of his Treatise on the Theory and Practice of Landscape Gardening to Asa Gray.” In 1859, after Downing’s death, his neighbor and è The better known example is Jonn Linpiey, A Natural System of Botany (First American edition, with an appendix by John Torrey, N. Y., 1831). ‘See Liebig and After Liebig: A Century of Progress in Agricultural Chemistry (Washington, 1942). 5 LINDLEY, Theory of Horticulture (American edition), 314 n. 6 HERBERT A. Kerrar, “Andrew Jackson Downing,” Dictionary of American Biography, V, 417—418. 7 Presentation copy in Gray Herbarium Library. 1956] Erskine,—Flora of St. Paul Island 245 follower, Henry Winthrop Sargent, brought out a sixth edition of the same work with a supplement describing Wodenethe, his own estate on the Hudson, and also the estate of this relative, Horatio Hollis Hunnewell, of Wellesley, Massachusetts.^ It was Henry W. Sargent who first kindled the interest of both Hunnewell and his younger relative, Charles Sprague Sargent, in horticulture and landscape gardening.’ ‘Thus of the men who played major roles in the institutional rearrangements of Harvard in 1872 out of which the Arnold Arboretum grew, three—Gray, Hunnewell, and C. 8. Sargent—had long acquain- tance with the tradition stemming from the work of Downing. Frederick Law Olmsted, an old friend of Gray’s and the co-worker with C. S. Sargent in laying out the Arnold Arboretum, also owed much to the influence of Downing. Hence the founders of the Arboretum in the 1870’s had the benefit of a horticultural tradition, of which they were a part, going back thirty years to New York state and Andrew Jackson Downing, and the collab- oration of the two co-editors of the American edition of Lindley’s Theory of Horticulature bore fruit after many seasons.—UNI- VERSITY OF CALIFORNIA, BERKELEY. ADDITIONS TO THE FLORA OF ST. PAUL ISLAND, NOVA SCOTIA J. S. ERSKINE Sr. PAUL ISLAND is the northernmost point of Nova Scotia, a cliffbound island three and one-quarter miles long by never more than one mile wide. The vascular flora was first listed by Dr. A. H. MacKay in 1883. Then in 1929 Miss L. M. Perry and Dr. Muriel Roscoe spent a month on the island and collected exhaustively the vascular plants and the algae, lists of which appeared in Rhodora in May, 1931. The present author spent from August 10th to 17th on the island, collecting plants for the Nova Scotia Museum of Science. Miss Perry’s lists and floristic descriptions are still valid today, although 8 See C. S. SARGENT, “Henry Winthrop Sargent," L. H. Bailey, ed., Standard Cyclopedia of Horticulture (N. Y.. 1915), LIT, 1594. » E. W. Sargent and C. S. SARGEN, eds., Epes Sargent of Gloucester and his De- scendants (Boston, 1923), 193. 246 Rhodora [Vor. 58 many oÍ her plant-names have been changed and some varieties discarded. As all these alterations can be traced through Gray's Manual of Botany, 8th Edition, I have not thought it worthy of space to comment upon them and I offer this list merely as a supplement to hers. When new varietal refinements have seemed of interest, I have included them in my list. The vascular collections have been checked completely by D. S. Erskine in the N. S. Museum, and in part by others, and I have accepted their determinations when their judgement has disagreed with mine. The chief differences between Miss Perry’s list and mine, apart from the fluctuating element of introduced and often impermanent weeds, suggest that she combed the bare headlands more thoroughly, I the woods and bogs. Her list showed 280 species and 17 still accepted varieties, mine 290 species and 6 varieties, the combined total being 320 species thus far known for the island. ADDITIONS TO Miss Pgerry’s List ji Lycopodium complanatum L., var. flabelliforme Fern. One station, Atlantic Cove by path to Big Lake. Picea mariana (Mill.) BSP. Many stunted trees in bogs. . Potamogeton confervoides Reichenb. Sterile in Little Lake. . Festuca elatior L. Scarce weed by road to Southwest. . Poa annua L. Clumps around barn, Southwest. . Poa palustris L. Headland, Southwest. . Agropyron repens (L.) Beauv., var. subulatum (Schreb.) Reichenb., f. Vaillantianum (Wulf. & Schreb.) Fern. Garden, Atlantic Cove. 8. Agropyron trachycaulum (Link.) Malte, var. novae-angliae (Seribn.) Fern. Headlands, Atlantie Cove & Petrie's Point. 9. Hordeum jubatum L. Salty hilltop, Northeast. 10. Elymus mollis Trin. Headland, Martin Power’s Cove. 11. Agrostis scabra Willd. Roadside to Southwest. 12. Agrostis tenuis Sibth. Wood-edge, Atlantic Cove. 13. Hierochloe odorata (L.) Wahl. Boggy slope, Trinity Cove. 14. Eleocharis halophila Fern. Brackish pool on NE Head. 15. Eleocharis Smallii Britt. Edge of Little Lake. Probably Miss Perry’s E. palustris was this alone and not 14. 16. Eriophorum spissum Fern. Occasional in bog, Petrie's Pond. 17. Carex cephalantha (Bailey) Biekn. Swamp by path to Northeast. 18. Carex leptalea Wahl. Wood near Little Lake. 19. Carex novae-angliae Schwein. Wood, Atlantic Cove. 20. Carex salina Wahl. (X C. paleacea Wahl. ? DSE) 21. Carex pallescens L., var. neo-gaea Fern. Petrie’s Pond. NO = eo t 1956] Erskine,— Flora of St. Paul Island 247 22 . Carex oligosperma Mx. Frequent in high bogs, Coggin's Mtn. 23. 24. . Habenaria Hookeri Torr. Wood, Atlantic Cove. . Habenaria blephariglottis (Willd.) Torr. Common on high bogs, Juncus pelocarpus Mey. Floating in bog-pool, SW Head. Smilacina trifolia (L.) Desf. Frequent in bog-holes. Soggin’s Mtn. . Goodyera repens (L.) R. Br., var. ophioides Fern. In woods, Nor- wegian Mtn., Coggin's Mtn., SW Head. s. Salix Bebbiana Sarg. Three sterile shrubs from scattered stations. . Myrica Gale L. Common around bog-holes. . Betula lutea Mx. f. Rare shrubs with unusual leaves. . Betula papyrifera Marsh., var. commutata (Regel) Fern. and var. macrostachya Fern. The species was reported by Miss Perry. These varieties are recognizable among a jumble of variants. . Betula borealis Spach. In alder scrub by path to NE and on high bogs, Coggin's Mtn. . Betula pumila L., var. renifolia Fern. Found by Miss Perry and later segregated to this variety. Rumex domesticus Hartm. Hen-run, Atlantic Cove. Cerastium arvense L., var. viscidulum Gremli. Found by Miss Perry and now segregated into this variety. . Neslia paniculata (L.) Desv. Hen-run, Atlantic Cove. . Drosera intermedia Hayne. Common around edge of bog-holes. . Pyrus americana Marsh. Miss Perry found only P. decora, but my collections were nearer to P. americana. 9. Prunus virginiana L. Occasional in wood, Atlantic Cove; forming serubby woods west of Big Lake. . Vicia Cracca L. Rare weed, Atlantic Cove. . Hypericum perforatum L. Beside road to Southwest. . Oenothera biennis L. Spreading weed, Atlantic Cove. . Sanicula marilandica L. Wood west of Big Lake. . Coelopleurum lucidum (L.) Fern. Headland, Martin Power's Cove. . Cornus X unalaschkensis Ledeb. Reported but not collected by Miss Perry. Fairly frequent. . Cornus alternifolia L. f. Wood near Little Lake. . Monotropa Hypopitys L. One patch on Coggin's Mtn. . Bartonia paniculata (Mx.) Muhl., var. sabulonensis St. J. & Fern. Not uncommon in bog, Petrie's Pond & Coggin's Mtn. . Lappuld echinata Gilib. By barn, Southwest. . Achillea. Millefolium L. was found by Miss Perry, but my specimens have been corrected to A. lanulosa Nutt. . Matricaria maritima L. By barn, Southwest. . Sonchus arvensis L. Headland, Northeast. . Sonchus asper (L.) Hill. Roadside to Southwest. . Lactuca biennis (Moench.) Fern. Occasional in east woods. . Hieracium aurantiacum L. One plant, Atlantic Cove. . Hieracium canadense Mx. Established by road to Southwest. 248 Rhodora [Vor. 58 The bryophytes of St. Paul Island have not previously been listed. The island, with its woods, bogs and raw headlands, is a better habitat for mosses than for vascular plants, yet only one species of moss to six of vascular plants was collected as compared to one to four on the mainland. This list, there- fore, cannot claim to be exhaustive. The more difficult mosses and hepaties were checked by Dr. A. L. Andrews of Ithaca, N. Y., the sphagna by Dr. H. L. Blomquist of Duke University, and my thanks are due to both of them. "Twenty species of lichens were collected also, but, as these have not been checked by an authority and as they do not differ appreciably from those of the Cape Breton barrens, the names are not listed here. Hepatics Barbilophozia barbata (Schmid.) Loeske. 2417, Atlantic Cove. Bazzania trilobata (L.) S. F. Gray. 2411, Woods, Little Lake. Blasia pusilla L. 2360, 2436. SW Head, Norwegian Mtn. Blepharostoma trichophyllum (L.) Dum. 2375, path to NE. Calypogeia Neesiana (M. & C.) K. Mull. 2415, Atlantie Cove. Calypogeia Trichomanis (L.) Corda. 2360, Southwest Head. Cephalozia bicuspidata (L.) Dum. 2436, Norwegian Mtn. Cephalozia connivens (Dicks.) Lindb. 2436, Norwegian Mtn. Cephalozia leucantha Spruce. 2436, Norwegian Mtn. Cephaloziella hampeana (Nees) Schiffn. 2438, Path to NE. Harpanthus scutatus (Web. & Mohr) Spruce. 2460, Southwest Head. Lepidozia reptans (L.) Dum. 2416, Atlantic Cove. Lophozia grandiretis (Kaal) Schiffn. Southwest Head. Scapania irrigua (Nees) Dumort. 2418, 2459, SW Head, Atlantic Cove, Mosses Aulacomnium palustre (Web. & Mohr) Schwaegr. 2368, South Shore. Amblystegium compactum (C. Müll.) Aust. 2420, woods, SW Head. Atrichum undulatum (Hedw.) Beauv. 2427, Norwegian Mtn. Brachythecium salebrosum (Web. & Mohr) Bry. Eur. 2372, 2379, swampy woods, South Shore. Bryum pallens (Web. & Mohr) Brid. 2380, wet rocks, path to NE. Ceratodon purpureus (Hedw.) Brid. 2351, 2358, 2400, 2425. Outerops, Norwegian Mtn., SW Head, Atlantic Cove. Dicranella heteromalla (Hedw.) Schimp. Humus in wood, SW Head. Dicranella subulata (Hedw.) Schimp. 2394, path to NE. Dicranum Bergeri Blan, 2398, 2450. Atlantic Cove, Coggin’s Mtn. Dicranum Bonjeani DeNot. Humus in wood, SW Head. Dicranum Drummondii C. Müll. 2354, 2379, 2445. Humus in wood, Norwegian Mtn.; SW Shore; slope by Petrie’s Pond. Dicranum scoparium Hedw. 2392, 2408, 2413, 2420, 2448. Woods, Path to NE, Little Lake, Trinity Cove, Coggin's Mtn. 1956] Erskine,— Flora of St. Paul Island 249 Drepanocladus fluitans (Hedw.) Warnst. 2422, 2426. Bog, SW Head, Norwegian Mtn. Drepanocladus revolvens (Sw.) Warnst. 2378, Bog by path to NE. Fontinalis novae-angliae Sull. 2406, in Little Lake. Grimmia maritima Turn. 2431, rocks, Trinity Cove. Hylocomium splendens Hedw. 2412, rare, woods by Little Lake. Hypnum pallescens (Hedw.) Bry. Eur. 2353, tree bole, Norwegian Mtn. Leptodictyum riparium (Hedw.) Warnst. 2390, bog-pool, SW Head. Mnium hornum Hedw. 2364, 2370, 2356. Rocks, SW Head. Mnium punctatum Hedw. 2414, 2451. Atlantic Cove, Coggin’s Mtn. Pohlia nutans (Hedw.) Lindb. 2377, 2397, 2444. Common. Plagiothecium striatellum (Brid.) Lindb. 2383, Rock talus, SW Head. Pleurozium Schreberi (Brid.) Mitt. 2407, by Little Lake, rare. Polytrichum gracile Smith. 2409, woods by Little Lake, rare. Polytrichum juniperiunm Hedw. 2388, 2446, Bog, SW; path, NE. Rhytidiadelphus squarrosus (Hedw.) Warnst. 2387, 2410, Little Lake, SW Head. Sphagnum capillaceum (Weiss.) Schrank., var. tenellum (Schimp.) An- drews. 2447, Coggin’s Mtn. Sphagnum centrale C. Jens. 2424, Big Lake. det. A. L. Andrews. Sphagnum compactum DC. 2366, SW Head. Sphagnum cuspidatum Ehrh. 2391, SW Head. Sphagnum fimbriatum Wils. 2369, boggy SW shore. Sphagnum fuscum (Schimp.) H. Klingr. 2395, 2449. Bogs. Sphagnum Girgensohnii Russ. 2374, wet woods, South Shore. Sphagnum imbricatum Hornsch. 2384, SW Head. Sphagnum Lindbergii Schimp. 2381, Big Lake. Sphagnum magellanicum Brid. 2428, Norwegian Mtn. Sphagnum papillosum Lindb. 2365, SW Head. Sphagnum recurvum Beauv. 2419, 2429. Atlantic Cove; Norwegian Mtn. Sphagnum riparium Angst. 2389, SW Head. Sphagnum squarrosum Crome. Wood, SW Head. Sphagnum tenellum Pers. 2386, SW Head. Sphagnum Warnstorfii Russ. 2396, bog by path to NE. Tetraphis pellucida Hedw. 2435, Norwegian Mtn. Ulota crispa (Hedw.) Brid. 2352, tree-trunk, Norwegian Mtn. Ulota phyllantha Brid. 2434, Norwegian Mtn. Of these only two are new Nova Scotian records. Sphagnum riparium Angst. is to be expected here but has not hitherto been reported. Sphag- num centrale C. Jens. has not usually been found near the coasts but is otherwise within its expected range. Dr. Andrews said that this specimen lacked the usual robustness of this species. I have to give grateful acknowledgement of assistance to the Nova Scotia Museum of Science for financing the expedition, to Dr. Andrews and Dr. Blomquist for determinations, and to Miss Perry for having provided an excellent list to build on.—WorrvrrLE, Nova Scorta. 250 Rhodora [Vor. 58 A CYTOTAXONOMIC STUDY OF THE GENUS HYMENOPAPPUS (COMPOSITAE) BILLIE L. TURNER (Continued from page 242) 3l. Hymenopappus filifolius var. idahoensis, var. nov. Herbae perennes, caulibus 0—2-foliatis 10-45 em. altis; foliis inferioribus bipinnatis sparse tomentosis glabratisve, 6-15 em. longis, 3-5 em. latis, segmentis remotis 8-20 mm. longis, 1-2 mm. latis; inflorescentiis 4-14 capitulatis; flosculis flavis 3.2-4.5 mm. longis, fauce 1.5-2 mm. longa; pappo 0.7-1.3 mm. longo. Plants perennial, 15-45 em. high, mostly glabrate to sparsely tomentose, except in the axils of the rosette leaves which are densely woolly; principal rosette leaves 6-15 cm. long, 3-5 em. wide, glabrous to sparsely tomentose, sparsely bipinnately dissected with relatively broad, flattened, linear divisions, mostly 8-30 mm. long, 1-2 mm. wide; stem leaves 0-2, much reduced; heads 4-14 per stem, campanulate, 25-70 flowered (rarely less), on nearly glabrate ultimate peduncles 1-6 em. long; principal involucral bracts 7-9 mm. long, 3-4 mm. wide, mostly glabrate, or nearly so, yellow- membranous (very rarely reddish) for 1-2 mm. from the acute tip; corollas bright yellow, 3.2-4.5 mm. long, the tube moderately glandular, 1.5-2.5 mm. long, the throat campanulate, 1.5-2 mm. long with lobes reflexed, 3-4 times as long as the lobes; achenes 4.5-6 mm. long, pubescent with hairs about 1 mm. long; pappus of 12-16 linear oblong scales, 0.6-1.3 mm. long; anthers partly exserted, 2.5-3 mm. long; n = 17. Typr (ws): Idaho. Lemhi Co.: Junction of Warm Spring Cr. and Salmon R. (about 14 mi. S. of Salmon); rusty-red, sandy-silty hill slopes, June 12, 1951, S. J. Preece & B. L. Turner 2378 (Isotypes to be distributed). DisrRiBUTION.— Known only from Idaho along the Salmon and Big Lost Rivers in Custer and Lemhi counties where it occurs on low rocky, gravelly, or sandy foothills (Fig. 28). June-July. Hymenopappus filifolius var. idahoensis is a poorly defined variety that has been segregated primarily on its geographical isolation and secondarily on its combination of rather minute differences (cf. Table 1, p. 218). The variety would probably be considered no more than a race of var. nanus were it included in the range of that taxon; however, it may be distinguished from that variety by being less pubescent throughout, having more numerous heads on shorter peduncles, shorter pappus, and longer, more flattened, thicker ultimate leaf segments. At the type locality, specimens of var. idahoensis that super- ficially resemble var. nanus (smaller, more pubescent plants, with fewer heads) may be found growing on rocky hilltops. 1956] Turner, —A Study of the Genus Hymenopappus 251 The character of the pappus is seemingly constant, however, and the degree of resemblance is merely an approach. REPRESENTATIVE SPECIMENS.—Idaho. cusrer co.: Below Clayton, R. J. Davis 486 (ps, ws); 10 mi. N. E. of Clayton, Hitchcock 15668 (RSA, uc, ws); low hills across river from Challis, Hitchcock & Muhlick 8958 (CAS, GH, MO, RM, SMU, ws); Challis Cr., Macbride & Payson 3338 (cas, GH, MO, NY, POM, RM, UC, US); Mackay, Nelson & Macbride 1561 (ps, GH, MO, RM). LEMHI CO.: about 15 mi. S. of Salmon, Hitchcock, Rethke & van Raadshooven 3726 (cas, Ds, GH, UC, ws); Junction of Box Cr. and Salmon R., 16 mi. S. of Salmon, Preece & Turner 2887 (SMU, ws). 3m. Hymenopappus filifolius var. cinereus (Rydb.) Johnston Hymenopappus cinereus Rydb. Bull. Torr. Bot. Club 27: 634. 1900. Hymenopappus filifolius var. cinereus Johnston, Contrib. Gray Herb. 68: 98. 1923. Type examined (Ny): Colorado. Walsenberg, 1800 m., June 5, 1900, P. A. Rydberg & F. K. Vreeland 5479. Hymenopappus arenosus Heller, Bull. Torr. Bot. Club 25: 200. 1898. Isotypes examined: New Mexico. Santa Fe Co.: near Espanola, 5600 ft., May 17, 1897, A. A. & E. G. Heller 3542. Hymenopappus ochroleucus Greene, Plantae Bakerianae 3: 30. 1901. Isotypes examined: Colorado, Cimarron, 6900 ft., June 28, 1901, C. F. Baker 269. In the original description Greene cited only the following specimens, “C. F. Baker, 25 and 269." Material of collection number 269 is widely distributed in herbaria and should serve as the type, instead of collection number 25 which was not found in the specimens examined by the author. Plants perennial, 15-40 em. high, sparsely grayish-green tomentose; principal rosette leaves 5-14 cm. long, bipinnately dissected with linear, filiform divisions, 0.5-1 mm. wide, conspicuously impressed-punctate; stem leaves (0-)2—4, much reduced upwards; heads 1-6 per stem, sub- turbinate to campanulate, 25-40 flowered, on ultimate peduncles 1-6 em. long; principal involucral bracts 6-9 mm. long, 2-4 mm. wide, gla- brous to densely tomentose, yellow or rarely white-membranous for 1-2(-3) mm. from the acute to obtuse tip; corolla yellow or rarely white (ochroleucous), 3-4.5 mm. long, the tube moderately glandular, 1.5-2.5 mm. long, the throat campanulate 1.5-2.5 mm. long with lobes reflexed, 2.5-4 times longer than the lobes, achenes 4—6 mm. long, evenly pubescent with conspicuous hairs 1-3 mm. long; pappus of 14-18 linear oblong scales, 1.5-2 mm. long; anthers partially exserted 2-3 mm. long; n — 17. DisrrisuTion.—Rocky Mountains, in exposed situations at mid- elevations (5,500—10,000 ft.), Colorado, eastern Utah, central and northern New Mexico, and northeastern Arizona, with outliers in the rocky, limestone canyons of the northern panhandle of Texas, and in White Sands National Monument, New Mexico (Fig. 37). May-September. Hymenopappus filifolius var. cinereus is a relatively widespread, vari- able taxon the races of which occur on several soil types at various eleva- 252 Rhodora [Vor. 58 tions: White Sands National Monument (gypsum); Espanola, New Mexico (deep sand); Cimarron, Colorado (gravelly-clay); Boulder, Colorado (black shale), ete. The name Hymenopappus ochroleucus Greene is based on a white-flowered race of var. cinereus (as treated in this paper) which shows evidence of apparent introgression or gene-flow from var. megacephalus, a variety found only a short distance to the west from this locality (Cimarron, Colo.). A white- flowered race is also known in var. lugens of the filifolius complex, but that is a tetraploid, differing in several characters from the present race, which is diploid. In the mountainous regions of northern New Mexico, var. cinereus appears to have two distinct altitudinal races: a leafy- stemmed ecotype at lower elevations occurring principally in deep sand (Heller 3542; Preece & Turner 2739, 2741), and a scapose ecotype at somewhat higher elevations occurring in rocky or shallow, sandy-clay soils (Heller 3555). No sharp breaks or lines can be drawn between these populations and attempts to recognize these groups on herbarium sheets has not met with success. At present it seems best to consider them only local races of a highly variable taxon. Specimens from the rather isolated panhandle region of Texas can be distinguished from the mass of the material of var. cinereus by several characters, such as its more subscapose habit, longer filiform leaf segments, and apparently white flowers. It seems best, in view of our present knowledge, to consider these Texas plants no more than relict races of a once more widespread cinereus complex, noting, however, that further study may justify the elevation of these races to varietal rank. Hymenopappus filifolius var. cinereus grades into var. luteus in Moffat Co., Colorado, into var. megacephalus in the western part of its range (F. H. Graham 8321), and into var. lugens and var. pauciflorus in the southwestern part of its range (north- eastern Arizona and southeastern Utah). Locally in Colorado it intergrades with var. parvulus (see discussion under that varlety). Hymenopappus arenosus Heller is the earliest name for this taxon; however, Johnston, in making the varietal combination, 1956] Turner,—A Study of the Genus Hymenopappus 253 chose Rydberg's species (he gave no reason) and consequently cinereus is now the earliest varietal name and must be used, according to the International Code of Botanical Nomenclature (1952). REPRESENTATIVE SPECIMENS.—Arizona. APACHE CO.: Luka-Chukai Mts., S. of View Point, Goodman & Payson 2881 (GH, MO, NY, UC). co- CONINO co.: between Winslow and Flagstaff, McKelvey 4505 (GH, POM). NAVAJO Co.: Mishongnovi, Hopi Ind. Res, A. A. Darrow (cas, NY). Colorado. ALAMOSA co.: Alamosa, F. Ramaley 12006 (RM). ARCHULETA co.: Arboles, C. F. Baker 688 (GH, MO, NY, POM, RM). BOULDER CO.: 7 mi. N, of Boulder, Preece & Turner 2848 (SMU, WS). CHAFFEE CO.: 4 mi. W. of Salida, Preece & Turner 2830 (smu, ws). DELTA CO.: Eckert, Osterhout 6124 (RM). DENVER co.: Inspiration Point, Clokey 3951 (cas, GH, MO, NY, POM, RM, UC, US, WS). DOUGLAS CO.: Gann, June 23, 1920, Osterhout & Clokey 3952 (CAS, GH, MO, POM, RM, UC, US, WS). EAGLE co.: Deep Creek, Killip 36451 (vs). ELBERT CO.: 2 mi. N. E. of Fondis, M. Ownbey 1282 (GH, MO, NY, RM, UC, WS). EL PASO CO.: Colorado Springs, June 25, 1879, M. E. Jones s.n. (809) (DS, NY, POM, RM, US). FREMONT co.: Penrose, A. Nelson 10544 (GH, MO, NY, RM, UC). GARFIELD CO.: Rifle, Osterhout 2128 (NY, RM). GRAND CO.: 6.5 mi. E. of Kremmling, Turner 2948 (ws). HUERFANO CO.: La Veta, Osterhout 6700 (MO, POM, RM). JEFFERSON CO.: Morrison, Osterhout & Clokey 3096 (CAS, GH, RM, UC, US). LAKE CO.: Twin Lakes, 1896, F. Clements 386 (NY). LA PLATA co.: Durango, Baker, Earle & Tracy 1028 (GH, MO, NY, POM, UC, US). LARIMER CO.: W. of Loveland, Osterhout 256 (NY, RM). LAS ANIMAS co.: Brantly Canyon, Osterhout 2044 (GH, RM). MESA CO.: Grand Junc- tion, Osterhout 6552 (pom, nM). This group of specimens presents good evidence that H. filifolius var. megacephalus hybridizes or introgresses into var. cinereus in this region; near coal mines, Grand Junction, June 15, 1900 (1901) S. G. Stokes (NY, UC, US). MONTEZUMA CO.: Mancos, Crandall 3205 (NY, RM). MONTROSE CO.: Cimarron, C. F. Baker 269 (Ds, GH, MO, NY, POM, RM, UC, US, WS). OTERO CO.: Rocky Ford, July, 1894, G. E. Osterhout (RM). ouRAY co.: Dallas, Preece & Turner 2788 (SMU, WS). PARK CO.:2 mi. W. of Glentivar, Beetle 224 (NY, RM). PUEBLO co.: Walsenburg, Rydberg & Vreeland 5479 (NY, RM). SAGUACHE CO.: Crestone, F. Ramaley 12083 (RM). New Mexico. BERNALILLO CO.: 2 mi. E. of Albuquerque, Krammerer 44 (MO, NY, US). COLFAX CO.: between Cimarron and Raton, McKelvey 2430 (GH, POM). DONA ANA CO.: White Sands, Wooton 167 (DS, MO, NY, POM, RM, UC, US). HARDING CO.: about 20 mi. N. of Mosquero, Eggleston 20177 (NY, Us). MCKINLEY co.: Gallup, Eastwood 5617 (cas). orEROo co.: Round Mt., along Tula- rosa Cr., Wooton (us). RIO ARRIBA CO.: near Lybrooks, Mathias 617 (MO). SANDOVAL CO.: 2 mi. N. of Jemez Springs, Preece & Turner 2746 (SMU, WS). SAN JUAN CO.: near Huerfano Peak, M. E. Mathias 630 (MO). SAN MIGUEL CO.: near Pecos, Standley 5059 (MO). SANTA FE CO.: hills at Santa Fe, 7300 ft., A. A. & E. G. Heller 3555 (Ds, GH, MO, NY, 254 Rhodora [Vor. 58 POM, US, WS). UNION CO.: Emery Gap, N. M. to Branson, Colo., Eggleston 20154 (NY, US). VALENCIA CO.: 7 ml. N. of Trechado, Cutler 2092 (cas, GH, MO, UC). Texas. ARMSTRONG CO.: 3 mi. S. W. of Palodura, Cory 13477 (GĦ). BRISCOE co.: chalk hill, edge of caprock about 5 mi. W. of Quitaque, highway 85, E. Whitehouse 10021 (SMU). HANSFORD CO.: 5 mi. S. E. of Gruver, Shinners 8232 (GH, RM, SMU, UC, WS). HEMPHILL co.: 5 mi. S. of Canadian, Shinners 8290 (SMU). OCHILTREE CO.: Š mi. S. 5. E. of Perryton on steep rocky bluffs only, Shinners 8264 (smu). POTTER CO.: Amarillo Cr., Reverchon 8326 (MO, SMU). RANDALL CO.: Palo Duro Canyon, Sept. 2, 1907 C. R. Bell 1229 (us). Utah. EMERY co.: Calf Springs Canyon, 8 mi. from road, San Rafael Swell, B. & R. M. Maguire 18313 (GH, NY, Us, Ws). GRAND co.: Grand River Canyon, below Moab, Rydberg & Garrett 8495 (NY, RM). SAN JUAN CO.: between Blanding and Kigalia Ranger Station, Holmgren & Hansen 3484 (au, NY, UC, US). WAYNE CO.: Thurber, M. E. Jones 5709C (POM, Us). 4. Hymenopappus mexicanus Gray Hymenopappus mexicanus Gray, Proc. Am. Acad. 19: 29. 1883. Rothia mexicana O. Ktze. Rev. Gen. 1: 361. 1891. Type examined (GH): Mexico. San Luis Potosi. ‘In montibus frigid[is] prope San Miguel," Sept. 1876, Dr. J. G. Schaffner 348. Hymenopappus integer Greene, Pittonia 3: 249. 1897. Isotypes examined: Mogollon Mts., (dry hills), Sept., 1881, H. H. Rusby 179 (180). Hymenopappus obtusifolius Heller, Bull. Torr. Bot. Club 26: 551. 1899. Type examined (ny): Arizona. Coconino Co.: “vicinity of Flag- staff," 7000 ft., July 5, 1898, Dr. D. T. MacDougal 240. A more specific locality was given in the original description by Heller as “Fort Valley, west of San Francisco mountains.” Hymenopappus petaloideus Rydb. N. Amer. Fl. 34: 54. 1914. Type examined (Ny): Arizona, Cochise Co.: “Head of Rock Creek Canyon,” Chiricahua Mts., 8000 ft., Oct. 6, 1907, J. C. Blumer 2215. Plants perennial, 20-90 em. high, stems slender, erect, unbranched, greenish-glabrate to densely white-tomentose; leaves alternate, forming a basal rosette, absent or becoming reduced up the stem, simple to once- pinnate, up to 20 cm. long and 2.5 cm. wide, nearly glabrous to densely tomentose, obscurely impressed-punctate, lobes (when present) broad, ovate to broadly lance linear, 1-7 mm. wide; heads several to numerous, discoid, 20-40-flowered, in flattish cymose panicles, on slender or short- thickened peduncles 0.5-10 em. long; involucre campanulate, principal bracts 9 to 11, nearly glabrate to densely tomentose, 7-9 mm. long, yellowish to white-membranous for 1-5 mm. from the acute to obtuse tip (rarely reddish-tinged); corollas yellow, 3-4.5 mm. long, the tube densely glandular, 2-2.5 mm. long, the throat campanulate, 1-2.5 mm. long, 3-4 times as long as the lobes; achenes obpyramidal, 4-sided, 4-6 mm. long, glabrous to sparsely puberulent (especially when immature), achene faces 2—-3-nerved, rarely somewhat rugose; pappus of 12 to 20 1956] Turner,—A Study of the Genus Hymenopappus 255 short, obtuse to spatulate-laciniate scales, 0.4 mm. long to nearly obsolete ; anthers partially exserted (rarely completely so) about 2.3 mm. long; = 17. D. —OỌOpen areas in igneous soils of yellow pine, spruce and aspen woods, central and southeastern Arizona, southwestern New Mexico, and south into the Sierra Madre Mts., states of Chihuahua, and San Luis Potosí, Mexico (Fig. 25). Late June-October. Hymenopappus mexicanus is a very distinct species through- out its range, but it shows a great deal of variability, especially in leaf shape. H. integer Greene is based on a series of specimens with mostly entire leaves; however, in the field, as well as on a wide selection of herbarium material, there is no constancy in this character, populations as well as individual specimens having leaves that range from completely simple to once-pinnate. Hymenopappus petaloideus Rydberg is a form of the species from the Chiricahua Mts., Arizona, which has more conspicuous, acute, membranous-tipped involucral bracts. This character is not peculiar to this region but occurs repeatedly throughout most of the range of the species, being quite variable even within the same general area (e.g., Mogollon Mts., New Mexico). However, material from or near the type locality of H. mexicanus (San Luis Potosí, Mexico) does tend to have consistently broader, more obtuse involucral bracts, this seemingly correlated with a more glandular corolla. The scarcity of material from central Mexico makes it difficult to weigh the constancy or variability of these characters in this region. Hymenopappus mexicanus, in so far as is known, does not hybridize or intergrade with other species of the genus although one such member (H. filifolius var. lugens) was observed growing close to this species in parts of Coconino Co., Arizona. Both H. filifolius var. lugens and H. mexicanus occur in pine woods, but the latter species tends to occur at somewhat higher eleva- tions, rarely, if ever, extending downslope to the juniper zone. Hymenopappus mexicanus apparently represents a reduced line that has evolved more or less separately from the main mass of the aboriginal stock that has produced the genus as it exists today. In leaf aspect it resembles H. artemisiaefolius and H. scabiosaeus; however, in its perennial habit and floral characters it approaches H. radiatus (disregarding the rays). In total characters H. mexicanus is closer to this latter species, 256 Rhodora [Vor. 58 but this does not mean that its ancestry can be traced through it. REPRESENTATIVE SPECIMENS.—Arizona. APACHE CO.: White Mts., Black R., Thompson’s Ranch, Goodding 598 (CAs, GH, NY, RM, US). CO- CHISE CO.: Chiricahua Mts., head of Rock Cr. Canyon, Blumer 2215 (GH, NY, UC). COCONINO co.: Flagstaff, M. E. Jones 3955 (cas, Ds, GH, NY, POM, RM, UC, US). GILA CO.: Natanes plateau, Goodding 1094 (NY, Us). NAVAJO CO.: Lakeside, Œ. J. Harrison 5481 (Us). PIMA co.: Rincon Mts., Neally 88 (Ny, us). YAVAPAI co.: Copper Basin, Towmey 713 (us). New Mexico. carron co.: Mogollon Mts., on Mogollon Cr., Metcalfe, 316 (DS, GH, MO, NY, POM, RM, UC, US). GRANT CO.: vicinity of Silver City, G. O. S. Ranch, M. E. Jones 28612 (ps, MO, UC). SIERRA CO.: Mimbres Mts., top of Hillsboro Peak, Diehl 432 (rom). Mexico. Chihuahua. Near Colonia Garcia, Townsend & Barber 187 (GH, MO, NY, POM, RM, UC, US). San Luis Potosí. Pelote, Purpus 4722 (Mo, vc); Minas de San Rafael, Purpus 4772a (an, NY, UC). SERIES BIENNES 5. Hymenopappus biennis sp. nov. Herbae biennes e radicibus simplicibus obconicis; caulibus foliatis ramosis erectis solitariis; foliis inferioribus sparse tomentosis bipinnatis 6-16 em. longis, segmentis remotis anguste linearibus 6-20 mm. longis, 1-3 mm. latis; inflorescentiis laxe eymoso-paniculatis 20-60 capitulatis ; capitulis radiatis, radiorum floribus 1-seriatis pistillis fertilibus, discorum hermaphroditis; involueris campanulatis, bracteis 5-8 mm. longis, 3-5 mm. latis, 2-3-seriatis; radiorum corollis ligulatis albidis 14-16 mm. longis, 6-8 mm. latis, disci flavis 3-3.5 mm. longis, tubo 1.5 mm. longo, fauce campanulata, lobis aequalibus triangularibus; achaeniis glabratis quad- rangularibus 4 mm. longis; pappo obscuro. Plants biennial, 60-100 em. tall; stems single from each tap-root, much-branched and leafy, tomentose to nearly glabrate, pithy at the center or often hollow; basal rosette leaves 6-16 em. long, 3-6 em. wide, bipinnately dissected into linear, mostly flattened, ultimate segments 6-20 mm. long, 1-3 mm. wide, sparsely tomentose to nearly glabrate, conspicuously impressed- -punctate; stem leaves 10—40, gradually reduced upward; heads 20-40 per stem, campanulate, 40-60- flowered, on ultimate peduncles 1-6 cm. long, inflorescence a large, much-branched, cymose panicle; receptacle dome-shaped, without chaff; principal involucral bracts 5-8 mm. long, 3-5 mm. wide, sparsely tomentose to nearly glabrate, yellow-membranous for 1-2 mm. from the acute to obtuse tip; ray flowers 8, pistillate and fertile, tubular at the base for about 2 mm., extending into a conspicuous white ligule 14-16 mm. long, 6-8 mm. wide, not cleft at the apex or obscurely so; disk flowers yellow, 3-3.5 mm. long, the tube densely glandular 1.5 mm. long, the throat campanulate 1.5-2 mm. long with lobes reflexed, 2-3 times longer than the acute lobes; achenes black, glabrous (or with a few sessile glands near the apex), 1956] Turner,—A Study of the Genus Hymenopappus 257 4 mm. long, obpyramidal, 4-sided, becoming incurved near the periphery of the head; pappus obsolete or nearly so (0-0.2 mm. long); anthers partially exserted, 2.5 mm. long; chromosome number not known. TYPE consisting of 2 sheets (GH): Texas. Culberson Co.: Guadalupe Mts., “wooded rocky ridge above McKittrick Canyon, 2300 m.," July 17, 1931, J. A. Moore & J. A. Steyermark 3484 (Isotypes cas, Ds, MO, NY, UC). DisrRiBUTION.—Principally in limestone soils in pine woods and protected eanyons of central and south-central New Mexico (Sandia, Oscuro, and Sacramento Mts.), extending into the Guadalupe Mts. of Trans-Pecos Texas, 7,000-10,000 ft. (Fig. 24). July-October. Hymenopappus biennis is a distinct species the total mor- phological relationships of which are undoubtedly with H. newberryi and H. radiatus. It differs from both these species, however, in several fundamental characters. Hymenopappus biennis is a tall, leafy biennial, whereas H. newberryi and H. radiatus are smaller, subscapose perennials. Since all three of these radiate taxa are perfectly distinct, being separated by morphological and geographical discontinuities which, insofar as is known, are not bridged, they are treated here as species. The discovery of H. biennis has made possible a better under- standing of the evolutionary trends which seem to have taken place within the genus Hymenopappus. It appears to be the ‘missing link" which ties a part of the biennial complex back to the perennial groups. Indeed, with rays, H. flavomarginatus would be a near perfect match for H. biennis. It is likely that H. biennis is an early off-shoot of the line that gave rise to the perennial radiate species. Thus, one might assume that the major evolutionary lines within the genus were established at a relatively early time, each of these lines retaining certain primitive characters of the supposed common progenitor. In the subsequent development, each of these lines seems to have gained new characters while modifying or losing old ones. REPRESENTATIVE SPECIMENS.—New Mexico. LINCOLN co.: Ruidoso, Fisher 65 (cas, vs); Mescalero Ind. Reservation, June 23, 1895, Wooton (vs); White Mts., Wooton 264 (Ds, Mo, NY, POM, RM, UC, US); White Mts., 5 mi. above Agency, Wooton & P. Standley 3552 (us). OTERO CO.: Sacramento Mountains: Clouderoft, Cockerell 35 (mw), E. D. Shulz 252 (us), Aug. 1914, H. D. Slater (us), Aug. 8, 1899, Wooton (Pom), Aug. 15, 1899, Wooton (us), Aug. 1, 1916, M. S. Young (uc); Rolland Canyon, Cloudcroft, Eggleston 14500 (us); Sacramento Mountains: James Canyon, July 23, 1899, Wooton (wv), July 6, 1899, Wooton (ps, 388 4a 42c | 42» | Morphology of Hymenopappus species, Series Biennes. Fig. 38. H. biennis (Isotype); (a) receptacle, long. sec., X 4; (b) radial floret, 8; (c) disk corolla, X 8; (d) style branches, X ca. 45; (e) upper surface of style branch, X ca, 45. Fig. 39. H. flavescens var. canotomentosus (Type), floret, part of pappus removed, x 8. Fig. 40. H. artemisiaefolius var. artemisiaefolius (Lindheimer 107, Fasc. 11), floret, part of pappus removed, X 8. Fig. 41. H. tenuifolius (T. 2895); (a) silhouette of basal leaf X 14; (b) floret, part of pappus removed, X 8. Fig. 42. H. Jlavomargina- tus (Type); (a) receptacle, X 4; (b) floret X 8; (c) anther, x ca. 40; (d) style branches, X ca. 40; (e) upper surface of style branch, x ca. 40. Fia. 38—42. 1956] Turner,—A Study of the Genus Hymenopappus 259 uc, Us); White Mts., above Mescalero, Aug. 4, 1901, Wooton (us); Tula- rosa Cr., 3 mi. S. of Mescalero Agency, C. B. Wolf 2762 (cas, Ds, GH, RSA). SANDOVAL CO.: Sandia Mts., near Oshan Springs, Aug.-Sept., C. E. Ellis 345 (Mo). SIERRA co.: Lookout Mines, south end of Black Range, Metcalfe 1176 (cas, MO, US). socoRRo CO.: Oscuro Mts., July 21, 1898, F. S. Earle (mo, Ny). Texas. CULBERSON CO.: Guadalupe Mts., in “The Bowl," summit of Pine Top Mt., Correll 13920 (smu); Guadalupe Mts., Oct., 1881, V. Havard (us); S. McKittrick Canyon, Guadalupe Mts., Hinckley 4472 (us); ridge above McKittrick Canyon, July 17, 1931, Moore & Steyermark 3484 (CAS, DS, GH, MO, NY, UC); Guad- alupe Mts., S. McKittrick Canyon, Muller 8287 (smu, ws). 6. Hymenopappus flavomarginatus Johnston Hymenopappus flavomarginatus Johnston, Contrib. Gray Herb. n.s. 68: 95. 1923. Type examined (aH): Mexico. Coahuila. “Can[ylon and elevated portion of Sierra Madre, 12 to 14 leagues south of Saltillo, Mexico; July 25 to August 1st—1880" (data from isotype label, vs), E. J. Palmer 650. Plants biennial, 30-100 em. high, sparsely canescent to glabrate; larger rosette leaves 6-12 cm. long, bipinnately dissected with linear ultimate divisions mostly 5-20 mm. long, 1-2 mm. wide, the petioles often reddish at base, comprising less than V5 the leaf length; stem leaves 10-20, becoming reduced up the stem; heads per stem 10-30, campanulate, 30-100-flowered, on ultimate peduncles 3-11 cm. long, these with con- spicuous membranous basal bracts, 4-9 cm. long, 3-7 mm. wide; inflores- cence an open cymose panicle; principal involucral bracts broadly obovate to oval, 6-9 mm. long, 4-8 mm. wide, yellow-membranous (often red- tinged) for 1-3 mm. from the broadly obtuse tip; corollas “yellow,” 2.7-3.5 mm. long, the tube 1.5-2 mm. long, densely glandular-pubescent to nearly glabrate, the throat campanulate, 1.2-1.5 mm. long with lobes reflexed, 1-2 times longer than the lobes; achenes obpyramidal, 4-sided essentially glabrous, 3-4 mm. long; pappus of 18-22 minute scales (0.1— -0.2 mm. long) or obsolete; anthers not completely exserted, about 2 mm. long; chromosome number not known. DisrrisuTion.—Known only from pine and oak woodlands in the mountains of north-central Mexico, states of Coahuila, Nuevo León, and San Luis Potosf, 7500-9000 ft. (Fig. 25). May-August. This is a well marked species, undoubtedly belonging close to the H. flavescens and H. tenuifolius groups as suggested by Johnston in his original description. Its closest relationship seems to be with H. biennis, however, for in total characters it resembles this species of Hymenopappus more than any other member of the genus, differing principally in the lack of rays and in the broader involucral bracts. It also has some characters of H. artemisiaefolius such as the large, membranous peduncular bracts of that species. In short, the species has characters 260 Rhodora [Vor. 58 which connect it to nearly all of the more primitive members of the genus, suggesting that it is of an old age and has developed along its own line, retaining some characters and losing others. REPRESENTATIVE SPECIMENS.—Mexico. Coahuila. 24 kilometers N. W. of Fraile, 2900 meters, July 15, 1941, Stanford, Retherford & Northcraft 421 (DS, MO, NY, Uc). Nuevo León. Municipio de Galeana, Haciendo Cieneguillas on Cerro Potosí, 8000 ft., Aug. 7, 1938, (students of) Mexican Biological Expedition, University of Illinois (aH, mo, NY). San Luis Potosí. Chareas, Santo Domingo Road, July-Aug., 1934, Lundell 5604 (ps, GH, MO, US). 7. Hymenopappus tenuifolius Pursh Hymenopappus tenuifolius Pursh, Fl. Am. Sept. 2: 742. 1814. Rothia tenuifolius O. Ktze. Rev. Gen. 1: 361. 1891. Type collected “in upper Louisiana” by Bradbury; no specimen could be found in the American herbaria consulted. It was probably collected in what is now South Dakota. The description, combined with the statement that “the flowers are white,” leaves little doubt as to its identity. Hymenopappus corymbosus var. nuttallii T. & G. Fl. N. Am. 2: 372. 1842. Type or isotype examined (GH): “Red River, Arkansas," without additional information, but probably collected by Nuttall in what is now southern Oklahoma. Hymenopappus tenuifolius Nutt. non Pursh, in T. & G. Fl. N. Am. 2:372. 1842, as synonym of above. Plants biennial, 40-150 em. high; stems from a single tap-root (rarely 2 erowns forming on a root, especially on injured plants), erect, much- branched, angled and grooved, sparsely white-tomentose to more fre- quently nearly glabrous; leaves alternate, forming the first year a basal rosette, the larger basal leaves 8-15 em. long, 4-5 em. wide, bipinnately dissected with linear, filiform, ultimate segments 0.5-1.5 mm. wide, conspieuously impressed-punctate, the stem leaves 8-30, becoming reduced upwards; heads numerous (20-200), discoid, 25-50-flowered, on pubescent to glabrate ultimate peduncles 1—5 em. long; inflorescence a flat-topped cymose panicle; involucre campanulate, the principal braets mostly glabrate (rarely tomentose) or densely glandular, 5-8 mm. long, 2-4 mm. wide, yellowish-membranous for 1-2 mm. from the acute to obtuse tip; corollas white, 2.5-3 mm. long, the tube glandular, 1.5-2.2 mm. long, the throat campanulate, 0.8-1.5 mm. long with lobes reflexed, 1.5-2 times longer than the lobes; achenes 3.5-4.5 mm. long, 4-sided, the faces 2-3-nerved, pubescent principally on the corners with hairs 0.5-1 mm. long; pappus of 16-18 linear oblong scales, 1—-1.5(-2) mm. long; anthers mostly completely exserted, 2-2.5 mm. long; n = 17. DisrRIBUTION.—Common on the southern Great Plains of the central United States from South Dakota to south-central Texas in sandy or, less frequently, gravelly or rocky limestone soils; also in New Mexico on intermontane plains (Fig. 43). Late May-August. 1956] Turner, A Study of the Genus Hymenopappus 261 Hymenopappus tenuifolius overlaps several related taxa in its range, but evidence of present-day hybridization and possible introgression is found in relatively few localities. One such area is in the Panhandle of Texas where small, less leafy, wooly- stemmed individuals are occasionally found, suggesting possible contamination of the species from disjunct, perennial, white- flowered races of H. filifolius var. cinereus, a taxon the normal range of which is in the Rocky Mountains (Fig. 37). "This is of interest since it has been suggested that the perennial, yellow-flowered tetraploid, H. filifolius var. polycephalus (a member of the northern Great Plains flora) is possibly the Ss m Esa MESI NEME SN m Au | X > the Tired oT ° | N D wi ee | Fria. 43. Distribution of Hymenopappus tenuifolius. 262 Rhodora [Vor. 58 result of amphiploidy between H. tenuifolius and H. f. cinereus (see diseussion under H. f. polycephalus). Natural hybridiza- tion of these two taxa in the Panhandle of Texas would seem to add support to such an hypothesis. The isolation of H. tenuifolius from the adjacent, more easterly taxon, H. scabiosaeus var. corymbosus, where the two species overlap, is perhaps partly seasonal, the latter entity having its principal flowering time 3-4 weeks before H. tenuifolius. REPRESENTATIVE SPECIMENS.— Colorado. saca co.: Springfield, Oster- hout 5054 (RM). BENT CO.: Las Animas, Osterhout 3915 (RM). CHEYENNE co.: Plains near Cheyenne Wells, July 12, 1887, C. H. Demetrio (am). ELBERT CO.: about 25 mi. E. of Kiowa, M. Ownbey 1293 (an, MO, NY, RM, UC, WS). KIOWA CO.: Eads, Baker, Earle & Tracy 1029 (wo, row). LAS ANIMAS CO.: Mesa de Maya, 60 mi. E. of Trinidad, Rollins 1838 (GH, MO, NY). LINCOLN CO.: E. of Limon, Osterhout 8226 (RM). OTERO co.: Apishipa Cr., Osterhout 2043 (aH, NY, RM). SEDGWICK co.: Julesburg, Osterhout 4907 (RM). YUMA co.: Wray, Osterhout 3983 (nw). Kansas. BARBER CO.: 5 mi. 8. W. of Medicine Lodge, Rydberg & Imler 684 (Mo, NY). CLARK co.: 6 mi. W. of Englewood, Rydberg & Imler 834 (NY). DECATUR co.: Jennings (no additional data) (Mo). ELLIS CO.: 12 mi. N. of Hays, E. Runyon 130 (CAS, GH, MO, RM). FINNEY co.: near Garden City, Coville 27 (vs). Grove co.: plains, A. S. Hitch- cock 287 (GH, MO, NY, RM, US). GRANT CO.: Ulysses, C. H. Thompson 24 (Mo, US). NORTON CO.: Almenz, Harshbarger (us). OSBORNE CO.: within 5 mi. of Osborne City, Shear 71 (GH, NY, RM). PHILLIPS CO.: Long Island, June 15, 1885, J. B. Hatcher (uc). RAWLINS co.: Atwood, May 30, 1891, L. Fry (Ny). RILEY co.: Manhatten, 1892, (w/o collector) (NY). mooks co.: Rockport, June 10, 1891, E. Bartholomew (uc). TREGO co.: Wakeeney, July 8, 1892, M. Reed & A. D. H. (NY). WALLACE co.: Wallace, Aug. 22, 1884, Letterman (wo, Ny, Us). Nebraska. ANTELOPE co.: Neligh, June 3, 1906, E. S. Bacon (am). BOONE CO.: 4 mi. W., 1.5 mi. N. of Lorette, B. Osborn 1216R (Mo). BROWN co.: Lone Pine, July 13, 1899, J. M. Bates (au, RM). CEDAR co.: Beaver Cr., F. Clements 2664 (Ny, UC, US). CHASE co.: S. E. of Enders, French- man Valley, Tolstead 41427 (wo). CHERRY co.: near Valentine, T'olstead 339 (aH). CHEYENNE CO.: N. of Sidney, Osterhout 7196 (wo, nM). CUSTER co.: Broken Bow, July 4, 1889, H. J. Webber (Mo, NY). DEUEL CO.: Rush Cr., Rydberg 194 (NY). FRANKLIN co.: Franklin, July 4, 1930, H. Hapeman (ws). GRANT co.: Whitman, June 20, 1938, E. Anderson (MO). GREELY CO.: Greely Center, July 4, 1889, T. A. Williams (vs). HITCHCOCK CO.: near Culbertson, Tolstead 411426 (Mo). HOLY co.: Ewing, June 19, 1899, J. M. Bates (GH). KEARNEY CO.: dry prairie, June 28, 1894, J. E. Bodin (NY, POM, RM). LANCASTER CO.: prairie formations, Sept., 1898, G. G. Hedgcock (wo). LINcoLN co. Hershey, C. D. Mill 73 (xy, vs). PHELPS co.: sand hills, Rydberg 194 (NY, vs). 1956] Turner,—A Study of the Genus Hymenopappus 263 New Mexico. BERNALILLO CO.: Sandia Mts., Tijiras Canyon, C. C. Ellis 459 (mo, Ny, US). CATRON Co.: Patterson, Aug. 15, 1900, E. O. Wooton (vs). coLFAX co.: Cimarron Canyon, Mathias 554 (Mo, POM). GRANT CO.: near San Lorenzo, July 26, 1906, £. Wooton (us). GUADALUPE co.: Anton Chico to Santa Rosa, Arsène & Benedict 16686 (cas). Mc- KINLEY co.: Gallup, Degener 4881 (Ny). Quay co.: Logan, May 31, 1911, E. Wooton (US). SAN MIGUEL CO.: vicinity of Las Vegas, Romero- ville, Arsène & Benedict 15480 (Us). SANTA FE CO.: 14 mi. S. of Santa Fe, July 7, 1951, Turner 2895 (smu, WS). UNION Co.: Emery Gap to Branson, Colorado, Eggleston 20156 (GH, NY). VALENCIA CO.: Alamositas Canyon, July 15, 1906, Wooton (us). Oklahoma. cappo co.: Cement, Demaree 12540 (MO, NY). CARTER co.: 4 mi. N. E. of Ardmore, G. E. Hall 123 (RM). CIMARRON co.: 11 mi. N. of Boise City, R. Stratton 454 (MO). CLEVELAND CO.: 4 mi. W. of Norman, Demaree 12772 (Mo, NY). COMANCHE CO.: Wichita Natl. Forest, June 12, 1926, A. J. Ortenburger (us). CUSTER CO.: 1 mi. W. & 1 mi. S. of Weatherford, Waterfall 5513 (au). ELLIs co.: Canadian R. valley, near Pack Saddle Bridge, Goodman 2598 (GH, MO, NY, RM). JACKSON CO.: near Snyder, G. W. Stevens 1275 (GH, NY). LOGAN CO.: near Guthrie, G. W. Stevens 3324 (NY). MCCLAIN CO.: Johnson's Pasture, Eskew & Barkley 1202 (Mo). OKLAHOMA CO.: 5 mi. E. & 4.5 mi. N. of Oklahoma City, Waterfall 1311 (NY). PAYNE co.: 14 mi. Š. W. of Stillwater, Stratton 3753 (CAS). ROGER MILLS co.: Roger, Antelope Hills, Ortenburger 114 (us). STEPHENS CO.: S. of Comanche, Waterfall 3680 (NY). woops co.: July 6, 1900, P. J. White (RM). South Dakota. Beapie co.: Huron, July 11, 1896, T. A. Williams (MO). CHARLES MIX CO.: Colvin, Aug. 29, 1892, E. T. & S. A. Harper (Us). MELLETTE CO.: Valley of White R., May 1855, F. V. Hayden (NY). SPINK co.: Northville, J. F. Brenckle 41-69 (CAS, GH, MO, NY, smu). TODD CO.: Highland, Antelope Cr., E. J. Wallace 22 (NY). Texas. ANTASCOSA CO.: 12 mi. N. of Pleasanton, Cory 19178 (Gu). ARMSTRONG CO.: Gamble’s Ranch, E. J. Palmer 13912 (Mo). BEXAR co.: 15 mi. Š. of San Antonio, Schulz 421 (us). BOSQUE co.: 12 mi. N. N. E. of Walnut Springs, Shinners 10067 (SMU). CALLAHAN CO.: Baird, Aug., 1882, Letterman (GH, Mo). CoMAL co.: New Braunfels, Dapprich 6207 (smu). COMANCHE co.: Round Top Mt., May 9, 1900, H. Eggert (mo). cookE co.: Tyler Bluff, western edge of county, D. S. & H. S. Correll 12996 (smu). crosBy co.: 3 mi. E. of Crosbyton, Shinners 8362 (SMU). DALLAS CO.: Dallas, May, 1876, J. Reverchon (Mo, NY). DAWSON CO.: between Lamesa and Tahoka, Small & Wherry 12130 (NY). DICKENS co.: 1.5 mi. E. of Dickens, Shinners 8379 (SMU). DOWLEY CO.: 7 mi. N. W. of Memphis, Innes & Moon 1013 (GH). GARZA co.: Post, E. J. Palmer 138856 (Mo, us). GONZALES co.: Waelder, July 9, 1889, M. Hopkins (us). HANSFORD co.: 5 mi. S. E. of Gruver, L. H. Shinners 8233 (smu). Hays co.: San Marcos and vicinity, Sept. 1, 1896, S. W. Stanfield (NY). HEMPHILL CO.: 5 mi. S. of Canadian, Shinners 8280 (GH, SMU, UC). HOOD co.: prairies N. of Granbury, May 4, 1900, H. 204 Rhodora [Vor. 58 Eggert (M0). HOWARD co.: prairies N. of Big Springs, June 11, 1900, Eggert (MO). HUTCHINSON CO.: 2 mi. S. of Borger, Shinners 8092 (sMU). LAMPASAS CO.: | mi. Š. of Lampasas, Whitehouse 15381 (SMU). LIPSCOMB co.: Lipscomb, A. H. Howell 41 (us). LUBBOCK co.: Posey Canyon, Demaree 7574, (GH, MO, WS). MEDINA CO.: 2.75 mi. S. W. of Devine, Cory 12814 (Gu). MorLEY co.: 16.4 mi. E. of Matador, Whitehouse 9914 (SMU). NOLAN CO.: 3 mi. E. of Sweetwater, Waterfall 6736 (an). OCHILTREE CO.: Š mi. S. S. W. of Perryton, Shinners 8265 (smu). oLD- HAM CO.: Magenta, Shinners 8158 (SMU). PARMER CO.: Bovina, F. S. Earle 684 (NY). rorrER co. 5 mi. S. of Canadian R., highway 287, B. & H. Jespersen 2695 (DS, MO, RM, UC, SMU, WS). RANDALL CO.: 15 mi. E. of Canyon, Palo Duro State Park, Shinners 8023 (smu, ws). ROBERTS CO.: 2.5 mi. S. W. of Miami, Shinners 8310 (SMU). SHERMAN co.: 25 mi. E. and 4 mi. S. of Stratford, Shinners 8223 (SMU). TAYLOR co.: Camp Barkeley, T'olstead 7024 (mo, smu, UC). TRAVIS co.: Glen Rose, divide between Cow and Sandy Creeks, R. T. Hill 9 (us). WICHITA co.: 10.7 mi. N. of Electra, Whitehouse 10489 (SMU). WILBARGER Co.: 6.5 mi. N. of Oklaunion, S. side of Red R., Whitehouse 10961 (smv). WILSON CO.: Kicaster School, Cory 15144 (Gu). wise co.: 3 mi. W. of Decatur, Shinners 7934 (suu). Wyoming. crook co.: Bear Lodge Mts., Aug., 1897, D. Griffiths s.n. (mo). This record should be checked since, if correct, it represents a considerable extension of range for the species. Sa. Hymenopappus flavescens Gray, var. flavescens Hymenopappus flavescens Gray, Mem. Am. Acad. Arts. Sci. n.s. 54 (1): 97-98. 1849. Rothia flavescens (Gray) O. Ktze. Rev. Gen. 1: 361. 1891. Type examined (GĦ): New Mexico. ‘Between San Miguel and Las Vegas, 10 miles W. of Vegas," Aug. 14, 1847, A. Fendler 464. Locality cited is that on isotypie material at Mo. Hymenopappus fisheri Wooton & Standley, Contrib. U. S. Natl. Herb. 16: 191. 1913. Type examined (us): New Mexico. Quay Co.: ‘“Col- lected at Nara Visa, clay and sand soil," Sept. 8, 1910, G. L. Fisher 16. Plants biennial, 45-90 cm. high, the stems single from each tap-root ; larger rosette leaves 6-14 em. long, sparsely canescent to glabrate above, densely tomentose below (rarely tomentose on both surfaces). bipinnately parted (except for the first 1—4 leaves) with broad ultimate segments mostly 2-6 mm. wide; stem leaves 15-40, becoming reduced upward; heads 30-100 per stem, campanulate, 30-70-flowered, on short ultimate peduncles 0.5-3 em. long; inflorescence a many-headed, mostly congested, cymose panicle; principal involucral bracts pubescent to glabrate, 4—5(-6) mm. long, 2-4 mm. wide, yellow-membranous for about 1 mm. from the acute or narrowly obtuse tip; corollas bright yellow, 2.5-3.5 mm. long, the tube densely to sparsely glandular, 1.5-2 mm. long, the throat abruptly campanulate to campanulate-funnelform, 0.8-1.5 mm. long, with lobes reflexed, as long as the lobes (rarely shorter); achenes obpyramidal, 4-sided, 3.5-4 mm. long, pubescent principally on the corners with 1956] Turner,—A Study of the Genus Hymenopappus 265 hairs 0.3-1 mm. long; pappus of 18-20 linear-oblong scales, 0.5-1(-1.2) mm. long; anthers mostly completely exserted, about 2 mm. long; » — 17. Disrripution.—Principally northwestern Texas on the Llano Estacado (Staked Plains) and its periphery; eastern New Mexico, western Okla- homa, southwestern Kansas and southern Colorado (known in this latter state by only one collection from along a roadside south of Trinidad), mostly in deep, red, sandy soils, commonly associated with *shinnery" (Quercus spp.) in the western part of its range (Fig. 46). Late May- September (See Fig. 44). In the eastern part of its range (Hemphill Co., Texas) there is considerable intergradation of this variety with Hymenopappus scabiosaeus var. corymbosus with respect to corolla and leaf shape, pappus length, and general pubescence; in the western part of its range (eastern New Mexico) there is a similar but strong intergradation of these same characters with those typical of H. flavescens var. cano-tomentosus. This, combined with the ranges of the taxa involved, makes it seem probable that the variety is of hybrid nature, having had its origin at some past time, perhaps during a pluvial period, when the ranges of H. scabiosaeus var. corymbosus and H. flavescens var. cano- lomentosus had considerable overlap. With subsequent with- drawal of the putative parents at a later time, the hybrids and their derivatives were left isolated and have since evolved more or less independently. A similar situation has been hypothesized for some species of Quercus by Muller (1951). An alternate hypothesis would be that the variety has evolved, without this factor of hybridization, from H. flavescens var. cano-tomentosus entirely as a result of mutation and selection in its appropriate habitat and has since this initial isolation come into contact and introgressed with the two peripheral taxa mentioned. Experimental crosses between these entities should do much to offer positive evidence. The taxon is placed as a variety within the same species as cano-tomentosus since it resembles this taxon phenotypically more than it does H. scabiosaeus var. corymbosus, perhaps indicating that it has drawn a larger number of characters from the latter by the hybridization and introgression visualized above. Figure 44 has been constructed from information tabulated by the method given in the footnote, p. 241. By comparison 266 Rhodora [Vou. 58 of the flowering dates of the taxa mentioned, it will be noticed that var. corymbosus begins flowering early, rapidly reaches a peak, then drops off less sharply leaving only a few late-flowering individuals which overlap into the longer, continuous flowering period of var. cano-tomentosus. Assuming that var. flavescens became partially isolated seasonally from H. scabiosaeus var. corymbosus, either as a result of hybridization or mutation, it is clear that present crossing and resulting introgression will be more with H. flavescens var. cano-tomentosus than H. scabio- saeus var. corymbosus since there would be more opportunities for crossing with the former. Hymenopappus flavescens was reduced to synonymy under H. corymbosus by Johnston (1923) with the following statement, "Gray's H. flavescens was based upon a good specimen of the present species, H. corymbosus collected by Fendler, and upon a few fragments, apparently of H. artemisiaefolius, which were collected by Wislizenus." However, H. flavescens as defined by Gray may be distinguished immediately throughout its range from corymbosus by its later blooming period, yellow flowers, and more campanulate corolla-throat; the latter is % IN FLOWER * ° — — '5 AN ^ Dy / n ` À 2 Ú SN / N \ ¥ / \ / Ë \ / / \ 10 $/ X / \ £ Y / \ A \ $ N \ / £ / N / \ 5 z / N \ uM «/ / ` sanm 2/ / S / -— — / / ~~ d — / u - L vo " 4 ' 20 ' 10 20 ' 10 20 ' 10 20 ' 10 ' 10 20 ' 10 MARCH APRIL MAY JUNE JULY AUGUST SEPT Fra. 44. Graph showing dates of flowering expressed in per cent of herbarium specimens examined from eastern New Mexico, Texas, and western Oklahoma. 1956] Turner,—A Study of the Genus Hymenopappus 267 white flowered and has a predominately funnelform throat. Fendler's collections (type and isotypes) are biologically typical specimens of H. flavescens as are Wislizenus’ fragments. H. artemisiaefolius is a species of the pine woods in eastern Texas, and certainly was not collected by Wislizenus on the “Santa Fe Road" in New Mexico. REPRESENTATIVE SPECIMENS.— Colorado. LOS ANIMAS Co.: 12 mi. 8. of Trinidad, highway 87, May 29, Brenckle 48139 (swv). Kansas. GRANT CO.: Ulyesses, Thompson 64 (Mo, US). HAMILTON CO.: sandhills, A. S. Hitchcock 607 (286) (GH, MO, NY, RM). SEWARD CO.: W. of Liberal, McKelvey 2488 (GH, POM). STEVENS CO.: sandhills, H. W. Norris 91 (wo). New Mexico. cHavES co.: 7 mi. N. E. of Boaz, Waterfall 4321 (Gu, wo, NY). DE BACA CO.: La Lande, Pohl 5036 (smu). EDDY CO.: near Loving, Standley 40361 (us). LEA co.: 5 mi. N. of Eunice, Turner 2947 (smu, ws). LINCOLN CcoO.: 35 mi. W. of Roswell, F. S. & E. S. Earle 508 (NY). Quay co.: Nara Visa, G. L. Fisher 16 (RM, US). ROOSEVELT CO.: 5 mi. N. E. of Portales, Goodman & Hitchcock 1123 (cas, DS, GH, MO, NY, RM, UC). SAN MIGUEL CO.: between San Miguel and Las Vegas (10 mi. W. of Vegas), Fendler 464 (417) (GH, MO). SANTA FE CO.: near La Glorieta, Brandegee 12068 (Mo, uc). UNION co.: Willow Bar of the Cimarron, Fendler 463 (417) (mo). Oklahoma. BECKHAM co.: Sayre, R. Stratton 338 (mo). DEWEY CO.: W. of Vici, Goodman 2577 (CAS, GH, MO, NY, RM). WOODWARD co.: Indian Cr. Station, 7 mi. S. E. of Woodward, A. &. R. Nelson 5651 (RM). Texas. BAILEY CO.: 5 mi. N. W. of Muleshoe, Correll 13107 (smu). CASTRO co.: Dimmitt, E. L. Reed 3558 (us). CHILDRESS CO.: 9 mi. N. of Childress along Red R., Whitehouse 18699 (SMU). CRANE CO.: near Crane, highway 51, L. Cutak 3 (mo). crosBy co.: 30 mi. S. W. of Spur along Blanco R., Erlanson 1196 (smu). DALLAM CO.: 7 mi. N. W. of Dalhart, Shinners 8176 (GH, RM, SMU, UC). DONLEY CO.: 9 mi. S. W. of Claredon, Innes & Moon 1019 (ps, GH). FISHER CO.: 7.5 mi. E. of Roby, Whitehouse 16724 (SMU). GAINES CO.: 15.1 mi. W. of Lamesa, highway 180, Whitehouse 16779 (smu). GARZA co.: near Double Mt. R., A. Ruth 1308 (vs). HALL co.: W. of Estelline, May 26, 1906, Reverchon (Mo). HARDEMAN CO.: 12 mi. N. of Chillicothe, Cory 18395 (GH). HARTLEY CO.: 10 mi. E. of Romero, Cory 16464 (GH). HEMPHILL CO.: prairies N. of Canadian, June 7, 1901, H. Eggert (Mo). HowarD co.: near Big Spring, E. J. Palmer 34006 (wo, NY). KENT co.: 1 mi. S. of Jayton, Shinners 8387 (SMU). LAMB CO.: 8 mi. S. of Olton, Cory 13550 (GH). LUBBOCK co.: Lubbock, Studhalter 1270 (us). MrrcHELL co.: N. of Colorado, June 9, 1900, Eggert (Mo). POTTER CO.: 1.4 mi. S. of Canadian R. Bridge, highway 287, B. & H. Jespersen 2686 (DS, MO, RM, SMU, UC, WS). RANDALL co.: Palo Duro Canyon, A. C. Martin 292 (us). RUNNELS CO.: Ballinger, E. J. Palmer 10326 (ps, Mo, US). TERRY CO.: Wellman, July 10, 1941, B. C. Tharp (GĦ, MO). wann co.: 3.5 mi. E. of Monohans, sand dunes, C. H. Muller 8528 (suu). 268 Rhodora [Vor. 58 sp. Hymenopappus flavescens var. cano-tomentosus Gray Hymenopappus flavescens var. cano-tomentosus Gray, Pl. Wright 2: 94. 1852. Type examined (an): “Sandhills near Frontera, New Mexico; April, May, " 1851-1852, C. Wright 1412. Quoted locality is that given in the type description. Hymenopappus canescens var. cano-tomentosus Rothrock, in Wheeler Exped. 6: 167. 1878. This name is apparently an error for Hymeno- pappus flavescens var. cano-tomentosus Gray. Hymenopappus robustus Greene, Bull. Torr. Bot. Club 9: 63. 1882. Probable isotype examined (aH): New Mexico. Grant Co.: “collected near Santa Rita del Cobre in 1877," E. L. Greene s.n. (22). A type was not designated by Greene, only the phrase “common on the sandy plains of New Mexico" being used to identify the collections from which the description was drawn. The specimen cited above has the annotation "forma robusta" on the collection label, apparently put there by Greene himself. There can be little doubt that the plant referred to is H. fla- vescens var. cano-tomentosus Gray. It is apparently common on the plains near Silver City, New Mexico, where Greene resided at the time of his publication, Plants biennial, 30-90 em. high, the stems typically densely tomentose, single from each tap-root (rarely 2 crowns forming on injury); larger rosette leaves 6-15 em. long, 2-4 em. wide, evenly pubescent on both surfaces, usually densely tomentose, but often merely canescent, bipin- nately dissected with the small, narrow, ultimate segments 1-2 mm. wide; stem leaves 10-50, becoming reduced upwards; heads 15-100 per stem, campanulate, 30-90-flowered, on ultimate peduncles 1-6 cm. long; principal involueral bracts (4-)5-8 mm. long, 2-4 mm. wide, densely tomentose to nearly glabrate, yellow-membranous for about 1 mm. from the aeute to rarely obtuse tip; corollas yellow, 2.5-3.5 mm. long, the tube glandular, 1.5-2 mm. long, the throat abruptly campanulate 1-1.5 mm. long, with lobes reflexed, 1-1.5 times longer than the lobes; achenes obpyramidal, 4-sided, 3-4.5 mm. long, evenly pubescent with hairs 0.5-1.5 mm. long; pappus of 16-22 linear oblong scales, 1-1.5 mm. long (very rarely less); anther usually completely exserted, 1.8-2.2 mm. long; n — 17. DisrRIBUTION.—Gravelly, rocky, sandy or sandy limestone soils on intermontane plains of Trans-Pecos Texas, western New Mexico, eastern Arizona and adjacent areas of Mexico (Fig. 46). Flowering at two principal times: late April to May and again in late July to early August (Fig. 44), apparently depending on rainfall. Hymenopappus flavescens var. cano-tomentosus can be dis- tinguished from the closely related var. flavescens by its more finely dissected leaves with narrow segments and by its generally shorter pappus and involucral braets. However, at the eastern periphery of its range, it intergrades almostly completely with 1956} Turner,—A Study of the Genus Hymenopappus 269 var. flavescens (see discussion under that variety). Otherwise, it is essentially free of contamination from other taxa. As an exception is a single collection which seems to represent a clear hybrid between var. cano-tomentosus and some member of the H. filifolius complex (probably var. cinereus or var. lugens): Arizona. Navajo Co.: Kayenta, 1922, John Weterill s.n. (NY). The specimen is tall and leafy as is var. cano-tomentosus, but the leaves (which are obviously atypical) and the woolly stems approach those of H. filifolius. The inflorescence is apparently retarded in its development, being represented by a tight, abnormal cluster of heads at the stem apex. REPRESENTATIVE SPECIMENS.—Arizona. APACHE CO.: Adamana, Sept. 1, 1909, H. M. Rusby (NY). GILA co.: Tonto Hill, Collom 163 (GH, Mo, NY, US). GRAHAM CO.: between Safford and Globe, Peebles 14601 (us). NAVAJO co.: 14 mi. E. of Holbrook, Goddard 698 (uc). PIMA co.: Santa Catalina Mts., Lemmon 218 (cu). New Mexico. BERNALILLO CO.: Albuquerque Mesa, Castetter 1228 (RM). carron cO.: 8 mi. S. W. of Horse Springs, Preece & Turner 2738 (SMU, WS). DE BACA CO.: Buchanan, Aug. f2, 1909, Wooton (Us). DONA ANA co.: Organ Mts., Wooton 139 (ps, MO, NY, POM, RM, UC, US). EDDY CO.: 10 mi. W. of Hope, Aug. 4, 1905, Wooton (us). GRANT co.: Whitewater Junction, Silver City Eastwood 8530 (cas, GH). LUNA co.: Nutt, Diehl 751 (POM). SAN JUAN co.: 45 mi. N. W. of Cuba, Preece & Turner 2753 (SMU, WS). SOCORRO co.: near Socorro, Aug., 1880 (1881), H. H. Rusby 180 (wo, NY, WS). TORRANCE CO.: Willard, Aug. 26, 1904, Wooton (US). VALENCIA CO.: E. of Laguna Pueblo, A. &. R. Nelson 2176 (mo, nM). Texas. CULBER- SON co.: 2 mi. W. of Van Horn, Waterfall 4409 (cas, GH, MO, NY, SMU). EcTOR CO.: Odessa, E. L. Reed 1907 (Us). EL Paso co.: E] Paso, May, 1881, G. R. Vasey (D8, GH, US). HUDSPETH CO.: 32 mi. W. of Sierra Blanca, highway 62, on pipe line road, Tharp 46149 (RM, RSA, UC). LOVING CO.: 10 mi. E. of Mentone, Turner 984 (smu). PRESIDIO CO.: 2 mi. W. of Marfa, Warnock 5592 (swv). REEVES CO.: Saragosa, Warnock 5271 (smu). WARD co.: Barstow, Earle & Tracy 43 (NY). Mexico. Chihuahua.: near Paso del Norte, Pringle 759 (GH, MO, NY, UC, US). (To be concluded) 270 Rhodora [Vor. 58 LYGODIUM PALMATUM vs. UrBANIZATION.—The climbing fern, Lygodium palmatum (Bernh) Sw., was once so plentiful within the limits of what is now the city of Hartford, Conn., that it was locally known as the Hartford Fern. But residents of Windsor, adjoining Hartford on the north, found it in such abundance there that they insisted on calling it the Windsor Fern. Citizens of both these communities recognized its com- mercial value and collected it extensively for ornamental dis- play. 'The Connectieut Legislature, in an attempt to preserve it from extermination, passed a law in 1869 making it an offense punishable by fine for any person to gather it without the con- sent of the owner on whose property it was growing. This law is still on the statute books of Connecticut. Despite this legal attempt at its preservation the advance of civilization caused human habitations to occupy the habitats of these ferns, so that for a quarter century or more no plant of this fern has been found growing wild in either Hartford or Windsor. Lygodium palmatum is now so rare in the Connecticut valley that even professional botanists confess that they have never seen it growing in its natural surroundings; and nursery- men in Connecticut, Massachusetts and Vermont offer it for sale as a rare species. Gray's Manual (8th edition) states that its range extends from Georgia north to southeast New York, southern New Eng- land, and west to Ohio and Kentucky. It would be of interest to both botanists and conservationists to have a report from the states outside New England as comprehensive and authorita- tive as that given by Dr. Shaver in his “Ferns of Tennessee." He also notes its appearance at four stations in Great Smoky Mountains National Park in West Virginia. Such a report would indieate to what extent this fern is withstanding the attacks of civilized humanity. Herewith is a partial report of its status in the Connecticut Valley based on observations covering the last decade. A small colony was discovered in a wooded area in Granby, Ct. by Mrs. Harry L. Oppenheimer, of Suffield, Conn., in 1950. Since that date the owner of the property has cleared his acres for building sites and eradieated the fern. 1956] Smith, Lygodium palmatum vs. Urbanization 271 A similar colony was discovered in Wilbraham, Mass., near the Springfield line, and reported to me by Mr. R. A. Doray, of Greenfield, Mass. This station I visited in 1950, but I noted that it was very close to the highway and that the city was extending its tentacles dangerously near to the spot. On my last visit, in October 1955, I found one of these modern houses (two bedrooms and a two-car garage) occupying the site. Mr. Doray reports that as recently as 1953 a sizeable colony persisted in a swamp off Tinkham Road, Springfield, a short distance from Wilbraham Road. Another station, in a pasture in North Amherst near the road to Sunderland, is apparently not yet threatened by real estate developments. In the spring of 1955 Mr. Perey E. Fellows of Wethersfield, Conn. brought me a spray of the fern which he had found close to the highway in Rocky Hill, Conn. On visiting that spot some weeks later, all vegetation, including the fern, had become a ghastly brown, the result of spraying along the highway by the State Highway Department in its successful effort to destroy roadside weeds. Whether the fern will survive this treatment and make another attempt to survive next spring is problematical. In Granby, Conn., along an abandoned wood road, a small colony was discovered in 1954. This station is known to the discoverer only. Also in Granby is another small colony on property owned by Johnson’s Nursery of Southwick, Mass. This station is being perpetuated for commercial purposes and will be well guarded by the owner. Near the Granby line, in Suffield, Conn., another colony was discovered in 1952. This is on property owned by a fern-lover who is determined to preserve this fern from destruction. Its location is known only to a few high-minded botanists. Its site will be revealed in the future only to those who respect the laws of Connecticut, who share the owner’s desire that no plant shall be molested, and are content to observe the modified adage: “a fern in the wood is worth two in the hand.” It is high time that botanists, fern-lovers and conservationists throughout the country enter into a compact to save this lone American species of Lygodium from extermination.—JEssE F. SMITH, SUFFIELD, CONN. 272 Rhodora [Vor. 58 CLARENCE R. Hanes.—The death of Mr. Hanes on February 10, 1956, brought to a close the career of one of Michigan’s most active field botanists. Clarence Robert Hanes was born March 26, 1874, in the family home in Schoolcraft, Michigan, where he lived for the rest of his life (except while away at school). He farmed the family property in the vicinity, and was very active in civic affairs. Upon graduation from Schoolcraft High School, where he concentrated in languages, he continued his studies at the University of Michigan, from which he graduated in 1898. In 1911, he married Florence Nutten, who later shared with him a quarter century of intensive study of the local flora. Started as an avocation in the early 1930’s, the Hanes’ floristic work was originally limited to the area within five miles of their home in Schoolcraft, Kalamazoo County, Michigan. Botanical authorities to whom they sent some of their unusual finds urged that they extend their study to the entire county. Subsequent years of thorough field work culminated in the publication of their flora of the county in 1947.1! This remains the finest county flora yet produced for the state of Michigan—and one of the few really good local floras for the entire Great Lakes region. Until Mr. Hanes’ death, he and his wife continued their collect- ing, with additional work in other counties of western Michigan. Mrs. Hanes, who did all of the mounting for the herbarium in addition to sharing fully in the field work, has been completing the mounting of some recent collections from other counties. Much of the area about Schoolcraft was covered by Mr. Hanes on his bicycle, while their old farm truck carried them to more distant areas. Other Kalamazoo County residents whose col- lections and interest aided the Hanes’ studies included H. R. Becker (who also died in February), Dr. L. A. Kenoyer of Western Michigan College, and Fred W. Rapp. The herbarium contains many critical species determined by specialists to whom material was sent. Of particular significance was Hanes’ col- laboration with Liberty Hyde Bailey on the genus Rubus, over 20 new species of which were described by Bailey from Kalamazoo County (including A. hanesii, R. florenceae [named for Mrs. Hanes], AH. schooleraftiamus, and R. kalamazoensis). Another ! CLARENCE R. Hanes AND FLonEkNCE N. Hawks, Flora of Kalamazoo County, Michigan. Vascular Plants. Schoolcraft, Michigan, 1947. xii + 295 pp. (Re- viewed in Rhodora 49: 143—144, 1947.) 1956] Steyermark,— Flora of Winnebago Co., Illinois 273 valuable collaboration was with F. J. Hermann on the genus Carex, many new county and state records having resulted from the Hanes’ field work in southwestern Michigan. The Hanes herbarium has been given to Western Michigan College in Kalamazoo, where it is housed in the ‘Clarence R. Hanes Room," dedicated May 2. The herbarium is to be kept as a unit, neither being distributed among other collections nor having additions inserted, until such time as a flora of Michigan is published. Given with the herbarium was a large record book in which notes and observations are recorded geographically, by the various localities in the vicinity. The discoverer of many species previously unknown from Michigan, Mr. Hanes published his findings primarily in the Papers of the Michigan Academy of Science, Arts, and Letters. He also submitted a number of notes to RHoporA, of which the most recent was published in January, 1956.—Epwarp G. Voss AND JAMES S. WILSON, UNIVERSITY OF MICHIGAN, ANN ARBOR. FLonA or WINNEBAGO County, ILuiNots!.—Local floras are valuable as supplements to state and regional floras. By confin- ing their data and subject matter to the local vegetation, they emphasize many aspects of a flora that are lost or otherwise not evident in state and regional manuals. Local floras further- more reflect the intensive concentrated work of an enthusiastic botanist who has devoted all his spare moments to these interests. To such a category of devout enthusiastic workers Dr. Fell belongs. By profession a physician, his avocational interest throughout his life has been centered on the collecting and study of the native flora of this northern Illinois county, in which the city of Rockford is located. The present volume represents the results of a life time of study and is an admirable contribution to add to other recent works relating to the Illinois flora. Actually, it is an expanded treatment of the Check list of the vascular plants of Winnebago County, Illinois, published in 1949 (Trans. Ill. Acad. Sci. 42: 68-79) by George D. Fuller, E. W. Fell, and G. B. Fell. The book under review is packed with the type of phytogeo- 1 PEnLL, Kapert W. Flora of Winnebago County, Illinois. Nature Conservancy. Washington, D. C. 1-207, illustrated. 1955. 274 Rhodora [Vor. 58 graphical and ecological information which furnishes the reader with data he could not easily find elsewhere if at all. In addi- tion to such topics as climate, geology and physical features, types of habitats within the county, there are interesting units discussed under the distribution of species, and plant refuges. The diseussion of the last topie refleets Dr. Fell's profound intimacy with the flora of the region. Altogether, 1210 species of vaseular plants (ferns and flowering plants), of which 1013 are native, are treated in the flora. The data are based upon the extensive herbarium specimens collected by Dr. Fell, as well as upon those of M. 8. Bebb, Evelyn I. Fernald, and various other local botanists and collectors. Eighty seven illustrations of various species, five photographs representing different kinds of habitats, one photograph of M. S. Bebb, a map of Winnebago County showing distribution of prairie and woodland, and of names of places mentioned in the flora, and a detailed index serve to widen the usefulness of the book. Categories below the species level are not recognized, a prac- tice with which many botanists will not be in agreement. The adherence to “species only" concept has led to the maintenance of Botrychium dissectum and B. obliquum, and of Dryopteris spinulosa and D. intermedia as distinct species, whereas most fern students have treated these as forms or varieties. Again, this style has led to the listing of Lespedeza longifolia DC. as a distinct species. However, as brought out in the eighth edition of Gray's Manual, L. longifolia DC. is a little known eastern plant, and the proper position of it is in varietal rank under L. hirta. It has no close taxonomic relationship with L. capitata, with which Dr. Fell has indicated it as *(L. capitata var. longifolia)." Many other similar cases could be specified, but this does not detract greatly from the general usefulness of the work. Dr. Fell is to be congratulated for having brought out this splendid book, which definitely fills an important gap in the literature on the local flora and on Illinois plant geography and ecology.—JULIAN A. STEYERMARK, CHICAGO NATURAL HISTORY MUSEUM, CHICAGO, ILLINOIS AND MISSOURI BOTANICAL GARDEN, ST. LOUIS, MISSOURI. Volume 58, no. 692, including pages 199-242, was issued 29 A ugust, 1956. ij 73 47/9/7906 (f « axvenenet vien RY a JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD CARROLL EMORY WOOD, JR, Vol. 58 October, 1956 No. 694 CONTENTS: The Identity of Calyptocarpus blepharolepis. S. F. Blake....... 275 Range Extensions in Northwestern Plants. F. J. Hermann...... 278 Nuttall Not the Author of Fraser’s Catalogue. Lloyd H. Shinners 281 Some Additions to the Orchid Flora of Puerto Rico. Harold F. Winters and Charles Schweinfurth ........ tp nn m 290 Growth Habits of Arabis perstellata. HM. Lucy Braun........... 292 A Cytotaxonomic Study of the Genus Hymenopappus (Compo- sitae m Mite I. urnen (concluded) E 295 Notes on Collinsia violacea. Glen S. W nterrtnger.............. 308 Tiarella cordifolia in Wisconsin. Paul F. Maycock............. 309 Medicago polymorpha var. vulgaris. Lloyd H.Shinners......... 310 The New England Botanical Club, Inc. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $6.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 60 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $5.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scientific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of pub- lication. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U.S. A. QTRbooora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 October, 1956 No. 694 THE IDENTITY OF CALYPTOCARPUS BLEPHAROLEPIS S. F. BLAKE THE type sheet of Calyptocarpus blepharolepis Robinson (Proc. Amer. Acad. 47 (Contr. Gray Herb. 39): 214. 1911) in the Gray Herbarium bears a printed label indicating that the material had been collected by S. M. Tracy (no. 8946) at Tensaw, Ala- bama, August 18, 1904, and identified as Calyptrocarpus! tampi- canus Small [= Calyptocarpus vialis Less. by E. L. Greene. Both specimens on the type sheet are characteristic Sanvitalia ocymoides DC., matching very well a number of collections in the Gray Herbarium, including Berlandier 2102 (type collection of S. ocymoides), Berlandier 2233 (type collection of S. tragiae- folia DC., a synonym of S. ocymoides), Pringle 1918, and others. In the same herbarium is a sheet of true Calyptocarpus vialis Less. with printed label stating that it had been collected by Tracy (no. 8942) at Corpus Christi, Texas, April 4, 1905, and determined by Greene as Sanvitalia ocymoides DC. The specimen of Tracy 8942 in the British Museum of Natural History, which I examined many years ago, agrees with that in the Gray Herbarium, and the specimen in the U. S. National Herbarium is like the others. "The label of the sheet of no. $946 in the latter herbarium, however, has only the heading printed, the remainder being in ink. It is labeled in Greene’s hand Calyptrocarpus tampicanus Small; the number $946 and the data —Brownsville, Texas, April 13, 1905—are in another hand- writing, which Mrs. Agnes Chase identifies as that of S. M. Tracy 1 Lessing's name Calyptocarpus has been persistently but improperly emended by authors to Calyptrocarpus. 276 Rhodora [Vor. 58 himself. The plant itself is Sanvitalia ocymoides, like all the other specimens of 8946 examined. This specimen (8946) is sheet no. 1,240,727 in the U. 8. National Herbarium, in a quite different group of numbers from the sheet of Tracy 8942 with the printed label, which is no. 513729. The latter is one in a broken series of nearly 300 numbers received from Tracy himself (between no. 7642 and no. 9618). The sheet of 8946 falls in a much smaller lot including about 70 specimens, which has no accession number, so that its source is not definitely recorded. The lot consists largely of Canadian plants (mostly Compositae) collected by Macoun and Spreadborough, but includes among others 9 Tracy specimens, all numbered in the 8900's. Of 7 of these that I have been able to trace in the herbarium, all 6 besides the Calyptocarpus have a mostly handwritten label like that of the Calyptocarpus (data in Tracy’s hand, name in Greene’s). One of them, labeled Car- phephorus corymbosus in Greene’s hand, is actually Polypteris integrifolia Nutt., and a duplicate with completely printed label, belonging in the lot of nearly 300 numbers referred to above, bears the same data. In this case, there is no reciprocal specimen of Carphephorus labeled as Polypteris in the herbarium and it would appear that Greene, in a moment of mental aberration, had actually made this misidentification. It seems reasonably certain that this lot of about 70 specimens was received from Greene himself, perhaps at different times, since there is no lot number. From all this it would appear that Greene had actually identi- fied no. 8946 as Calyptocarpus, although it is Sanvitalia. Such a misidentification, although unlikely, is possible, but it seems most improbable that Greene would have misidentified the other component of the pair in a precisely reciprocal way. If a trans- position somehow occurred, it seems impossible to determine from the available evidence whether the locality was switched as well. Mrs. Chase tells me that when she looked through the grasses of Tracy's herbarium in 1907, at his invitation, she found the arrangement of the material far from orderly, so that it would be easy to believe that an earlier transposition of numbers or labels might have occurred there. Whatever the truth of this matter, the fact that evidently both 8942 and 8946 were 1956] Blake,—Calyptocarpus Blepharolepis 277 collected in Texas makes it possible to discard the locality Tensaw, Alabama, attributed to the type of C. blepharolepis on its printed label and in the original description. Whether it actually came from Brownsville or from Corpus Christi is un- certain. Dr. Albert L. Delisle informs me that there 1s no material of either Calyptocarpus or Sanvitalia in the E. L. Greene Herbarium at Notre Dame University, and Dr. F. W. Gould writes that the Tracy Herbarium preserved at the Texas A. & M. College at College Station, Texas, does not have a specimen of either of the two Tracy exsiccatae in question. Aside from the evidence supplied by the Texas locality on the handwritten label of no. 8946, it is extremely unlikely that either Sanvitalia ocymoides or Calyptocarpus vialis could have been found in Alabama in 1904. So far as the Sanvitalia is con- cerned, there are at the present time no specimens from east of Texas in the herbaria at Washington, Beltsville, Cambridge, and New York, and I know of no records for it. Calyptocarpus vialis, although now known from Texas to Florida, and south into Mexico, does not appear to have begun its eastward march so early. Small (Flora, ed. 1, 1903 and ed. 2, 1913) gave it only from southern Texas, and it is not listed in Mohr's Plant Life of Alabama (1901), in Lowe's Plants of Mississippi (1921), or in any of the local floras of these states and of Louisiana and Florida listed in part 1 of the Geographical Guide to Floras of the World (1942). It is represented in the U. S. National Herbarium from the states east of Texas only by a sheet from New Orleans, La., collected by H. R. Reed in 1927. I collected it myself in waste ground in Audubon State Park in that city in the spring of 1955. In the herbarium of the U. S. National Arboretum are specimens collected by O. M. Freeman in lawns at Miami, Florida, in Jan. 1940 and again in Feb. 1948. In the herbarium of the New York Botanical Garden (data supplied by R. C. Cowan) are specimens from LOUISIANA: Waste lots, New Orleans, 1930, Small and Alexander. ALABAMA: Campus of the University, on damp, shaded, wooded hillside, Tuscaloosa, 1951, Bert C. Williams. FLORIDA: On edge of lawn, growing in dense mat for about 100 feet, Coral Gables, 11 Jan. 1946, W. M. Buswell. 278 Rhodora [Vor. 58 To summarize: It is evident that Dr. Robinson, observing that the material distributed as ‘‘Calytrocarpus tampicanus" under Tracy’s no. 8946 was not that species, proceeded to describe it as new without sufficiently checking the generic characters. It is also quite evident that the specimens of Sanvitalia ocymoides on which the new specific name was based were collected in 'Texas, not in Alabama, but it is not certain whether they came from Brownsville or Corpus Christi. The name Calyptocarpus blepharolepis Robinson becomes a synonym of Sanvitalia ocy- moides DC., and the genus Calyptocarpus remains monotypic.— PLANT INDUSTRY STATION, BELTSVILLE, MD. RANGE EXTENSIONS IN NORTHWESTERN PLANTS F.J. HERMANN DuniNc the late summer of 1955 Bernard M. Leese and the writer were detailed by the Plant Introduction Section, Horticul- tural Research Branch, U.S. Department of Agriculture to collect seed of native plants in northwestern United States for forage experimentation. A number of the plants collected represent appreciable extensions in range, new state records, ete., which seem to be worth recording. Collection numbers cited in the list are those of the writer. Herbarium abbreviations, taken from Lanjouw's Index Herbariorum (Regnum Vegetabile Vol. 2, Part 1, 1954), are as follows: CA—California Academy of Sciences, San Francisco; Mont— Montana State College, Boze- man; NA—National Arboretum Herbarium, Washington, D. C.; RM—Rocky Mountain Herbarium, Laramie, Wyo.; and US— National Herbarium, Washington, D. C. PLANTS NEW TO MONTANA MUHLENBERGIA GLOMERATA (Willd.) Trin. Salix-Potentilla bog, 2 miles NE. of Upper Red Rock Lake, Beaverhead Co., Sept. 10 (12491, Mont, U. S). CAREX LEPTOPODA Mack. Springy roadside embankment bordering Thuja woods E. of Lake McDonald, Glacier National Park, Flathead Co., Sept. 4 (12440, Mont, NA); boggy edge of streamlet, clearing in larch- hemlock woods, SW. of Hungry Horse Dam, ca. 5 miles SE. of Hungry Horse, Flathead Co., Sept. 6 (12466, Mont.). Reported from Glacier National Park by Standley but not recorded from the state by Mackenzie. 1956] Hermann,—Extensions in Northwestern Plants 279 JUNCUS COMPRESSUS Jacq. Marshy roadside ditch, 9 miles S. of Wisdom, Beaverhead Co., Sept. 9 (12484 CA, Mont, NA, US). Ap- parently not previously reported west of eastern Ontario. Additions to the Flora of Glacier National Park The following species are not recorded in Standley’s Flora (Contrib. U. S. Nat. Herb. 22: 235-438. 1921) or in the supplementary reports by Graff (Bull. Torr. Bot. Club 49: 175-181. 1922), Maguire (Rhodora 36: 305-308. 1934 and 41: 504-508. 1939), McLaughlin (Rhodora 37: 362- 365. 1935) and Harvey (Proc. Mont. Acad. Sei. 14: 23-25. 1954). ELvMvus iNNovATUS Beal. Moist open bank of Belly River, 3 miles 8. of Canadian border, Glacier Co., Sept. 3 (12432, Mont, US). CAREX CUSICKHI Mack. Marshy border of pond, end of old logging road No. 20, alt. ca. 3500 ft., Flathead Co., Sept. 5 (12445, Mont, NA; 12447, US). CAREX MICROPTERA Mack. Low moist edge of log trail through spruce- fir woods, alt. ca. 4900 ft., Piegan Pass Trail ca. 1⁄4 mile N. of Gunsight Lake Trail, Glacier Co., Sept. 2 (12413, Mont, NA). CAREX PHYSOCARPA Presl. Marshy clearing in willow thicket, W. shore of Swifteurrent Lake, alt. 4860 ft., Glacier Co., Aug. 30 (12368, Mont, NA). Standley’s report of C. miliaris Michx. is probably referable to this species. CAREX sTIPATA Muhl. Springy roadside embankment bordering Thuja woods E. of Lake McDonald, Flathead Co., Sept. 4 (12439, Mont, NA); edge of Lime Springs Beaver Pond, alt. 3500 ft., 5. of Apgar Mts., Flathead Co., Sept. 4 (12443, Mont, US). PLANTS NEW TO WYOMING CAREX HETERONEURA W. Boott. Wet mossy border of streamlet in narrow ravine, alt. 8700 ft., W. slope of Mt. Washburne, ca. 4 mile N. of Dunraven Pass, Yellowstone National Park, Sept. 12 (12535, NA; 12536 CA, Mont, RM, US). Previously known only from California and western Nevada. CAREX LEPORINELLA Mack. Dried bed of draw in meadow, Conti- nental Divide, alt. 8300 ft., 4 miles NW. of West Thumb, Yellowstone National Park, Sept. 12 (12519, CA, Mont, NA, RM). Recorded only from the Pacifie States by Mackenzie; recently reported (Amer. Midl. Nat.51:272. 1954) also from Nevada. CAREX LIMNOPHILA F. J. Hermann (Leafl. West. Bot. 8: 28 1956). Gravelly edge of road bordering willows on north shore of Half Moon Lake, alt. ca. 7500 ft., Wind River Range, 7 miles NE. of Pinedale, Sub- lette County, Aug. 21 (12252, CA, NA, RM, US); abrupt mossy shore of Bridge Bay, Yellowstone Lake, alt. ca. 7800 ft., 3 miles SW. of Lake Junction, Yellowstone National Park, Sept. 12 (12527, Mont, NA, US). Endemie to Wyoming so far as known. CAREX PRAECEPTORUM Mack. Willow fringe of Little Bear Creek, alt. ca. 9000 ft., in meadow 214 miles ENE. of Beartooth Lake, Park 280 Rhodora [Vor. 58 Co., Sept. 13 (12552, Mont, NA, RM). Known to Mackenzie only from the Pacific States; recently reported (Amer. Midl. Nat. 51: 274. 1954) also from Nevada. CAREX STENOPTILA F. J. Hermann (Leafl. West. Bot. 4: 194. 1945). Rocky bank of Isa Lake, Craig Pass, alt. 8300 ft. (12512, CA, Mont, NA, RM) and dry open lodgepole pine-fir woods, alt. 8700 ft., W. slope of Mt. Washburne ca. 14 mile N. of Dunraven Pass, Yellowstone National Park, Sept. 12 (12537, CA, NA, US). Previously known only from Colorado. CAREX SUBSTRICTA (Kükenth.) Mack. Sedge meadow, Fiddler’s Lake, Wind River Range, alt. ca. 10,000 ft., 24 miles SW. of Lander, Fremont Co., Aug. 23 (12280, NA). Juncus BIGLUMIS L. Edge of cold streamlet on tundra below Beartooth Pass, alt. 10,500 ft., Park Co., Sept. 13 (12562, CA, Mont, NA,RM, US). Although Rydberg (Flora of the Rocky Mountains and Adjacent Plairs, p. 152. 1922) indicates the range of this arctic-alpine species as “Greenl. —B.C.—Alaska; Eurasia," Buchenau (in Engler, Das Pflanzenreich 4 (36): 223. 1906) cites a collection of Theodore Holm’s from Long’s Peak, Colorado, and Standley (Contrib. U. S. Nat. Herb. 22: 307. 1921) reports the plant from Gunsight Pass, Glacier National Park, Mont. Juncus TRACYI Rydb. Dominant on wet open springy slope, edge of Douglas fir-aspen woods, alt. ca. 7000 ft., 5.6 miles W. of Tower Junction, Yellowstone National Park, Sept. 13 (12541, CA, Mont, NA, RM, US). Juncus vAsSEYI Engelm. Damp crevices on rock shore, head of Half Moon Lake, alt. ca. 7500 ft., Wind River Range, 8 miles NE. of Pinedale, Sublette Co., Aug. 21 (12262, NA). Additions to the Flora of Yellowstone National Park McDougall and Baggley’s “Plants of Yellowstone National Park" (U. 8. Govt. Printing Office, Washington, 1936) does not include the following plants. CAREX HETERONEURA W. Boott. (Data under Wyoming.) Carex ILLOTA Bailey. Boggy edge of Isa Lake, Craig Pass, alt. 8300 ft., Sept. 12 (12509 NA, RM); meadow, Continental Divide, alt. 8300 ft., 4 miles NW. of West Thumb, Sept. 12 (12522, US). CAREX KELLOGG W. Boott. Rocky bank of Isa Lake, Craig Pass, alt. 8300 ft., Sept. 12 (12511, NA, RM). CAREX LEPORINELLA Mack. (Data under Wyoming.) CAREX LIMNOPHILA F. J. Hermann. (Data under Wyoming.) CAREX MICROPTERA Mack. Rocky bank of Isa Lake, Craig Pass, alt. 8300 ft., Sept. 12 (12510, NA, RM, US). CAREX NEUROPHORA Mack. Wet mossy border of streamlet in narrow ravine, alt. 8700 ft., west slope of Mt. Washburne, ca. 14 mile N. of Dunraven Pass, Sept. 12 (12534, NA, RM, US). CAREX STENOPTILA F. J. Hermann. (Data under Wyoming.) CaREX TOLMIEI Boott. Meadow, Continental Divide, alt. 8300 ft., 4 miles NW. of West Thumb, Sept. 12 (12521, NA, RM). Juncus HALLI Engelm. Plentiful in meadow, Continental Divide, alt. 8300 ft., 4 miles NW. of West Thumb, Sept. 12 (12520, NA, RM, US). 1956] Shinners,—Fraser’s Catalogue 281 Juncus Tracy! Rydb. (Data under Wyoming.) LuzuLA PIPERI (Coville) Jones. Abrupt mossy shore of Bridge Bay, Yellowstone Lake, alt. ca. 7800 ft., 3 miles SW. of Lake Junction, Sept. 12 (12528, NA). ASTER LEUCANTHEMIFOLIUS Greene. Open rhyolite slope on Norris Basin-Mammoth Hot Springs Road, alt. ca. 7800 ft., Sept. 11 (12502, NA). PLANT NEW TO UTAH THERMOPSIS RHOMBIFOLIA Nutt. Open lodgepole pine forest W. of Sheep Creek Canal, alt. ca. 8500 ft., Uinta Mts., Aug. 14 (12129, NA, US). Additions to the Flora of the Uinta Mountains Two species not reported in Graham’s “Botanical Studies in the Uinta Basin of Utah and Colorado” (Annals Carnegie Mus. (Pittsburgh) vol. 26. 1937) are the following. AGROPYRON LATIGLUME (Scribn. & Sm.) Rydb. Meadow along Middle Fork of Sheep Creek, atl. 9000 ft., on road to Spirit Lake, Aug. 14 (12136, US). THERMOPSIS RHOMBIFOLIA Nutt. (Data under Utah.) NEVADA RANGE EXTENSION The occurrence of Trifolium monanthum A. Gray in eastern Nevada, while not a new record for the State, is a rather notable extension in range since the species has previously been known only in western Nevada. This was collected on the moist bank of a stream in aspen woods, Lamoille Canyon, alt. ca. 7800 ft., Ruby Mts., 7 miles SE. of Lamoille, Elko Co., Aug. 10 (12087, NA)—PLANT INDUSTRY STATION, BELTSVILLE, MD. NUTTALL NOT THE AUTHOR OF FRASER’S CATALOGUE Luoyp H. SHINNERS NUTTALL in 1818 unequivocally acknowledged only 13 out of 71 new names in Fraser’s 1813 Catalogue as his. Ten of the 13 had appeared as nomina nuda. Two were placed in synonymy, and two were misquoted (including one of those placed in synonymy). In addition he published as new 10 species listed in the Catalogue without mention of the fact that they had appeared there. He obliquely claimed responsibility for another which he does not cite by name, and for which he adopts Pursh’s binomial, though pointing out that Pursh had described a mixture representing two genera (Astragalus crassicarpus, called by Nuttall A. carnosus Pursh, the discordant element being named Sophora sericea 282 Rhodora [Vor. 58 Nuttall). In 1840 (Trans. Amer. Philos. Soc. 7: 301, in com- ments under Dieteria sessilifolia) he directly claimed responsi- bility for one or presumably two more (‘‘A plopappus spinulosus, to which I applied the name of Sideranthus in Fraser's Cata- logue": two species are there listed under this generic name). Two names which had appeared in the Catalogue were credited by Nuttall in 1818 to the Botanical Magazine (Bartonia decape- lala, there credited to Pursh) and its editor, Sims (Allium stellatum, there credited to Ker-Gawler), and a third (Oenothera macrocarpa) was credited to Pursh. "The implication is that the Catalogue (with 89 names altogether) was prepared by someone who used a small number of Nuttall's names and in addition a much larger number of names coined by other persons. This seemed so obvious that the point was not elaborated in my previous article. It was surprising to read Dr. Graustein's statement that “there is no doubt" that Nuttall authored the Catalogue, and that I had furnished ‘‘no convincing evidence" to the contrary. The assertion of Cronquist, Keck and Maguire that “it is universally acknowledged that many or all of the new names contained in it were those of Thomas Nuttall” is untrue, and is an attempt to present hearsay as if it were scientific evidence. Under English and American common law, a man is judged innocent until proved guilty. The ready assumption that Nuttall authored Fraser's Catalogue is therefore legally unsound. I must protest being put on the defensive when in reality it is the opposition who must prove their case. Instead of uncritically adopting a plausible supposition, we should assemble evidence about it. By a detailed comparison of the Catalogue and the works of Nuttall and Pursh, I obtained what seemed to me adequate evidence for rejecting the thesis that Nuttall wrote the Catalogue. It did not seem adequate to others, and for that reason I am going to the unwelcome trouble of presenting further arguments. Let me insist again that this is really uncalled for. As a matter of basic law, my side of the case requires no argument until seriously challenged. The burden of first proof rests with the opposition. Let me begin by playing Devil's advocate and summarizing their evidence, since they have not made any effort 1956] Shinners,—Fraser’s Catalogue 283 to do so. First is the fact that Nuttall wrote “by T. Nuttall” on the Philadelphia copy of the Catalogue. This I have already stated does not have to mean that he claimed to be its author, but can mean instead merely that he claimed to have brought back the plants listed in it, which he unquestionably had. The philosophical Doctrine of Parsimony asks that in propounding a theory, we adopt one that requires the least possible amount of assumption. Stated more pithily, that theory is best which theorizes least. It better accords with this doctrine to uphold the second interpretation, which asks us to assume nothing that is not already known to be true, than it does to uphold the first one, which is wholly supposititious. Further, Nuttall’s own later actions are strangely inconsistent with the first interpretation, but not at all with the other. I hold therefore that the first piece of evidence to support the claim of Nuttallian authorship has not proved it. The second piece of evidence is the fact that Nuttall in later publications (chiefly in his Genera, 1818; one item in 1840, mentioned above) claimed responsibility for certain names in the Catalogue. I have already pointed out that he claimed remarkably few of them, and that he specifically credited three of them to other authors. If he was the direct author of the catalogue, why did he ignore most of the names in it, and why did he credit some to persons other than himself? Again I choose the explanation that demands least in the way of gra- tuitous assumption: that the Catalogue was prepared by someone else who adopted a rather small number of names coined by Nuttall, along with many more which were not. Again, the thesis of Nuttallian authorship is not proved. And let me re- peat that what we must first require is that it be proved, not that it be disproved. Let me extend my role of Devil's advocate by indulging in further suppositions like that of Nuttallian authorship for the Catalogue, but in another direction. I suggest that the 10 names which Nuttall published in 1818 as new, without men- tioning that they had appeared in the Catalogue, had not originated with him at all, but were the inventions of someone else; that Nuttall was expropriating them in exactly the same free manner that Sims, Ker-Gawler and Pursh had done before him. Such was the practise of the time! There has been too 284 Rhodora [Vor. 58 much repetition of the theme that Pursh was an exceptional blackguard. Let us see how his contemporaries behaved. John Sims described Oenothera missouriensis with Plate 1592 of Curtis’s Botanical Magazine, dated Nov. 1, 1813. Sims states that it was “found by Mr. Nuttall in the neighbourhood of the Missouri,” and adds, “We do not find that this species has been before noticed: it seems to differ from every one de- scribed by Micnavx or by Pursn, whose valuable Flora, speedily to be published, we have been favoured with the opportunity of consulting. . . . Communicated from the Sloane-Square Nurs- ery by Messrs. J. and J. T. Fraser.” It was not relayed to Pursh for him to publish in his Flora! In Fraser’s Catalogue it appears as O. macrocarpa (which name must be adopted if the Catalogue names are considered valid). It was described as new by Pursh under the same name, and by Nuttall in 1818 as his new species O. alata, with “O. macrocarpa PH." as synonym, but no mention of O. missouriensis. In view of Nuttall’s several references to the “Bot. Mag.," his disregard of Sims’s binomial is very odd, and possibly not wholly “innocent,” to borrow from Dr. Graustein’s quotation. And how are we to regard his bald rejection of both the original Catalogue name and that of Pursh? John Bellenden Ker, or Ker-Gawler as he styled himself (identi- fied only by the cryptic initial **G."), described Scilla esculenta with Plate 1574 of the Botanical Magazine, dated Aug. 1, 1813, citing as synonyms Phalangium esculentum **Fraser's Catalogue, &e.,” and Phalangium Quamash Pursh, “nondum evulgata." He says, “We have added the synonym from the work of Mr. Pursn, in consequence of a communication that gentleman was so oblig- ing as to make to us, in which he assured us, that Mr. Fraser’s plant, from which our drawing has been made, was of the same species as that he had in view. . . . We have not had the op- portunity of seeing the figure in Mr. PunsH's work, which is not yet published. . . . Our drawing was made from a plant imported by Mr. NurrALL, which flowered at Mr. Fnaskn's Nursery, in Sloane-Square.”’ Instead of belaboring Pursh so persistently for his treatment of Nuttall, should we not extend our sympathies for the treat- ment he received from Sims and Ker-Gawler? Perhaps not. They were birds of a feather, and on at least one notorious 1956] Shinners,—Fraser’s Catalogue 285 occasion joined forces in a deed now regarded as nefarious, but taken more casually then. In publishing Bartonia decapetala (Bot. Mag. t. 1487, Aug. 1, 1812), Sims declares, “Living plants have been brought to this country, by Mr. THomas NUTTALL, who collected them. . . . It has not however as yet flowered here, on which account our drawing was taken from dried speci- mens. It is by particular request only that we have been induced to publish such, contrary to our rule, seldome deviated from, and never without mentioning it, of admitting none but drawings from the life. . . . For the above generic and specific characters, and indeed for the whole communication, we are indebted to Mr. FREDERICK PunsH, author of a new Flora of North-American plants, now in the press.” Considering the very free and easy practices of the time, it does not seem to me at all unreasonable to think that Nuttall in 1818 sometimes indulged in the same kind of thing. I think it plausible to explain the 10 names lifted from Fraser’s Catalogue without acknowledgment as evi- dence that he did. True, I have no proof of it. But it is just as true that there is no real proof of the contrary. A point in the above examples calling for particular emphasis is the fact that none of the authors accepted Fraser’s Catalogue names as having legal status. Pursh and Nuttall both quote them, sometimes adopting and publishing them as new. Sims and Ker-Gawler regularly quote as inconsequential synonyms names from ‘Fraser’s Catalogue, &c." when describing new species. What is meant by the unexplained abbreviation *&e."? Could there have been other, less renowned nursery lists in which names appeared? Could one or more such have come out ahead of Fraser's? For the benefit of the opposition, I declare that both suppositions are true, and that we must search for still older and unknown authors of the names, and perhaps for addi- tional descriptions for many which were nomina nuda in the one catalogue which has come down to us. Fantasy, yes, but surely plausible. In accordance with the practise of the opposition, my thesis is therefore valid, and must be accepted until someone can absolutely disprove it. Dr. Graustein's assertion that Nuttall did not consider the Catalogue names validly published should dispose of the matter for those who believe that he wrote it. But Cronquist, Keck 286 Rhodora [Vor. 58 and Maguire have put forth an astonishingly incoherent argu- ment for validity of **Nuttall's" names containing the statement that authorship is irrelevant. "Their assertion that the names “were accepted in that published work by whoever wrote it” is dogmatic and quite meaningless. As pointed out above, none of the four known botanical authors involved (at least three of whom reportedly coined names used in the Catalogue) accepted those names as legally published. We must look at the case with our present rules of nomenclature in mind. Despite the seemingly scandalous customs of the time, our four early botanists behaved in this ease remarkably in keeping with our current views on valid publication and formal acceptance by an author of new names. I see no justification for acting contrary to both early custom and modern rules. It is really supererogation to bring in anonymity as equivalent to non-acceptance in order finally to reject the Catalogue names, but I repeat that this is logical and legitimate. "The three contending authors have pre- sented no reason to refuse Rousseau’s suggestion. They state flatly in one sentence that “anonymity of the author is no bar, under the Rules, to validity of publication of a name," but add lamely in the same paragraph that they “do not wish to comment on the status" of Rousseau's example, and follow this with the dogmatic statement about the Catalogue (wholly unproved, and refuted by contemporary evidence) which has just been quoted. Still we have not done with this much belabored subject. If so much in it is controversial, we may well turn to established usage, to glean any help we can. Let us review in detail the names which appeared in Fraser's 1813 Catalogue with enough description to require consideration. Of the twenty, three were validly described by Nuttall in 1818 with the same binomials, were not named by anyone else in the interim, and therefore need not concern us further: Cactus (now Mammillaria) viviparus, Lilium andinum, Rudbeckia purpurea var. serotina. A fourth had an older name: Vitis campestris was V. riparia Michx. (V. vulpina of authors). The two species of S?deranthus may be disregarded, since they were given a combined description and were not differentiated. (The generic name may also be re- jected as an inextricable mixture, though as Britten pointed out, it should have been adopted under the American Code for 1956] Shinners,—Fraser’s Catalogue 287 Chrysopsis because the first species to be listed when the two were differentiated was in the later-described genus: a case for so doing could even be made out now, if my transfer of Chrysopsis to the older Heterotheca is no more acceptable than my views on Fraser's Catalogue.) For three species, the only question is one of authorship: Hriogonum flavum (published by Pursh, not claimed by Nuttall), Malva (now Sphaeralcea) coccinea (pub- lished by Pursh, later claimed by Nuttall), and Phalangiwm (now Camassia) esculentum (described by Ker-Gawler as Scilla escu- lenta, next by Pursh as Phalangiwm Quamash, then claimed by Nuttall with the original Catalogue name). For the last-named species, Gould unaccountably uses the binomial Camassia Quamash (Pursh) Greene in his revision of the genus (Amer. Midl. Nat. 28: 728, 1942). "This must give way to C. esculenta, with (Nutt.) Lindley as authorities if dated from the Catalogue (very precisely with “J. & J. T. Fraser" in parentheses, as publishing authors), or (Ker-Gawler) Lindley if dated from the Botanical Magazine. If we reject the Catalogue names, we must abandon Amorpha nana Nutt., Astragalus crassicarpus Anon. (Nutt.?—he does not actually admit responsibility for this binomial, though obliquely claiming the species was his), Dalea enneandra Anon., Agastache anethiodora (Anon.) Britton, Penstemon grandiflorus Anon., Ratibida columnifera (Anon.) Wooton & Standley, and Yucca glauca Anon. (Names which were never acknowledged by Nuttall as his are marked *Anon."; they would have to be credited to J. & J. T. Fraser if to anyone.) These would have to be called respectively Amorpha microphylla Pursh, Astragalus carnosus Pursh emend. Nutt., Dalea laxiflora Pursh, Agastache Foeniculum (Pursh) Kuntze, Penstemon Bradburii Pursh, Ratibida columnaris (Pursh) Rafinesque, and Yucca angustifolia Pursh. It should be pointed out again that Nuttall himself accepted the second, third, fifth, and sixth of the latter names, in contemporary combinations; that he credited Penstemon grandiflorus to “Fras. Cat.” and not to himself; that he never mentioned Hyssopus (later Agastache) anethiodorus; and that he claimed only Amorpha nana as his. If we accept the Catalogue names as validly published, we must abandon Oenothera missouriensis Sims, O. caespitosa Sims, 288 Rhodora [Vor. 58 Grindelia squarrosa (Pursh) Dunal, and Gaillardia aristata Pursh. These must be called respectively Oenothera macrocarpa Nutt. (or J. & J. T. Fraser), O. cespitosa Nutt. (or J. & J. T. Fraser), Grindelia (new combination based on Thuraria herbacea Anon. in J. & J. T. Fraser), and Gaillardia (new combination based on Virgilia grandiflora Anon. in J. & J. T. Fraser). Under present extremist rules, since publishing authors are held more important, we need cite only the names of the Frasers if we choose, crediting none to either Nuttall or Pursh. In mentioning only the case of Penstemon grandiflorus vs. P. Bradburii (“bradburyi”), the three New York authors were telling us a great deal less than the whole truth. If their views are accepted, we shall be required to replace three very familiar names of widespread and economically important species (Oeno- thera missouriensis, cultivated; Grindelia squarrosa, weed; Gail- lardia aristata, cultivated) with much more offensive substitutes than Penstemon Bradburii, two of them new combinations! How much better to take the course which requires no new names, and the displacement of none that are nearly so common and widely known as the three just cited. (I discount Astragalus carnosus and Dalea laxiflora because each of these species appears in current works not under one name but under three. No possible course of action can avoid the rejection of names for them which have become well known for some parts of the country.) The three authors claim to “believe that the interests of nomenclatural stability would be better served by the admis- sion of Fraser’s Catalogue as a proper publication than by its rejection." I ought to have been utterly crushed by the weight of so much eminent authority. Instead I survive to observe mildly that authority is, after all, an imprecise term. My earlier conclusions still stand. There is no proof that Nuttall himself wrote Fraser’s Catalogue. This was compiled by a person or persons unknown, using some unpublished names coined by Nuttall (possibly even with his assistance in so doing), but using a much larger number of names originated by others. The Catalogue names were not accepted as validly published in it by Nuttall, nor by Pursh, nor by Sims, nor by Ker-Gawler. There is no necessity to accept them today. By rejecting them, we keep name changes to a minimum. 1956] Shinners,—Fraser’s Catalogue 289 A last additional note on Astragalus carnosus. It is known that Pursh himself was describing primarily a specimen of Sophora sericea. He added the fruit characters from an As- tragalus collected by Nuttall, and chose an epithet describing the fruit in particular. The binomial therefore rests on dis- cordant syntypes, regardless of the fact that Pursh may not have had in his hands material of anything but the Sophora. Under the rules, the first author to select a lectotype for a species based on a mixture fixes the application of the binomial. This Nuttall did in 1818, by restricting the binomial to the Astragalus ele- ment. If we reject the Catalogue name (as Nuttall himself did, and as I argue we therefore must do), and do not split up the genus Astragalus, the plant must be Astragalus carnosus and cannot legally be called anything else. Pursh’s primary type material ceased to have any bearing on the application of the binomial when Nuttall explicitly defined it as belonging to an Astragalus. PERTINENT LITERATURE BRITTEN, JAMES. 1899. Bibliographical notes. XXI.—Frasers' Cata- logues. Journ. Bot. 37: 481-487. (The author insists on ''Frasers' Catalogues" instead of ‘‘Fraser’s Catalogues," but the two younger Frasers and their contemporaries invariably used the latter form, which is retained in my articles.) CRONQUIST, ARTHUR, Davi» D. Keck AND Bassetr MAGUIRE. 1956. Validity of Nuttall’s names in Fraser's Catalogue. RHODORA 58: 23-24. GRAUSTEIN, JEANNETTE E. 1956. Nuttall’s quarrel with Pursh. RHODORA 58: 20-22. GREENE, Epwarp L. 1890. Reprint of Fraser's Catalogue. Pittonia 2: 114-119. Hooker, W. J. 1837. Biographical sketch of John Fraser, the botanical collector. Comp. Bot. Mag. 2: 300-305. (Includes a list of ‘North American plants, discovered and introduced by John Fraser, between 1785 and 1799; and by J. Fraser, Jun., from 1799 to 1817." For 74 species, reference is made to published plates, mostly in the Botanical Magazine, some in the Botanical Register, Hooker’s Exotic Flora, and Loddiges’ Botanical Cabinet. None is cited after Oenothera missouri- ensis, Sims's plate of this being listed a line above after O. macrocarpa, perhaps from a clerical or typographical error.) NUTTALL, THomas. 1818. The Genera of North American Plants. 2 vols. Punsu, FREDERICK. 1814 (mid-December 1813, fide Graustein). Flora Americae Septentrionalis. 2 vols. SHINNERS, Luoyp H. 1956. Non-validity of Nuttallian names in Fraser's Catalogue. Ruopora 57: 290-293. 290 Rhodora [Vor. 58 SOME ADDITIONS TO THE ORCHID FLORA OF PUERTO RICO Hanorp F. WINTERS AND CHARLES SCHWEINFURTH! THE Island of Puerto Rico, the smallest and easternmost of the Greater Antilles, is one of the more thickly populated areas of the world. It has a minimum of undisturbed vegetation and is not particularly rich in orchid species. The orchids of Puerto Rico have been studied by various well-known taxonomists. Among them are Alfred Cogniaux whose studies were published in conjunction with Ignatius Urban (1903, 1909-10) at the beginning of the century. The col- lections and writings of Britton and Wilson (1924) comprise the most recent comprehensive taxonomic investigation. Fifty-four genera of orchids with 95 species are reported in their Island flora. Subsequent exploration by various collectors (Winters, 1949) has brought to light additional orchid species. Some of these are widespread in the West Indies, but have not been reported previ- ously from Puerto Rico. Others are rare in the area. The purpose of this report is to extend the recorded range of the several species to include Puerto Rico. Unless otherwise indicated, the identi- fications were made by Mr. Charles Schweinfurth. Basiphyllaea angustifolia Schlter. Winters 88, collected Oct. 8, 1955, in late flowering and early fruiting stage. Terrestrial in deep leaf mold over serpentine rock. Maricao Insular Forest, at 2,000 feet altitude. Rare. Formerly known only from Prov. of Matanzas, Cuba. Cranichis tenuis Reichb. f. Winters 10, collected Nov. 11, 1953. Terrestrial in organic accumulation over rocks. Maricao Insular Forest, on lee (dry) side of mountain toward Sabana Grande, at 2,000 feet altitude. Recorded only from Cuba. Epidendrum cochleatum L. var. triandrum Ames. Collected by Rev. Donald Dod along Fish Hatchery Trail, Maricao Insular Forest. Identi- fication by H. F. Winters. Recorded only from southern Florida. See plate 113, Correll, D. S., Native Orchids of North America, 399 pp., illus. Waltham, Mass. Epidendrum jamaicense Lindl. Winters 17, collected Sept. 9, 1947, Dec. 17, 1954, epiphytie, Maricao Insular Forest, at 3,000 feet altitude; Nov. 22, 1954, La Mina Recreation Area, Luquillo Mts., at 2100 feet altitude; July 1939, collected by A. G. Kevorkian, Maricao Insular Forest. Also collected by F. H. Sargent No. 355, at El Yunque, on a tree, at 1969 feet altitude, April 10, 1938; and by Claud L. Horn No. 5862, on ! Respectively, Horticulturist, Federal Experiment Station, Mayaguez, Puerto Rico, and Research Fellow in Botany, Harvard University. 1956] Winters and Schweinfurth,—Orchid Flora 291 old tree trunks, at 2500 feet altitude, in Maricao Insular Forest, June 1939. Apparently fairly widespread in the West Indies. Epidendrum oncidioides Lindl. var. gravidum (Lindl) A. H. & 8. Collected Jan. 9, 1940, by A. G. Kevorkian, Maricao Insular Forest. Recorded from Middle America, from Mexico to Panama and Colombia. Epidendrum vincentinum Lindl. Winters 38, collected Nov. 22, 1954, from slope of Mt. Britton, Luquillo Mts., epiphytic on trees and shrubs, 2900 feet altitude. This species is recorded from several of the West Indies, from Costa Rica, and Peru. Habenaria strictissima Reichb. f. var. odontopetala (Reichb. f.) L. O. Wms. Collected Nov. 27, 1939, by A. G. Kevorkian, Maricao Insular Forest. This species occurs from Mexico to Costa Rica and in Cuba. Lepanthopsis melanantha (Reichb. f.) Ames. Winters 48, collected Dec. 7, 1954, epiphytie along trail to San German from Camp Buena Vista, Maricao Insular Forest, at 3000 feet altitude. Rare. This species has been recorded from Cuba, Santo Domingo, and the west coast of Florida. See plate 99, D. S. Correll, 1. c. Liparis Saundersiana Reichb. f. Winters 86, collected Sept. 17, 1955, from Las Mesas near Mayaguez, 1100 feet elevation, terrestrial; Oct. 8, 1955, Maricao Insular Forest, at 2000 ft. altitude; Rev. Donald Dod, Oct. 1, 1955, Maricao Insular Forest. Previously recorded only from Jamaica. Maxillaria rufescens Lindl. Collected by Rev. Donald Dod, El Verde area Luquillo Mts.; Maricao Insular Forest, at 2900 feet altitude. A widespread species in Central and South America and in the West Indies. Pleurothallis Ghiesbreghtiana Rich. & Gal. (P. longissima Lindl.) Winters 51, collected Feb. 22, 1955, saxicolous on exposed mountain ridge, Maricao Insular Forest, at 2500 feet altitude. A species frequent in Middle America and the West Indies. Polystachya foliosa (Hook.) Reichb. f. Winters 11, collected Oct. 8, 1955, epiphytie at base of shrubs. Maricao Insular Forest, at 2000 ft. altitude. An extremely variable species extending through the American Tropics. Stelis pygmaea Cogn. Winters 83, collected in flower and fruit, July 31, 1955, epiphytic along mountain ridge, El Verde area, Luquillo Mts., at 2800 ft. altitude. A little plant confined to the West Indies. Vanilla barbellata Reichb. f. Winters 80, collected July 2, 1955, from several locations in Guanica Forest, terrestrial, with vines clambering over shrubs and trees, sea level to 250 ft. altitude. Identified by Donovan 8. Correll. Records of this species are from the Bahamas, Florida, and Cuba. See plate 57, D. 8. Correll, l. c. Vanilla Dilloniana Correll. Winters 79, collected Nov. 21, 1954, in fruit, hilltop, Cambalache Experimental Forest; June 28, 1955, common on mountain ridges, Maricao Insular Forest, at 2300 ft. altitude. "Terrestrial, with vines clambering over shrubs and trees. Identified by Donovan S8. Correll. This species is recorded from Florida, Cuba, and Santo Domingo. See plate 58, D. S. Correll, l. c. 292 Rhodora [Vor. 58 LITERATURE CITED Brirron, N. L., anb P. Wirsow. 1924. Scientific Survey of Porto Rico and the Virgin Islands. Botany of Porto Rico and the Virgin Islands. N. Y. Acad. Sei. 5: 180-217. COGNIAUX, ALFRED, in URBAN. 1909-10. Symbolae Antillanae. 6: 293-696. URBAN, IGNATIUS, AND A. Coaniaux, in URBAN. 1903. Symbolae Antillanae. 4: 162-184. Winters, H. F. 1949. Native Orchids of Puerto Rico. American Orchid Society Bulletin 18: 355-361. GROWTH HABITS OF ARABIS PERSTELLATA E. Lucy Braun Since describing Arabis perstellata in 1940 (RuoponA 42: 47- 48), I have had plants of this species almost continuously under observation, noting growth habit year by year. In the eighth edition of Gray’s Manual (Fernald, 1950), this and what was known as Arabis dentata T. & G. are included in one species, A. perstellata E. L. Br., made up of the “typical” variety and var. Shortii Fern. (A. dentata (Torr.) T. & G.). In the New Britton and Brown Illustrated Flora (Gleason, 1952), these are main- tained as distinct but related species, A. perstellata Braun and A. Short (Fern.) Gl. Neither manual adequately states the habit of growth of A. perstellata, the former stating “perennial, with strong branching base and definitely perennial leafy basal offsets”; the latter, with no statement, figures a single erect stem arising terminally from an ill-defined basal rosette. Flowering plants may have from one to thirty flowering stems, all axillary; the central axis of the plants is a short leafy stem, very short at first, later elongating 1-3 cm. a year, with leaves more closely placed toward apex (Figs. 1, 2). This central or vegetative axis persists throughout the life of the plant. The first spring, the plant (from seed germinating sometime during the previous summer or fall) is a loose rosette of petioled leaves (Fig. 3). The central axis elongates, and by the second spring is 1-2 em. long, the shriveled petioles of leaves of the previous season persisting near its base (Fig. 4). If flowers are produced the second spring, the flowering stem is definitely lateral, from the axil of a leaf of the previous season (Fig. 2). Year by year, henceforth, the central or vegetative axis elongates, and each 293 Braun,—Habits of Arabis perstellata 1956] ‘ILE exm-epeos pes1e[Use X19915 '6 'DI4 'oo'€egins Jeo[ JOMO] JO SIPE 91%[[91S `8 'D14 'oovjuns Jeol, Joddn Jo sarey o[durrs pu 91%[[91S `¿ `DIa "pese[ue ‘pees ssə[gurAa pue ənDIIS "9 'DI4 'poz1e[uo 'IOMO[4d 'G 'DI4 -'OZIS “yeu % “ed +-I 'DI4 '1eo4 snorxəid Jo (7) so[orjed porum ura ‘suLids puooos Jo LIN 94k] Ur 9339SO3] “fF 'ÐIA ‘yD ye Suopo[4309 pOu ‘BUAS ysiy Jo Av 91lV%[ ur 93390803] `€ 'ÐIA ‘IVIL snorAo1d 9u3 JO soAvo[ Jo so[orjod po[eArrus 'p 3e pue 'qoue1q ZULIGMOY euo pue STXe ƏA1191939A [e1JUGD uA ‘ULed p[o Ie94-OA^4 V `ç ‘DIX '39sgo 9411939204 V JO WS əu1 JO 31ed 'opou [CET JV ‘PUL ‘SILVIA snorA9:d Jo seqouedq ZULIOMOY Jo SNIS 'sopou 1op[o A[9AISSOOOnS V ‘MOTO :u340418 PEGI JO seqouguq SULIOMOP 991u1 pue SIX? 9A1101939A [ve13u99 ZUTMOYS ‘PJO S1e9& G ISVLA JV YURI '[ ‘NIA '070jj27$42d S1QD4V ‘G-I 'S5Idq 294 Rhodora [Vor. 58 year flowering stems arise from nodes of the year-old part of the stem. The old stem becomes hard and woody, and the bases of fruiting branches persist for several years, a group for each year of growth (Fig. 1). Occasionally, a vegetative axillary branch is formed (as at the 1951 node in Fig. 1), and henceforth elongates slowly as does the main axis. Very often, the old vegetative axis is decumbent and covered by an accumulation of humus, only the year-old leafy stem arising vertically above the ground surface. Arabis perstellata, then, is definitely perennial by a slowly elongating central axis; axillary offsets occasionally arise; all flowering stems are lateral, and arise near the apparent base of the plant, i. e., from the year-old part of the stem. Examination of herbarium specimens of the “chiefly biennial”’ A. Shortii suggests that the growth habit may be similar, and that at least some plants may live for more than two years. A. perstellata differs from A. Shortii in being densely stellate- pubescent throughout, so densely that stems and foliage are whitish to grayish-green. Forked hairs of the calyx and pedicels (Fig. 5) are long and ascending; those of the silique (Fig. 6) very small and inconspicuous (similar to those that may be present on A. Shortit). Long stiff ascending hairs (like those of the upper leaf surface of A. Shortii) are scattered among the more numerous stellate hairs of the upper leaf surface (Fig. 7). The lower leaf surface (Fig. 8) and basal parts of stems are whitened with closely placed and interlocking stellate hairs. Occasional much larger scale-like hairs, with papillate arms, occur on the lower leaf surface (Fig. 9). Pedicels of A. perstellata at anthesis are nearly twice as long as the flower (those of A. Shortit about one- half as long) and later elongate to 10 mm.; petals are pink (instead of yellowish-white), 3-4 mm. long and half again as long as the sepals (instead of 2-3 mm. long and barely exceeding the sepals). Cauline leaves of A. perstellata are more deeply and irregularly toothed than those of A. Shortii, auricled or clasping on the middle and upper parts of the branches, tapering toward the base or slender-petioled and lyrate-pinnatifid on the lower part of branches; at least some of the leaves of the vegetative axis lyrate-pinnatifid, all slender-petioled. The two taxa are evidently closely related, but apparently distinct species. A. perstellata is very abundant at and near the 1956] Turner,—A Study of the Genus Hymenopappus 295 type locality—wooded calcareous bluffs of Elkhorn Creek, Franklin County, Kentucky, in the Bluegrass region of that state.— UNIVERSITY OF CINCINNATI. A CTYOTAXONOMIC STUDY OF THE GENUS HYMENOPAPPUS (COMPOS.TAE) BILLIE L. TURNER (Contanued from page 269) 9a. Hymenopappus scabiosaeus L'Hér. var. scabiosaeus Hymenopappus scabiosaeus L'Hér. Hymenop. 1: 1788. Rothia caroliniensis Lam. Jour. Hist. Nat. 1: 17. 1792. Hymeno- pappus caroliniensis Porter, Mem. Torr. Bot. Club 5: 338. 1894. Photo- graph of type examined (GH): without date, sheet from Lamarck’s her- barium. Hymenopappus laxiflorus L'Hér. in DC. Prodr. 5: 658. 1836, as synonym. Plants biennial, 40-150 em. tall, the stems single from each tap-root, erect, much-branched, angled and grooved, glabrous to sparsely pubescent ; leaves alternate, mostly glabrous above and variously pubescent beneath, in the first year simple or mostly once-pinnate, up to 25 cm. long, 4-5 cm. wide, forming a basal rosette; later formed stem-leaves mostly glabrous, 15-50 in number, not much reduced, once-pinnate to bipinnately dissected, with broad to narrowly linear segments; heads numerous, discoid, 25-80-flowered, on densely strigose or glabrate ultimate peduncles mostly 1 to 5 em. long, these having at their bases conspicuous, mem- branous, petaloid bracts 5-14 mm. long, 3-10 mm. wide (rarely much reduced); involucral bracts white, petaloid, showy, equal or subequal, 7-15 mm. long, 4-8 mm. wide, membranous for half or more of their length; corollas white or creamy white, sweet scented, 3-5.5 mm. long, the tube stipitate-glandular, 2-3 mm. long, the throat funnelform, 1.2-3 mm. long, with lobes reflexed, about equaling the lobes; achenes obpyrami- dal, 4-sided, 3.5-5 mm. long, short-pubescent, principally on the corners, with hairs 0.1-0.4 mm. long, the faces with 2-3 nerves; pappus of 14-18 small obovate scales 0.1-0.6 mm. long; anthers completely exserted, 2-2.5 mm. long; chromosome number not known. DistRIBUTION.—Scattered, apparently rare, principally in the Mis- sissippi Valley region and southeastern United States: known by rela- tively few collections from Indiana, Illinois, Missouri, Oklahoma, Arkan- sas, and eastern Mississippi, Georgia, adjacent South Carolina, and Florida. Inthe southern part of its range it is commonly found in sandy pine woods; in the northern part it occupies rocky, sandy barrens and open disturbed areas (Fig. 45). April-June. 296 Rhodora [Vor. 58 Hymenopappus scabiosaeus var. scabiosaeus is difficult to characterize because of the inadequacy of herbarium material from the states outside of Florida. Taken throughout its range the variety demonstrates considerable variability. In eastern Oklahoma and western Arkansas the taxon has races with more pinnately dissected, glabrate leaves than are typical for the variety in South Carolina and Florida. Likewise, the races in Illinois and Indiana tend to be distinguished by extremely large, ovate peduncular bracts. To apply formal nomenclature to these races would be unjustified on present-day evidence, since these same characters may reoccur (with much less fre- quency) in material from the Southeastern States. In addition, this regional variability is complicated by the seemingly com- plete intergradation of var. scabiosaeus with var. corymbosus at their area of contact in eastern Oklahoma and Kansas. Typical Hymenopappus scabiosaeus var. scabiosaeus can be distinguished from var. corymbosus by the key characters listed. In this region, however, clearly intermediate specimens are found. This rather complete intergradation between the two taxa, combined with their close, overall resemblance has been the principal reason for the reduction of H. corymbosus to varietal rank. Early in the course of this study the author was inclined to treat Hymenopappus scabiosaeus in a broad sense, including within this species all the funnelform-throated, biennial taxa of the genus. Such a treatment, although perhaps justifiable on purely morphological grounds, would ignore the physiological, genetical, and ecological isolation of some included members. When such isolational barriers can be demonstrated or inferred, even when the morphological characters are somewhat inter- mediate between the specific and infraspecifie level (as in H. artemisiaefolius), these other factors should be considered in drawing taxonomic lines. For this reason H. scabiosaeus and H. artemisiaefolius have been treated as two specific taxa in- stead of included together under one. REPRESENTATIVE SPECIMENS.—Arkansas. BENTON co.: Along R.R., Sulphur Springs, E. J. Palmer 2938 (Mo). CARROLL co.: Beaver, E. J. Palmer 5586 (Mo, rom, vs). Florida. ALACHUA co.: 10 mi. W. of Gaines- ville, Blanton 6398 (pom). JACKSON co.: near Marianna, Curtiss 6806 (GH, MO, NY, UC). LEON co.: near Tallahassee, Apr. 1843, Rugel (mo, 1956] Turner,—A Study of the Genus Hymenopappus 297 NY). MARION CO.: E. of Flemington, Moldenke 1084 (MO, NY, US). MARTIN co.: dry place, May 2, 1941, P. O. Schallert (Ny, UC). SUWANEE CO.: (w/o locality), June-July, 1898, A. S. Hitchcock (Mo). WAKULLA Co.: Sopchoppy, F. H. Sargent 6095 (smu). Georgia. JENKINS co.: dry sandy soil near Millen, R. M. Harper 762 (GH, Mo, Ny, Us). Illinois. cass co.: Beardstown, July, 1842, C. A. Geyer (GH, MO). KANKAKEE CO.: along R. R., St. Anne, G. N. Jones 17280 (MO). MASON CO.: sandy barrens and prairie, June 18, 1845, S. B. Mead (au, ny). Indiana. STARKE CO.: ‘frequent in three dry sandy fallow fields," 3 mi. N. and 1.5 mi. E. of North Judson, Deam 49006 (an). Mississippi. LOWDENS co.: Columbus, S. M. Tracey 1400 (am, Mo, vs). Missouri. BARRY CO.: Eagle Rock, B. F. Bush 104 (GH, MO, US). MISSISSIPPI CO.: cemetery 2 mi. W. of Charleston, Steyermark 10259 (Mo). scorT Co.: sandy open places, May 20, 1895 (1894), H. Eggert (GH, MO). STONE co.: Baxter, Bush 15587 (Mo). Oklahoma. CHEROKEE Co.: 20 mi. N. E. of Tahlequah, N. T. Knodos 71 (smu). LE FLORE co.: Pine Valley, Goodman 2537 (GH, MO, NY, POM, RM). MAYES CO.: 2 mi. N. of Locust Grove, Stratton 3718 (CAS). MCCURTAIN CO.: Bethel, dry pine hills, Demaree 12691 (mo, ny). South Carolina. aiken co.: Aiken, May 1869, W. M. Canby (DS, GH, MO, NY, US). 9b. Hymenopappus scabiosaeus var. corymbosus (T. & G.), comb. nov. Hymenopappus corymbosus T. &. G. Fl. N. Amer. 2: 372. 1842. Rothia corymbosa O. Ktze. Rev. Gen. 1: 361. 1891. Type examined (ny): *Arkans., Leavenworth." Probably collected in what is now present- day Oklahoma. (For an outline of Leavenworth's activities in this region, see McVaugh, 1947.) Hymenopappus engelmannianus Kunth, Ind. Sem. Hort. Berol. 15. 1848. Ann. Sci. Nat., Series 3, 11: 229, 1849. The plant from which this description was taken was grown from seed that had been collected in Texas and sent to Kunth by Engelmann. In the United States Na- tional Herbarium there is a sheet of a Lindheimer collection (Fasc. III. Texas. Comal Co.: New Braunfels along Guadalupe R., April. May, 1846, Lindheimer 438) which has the name H. Engelmannianus written in by hand on the original label. It is likely that this specimen represents a collection of the plant from which the seed was taken, Lindheimer being an “intimate friend" and botanical correspondent of Engelmann (Geiser, 1948, p. 137). The original description is undoubtedly that of H. scabiosaeus var. corymbosus. Hymenopappus sulphureus Rydb. in Britton, Man. 1007. 1901. Type examined (ny): Kansas. Riley Co.: ‘Stoney hills," May 23, 1895 J. B. Norton 285. Plants biennial, 40-100 cm. high, the stems single from each tap-root (rarely two crowns forming, especially on injured plants), erect, usually much branched, angled and grooved, tomentose to nearly glabrate; leaves alternate, forming the first year a basal rosette, the primary leaves entire to variously lobed, the later formed rosette leaves becoming pin- 298 Rhodora [Vor. 58 emm at. PX as ar EX HII I HH u J ° A= 7 46 sas ] Í Fig. 45—47. Distribution of Hymenopappus species. Fig. 45. Hymenopappus scabiosaeus var. scabiosaeus (triangles); H. s. corymbosus (disks); intermediate speci- mens (X). Fig. 46. H. flavescens var. flavescens (triangles); H. f. cano-tomentosus (disks). Fig. 47. H. artemisiaefolius var. artemisiaefolius (disks); H. a. riograndensis (triangles). nately dissected, 5-15 cm. long, 5-7 cm. wide, mostly glabrate or sparsely canescent above, more tomentose below, the ultimate divisions linear, acute, 2-8 mm. wide; stem leaves (on well developed plants) 15-30, becoming reduced upward; heads campanulate, numerous (40-100), discoid, 20-60-flowered, on densely strigose ultimate peduncles 0.5-3 1956] Turner,—A Study of the Genus Hymenopappus 299 em. long, these without conspicuous petaloid bracts at their bases, in- voluere campanulate, principal bracts 5-9 mm. long, 2-4 mm. wide, glabrate above, pubescent below, yellowish-white or white membranous for 2-4 mm. from the acute to narrowly obtuse tip; corollas white or creamy white, 3-4 mm. long, sweet scented, the tube sparsely glandular, 2-3 mm. long, the throat funnelform 1-1.5 mm. long, with lobes reflexed, the lobes equal to or nearly as long as the throat; achenes obpyramidal, 4-sided, 3-4 mm. long, short-pubescent (especially on the corners) with hairs 0.1-0.5 mm. long, the achene faces 2—3-nerved; pappus of 14-18 obtuse or spatulate scales, 0.2-0.8(-1) mm. long; anthers completely exserted, 1.6-2.5 mm. long; n = 17. DisrRIBUTION.— Open prairies in heavy, calcareous soils and on lime- stone outcrops from southern Nebraska, central Kansas, Oklahoma to southern Texas and adjacent Mexico (Fig. 45). Late March-July. The variety is well marked throughout the greater portion of its range, but where it comes in contact with the closely related var. scabiosaeus in eastern Oklahoma and Kansas there is complete intergradation between the two taxa (intermediate specimens are designated by x on the distribution map). Typical var. corymbosus may be distinguished from var. scabiosaeus by its short, narrow involucral bracts, and by its ebracteate pedun- cles. Typical var. scabiosaeus has long, broad involucral bracts, and conspicuous white, membranous peduncular bracts. Like- wise, in western Texas, where var. corymbosus comes in contact with the yellow-flowered H. flavescens var. flavescens there appears to be considerable intergradation of their characters, especially as regards leaf- and corolla-shape, perhaps indicating considerable hybridization and introgression in the past between these species. (For implications of such hybridization see discussion under H. flavescens var. flavescens.) Evidence of present-day hybridization between these two taxa has been found only in a few specimens from Tom Green Co., Texas (Havard 22), where their ranges overlap. H. scab?osaeus var. corymbosus and H. flavescens var. flavescens are kept genetically separate, for the most part, at their regions of overlap, by dis- tinet seasonal differences in principal flowering times, the former reaching its peak 3-4 weeks before the latter (Fig. 44). In Texas, the distribution of H. scabiosaeus var. corymbosus is especially interesting. It is found almost exclusively on heavy clay prairies, its range stopping abruptly at the margins of the 360 Rhodora [Vor. 58 sandy soils of pine and post oak woods.” H. artemisiaefolius var artemisiaefolius occupies these latter habitats. In spite of their extensive region of apparent contact, there is little evidence of intergradation, hybridization or introgression between the two taxa, there presumably being almost complete ecological isolation (possible exceptions discussed below). Lindheimer, an early botanical explorer in Texas, was probably the first to note the ecological distinction between these species, noting in 1846 on a field label of one of his collections the following (Lindheimer 438: an): Hymenopappus corymbosus In patches on Prairies and margin of wood in fertile rather heavy soil (H. artemisiaefolius in sandy soil) 4.5 F. L. n. Br. 4.6 In the Rio Grande Valley of Texas H. scabiosaeus var. corym- bosus becomes quite variable as regards several characters, possibly the result of contamination from H. artemisiaefolius var. artemisiaefolius or its var. riograndensis. This area of Texas may prove to hold the answer to some of the evolutionary problems of the H. artemistaefolius—H. scabiosaeus complex. The lack of herbarium material from this region has made the task of evaluating variability of the represented taxa impossible, but it seems clear that the area occupies a critical position as concerns the origin, migration, and subsequent evolution of the funnelform-throated Hymenopappi. A peculiar group of specimens from the Gulf Coastal Prairies in Matagorda and Ft. Bend counties, Texas (Fisher s.n.; Palmer 9668; Fisher 40093), perhaps points to the existence of yet an- other closely related variety within the H. scabiosaeus complex. These specimens are like var. corymbosus except for their dwarf- ish, scapose habit (scarcely exceeding 30 cm. in height). How- ever, in light of our present knowledge, it seems best to consider these specimens as representatives of a rather distinet race of 12 One must be familiar with the peculiar long-distance interfingering of prairie with pine and oak woodland in south-central Texas to appreciate fully this ecological isolation. Dots, which seem on the map to be included in the range of H. artemisi- aefolius var. artemisiaefolius, are, in actuality, on well defined prairie strips within this region. 1956] Turner,— A Study of the Genus Hymenopappus 301 the variety, at least until further field and variability studies are made on the populations represented in this area. REPRESENTATIVE SPECIMENS.— Kansas. ANDERSON CO.: 1.5 mi. N. of Welda, R. L. McGregor 4277 (GuH)* BARBER co.: 6 mi. W. of Hardtner, Rydberg & Imler 626 (NY). BUTLER co.: August, Š. F. Poole 1302 (us). CLOUD CO.: Miltonvale, Benke 5181 (aH). COWLEY co.: (w/o lceelity), May, 1898, M. White (Mo). DICKINSON co.: Solomon, Benke 4313 (us). EDWARDS CO.: prairies, A. Finch 136 (Mo). GEARY co.: Ft. Riley, E. E. Gayle 469 (NY). HARPER co.: 2 mi. W. of Athony, Rydberg & Imler 611 (NY). LABETTE CO.: Oswego, bluffs 4 mi. W. of town, Rydberg & Imler 326 (MO, NY).* MARION co.: Florence, June 18, 1885, Pringle (Gu). RILEY CO.: Stony hills, J. B. Norton 285 (GH, MO, NY, RM, US). SALINE CO.: 6 mi. N. of Salina, Hancin 2162 (NY). WILSON co.: 6 mi. S. W. of Neo- desha, W. H. Horr E526 (RM, smu, us, ws). Nebraska. LANCASTER co.: Lincoln, Aug. 26, 1896, G. G. Hedgcock (wo). Oklahoma. ADAIR co.: Westville, May 19, 1921, Ensign (Ny).* ATOKA co.: 8 mi. S. of Atoka, M. Hopkins & A. & R. Nelson 1073 (MO, RM). BLAINE CO.: near Longdale, G. W. Stevens 831 (GH). cappo co.: 3 mi. E. of Hydro, Hubricht, Shoop, & Heinze B1387 (Mo). CARTER co.: S. of Turner Falls, Demaree 12298 (Mo, NY, UC, Us). CLEVELAND CO.: Norman, W. H. Emig 493 (MO, US). COMANCHE co.: Fort Sill, J. Clemens 11843 (cas, GH, MO, RM). ELLIS CO.: near Shattuck, Clifton 3452 (GH, MO, NY). GRADY CO.: on False Washita between Fort Cobb and Fort Arbuckle, 1868, E. Palmer 447 (NY, US). JOHNSTON co.: Tishomingo, Cory 58892 (SMU). LATIMER CO.: Limestone Gap, June, 1875, G. D. Butler (GĦ, M0).* LOGAN CO.: Guthrie, G. W. Stevens 3224 (GH). MAJOR co.: near Cleo, (z. W. Stevens 786 (ps, GH, MO). MURRAY co.: Platt Nat’l. Park, Demaree 12230 (MO, NY, UC). PoNTOTOC CO.: Š. of Ada, near city limits, G. T. Robbins 2483 (SMU, UC). PUSHMATAHA CO.: near Finley, E. J. Palmer 39397 (GH).* ROGERS CO.: Catoosa, B. F. Bush 1140 (Mo, NY).* TULSA co.: R. R. near Tulsa, May 10, 1940, H. A. Hawk (MO). woops co.: near Alva, May 22, 1913 (G. S. Stevens 547 (ps, MO, NY, Us). Texas. ARKANSAS CO.: W. of Tivoli on coastal prairie, Whitehouse 12075 (smv). BEE CO.: Beeville, M. E. Jones 29436 (POM). BELL co.: near Temple, S. E. Wolff 503 (us). BEXAR co.: 4 mi. N. W. of San Antonio, M. Clara 650 (CAS, POM, UC). BLANCO CO.: between Johnson City and Marble Falls, C. L. & A. A. Lundell 14544 (SMU, US). BREWSTER Co.: 38 mi. Š. of Mara- thon, Ferris & Duncan 2882 (cas, DS). BROOKS co.: roadside, Clover 823 (NY). BURNET CO.: 10 mi. S. E. of Marble Falls, Rd. to Bee Caves, L. H. Shinners 7239 (SMU). CALLAHAN co.: Clyde, E. J. Palmer 13683 (MO). COLLINS CO.: 2 mi. S. of McKinney, Timmons 436 (NY). COMAL co.: New Braunfels, Lindheimer 929 (GH, MO, NY, SMU, UC, US). COOKE co.: 7 mi. N. of Gainesville, Cory 56131 (smu). DALLAS co.: rocky prairies near Dallas, Reverchon 1508 (GH, MO, NY, US). DENTON CO.: 13 Specimens intermediate toward var. scabiosaeus are indicated by an asterisk (*); these are shown on the distribution map (Fig. 45) with an “x.” 302 Rhodora [Vor. 58 2,5 mi. S. of Sanger, Cory 53240 (smu). EDWARDS co.: Ranch Expt. Station, Cory 19029 (aH). ERATH co.: 8 mi. N. E. of Stephenville, F. W. Gould 5666 (smu). FANNIN co.: Bonham, May, 1896, J. M. Milligan (us). FoRT BEND co.: Richmond, G. L. Fisher 40093 (cas, US). GILLESPIE CO.: ] mi. E. of Willow City, Cory 53625 (SMU). GOLIAD co.: prairies N. of Goliad, Apr. 8-9, 1900, H. Eggert (mo). GRAYSON co.: 1.7 mi. N. of Collinsville, Shinners 12402 (smu). KARNES CO.: 6 mi. N. of Tulsita, J. F. Hennen 605 (SMU). KAUFMAN co.: 2 mi. E. of Terrell, Shinners 10101 (smu). KERR co.: Kerrville, A. A. Heller 1638 (GH, MO, NY, UC, US, Ws). LAMAR CO.: Arthur City, along Red R., R. R. embankment, Cory 56076 (sMu, US).* LASALLE Co.: sands, Cotulla, Apr. 29, 1905, J. Reverchon (SMU). LLANO co.: near Llano, S. E. Wolff 1586 (US). MCLENNAN CO.: gravel pit, mouth of White Rock Cr., L. D. Smith 392 (us). MATAGORDA CO.: Bay City, E. J. Palmer 9668 (ps, Mo, US). MEDINA CO.: 1 mi. W. of D’Hanis, Shinners 7301 (SMU). MILLS CO.: 6.5 mi. S. of Goldthwaite, Cory 13081 (GH). MITCHELL co.: red-bed slopes above Lake Hollywood, Pohl 4745 (smu). MONTAGUE co.: W. of Nocona, highway 82, Whitehouse 15050 (smu). NOLAN co.: Eagle Cr., Blackwell, Studhalter 1198 (us). PALO PINTO co.: vicinity of Mineral Wells, Gillespie 5220 (ps, us). PARKER co.: Weatherford, Tracy 8544 (GH, MO, NY, US). PECOS CO.: 7 mi. N. of Iraan, cut-off from highway 290, near Sheffield, Warnock & Turner 807 (sMv). REAL co.: 29 mi. E. of Rocksprings, Shinners 7338 (sMU). SAN PATRICIO CO.: near Mathis, McKelvey 1711 (GH, POM). SOMERVELL co.: 5 mi. W S. W. of Glen Rose, Shinners 10072 (SMU). SUTTON co.: Sonora Expt. Station, Eggleston 16700 (NY). TARRANT CO.: Polytechnic, Ruth 301 (Mo, RM, US). TAYLOR co.: 2 mi. W. of Buffalo Gap, Tolstead 7036 (Mo, UC). TERRELL CO.: 7 mi. E. of Longfellow, along highway 90, Warnock & Turner 592 (smv). THROCKMORTON CO.: 11 mi. N. of Throckmorton, Cory 37257 (GH). TOM GREEN CO.: Knickerbocker Ranche, Dove Cr., Tweedy 316 (vs). TRAVIS co.: Austin, Tharp (cas, MO, SMU, UC). UVALDE co.: W. of Uvalde, M. E. Jones 28164 (ps, POM, UC). VAL VERDE CoO.: N. of Del Rio, M. E. Jones 28165 (DS, POM, UC). VICTORIA CO.: Victoria, Tracy 9048 (GH, Mo, NY, US). WALLER CO.: dry prairies, Hempstead, E. Hall 357, in part (GH, POM, US). WASHINGTON CO.: (w/o locality), April 21, 1939, E. Brackett (aH). WEBB co.: E. of Laredo, A. C. Martin 102 (Us). WICHITA Co.: 1.6 mi. W. of Electra, Whitehouse 9775 (Us). WILBARGER CO.: 16.9 mi. W. of Electra, Waggoner pastures, Whitehouse 9845 (sMU). WISE CO.: 2 mi. W. S. W. of Chico, Shinners 1232 (smu). Mexico. Coahuila. Municipio de Muzquiz, hacienda La Rosita, June 26, 1936, Wynd & Mueller 294 (an, MO, NY, vs). 10a. Hymenopappus artemisiaefolius DC. var. artemisiaefolius Hymenopappus | artemisiaefolius DC. Prod. 5: 658. 1836. Rothia artemisiafolia O. Ktze. Rev. Gen. 1: 361. 1891. Photograph of type examined (us): “Texas, fl. sordide albi, 1832, M. Berlandier 1532." Type in the Delessert Herbarium. 1956] Turner,—A Study of the Genus Hymenopappus 303 Plants biennial, 40-90 em. high, tomentose to nearly glabrate; larger rosette leaves, 8-18 em. long, simple to once-pinnate with broad primary divisions, 6-30 mm. wide, mostly densely tomentose on the lower surface, becoming glabrate above; stem leaves (6—)8-16, not much reduced up- ward; heads 30-60 per stem, campanulate, on ultimate peduncles 1—4(-6) em. long; principal involucral bracts broadly elliptic to ovate, 6-12 mm. long, 3-7 mm. wide, snowy-white-membranous for about 1⁄4 their length or more (often tinged with red); corollas rosy-vinaceous to rarely com- pletely white, 3.5-5 mm. long, the tube moderately glandular, 2.5-3 mm. long, the throat funnelform 1-1.5 mm. long, with lobes reflexed, as long as, or 1.5 times longer than the lobes; achenes obpyramidal, 4-sided, 3.5-4 mm. long, principally pubescent on the corners with short white hairs 0.3-0.6 mm. long, the faces 2—3-nerved; pappus of 16-18 oblong scales 0.5-1(-1.5) mm. long; anthers completely exserted, 2-2.5 mm. long; chromosome number n = 17. DisrRIBUTION.—Confined to the sandy pine and post-oak woods of eastern Texas and adjacent Louisiana (Fig. 47). March-May. The closest relationship of this species is undoubtedly with the Hymenopappus scabiosaeus complex (probably var. scabiosaeus). It is not treated as a variety of that species because it is mor- phologically distinct throughout its range and apparently does not intergrade or hybridize with peripheral taxa (except possibly on a local scale with H. scabiosaeus var. corymbosus and to a limited degree with its own var. riograndensis in southern Texas). This treatment is justifiable on morphological, ecological, genetical, as well as practical grounds. Although the taxon is in apparent close contact with H. scabiosaeus var. corymbosus over at least 1000 miles (linearly along the peripheral area), along this entire line of contact the two taxa remain distinct, apparently as a result of edaphic isolation. H. scabiosaeus var. corymbosus occurs principally in heavy, clay soils while H. artemisiaefolius, as mentioned above, occurs principally in sandy soils. (For further information see discussion under H. scabiosaeus var. corymbosus.) H. artemisiaefolius var. artemisiaefolius intergrades to a large extent with H. a. riograndensis, of southern Texas. This is reflected in the tendency toward a larger pappus, less membran- ous involucral tips, and fewer-leaved stems as one approaches this region from the north. H. artemisiaefolius var. riograndensis could possibly represent the progenitor of this whole eastern complex (see discussion under that variety). Future field work should do much to 304 Rhodora [Vor. 58 clarify the limits of the three related taxa of this area: H. ar- temisiaefolius var. artemisiaefolius, H. artemisiaefolius var. riograndensis, and H. scabiosaeus var. corymbosus. REPRESENTATIVE — SPECIMENS.—Louisiana. NATCHITOCHES PARISH: Chopin, E. J. Palmer 7328 (cas, MO, NY, US). RAPIDES PARISH: Alex- andria, Aug., 1840, Hale (Ny). Texas. ANDERSON co.: Palestine, Apr. 14, 1929, Tharp (Ny). AUSTIN co.: Mills Cr., 16 mi. W. of San Felipe, Mar., 1844, Lindheimer 107 (46) (GH, MO, SMU, UC, US). BASTROP CO.: Bastrop, Apr. 4, 1939, Tharp (Ny, ws). BEXAR co.: San Antonio, E. H. Wilkinson 110 (Mo). Bowie co.: pine woods N. of Texarkana, June 12, 1898, Eggert (MO, NY). BRAZOS CO.: College Station, Apr. 28, 1927, H. B. Parks (RSA). BURLESON co.: 7 mi. Š. E. of Caldwell, Gould & Celerier 5447 (SMU, UC). CALDWELL CO.: 6 mi. S. E. of Luling, Cory 48875 (sMU). cass co.: Bivins, May 12, 1941, O. McGinnis (cas, aH, MO, SMU, UC). COLORADO co.: Columbus, June 11, 1910, H. H. Rusby (NY). COMAL co.: New Braunfels Dapprich 6740 (SMU). DALLAS CO.: sands, Dallas, J. Reverchon 527 (Us). DENTON CO.: 7 mi. N. E. of Roanoke, Whitehouse 15979 (smu). DEwrrr co.: Cuero, Bray 136 (us). FAYETTE co.: Colony, (8 mi. N. of Flatonia), E. W. Crawford 31 (Us). FREESTONE co.: 12 mi. S. of Fairfield, Shinners 7108 (smu, WS). GONZALES CO.: Ottine, Tharp 44472A (RSA). GUADALUPE CO.: swampy roadside between Luling and Seguin, M. C. Metz 3066 (NY). HARDIN co.: Fletcher, E. J. Palmer 9541 (DS, MO, US). HARRIS co.: Houston, Bush 35 (aH, Mo, NY). HARRISON CO.: Marshall, E. J. Palmer 5318 (Mo, POM, US). HENDERSON co.: 14.6 mi. S. of Athens, Hennen 274 (smu). HOUSTON co.: Grapeland, Tharp 926 (GH, US). JASPER co.: Š. of Jasper, C. L. & A. A. Lundell 11200 (SMU). JEFFERSON co.: Port Arthur (&) Beaumont, May 15, 1927, I. Kolthoff (us). LEON co.: Normange, Fisher 41269 (cas, vs). MCLENNAN CO.: Waco, L. Pace 96 (Mo). MARION co.: Jefferson, B. B. Harris 459 (Us). MATAGORDA co.: Bay City, E. J. Palmer 9629 (ps, Mo, US). MONTGOMERY CO.: 7 mi. S. of Conroe, Shinners 7742 (SMU). MORRIS co.: Daingerfield St. Park, D. S. & H. B. ( 'orrell 12440 (SMU). NEWTON co.: 4 mi. S. E. of Newton, Shinners 7656 (BM, SMU, UC). ORANGE CO.: Vidor, M. B. Wood (wo, vc). PANOLA co.: 18 mi. N. W. of Carthage, Shinners 7595 (GH, SMU). POLK CO : Livingston, E. J. Palmer 5249 (MO, US). REFUGIO CO.: (w/o locality), H. C. Benke 5439 (GH). ROBERT- SON CO.: 3.75 mi. S. E. of Hearne, Cory 55732 (smu, US, we). RUSK CO.: Henderson, Apr. 16, 1943, M. Riedel (mo), SABINE co.: 15 mi. N. of Jasper, Shinners 7637 (smu). SAN AUGUSTINE CO.: San Augustine, (w/o date), G. L. Crocket (Us). SHELBY co.: 16 mi. S. E. of Center, Shin- ners 7627 (SMU). SMITH CO.: Tyler St. Park, Cory 63392 (Gu, SMU). TRAVIS CO.: Austin, Tharp 1378 (uc, Us). TYLER CO.: 7 mi. S. of Wood- ville, Whitehouse 23303 (smu). VAN ZANDT CO.: 6.7 mi. E. of Grand Saline, Van Vleet 1105 (smu). VICTORIA co.: 10.5 mi.W. of Victoria, Cory 55117 (SMU, US). WALKER co.: (w/o locality), R. A. Dixon 512 (CAS, US). WALLER CO.: Hempstead, E. Hall 357 (aH, Mo, NY). wasH- 1956] Turner,—A Study of the Genus Hymenopappus 305 INGTON CO.: Brenham, Apr. 21, 1935, V. Lehman (NY). WILSON Co.: Sutherland Springs, M. E. Jones 29436 (Mo). woop co.: Mineola, Reverchon 2575 (Mo, SMU). 10b. Hymenopappus artemisiaefolius var. riograndensis var. nov. Herbae biennes, caulibus 2-8 foliatis, 45-100 em. altis; foliis inferiori- bus integris l-pinnatisve, 10—20 em. longis, 2-5 em. latis; pedunculis ebracteatis, 2-5 em. longis; inflorescentiis magnis laxe cymoso-paniculatis, 20-60-capitulatis; involucris campanulatis, bracteis 5-7 mm. longis, 2-4 mm. latis, apice ad 1 mm. membranaceis; flosculis 3-3.5 mm. longis, tubo 2 mm. longo, lobis acutis ad 1.5 mm. longis; achaeniis villosis 4-5 mm. longis, capillis albidis 1-1.2 mm. longis; pappo conspicuo, squamellis lineari-oblongis, 1.5-2 mm. longis. Plants biennial, 45-100 em. high; principal rosette leaves 10-20 cm. long, 2-5 em. wide, tomentose on both surfaces, simple to once-pinnate with broad, coarsely toothed divisions; stem leaves 2-8, much reduced upward; heads 20-60 per stem, campanulate, 40—-60-flowered, on slender, bractless ultimate peduncles 2-5 cm. long; inflorescence a large, open cymose panicle, often comprising 24 of the plant height; principal in- volueral braets 5-7 mm. long, 2-4 mm. wide, the apex narrowly obtuse or aeute with an inconspicuous, vellow-membranous tip 1(2) mm. long or less; corollas *'yellow-reddish," 3-3.5 mm. long, the tube sparsely glandular, 2 mm. long, the throat campanulate-funnelform, 1.5 mm. long with lobes reflexed, 1-1.2 times longer than the acute lobes; achenes 4-5 mm. long, pubescent principally on the corners with white hairs (0.8-)1-1.2 mm. long; pappus of 16-18 linear oblong scales, 1.5-2 mm. long; anthers exserted, 2-2.5 mm. long; chromosome number n = 17. Type: Texas. Brooks Co.: 2 mi. S. of Falfurrias, in deep sandy soil. March 20, 1952, F. B. Jones 695 (smu). Phototypes ws, TEX. DisrRiBUTION.—Rio Grande Valley of Southern Texas in open gravelly or sandy-clay soils of the lower Gulf Coastal Plain. (Fig. 47). This plant is a distinct variety which differs from Hymeno- pappus artemisiaefolius var. artemisiaefolius in a number of characters, including its more pubescent achenes, larger pappus, smaller corollas, fewer stem leaves, involucral bracts with in- conspicuous membranous tips, and ebracteate peduncles. The latter two characters seem to link the variety to H. scabiosaeus var. corymbosus, but the total characters are those of H. artemz- siaefolius. Further, the morphological characters listed above intergrade to a large extent with those of H. artemisiaefolius var. artemisiaefolius where the peripheral ranges of these two varieties approach each other. As yet, no similar intergradation has been found with H. scabiosaeus var. corymbosus. It is unlikely that the specimens cited represent hybrids or hybrid 306 Rhodora [Vor. 58 swarms between this species and H. artemisiaefolius var. ar- temisiaefolius, since the variety has characters which are peculiar to itself, such as its long pappus and achenial hairs (remarkably similar to those of the H. filifolius complex). Hymenopappus in the Rio Grande Valley of southern Texas is in critical need of field study both because of the scarcity of material from the area and because of the variability shown by the few specimens of the genus collected in this region. With the accumulation of more specific information, present concepts as to the relationship and status of var. riograndensis may have to be changed, but from the evidence available it seems best to consider it no more than a well marked variety of H. artemisiae- folius. REPRESENTATIVE SPECIMENS.—Texas. BROOKS CO.: 2 mi. S. of Fal- furrias, F. B. Jones 695 (smu); near Fulfurrias, Perkins & Hall 2857 (POM). CAMERON CO.: Point Isabel, R. Runyon 224 (us). KENEDY CO.: Norias Division of King Ranch, M. C. Johnston 54595 (TEX); KLEBERG co.: Santa Gertrudis Division of King Ranch, M. C. Johnston 54483 (TEX). HIDALGO CO.: 10 mi. N. of San Manuel, highway 281, R. Runyon 2634 (US). REFUGIO CO.: 5.5 mi. S.E. of Austwell, Cory 49078 (GH, sMU). (This specimen approaches var. artemisiaefolius.) W1ILLACY co.: loose sand prairle a few mi. W. of Redfish Bay, M. C. Johnston 54562 (TEX). EXCLUDED SPECIES Hymenopappus anthemoides Juss. Ann. Mus. Par. 2: 426. 1803. = Hy- MENOXYS ANTHEMOIDES (Juss.) Cass. Hymenopappus douglasii Hook. Fl. Bor. Am. 1: 316. 1834. = CHAENACTIS DOUGLASII (Hook.) Hook. & Arn. Hymenopappus glandulosus (S. Wats.) Rydb. N. Amer. Fl. 34: 38. 1914. = HYMENOTHRIX GLANDULOSUS S, Wats. Hymenopappus glaucus Spreng. Syst Veget. 3: 449. 1826. = CkPHALO- PHORA GLAUCA Cav. Hymenopappus integrifolius (Nutt.) Spreng. Syst. Veget. 3: 449. 1826. = POLYPTERIS INTEGRIFOLIUS Nutt. Hymenopappus ligulaeflorus Nelson, Wyoming Expt. Sta. Bull. 28: 135. 1896. = HYMENOXYS RICHARDSONII (Hook.) Cockerell. Hymenopappus matricarioides Spreng. Syst. Veget. 3: 450. 1826. = AGERA- TUM MATRICARIOIDES (Spreng.) Less. Hymenopappus nelsoni (Greenman) Rydb. N. Amer. Fl. 34: 49. 1914. = HYMENOTHRIX GLANDULOSA Var. NELSONII Greenman. Hymenopappus nevadensis Kellogg, Proc. Calif. Acad. 5: 46. 1873. = CHAENACTIS NEVADENSIS (Kellogg) Gray. Hymenopappus palmeri (Gray) Hoffm. in Engler & Prantl. Nat. Pfl. 45: 256. 1897. = HYMENOTHRIX PALMERI Gray. Hymenopappus pedatus Cav. ex Lag. Gen. et. Sp. Nov. 28. 1816. = Fro- RESTINA PEDATA (Cav.) Cass. 1956] Turner,—A Study of the Genus Hymenopappus 307 Hymenopappus wislizeni var. setiformis M. E. Jones, Contrib. West. Bot. 12: 47. 1908. = HYMENOTHRIX WISLIZENII Gray Hymenopappus wrightii (Gray) H. M. Hall, Univ. Calif. Publ. Bot. 3: 179. 1907. = HYMENOTHRIX WRIGHTII Gray. Hymenopappus wrightti var. viscidulus Jepson, Man. Fl. Pl. Calif. 1128. 1925. = HYMENOTHRIX WRIGHTII Gray. In addition, the following species of Rothia Lam., sensu O. Kuntze (1898) are not Hymenopappus: Rothia degenerica, Rothia intermedia, Rothia pinnata, Rothia pinnata pallida, Rothia pinnata purpurescens, Rothia pusilla. THE PLANT RESEARCH INSTITUTE, THE UNIVERSITY OF TEXAS, AUSTIN, TEXAS. BIBLIOGRAPHY AXELROD, D. I. 1948. Climate and evolution in western North America during middle Pliocene time. Evolution 2: 127-144. BALTZER, E. A. 1944. A monographie study of the genus Palaforia and its immediate allies. Ann. Mo. Bot. Gard. 31: 249-278. BENTHAM, G. 1873. Notes on the classification, history, and geographical distribution of Compositae. Jour. Linn. Soc. Bot. 13: 335-582. . and J. D. Hooker. 1873. Compositae. Genera Plantarum 2: 163-554. BIDDULPH, S. F. 1944. A revision of the genus Gaillardia. Res. Studies State College of Washington 12: 195-250. BRITTEN, J. and B. B. Woopwarp. 1905. Bibliographical notes XXV. L'Héritier's botanical works. Jour. Bot. 43: 266-273. CHANEY, R. W. 1947. Tertiary centers and migration routes. Ecol. Monogr. 17: 139-148. Constance, L. 1937. The genus Hriophyllum Lag. Univ. Calif. Publ. Bot. 18: 69-135. Cronauist, A. 1955. Phylogeny and taxonomy of the Compositae. Am. Midl. Nat. 53: 478-511. Crum, E. 1940. A revision of the genus Monolopia. Madrofio 5: 250-270. EvEsRLy, M. L. 1947. A taxonomic study of the genus Perityle and related genera. Contrib. Dudley Herb. 3: 375-390. GEISER, S. W. 1948. Naturalists of the Frontier. Dallas. Southern Methodist University Press. 296 pp. GLEASON, H. A. 1906. A revision of the North American Vernonieae. N. Y. Bot. Gard. Bul. 4: 144—243. . 1923. Evolution and geographical distribution of the genus Vernonia in North America. Amer. Jour. Bot. 10: 187-202. Gray, A. 1849. Plantae Fendlerianae. Mem. Am. Acad. Arts Sci. n.s. 4: 102. . 1886. Compositae. Synoptical Flora of North America. 2d ed. 12: 48-455. Heiser, C. B. Jg. 1944. Monograph of Psilostrophe. Ann Mo. Bot. Gard. 31: 279-301. . 1945. A revision of the genus Schkuhria. Ann. Mo. Bot. Gard. 32: 265-278. Horrman, O. 1897. Compositae. In Engler and Prantl, Die Natürlichen Pflanzenfamilien 4°: 87-391. INTERNATIONAL Cope oF BOTANICAL NOMENCLATURE. 1952. Regnum Vegetabile 3: 1-228. 308 Rhodora [Vor. 58 JOHANSEN, D. A. 1940. Plant Microteehnique. New York. MeGraw- Hill Book Co. 523 pp. JonNsTON, I. M. 1923. Diagnoses and notes relating to the Spermatophytes chiefly of North America. Contrib. Gray Herb. n.s. 68: 92-98. KuNrzg, O. 1891. Compositae. Revisio Generum Plantarum 1: 303-374. - 1898. Compositae. Revisio Generum Plantarum 3: 127-185. LAMARCK, J. 1792. Jour. Hist. Nat. 1: 16-19. Lansouw, J. and F. A. SrarLEUv. 1952. Index Herbariorum. Regnum Vegetabile 2: 1—167. MceCuiNTOCK, D. A. 1929. A method for making aceto-carmine smears permanent. Stain Tech. 4: 53-56. McVauau, R. 1947. The travels and botanical collections of Dr. Melines Conkling Leavenworth. Field and Lab. 15: 57-70. Mu.LER, C. H. 1952. Ecological control of hybridization in Quercus: a factor in the mechanism of evolution. Evolution 6: 147-161. Rose, J. N. 1891. List of plants collected by Dr. E. Palmer in Arizona in 1890. Contrib. U. S. Nat. Herb. 1: 117-127. RYDBERG, P. A. 1914. Helenieae. Hymenopappanae. N. Amer. Flora 34: 43-63. SMALL, J. 1919. The origin and development of the Compositae. London. William Wesley & Son. 334 pp. (Reprinted from the New Phytologist Vol. 16-18, 1917-1919.) SrkESE, B. M. and J. T. BALDWIN, JR. 1952. Chromosomes of Hymenorys. Amer. Jour. Bot. 39: 685-688. STEBBINS, G. L. 1950. Variation and Evolution in Plants. New York. Columbia University Press. 643 pp. —————. 1952. Aridity as a stimulus to plant evolution. Amer. Natur. 86: 33-44. STOCKWELL, P. 1940. A revision of the genus Chaenactis. Contrib. Dudley Herb. 3: 89-168. WALTERS, J. L. 1952. Ileteromorphie chromosome pairs in Paeonia cali- fornica, Amer. Jour. Bot. 39: 145-151. NOTES ON COLLINSIA VIOLACEA. —The designation of Collinsia violacea Nutt. as a spring annual, in spite of its very early flower- ing habit, has seemed worthy of observation to the writer. Both C. violacea Nutt. and C. verna Nutt. are found in Illinois, and the former has been collected only in Shelby County in the central part of the state where it has been reported for 20 years. The latter species is a rather common early spring flowering plant in widely separated moist woodlands throughout Illinois. Flowering records for Collinsia violacea in this state indicate that it is in bloom from late April through May. In his Scrophu- lariaceae of Eastern Temperate North America, Francis W. Pennell says, “Our species are spring-annuals of deciduous forests, dying when the trees have reached full foliage." (Page 290). In 1956] Maycock,—Tiarella cordifolia in Wisconsin 309 April 1955 the writer, with Dr. G. D. Fuller of Chicago, trans- planted some Shelby County seedlings of C. violacea, then in bud, to a garden in Springfield, Sangamon County, Illinois. These “garden” plants thrived, flowered and produced seed. During November of the same year it was observed that Collinsia seedlings had appeared in abundance. Such observation later prompted a visit to the Shelby County locality, and on March 6, 1956, vigorous seedlings of C. violacea were seen. Thus it appears that *spring-annual" is not a good designation for this species, and use of “winter annual" is suggested as a better description. No data have been personally collected for C. verna, but Mr. Harry E. Ahles of the Botany Department, University of North Carolina, reports (in correspondence) that he observed seedlings of C. verna in Brownfield Woods, Champaign County, Illinois, behaving in the same manner as that described for C. violacea. Collinsia violacea propagates easily from seed and probably will thrive in suitable habitats in many localities.—GLEN S. WINTER- RINGER, ILLINOIS STATE MUSEUM, SPRINGFIELD, ILLINOIS. TIARELLA CORDIFOLIA IN WiscoNsiN.— False miterwort or foamflower (Tiarella cordifolia L.) has not previously been re- ported for Wisconsin. The late Dr. N. C. Fassett in the pre- liminary report on the Saxifragaceae (1932) referred to the range of this species as reported in Gray's Manual, edition seven, as west to Minnesota and indicated that it should be sought in Wisconsin. Two specimens are contained in the University of Wisconsin Herbarium: G. S. Cornwall, Wis. (?), not dated and C. Goessl, Wisc., June 7, 1922. The poor data present and the fact that Fassett omitted any reference to these sheets is sufficient to suggest that they did not constitute a satisfactory first record. There is the possibility that they were not native collections. The writer, while examining specimens in the Plant Ecology Collections, encountered a fine specimen, collected in flower by Messrs. G. Cottam, R. T. Brown, and R. P. McIntosh. The plant was collected June 20, 1950, in a hemlock-hardwood stand in Florence County (T39N, RI7E, S8). The stand was composed mainly of sugar maple, hemlock, basswood, yellow 310 Rhodora (Vor. 55 birch and hornbeam with a scattering of balsam fir and slippery elm. Dominant understory associates were Aralia nudicaulis, Aster macrophyllus, Erythronium americanum, Galium triflorum, Mitella nuda, Maianthemum canadense, Streptopus roseus and Trientalis borealis. The genus Tiarella in North America, has been monographed by Lakela (1937) and the specimen in question conforms most closely to the description of Tiarella cordifolia, typical variety. Manistee, Southern Michigan was apparently the most north- western point of collection for the species known to Lakela. This report provides an authentic first record for this species in Wisconsin and an extension of approximately 200 miles in a northwesterly direction of the range as cited in Lakela (1937). The specimen is deposited in the University of Wisconsin Herbarium.—PAvr F. Maycock, DEPT. OF BOTANY, UNIVERSITY OF WISCONSIN. LITERATURE CITED Fassett, N. C. Preliminary reports of the flora of Wisconsin, XIX, Saxi- fragaceae. Trans. Wise. Acad. Se. 1932. LAKELA, O. A monograph of the Genus Tiarella L. in North America. Amer. Jour. Bot. 24: 344-351. 1937. RoBiNsoN, B. L. anb M. L. FERNALD. Gray’s New Manual of Botany, 7th edit. Amer. Bk. Co., 926 pp. 1908. MEDICAGO POLYMORPHA VAR. VULGARIS.—There is nothing like intense concentration on an intricate and perplexing problem to blind one to the very obvious. Medicago polymorpha L. var. ciliaris L., 1753, of course makes illegitimate my new combination based on M. denticulata var. ciliaris Ser. (Rhodora 58: 9). Our commonest bur clover, so-called M. hispida, must be M. polymorpha var. vulgaris (Benth.) comb. nov., based on M. denticulata var. vulgaris Benth., Cat. Pl. Ind. Pyr. p. 103, 1826. Under this go f. apiculata (Willd.) comb. nov., based on M. apiculata Willd., Sp. Pl. 3: 1414, 1802; and f. tuberculata (Godron) comb. nov., based on M. polycarpa var. tuberculata Godr. in Gren. & Godr., Fl. France 1: 390, 1848.—LLovp H. SHINNERs. Volume 58, no. 693, including pages 243-274, was issued 26 September, 1956 79 DEC 2 0 1956 dd Ow REFKRAENGE LIBRARY Hodova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD CARROLL EMORY WOOD, JR, Vol. 58 November, 1956 No. 695 CONTENTS: The Binomials of Helenium in Walter’s Flora Caroliniana. Howard S ee ee Oi e oues su e SERE n sa ee eae Sa 311 Notes on a New Hybrid Carex. Egbert W. Fell.......... sese 318 Additions and Extensions to the Flora of Manitoba. J.C. Ritchie 321 Andropogon Praematurus, a Fungus-induced Monstrosity. Richard TYME OT n ou) hy th un. nsi Sis EE 326 Rubus Chamaemorus in the Canadian Lakehead. Olga Lakela... 327 Prunus Americana var. lanata a Synonym of P. nigra. Lloyd HES kenner. O... ... ce ben esses trs a ho ERE 330 Hedyotis rosea in Arkansas. Jewel E. Moore........ s 331 Note on Echinochloa muricata. John R. Reeder............ e 331 Scientific Books, Libraries and Collectors. Gordon P. DeWolf, Jr. 332 The New England Botanical Club, Inc. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $6.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 60 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $5.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scientific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver's blocks is met through the author or his institution. Forms may be closed five weeks in advance of pub- lication. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. ——— CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U. S. A. Rhodora Plate 1227 - rarmrgr kr x Ld PLATE 1227. Page 52 of the Herbarium of Thomas Walter at the British Museum. Fig. A. Lectotype of Helenium vernale Walt. Fig. B. Syntype of Helenium aestivale Walt. Fia. O. Lectotype of Helenium aestivale Walt. Fia. D. Lectotype Helenium serotinum Walt. Rhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 November, 1956 No. 695 THE BINOMIALS OF HELENIUM IN WALTER’S FLORA CAROLINIANA Howanp F. L. Rock IN the course of a recent revision (unpublished) of the decur- rent-leaved species of the genus Helenium L., it became necessary to critically evaluate the binomials used by Thomas Walter in the genus for his Flora Caroliniana (1788). The general importance of the Flora Caroliniana in North American taxonomy and particularly in that of the southeastern United States has been noted by Blake (1915), Maxon (1936) and more recently by Totten (1956). The significance of Walter’s treatment of the genus Helenium lies in the fact that he created three new binomials within the genus. The identity of these binomials has long been intriguing, the more so in that they ostensibly did not apply to the even then highly segregated species, Heleni- um autumnale L. In evaluating these binomials there are two direct sources of evidence and data. The first of these, the descriptions provided in the Flora Caroliniana, is unfortunately rather meagre, the descriptions being quite short. Therefore, for clarity of ex- position and ease of reference, the text of Walter’s treatment is reproduced below. The second of these sources is Walter’s herbarium, which is preserved at the British Museum of Natural History. Inasmuch as John Fraser carried Walter’s personal herbarium to England at the same time as he did the manu- seript of Flora Caroliniana, it can be presumed that one was to serve as a supplement to the other. Nevertheless, it seems that the specimens of Walter’s herbarium have either been 312 Rhodora [Vor. 58 ignored or else overlooked in previous evaluations of the bino- mials. To these two sources can be added the modern knowledge of the Southeastern flora and these provide the basis for the interpretation and disposition presented herein. 327. HELENIUM. Receptaculum disci nudum, radii paleaceum. Pap- pus margo 8-partitus. Calyx simplex multipartitus. Corollulae radii 3 4 vefidae discus semiglobosus. vernale 1. foliis oblongis alternis integris, floribus solitariis terminalibus luteis, petalis radii (24) trifidis. aestivale 2. foliis oblongis integris, flosculis radii sex luteis quadrifidis, disco purpureo, pappo subsetaceo. serotinum 3. foliis serratis sessilibus caule laevi, floribus luteis, corollulis radii duodecim. autumnale foliis serratis decurrentibus, caule ramoso alato, floseulis 4. radii 12 luteis, disco purpureo aliquando geminato. The description of Helenium vernale sensu Walt. can be applied to two of the vernal species of Helenium found in the Southeastern Coastal Plain, Helenium Nuttallii A. Gray and Helenium pinnatifidum (Nutt.) Rydb. The description fits either one equally well, with the exception of the adjective "integris." The entire-margined condition of the leaves is more characteristic of H. Nuttallit than it is of H. pinnatifidum. Nevertheless, H. pinnatifidum very often occurs with entire- margined basal leaves, particularly so in the Florida-Georgia area of its distribution. Over the course of the years, however, H. vernale Walt. has been interpreted as applying to the same taxon as H. pinnatifidum (Nutt.) Rydb. Morphologically, Helenium pinnatifidum is characterized by having pubescent achenes, peduncles and involucral bases that are pubescent, mid-cauline leaves that are only shortly decurrent (4 mm. or less) along the stem and by radical leaves that are usually pinnatifid in outline and not with petioloid bases. "The distribution of H. pinnatifidum is in the outer portions of the Costal Plain, more so than H. Nuttallii, and extending farther southeastward into peninsular Florida and northeastward into southeastern North Carolina. However, a simple check of the specimen (Fig. A) in the Walter Herbarium that most closely fits his description! reveals that the achenes are glabrous and ! Morphological data provided by Mr. W. T. Stearn of the British Museum of Natural History. 1956] Rock,—Binomials of Helenium 313 that the peduncle is glabrous as well. These characters are diagnostic of H. Nuttallii and the author is convinced that if the midcauline leaves were present on the specimen they would be conspicuously decurrent and that the radical leaves would be more entire rather than pinnatifid and would have petioloid bases. In addition, H. Nuttallii has a distribution that is more toward the inner portion of the Coastal Plain than H. pinnati- fidum and is the only species of the two in question that occurs with regularity in Berkeley County, South Carolina. Moreover, the pappus scales of the Walter specimen correspond more closely to the stereotype of the pappus scales of H. Nuttall than they do to that of H. pinnatifidum. Regrettably, the vari- able nature of the pappus scales in both taxa is such that there is a degree of overlapping and intergradation in the size and shape of the scales to the extent that the pappus scale char- acteristics are not as differential criteria as those presented above. In the course of time, then, the binomial Helenium vernale has come to be misapplied to the taxon which was rightfully named H. pinnatifidum (Nutt.) Rydb., while the name H. Nuttallii A. Gray, which now becomes a synonym, has usurped the rightful position of H. vernale sensu Walt. The importance of the specimen in the Walter Herbarium and its preservation cannot be overstressed in relation to the binomial, for were it not for the specimen, the binomial would have to be declared a nomen dubium inasmuch as neither the epithet nor the brief description are sufficient to resolve the problem of its correct application. It is the author’s considered opinion that the binomials Helenium aestivale Walt. and Helenium serotinum Walt., long held as nomina dubia, do not have application to any of the species of Helenium, and most particularly not to any of those species encountered in the flora of the Southeastern United States. It is my belief that these two binomials apply instead to the genus Gaillardia, and that Walter made no distinction between the genus Helenium L. and the genus Gaillardia Fougx. (1787)2 This is not so surprising, for Walter classified his 2 Walter was probably not aware of Fougeroux’s paper, due in part to the time lapse in communication between Europe and North America and in part to the obscurity of the original publication in relation to botany. 314 Rhodora [Vor. 58 H. vernale along with the other species of Helenium in the Flora under the class Syngenesia Polygamia Superflua, even though by virtue of the neutral and sterile ray florets it patently belongs to the class Syngenesia Frustrania. Indeed, according to the interpretation presented herein, all the binomials of Walter in the genus Helenium, whether they be truly members of Helenium or Gaillardia, would be classified as pertaining to the Syngenesia Frustrania for both the vernal species of Helenium and the species of Gaillardia are characterized by neutral and sterile ray florets. The three remaining specimens in the Walter Herbarium labeled Helenium are in fact specimens of Gaillardia Fougx. These specimens belong to the Section HOL!ANDIA Biddulph (1944) of the genus Gaillardia. This section of the genus is based on the characters listed below: l. Receptacle with small dentiform setae; 2. Style branches long and hispidulous; 3. Lobes of the disk corollas eaudate-acuminate. It is to be noted that the setiferous character of the receptacle usually associated with Gaillardia as a diagnostic generic char- acter becomes less than distinctive within this section, especially in relation to the character of the receptacle in Helenium. In fact and in practice, no real distinction can be made between the genus Helenium and the section HOLLANDIA (Gaillardia) solely on the basis of the receptacular surface and its projections. Of these three Gaillardia specimens, two belong to the species Gaillardia lanceolata Michx. while the remaining one would be referred to G. lutea Greene according to present usage. Whether these specimens were collected in Berkeley County, South Carolina or not, is a moot question. Most likely they were collected to the southeast or southwest of Walter’s area and probably by John Fraser who traveled in both South Carolina and Georgia (fide Maxon). Biddulph records G. lanceolata from Aiken County, South Carolina? as well as from Georgia, Florida and thence westward into Texas. East of the Mississippi River, she records G. lutea from Baldwin County, Alabama and Brevard County, Florida. š Specimens of G. lanceolata are on deposit in the Duke University Herbarium from Sumpter Co., S. C. and Columbus Co., N. O. 1956] Rock,—Binomials of Helenium 315 The description of Helenium aestivale Walt. can certainly be applied to Gaillardia lanceolata Michx. The characters of oblong, entire leaves; few yellow rays; purple disk; and sub- setaceous (long-awned) pappus have no real meaning in relation to any species of Helenium that occurs in the Southeastern United States but do apply to G. lanceolata. In the same manner, if Helenium serotinum Walt. is equated with G. lutea Greene, then the description provided by Walter agrees with the particulars that Biddulph gives for that taxon. The author is aware of the dangers and fallacies that can arise in interpreting descriptions in terms of specimens and vice versa but if the binomial Helenium vernale Walt. is to be maintained then H. aestivale Walt. and H. serotinum Walt. should also be maintained, for certainly the latter two binomials are no more dubious than the former. In view of the priority of the epithets of Walter over those of later authors and according to Article 65 of the International Rules of Botanical Nomenclature (7th ed.), which states that prior epithets must be reinstated, there being no obstacle to their reinstatement, the following new combinations are proposed: Gaillardia aestivale (Walt.) H. Rock, comb. nov.' Helenium aestivale Walt. Fl. Carol. 210. 1788. Gaillardia lanceolata Michx. Fl. Bor.-Am. 2: 142. 1803. Gaillardia serotinuim (Walt.) H. Rock, comb. nov.? Helenium serotinum Walt. Fl. Carol. 210. 1788. Gaillardia lutea Greene Pittonia 5: 57. 1902. The remaining binomial used by Walter in the Flora is Heleni- um autumnale sensu Walt. (non Linnaeus). This binomial is not represented by a specimen in the Walter Herbarium. However, from the description and from the epithet, it is evident that what Walter considered to be Helenium autumnale L. is Helenium flexuosum Raf. (H. nudiflorum Nutt.) The decisive factor in this interpretation is the phrase “disco purpureo", one of the most obvious characters that serves to distinguish H. flexuosum from H. autumnale. However, in the light of the widely variable and descriptive polynomial synonymy 4 Lectotype: Helenium aestivale Walt., specimen No. 235, Walter Herbarium (BM), (Fig. O). > Lectotype: Helenium serotinum Walt., specimen No. 243, Walter Herbarium (BM), (Fig. D). 316 Rhodora (Vor. 58 listed under H. autumnale by Linnaeus (1753), Walter’s inclusion of the later defined H. flexuosum with H. autumnale is indeed warranted. Even though H. flexuosum is classified on technical characters along with the vernal-flowering species of Helenium (Leptopoda Nutt.), its actual flowering period in Walter’s area ranges from mid-June through late October. It is the author’s personal experience that H. flexuosum is one of the conspicuous and most abundant composites in the immediate vicinity of Walter’s grave and former garden and of that area of South Carolina in general. An alternate possibility is that Helenium autumnale sensu Walt. is the same as H. brevifolium (Nutt.) Wood. However, H. brevifolium is distinctly vernal in its flowering habit and while rather endemic in the nature of its distribution northward of the Gulf Coastal Plain, it is unreported from South Carolina. Moreover, the small number of ray florets cited by Walter would rule against H. brevifolium as a possibility, for in that taxon the number of ray florets is usually well in excess of twenty rays. Even though Walter’s descriptions are short and concise, the pattern of the phrasing and the order of arrangement. of the species appears to have a significance that exceeds that of just mere description. If one allows some liberality in transla- tion and a slight amount of change in the sequence of the phrases, but maintaining the order of the species, the following key can be constructed solely on the basis of the descriptive material provided by Walter. 1. Leaves entire and oblong 2. Ray florets 24 in number, trifid; heads solitary, terminal and yellow; leaves alternate...................... Helenium vernale. 2. Ray florets 6 in number, 4-fid; disk purple; pappus sub- setaceous (long-awned)......................... Helenium aestivale. 1. Leaves serrate 3. Leaves sessile (not decurrent) ; stem smooth (not winged) ; heads yellow; ray florets 12 in number........ Helenium serotinum, 3. Leaves decurrent; stem winged and branched; disk purple; ray florets yellow and 12 in number. . . Helenium autumnale. It is just such an analytical device that perhaps contains the answer to the contention raised by Dr. Baldwin in a letter? to Stephan Elliott: 5 An undated letter from Dr. Baldwin of Georgia to Stephan Elliott of Charleston, South Carolina, The letter is now mounted with the lectotype of Leptopoda puberula 1956] Rock,—Binomials of Helenium 317 That the first mentioned species of Galardia is the Helenium vernale of Walter I contend—Walter would scarcely have omitted the fol. [iis] decurrent. [ibus] It is probable that he was ignorant of the species with decurrent leaves. . From the schema above it can be seen that the decurrent-leaved condition has no bearing on the relation of Helenium vernale to the other three species of Walter's Flora. Rather, the question of whether the leaves are decurrent or not is raised only in effecting a decision between Helenium serotinum and Helenium autumnale. The fact that Walter did not state that Helenium vernale has decurrent leaves cannot be taken to infer that the leaves are non-decurrent, for by the same token, neither did he state that the leaves are sessile as he explicitly does for Hele- nium serotinum. Therefore, in view of the pattern and the phraseology of Walter's descriptions, it seems reasonable to assume that Walter intended the descriptive material to be diagnostie as well, even though he did not adopt a key-like for- mat, an innovation which apparently originated in the same year (1788), with the publication of Lamarck’s Flore Frangoise.— DEPARTMENT OF BOTANY, DUKE UNIVERSITY. LITERATURE CITED BIDDULPH, S. F. 1944. A Revision of the Genus Gaillardia. Res. Stud. State Coll. Wash. 12: 195-256. - Buake, S. F. 1915. Some Neglected Names in Walter's Flora Carolini- ana. Ruopora 7: 9-15. FoucErroux, DE B. 1787. Contenant la Description d'un Nouveau Genre de Plante. Obs. sur La Physique 29: 55. Lamarck, J. B. pe 1788. Flore Françoise. Paris. LINNAEUS, C. von. 1753. Species Plantarum 2: 886. Stockholm. Maxon, W.R. 1936. Thomas Walter, Botanist. Smithsonian Miscell. Coll. 95(8). Washington. Torren, H. P. 1956. Early Contributors to Southern Botany. The ASB Bulletin 3: 25-26. WALTER, T. 1788. Flora Caroliniana. 210. London. MacBride in Ell. contained in the Elliott Herbarium of the Charleston Museum. In the above quotation, the Galardia referred to is Galardia fimbriata Michx., a nomen rejiciendum, which is synonymous in part with Helenium pinnatifidum (Nutt.) Rydb. 318 Rhodora [Vor. 58 NOTES ON A NEW HYBRID CAREX! Eapert W. FELL Tue following is an extract from a paper entitled “Carex of Rock River Valley in Illinois” which Dr. Fuller, Mr. Ahles and I presented at the meeting of the Illinois State Academy of Science at Carbondale in 1955. It has not been published. The Carex under discussion was discovered by me in April 1949 growing on the wooded gravel bluff of Rock River south of Rockford at the north edge of Camp Grant. The bluff, about 50 feet high, a part of Rock River terrace, is water-lain sand and gravel without loess or ice-borne till. It has a western exposure and is lightly shaded. It supports the remains of a prairie vegetation but is now wooded with Quercus velutina, Q. alba and Q. macrocarpa, a transition from prairie to mixed oaks, the trees having advanced upward from the very narrow flood plain. Associated plants are, beside a few choke-cherry bushes (Prunus virginiana), a mixture of prairie plants and mesophytes: Bouteloua curtipendula, Leptoloma cognatum, Pani- cum scribnerianum, Aristida basiraema, A. oligantha, Carex meadii, C. pensylvanica, C. muhlenbergii, C. sprengelii, C. normalis, Claytonia virginica, Arenaria stricta, Dicentra cucullaria, Cory- dalis flavula, Draba reptans, Oxalis violacea, Dodecatheon meadia and Penstemon pallidus. The Carex grows in a patch about 75 feet in diameter on the crest of the bluff. During the past 7 years it has increased in size about one quarter. In the average winter the basal part of the plant stays green. Spring growth starts early, the thick staminate spike becoming conspicuous only a little later than C. pensylvanica. Achenes develop incompletely and infre- quently, varying in different years. I have found no mature achenes. I asked Mr. Ahles to write this description of the plant naming it for Dr. George Damon Fuller, who has been familiar with it from the time of its discovery. Last spring I sent Mr. Ahles, in Urbana, living plants of the hybrid, of C. pensylvanica, C. laeviconica and C. meadii from this area, but it was too late to ! A contribution from the Evelyn I. Fernald Memorial Herbarium of Rockford College, Rockford, Illinois. 1956] Fell,——Notes on a New Hybrid Carex 319 “Ç CAREX X FULLERI H.E.Ahles Q. plont X 4 D. perigynia x5 C. pistillote scale x 5 Fig. 1. Carex X FULLERI H. E. Ahles, hyb. nov. Plant x 1⁄4. Pistillate scale and perigynium X 5. 320 Rhodora [Vor. 58 do crossing experiments. It was expected that this would be undertaken during the spring of 1956, but the move of Mr. Ahles from University of Illinois to University of North Carolina interrupted the plan. Dr. Frederick J. Hermann and Dr. Theodore Just are familiar with the plant through herbarium specimens. There is apparent agreement as to its probable hybrid origin, though its location, just where the service area of the Base Hospital was during the first world war, suggests the possibility of its being an accidental introduction. Carex x fulleri H. E. Ahles, hyb. nov. A sterile hybrid possibly of Carex laeviconica Dewey and C. pensylvanica Lam. The plant forms small tufts of 2 to 5 erect culms at the end of long, slender rhizomes which reach a length of 1-2 dm. The rhizomes are covered with long, narrow, overlapping, fibrillose scales 12 mm. in length. The internodes on the rhizome measure about 10 mm. The leaves, 4 mm. wide or less, are glau- cous, scabrous on both surfaces, chiefly basal, and overtop the inflorescence by as much as 1 dm. The sharply angled culms, 5 dm. tall, are very thin, less than 1 mm. thick, and are scabrous at the top. The lowest pistillate spike is sessile or very short stalked, usually distant, upper one or very rarely two, crowded at base of staminate spike, sessile; all oblong- cylindric, 10-25 mm. (in general nearer 10 mm.) long, 5-10 mm. broad, densely flowered; dark purplish-brown scales with lighter central band, acuminate, 5 mm. long and 2 mm. wide, midrib prominent; staminate spike terminal, 15-30 mm. long, 3-4 mm. wide, straw colored; perigynia brown, dull, 3 mm. long, ovoid, gradually contracted to the short, smooth, bidentate beak, 0.2 mm. in length. Pistil three-parted; achene not developed, but the immature achene sharply triangular. Fig. 1. Culmis 5 dm. altis, erectis; foliis 4 mm. sive minus latis, glaucis, scabrosis. Spicis femineis sessilibus sive subsessilibus, oblongo-cylindricis, 10-25 mm. longis, 3-10 mm. latis, dense floribus intertextis. Perigyniis fulvis, surdis, 3 mm. longis, ovoideis, gradatim in breve rostrum, glabrum, bidentatum 0.2 mm. longum contractis. Pistillo tri-partito. Achenio non evoluto sed acute triangulari.? TYPE SPECIMEN: Illinois, Winnebago County, wooded gravel bluff of Rock River at Camp Grant, south of Rockford, June 16th, 1954, E. W. Fell 504—489. Additional collections from the same area are: April 29th, 1949, E. W. Fell 49-115; May 23rd, 1953, E. W. Fell 53-147; July 6th, 1953, E. W. Fell 53- 679. Specimens cited are all in the University of Illinois Herbarium. Dupli- cate specimens are in the Illinois State Museum Herbarium.” —ROCKFORD, ILLINOIS. 2 Assistance was given by Professor J. L. Heller, Head of the Department of Classics, University of Illinois, in preparing the Latin description. 1956] Ritchie,—Flora of Manitoba 321 ADDITIONS AND EXTENSIONS TO THE FLORA OF MANITOBA J. C. RITCHIE AN opportunity to spend part of the summer of 1955 in north- west Manitoba enabled the writer to make detailed local studies in two regions which had not been explored by any botanist. Four additions, and certain notable extensions to the known flora which resulted from this work are reported here. Detailed floristic and ecological accounts will be published elsewhere. The information which is reported here was not available in time to be included in the forthcoming flora of the province by Dr. H. J. Scoggan. The northwest region of Manitoba has received little attention from botanists, and only the collection of Baldwin (1953) provides representative information about the area. The present material was collected from two main areas, at MacBride Lake (56° 52’ N., 99° 57’ W) and Tod Lake (56° 45’ N., 101° 47’ W). In the following account the writer's collection numbers are shown in italics and the Herbaria of the University of Manitoba and the Department of Agriculture, Ottawa are referred to as MAN and DAO respectively. Woodsia alpina (Bolton) S. F. Gray. Tod L. 1264. Only one locality for this plant was established; it was found in a deeply shaded cliff ledge on the north-facing side of a high outcrop ridge on the northwest shore of Tod Lake. This is the first authentic record for the province; reports by Macoun (1890) and Lowe (1943) have not been substantiated by any specimens. Cystopteris dickieana Sim. Tod L. 1263. Growing in close association with Woodsia alpina (1264), this species was recorded from only that local- ity. This is the fourth record of the plant from Manitoba, the others being at Churchill (Ritchie, 1956) and from two localities near The Pas, about 300 km. south of the present locality (cf. map in Léve and Freedman, 1956). Dryopteris phegopteris (L.) Christens. Tod L. 1257. The only record, it was noted in rich wet soil below a small beaver dam under black-spruce and poplar. This is the first specimen collected in the province; Macoun’s (1890. p. 269) report that it is “rather scarce along Lake Manitoba and the Porcupine Mountains, Man.” is unsupported by any specimens. Potamogeton robbinsii Oakes. Macbride L. 1214. This aquatic was found in local abundance in certain shallow arms of MacBride Lake where it forms a broad submerged (at about 1 m.) zone of vegetation with P. alpinus var. tenuifolius. This is the second record from Manitoba; a 322 Rhodora (VoL. 58 specimen was collected by B. J. Marck in 1936 (DAO) from Bissett, in the southeast corner of the province, about 750 km. from the present station. Schizachne purpurascens (Torr.) Swallen. Tod L. 1247. Recorded in well drained mineral soil on a south-facing, unshaded site on the summit of a high outerop ridge, this extends northwards the area of this species in the province; the previous northern limit was at Cross Lake (Scoggan, 1950), about 250 km. southeast from the present station. Carex heleonastes Ehrh. MacBride L. 968, 1024, 1147, 1185. Never found in local abundance, this apparently rare sedge was discovered with surprising ease in several peat bogs in the area. Here it is confined to large cushions of Sphagnum warnstorfianum and Camptothecium nitens, where it is associated with Andromeda polifolia, Vaccinium oxycoccus, Rubus acaulis, Salix pedicellaris var. hypoglauca and Carex limosa. Due to the lack of adequate material from the province—the only other speci- mens of the C. heleonastes-amblyorhyncha complex are of C. amblyorhyncha from Churchill (Bécher, 1952)—it is not yet possible to provide represent- ative information about this species. Carex leptalea Wahl. MacBride L. 1143. Apparently a plant of restricted ecological amplitude, it was collected from the peaty banks of a small muskeg stream where it was associated with Carex paupercula and Caltha pal stris. This record merely extends the northern limit of the plant in Manitoba (it is known from Keewatin) from The Pas and Churchill into the northwest corner of the province. Carex abdita Bickn. Tod L. 1242. A new record for the province, a single colony was found on the southern slope of the high outerop ridge which bounds the northwest shore of Tod Lake. This record comprises a remarkable northern extension of this transcontinental American plant whose main area is in temperate latitudes. Carex lasiocarpa Ehrh. var. americana Fern. MacBride L. 1219, 1221. Both locally abundant and ecologically important this sedge forms a narrow zone at the periphery of certain lake-bogs, associated with C. rostrata, Calla palustris and Menyanthes trifoliata. This record extends northwards the area of the plant within the province from near Cedar Lake (Scoggan, 1950), a distance of about 400 km. Carex oligosperma Michx. Tod L. 1270. In local abundance at a single locality, it forms a zone of dominance in a small swampy depression over clay. This is the second authentic record and makes an extension northwards of 800 km. from the first locality which is in the southeast corner of the province. A new variety of this species has been discovered at Churchill, on Hudson Bay (described by Raymond in Ritchie, 1956). Acorus calamus L. MacBride L. 1228. A single colony was found, growing in the marginal bog vegetation of a shallow arm of MacBride Lake. This record extends the northern limit of the plant by 250 km. from Cross Lake where it was collected by Scoggan (1950). Salix pseudocordata (Anderss.) Rydb. MacBride L. 1113. A single plant of this was noted, growing in a muskeg where it was concomit- 1956] Ritchie,—Flora of Manitoba 323 ant with Picea mariana, Salix myrtillifolia, Rubus chamaemorus, Sphagnum fuscum, S. warnstorfianum and Aulacomnium palustre. This is the first record for Manitoba and it comprises a considerable (1000 km.) north- eastern extension from the nearest known station, at Cypress Hills in southwest Saskatchewan where it occurs as an outlier of the main area of the plant which is in the southern Cordillera (Breitung, 1954). Salix hebecarpa Fern. (= ? S. athabascensis Raup) MacBride L. 990, 1029. Two localities were established for this willow, and in both in- stances it grew in muskeg vegetation, associated with Picea mariana, Betula glandulosa, Ledum groenlandicum and other typical muskeg plants. The differences between this species and S. athabascensis are unconvincing, and the latter scarcely merits specific rank. S. hebecarpa is known only from Alaska (Hultén, 1943 p. 520-1) and the Gaspé Peninsula (Scoggan, 1950a p. 173). Possibly these are eastern and western sub- species of a single species, together forming a more or less continuous area in N. America. This record extends the provincial area northwards by about 600 km., from Riverton (Scoggan, 1953). Salix gracilis Anderss. var. textoris Fern. MacBride L. 1205. This was recorded only from certain rare stands of white-birch on raised peat ridges. It extends northwards by 300 km. the area of the plant from The Pas (collected by W. Krivda, MAN). Salix arbusculoides Anderss. MacBride L. 985, 989. This was recorded from muskeg in two localities. It is the second record for the province, the other being at Churchill, the eastern limit of the entire area of the plant. Arabis holboellii Hornem. var. collinsii (Fern.) Rollins. Tod L. 7253. Only one colony of plants was found, in dry soil on the southern slope of a high outcrop ridge at Tod Lake. This extends further the northern limit of this plant in the province from the locality established by Freed- man (MAN) near Flin Flon, about 250 km. from the present area. Potentilla arguta Pursh. Tod L. 1246. The only specimen collected from the area, it was found in dry mineral soil on the south-facing slope of the summit of a high outcrop ridge. This extends northwards by 300 km. (from The Pas) the area of the plant in northwest Manitoba. Potentilla pensylvanica L. var. pensylvanica. Tod L. 1251. In the same locality as the previous two species, this reeord extends the area, by 250 km. from the previous northern limit for the province at Cross Lake (Scoggan, 1950). Oxytropis splendens Doug]. Tod L. 1241. "This was the only locality in the areas which were surveyed; here it was concomitant with the three previous species and Artemisia caudata, This record extends the area in the province by 600 km. northwards from the locality at Cowan, Duck Mountain (Scoggan, 1952). Artemisia caudata Michx. Tod L. 1245. Associated with the last four species, it was confined to dry, open sites on the summit of the ridge on the northwest shore of Tod Lake. This record extends the area northwards from Knee Lake (Scoggan, 1951) by about 300 km. 324 Rhodora [Vor. 58 Certain of these additions and extensions are of phytogeo- graphical interest. While it is apparent that the flora of northern Manitoba is incompletely known at present, several of the records which have been enumerated constitute remarkable outliers of the main areas of the plants, while others are rare or local, even in regions which have been explored in greater detail. Of these plants, Woodsia alpina, Cystopteris dickieana, Schizachne purpurascens, Carex abdita, Arabis holboellii var. collinsii, Potentilla arguta, Potentilla pensylvanica var. pen- sylvanica, Oxytropis splendens and Artemisia caudata were recorded only from the relatively high ridge which runs parallel to the northwest shore of Tod Lake. This ridge is of a some- what exceptional geological nature, having arisen by considerable faulting between two series of sedimentary rocks (Milligan, 1952). Otherwise, by far the greater area of this entire region is of low relief, consisting of undulating glacial till ridges with the intervening depressions filled by lakes, muskegs and bogs, the whole showing little diversity of habitat and bearing a relatively poor flora, typical of much of the northern forested regions of the Canadian Shield. On this ridge on the excep- tionally well drained upper slopes with a southern aspect, above small cliffs and therefore relatively unshaded by trees, were found Schizachne purpurascens, Carex abdita, Arabis holboellii var. collinsii, Potentilla pensylvanica var. pensylvanica, Oxytropis splendens and Artemisia caudata. Without exception these occurrences, near the northern limit of the continuous coniferous forest, are considerably beyond the main areas of these species, all of which are more typical of the southern deciduous forest and prairie vegetation zones. Further, the presence of exceptional cliff habitats provides, on the north- facing slopes, conditions of shelter and moisture which are suitable for such ecologically specialized ferns as Woodsia alpina and Cystopteris dickieana. It is suggested that these records provide some corroboration, at the regional flora level, of the thesis of Böcher (1951) that many discontinuities of area can be explained satisfactorily in terms of ecological factors. The writer wishes to record his thanks to Dr. H. J. Scoggan who kindly provided much of the information about distribution within Manitoba. This work is being supported by generous 1956] Ritchie,—Flora of Manitoba 325 grants from the National Research Council of Canada.—pr- PARTMENT OF BOTANY, UNIVERSITY OF MANITOBA, WINNIPEG, CANADA, LITERATURE CITED Batpwiy, W. K. W. 1953. Botanical investigations in the Reindeer- Nueltin Lakes area, Manitoba. Nat. Mus. Canada, Bull. 128: 110-142. BREITUNG, A. J. 1954. A botanical survey of the Cypress Hills (Sas- katchewan and Alberta). Can. Field-Nat. 68: 55-92. Bocuer, T. W. 1951. Distribution of plants in the cireumpolar area in relation to ecological and historical factors. J. Ecol. 39: 376-395. 1952. A study of the cireumpolar Carex heleonastes- amblyorhyncha complex. Acta Arctica 5: 1-31. HuurúsN, E. 1943. Flora of Alaska and Yukon, Vol. HI. Lunds Univ. Arssk. N. F. Avd. 2, Bd. 39 (1). LóvE, D. and N. J. FREEDMAN. 1956. A plant collection from S. W. Yukon. Bot Notiser, 1956: 153—211. Lown, C. W. 1943. List of the Flowering Plants, ferns, club mosses, mosses and liverworts of Manitoba. Nat. History Soc. of Manitoba, Winnipeg. Macoun, J. 1890. Catalogue of Canadian Plants, Part V. Montreal. MiLLiGAN, G. C. 1952. Geology of the Laurie Lake Area, Granville Lake Division, Northern Manitoba. Manitoba Mines Branch, Rept. 50-7: 1-30. Rrrcuig, J. C. 1956. The native plants of Churchill, Manitoba, Canada. Can. J. Bot. 34 (2): 269-320. Scoaaan, H. J. 1950. Botanical investigations in Central Manitoba. Nat. Mus. Canada, Bull. 118: 84-92. — 950a. The flora of Bie and the Gaspé Peninsula, Quebec. Nat. Mus. Canada, Bull. 115. 399 pp. 1951. Botanical investigations along the Hayes River route, Northern Manitoba. Nat. Mus. Canada, Bull. 123: 139-161. — ————, 1952. Botanical investigations in Manitoba in 1950. Nat. Mus. Canada, Bull. 126: 208-227. —. 1953. Botanical investigations in the Glacial Lakes Agassiz-Souris Basins, 1951. Nat. Mus. Canada, Bull. 128: 103-107. 326 Rhodora [Vor. 58 ANDROPOGON PRAEMATURUS, A FUNGUS-INDUCED MONSTROSITY! RicHarp W. Pour Andropogon praematurus Fernald (Ruopora 42: 413-415, 1940) was described as a new species of the section Schizachyrium, differing from A. scoparius in the following features: 1. Possession of only one pedicel or pedicellate spikelet at each node of the rachis—‘Most remarkable, the sessile perfect spikelets, instead of being accompanied by a pair of truncate bearded pedicels or these terminated by mere rudiments, usually are accompanied by a single well developed staminate pedicelled spikelet." (p. 414) 2. Well-developed, staminate pedicellate spikelets. 3. Extremely early flowering. 4. Small size of plant. Through the courtesy of Dr. Reed Rollins, it has been possible to examine the type specimen of A. praematurus (Fernald and Long 10,092, au) and the other cited material of this species. As deseribed by Fernald, all specimens show one sessile and one pedicellate spikelet at each node of the rachis. This dis- position of the spikelets is exactly the same as in all other species of Andropogon known to the author. Professor Fernald's statement that other species of this genus are different in having two sterile pedicels accompanying each spikelet is a misinterpre- tation. A detached unit of the inflorescence in the genus Andropogon consists of: 1. a sessile spikelet, 2. a pedicel bearing a pedicellate spikelet or a rudiment, and 3. the rachis internode which would normally bear the next distal similar unit. Such a detached unit might be interpreted as having two pedicels, but this is not the case. A. praematurus is therefore not excep- tional in its inflorescence structure, a fact which has been pointed out by Agnes Chase (Man. Grasses U. S., Rev. Ed. 1950, p. 755). The pedicellate spikelets of the type of A. praematurus and all other specimens of this “species” available to me do indeed have well-developed pedicellate spikelets. This development, however, is occasioned by the fact that all of them are heavily infected by a kernel smut (Sphacelotheca sp.). This fungus ! Journal paper No. J-2972 of the Iowa Agricultural Experiment Station, Ames, Iowa. Project No. 1136, 1956] Lakela,—Rubus Chamaemorus 327 was kindly identified for me by Dr. J. C. Gilman. The pedicel- late spikelets are not staminate, as stated by Fernald. The entire flowers of both the sessile and pedicellate spikelets are completely destroyed by the fungus infection. Fernald’s plate of the inflorescence of the type specimen (RHODORA, 42: pl. 626, fig. 3) well illustrates the typical smut pustules formed by Sphacelotheca sp. Probably no plant species has ever been described from an entirely healthy individual, so the mere presence of a disease on the type specimen would not necessarily invalidate a new species. In the case of A. praematurus, however, all of the structural and physiological peculiarities upon which the species was differentiated may be ascribed to the well-known effects of systemic smut infection—malformation of organs, small stature, and early blooming. The type of A. praematurus is to be regarded as a monstrosity induced by disease, and hence the species is to be rejected under the provisions of article 77 of the 1952 edition of the Interna- tional Rules of Botanical Nomenclature. It could also be rejected under the provisions of article 76, since the appearance of the type is due to the combination of host and parasite, which are two genetically incongruous elements. Some of the specimens which have been designated as 4. praematurus have pubescent sheaths. These have been desig- nated A. praematurus f. hirtivaginatus Fernald (Ruopora 44: 383, 1942). These are also smutted individuals of Andropogon, section Schizachyrium and this name must also be rejected upon the same basis as that of the species. There is considerable variation in both vegetative and para- floral structure of the specimens which have been denominated A. praematurus. It is possible that several species of Andropogon are involved, but in view of the fact that systemic smut infec- tions may have altered all of the structures, it is probably point- less to attempt to determine the species. RUBUS CHAMAEMORUS IN THE CANADIAN LakEHEAD.— The publication of the Occurrence of Rubus Chamaemorus L. in Minnesota, brought to light additional information pertaining 1 LAKELA, O., Ruopona, 56: 272-273. 1954. 328 Rhodora [Vor. 58 to its midcontinental distribution. In a communication from Mr. Claude E. Garton, of Port Arthur, references were made to his own field observations and collections of the species and to those of Dr. T. M. C. Taylor in the Canadian Lakehead. Dr. James H. Soper, University of Toronto, made available a list of collections of this species including those of Dr. Taylor and his associates with an unpublished manuscript map. With this information at hand, an attempt has been made to redefine the southern limit of the range of this arctic species and to relate to the distribution pattern on the Canadian side the single known station at Basswood Lake, Minnesota, some 120 miles west from Lake Superior. The Canadian botanists collected the plant on Lake Superior north of the border more than 20 years ago. Mr. Garton, in 1952, discovered it on the north side of Pigeon Bay astride the International Boundary. Its absence on the south side is notable. The species is unknown in parts of Minnesota ad- joining the Thunder Bay District. Obviously the center of the interior distribution parallels the north shore of Lake Superior. The disjunct sites to the north and northeast are in Patricia and Cochrane Districts respectively. Westernmost is the Basswood Lake colony at 48° 07’ N. lat. and 92° 43’ W. long., the latitude approximating that of the Pigeon Bay colony, defining the southerly limit of the range. With the exception of the first citation, from Mr. Garton’s letter the collections herewith listed are deposited in the Her- barium of the University of Toronto, with the original distribu- tion map, presumably plotted by Dr. Taylor, which was amended and adapted for:its present use. THUNDER BAY DISTRICT: In a sphagnum swamp at Amethyst Harbor, 23 mi. n.e. of Port Arthur, June 17, 1933. Claude E. Garton. Open wet sphag- num bog, north side of Pigeon Bay, Lake Superior, Crooks Twp., July 16, 1952. Claude E. Garton 1865. In sphagnum in spruce bog, Fork Bay, Sibley Twp., (lat. 48° 20’ N. long. 88° 50’ W.), June 23, 1936. T. M. C. Taylor et al. 1286. Spruce-cedar swamp, Rossport, (appr. 48° 45’ N. 87° 15’ W.), Aug. 11, 1937. R. C. Hosie et al. 1996. In black spruce, Bluejay muskeg, Schreiber, (appr. 48° 45’ N. 87° 15’ W.), June 30, 1937. R. C. Hosie et. al. 1998. Inland lake, east side of Patterson Is., Slate Islands, July 28, 1937. R. C. Hosie et al. 1999. Open spruce swamp, Marshy Lake, vicinity of Peninsula, (appr. 48° 45’ N. 86° 15’ W.), July 13, 1939. T. M. C. Taylor et al. 1445. Sphagnum bed, Onaman Lake, June 18, 1950. W. L. Klawe 60. Algoma District: “Clay Belt," June, 1905. A. Henderson. COCHRANE 1956] Lakela,—Rubus Chamaemorus 329 MAP 1. Runus CHAMAEMORUS in the Interior of North America, DISTRICT: In bog, among sphagnum, Smoky Falls, (appr. 50° 05’ N. 82° 10’ W.), July 24, 1935. Jim Whelan 79. PATRICIA DISTRICT: Spruce swamp near Nikip Lake, June 20, 1950. A. T. Cringan, P. 18. Clearing, edge of bush, Lake Attawapiskat, Aug. 18, 1939. W. B. Scott 134. The writer wishes to express gratitude to Mr. Claude E. Garton, who first brought to attention the occurrence of the species in the Canadian Lakehead, to Dr. James H. Soper, who kindly gave the listing of the collections of the Toronto Her- barium and to Dr. J. B. Carlson, University of Minnesota, Duluth Branch, for adapting the distribution map.—OLGA LAKELA, UNIVERSITY OF MINNESOTA, DULUTH BRANCH. 330 Rhodora [Vor. 58 PRUNUS AMERICANA VAR. LANATA A SYNONYM OF P. NIGRA. As originally published, Prunus americana var. lanata Sudworth (U.S.D.A. Div. For. Bull. 14: 237, 1897) was a substitute name for P. americana var. mollis “Torrey & Gray . . . not P. mollis Torr.” This was in accord with the American Code regulation which treated species, subspecies, and varieties as of equal rank for nomenclatural purposes; the epithet mollis could not be used twice in the genus Prunus, even though the second use might be in a different category. Sudworth gave no descrip- tion and no distribution for the plant. Since P. americana var. mollis T. & G. (Fl. N. A. 1: 407, 1840) and P. mollis Torr. (Comp. Fl. N. & Middle States p. 199, 1824) are both taxonomically and nomenclaturally identical (the latter is cited as synonym and basinym for the former, the same differentiating features are stated, and no particular distribution is assigned for either), the “not P. mollis" of Sudworth’s publication cannot be taken at face value. It meant only that the identical epithet was not to be used in both categories. A year later, Sudworth (Div. For. Bull. 17: 75, 1898) added a statement of distribution, “Mis- souri to northern Mexico," which excluded the type region of Torrey’s original publication. An emended circumscription is implied, but since no description or reference to a description is given, the emendation is not actually made. If it had been, the newly defined taxon could not bear the name given by Sudworth, because the ultimate type of lanata is P. mollis Torrey, which for over 60 years has been unanimously treated as a synonym of P. nigra Aiton. On the basis of type, then, and regardless of varying applications of the names, Prunus americana var. lanata and P. lanata (Sudw.) Mackenzie & Bush are both synonyms of P. nigra. The plant which passes as lanata is common in most of Texas, chiefly in limestone areas (in the eastern Pine Belt are forms with thinner and narrower leaf blades, suggesting introgression with P. wmbellata Ell.). Itis certainly not to be referred to P. americana Marshall. It is a solitary small tree, not forming thickets from rhizomes; it rarely has spinescent branchlets; the leaf blades are acute to abruptly acuminate, sparsely to densely pubescent beneath even in age; the fruit is purple with a heavy bloom, not red or yellow. I can find no satisfactory grounds for separating it 1956] Reeder,—Note on Echinochloa muricata 331 from P. mexicana S. Wats., Proc. Amer. Acad. 17: 353, 1882, described from northern Mexico (Coahuila). Detailed informa- tion on the northeastern limit of this species, and possible introgression with others, must await local studies in the Ozark and Mississippi Valley regions.—Lrovp H. SHINNERS, SOUTHERN METHODIST UNIVERSITY, DALLAS, TEXAS. HEDYOTIS ROSEA IN ARKANSAS.'—On March 20, 1953, collec- tions of Hedyotis rosea Raf. were made from scattered patches in the low fields and lawns at Conway, Faulkner County, Arkansas. 'The plants had large pink corollas, ranging in length from 7.5 to 12.5 mm. and with length of corolla tubes ranging from 4-7 mm. The calyx length ranged from 2.5-4 mm. with lobes 0.25-0.5 mm. in length. "The throats of the corollas were hairy. The seeds were about 1 mm. long and 0.2-0.3 mm. wide. The upper surfaces of the corolla lobes were densely clothed with minute, short, flat, broad-based enations easily seen with mag- nification. The basal leaves were spatulate. The joined stipules were almost truncate, to rounded, to very short- and broadly-triangular at the summit. Originally described by Rafinesque? from Louisiana (probably), Hedyotis rosea has recently been reported from Oklahoma by U. T. Waterfall? This evidently is the first report from Arkan- sas.—JEWEL E. MooRE, ARKANSAS STATE TEACHERS COLLEGE. Nore oN ECHINOCHLOA MURICATA.—In a recent issue of this journal (RHoporA 58: 48. 1956) Fairbrothers indicates that the correct binomial for the indigenous barnyard grass is Echt- nochloa muricata (Beauv.) Fernald. This conclusion appears to be correct, but there are some errors in citations of authority among the binomials he cites as synonyms. As the author has indicated, the first legitimate use of the epithet muricata for this taxon was by Beauvois, who published the binomial Setaria muricata in 1812. According to Art. 81 1 Contributions from the Botanical Laboratory, the University of Tennessee, N. En rpl EN C. Florula Ludoviciana 77. 1817. 3 WATERFALL, U. T. The Identity of Hedyotis rosea Raf. RHODORA. 55: 201- 203. 1953. 332 Rhodora [Vor. 58 of the Stockholm Code, this must be considered as the starting date for the epithet for purposes of priority. In 1829 Kunth made the combination Oplismenus muricatus which was, in effect, a transfer of Beauvois’ epithet. The corrected synonymy for this taxon, as far as the epithet muricata is concerned, is as follows: Panicum muricatum Michx. 1803. non Retz. 1786. Setaria muricata Beauv. 1812. Oplismenus muricatus (Beauv.) Kunth 1829. Echinochloa muricata (Beauv.) Fern. 1915. To place the name of Michaux in parenthesis, as the original author, gives an erroneous impression. As pointed out above, the epithet under discussion dates from 1812 when Beauvois used it in combination with Setaria. When muricata is used for this taxon, but in combination with other generic names, the citation of authority should show Beauvois (in parenthesis) as the original author. As long as the Homonym Rule stands in its present form, it is important for purposes of priority that care be used in the citation of authorities for those names con- taining an epithet which had been previously used for the same taxon but in an illegitimate combination.—Jonw R. REEDER, YALE UNIVERSITY, NEW HAVEN, CONN. SCIENTIFIC Books, LIBRARIES AND CoLLECTORS:—For a long time few people seemed to be very much interested in the history of science. Of late years, due primarily to the efforts of a few devoted scholars, the study of science history has taken on a new air of respectability. The price that is paid for respectability, however, in this study as in all others, is the production of a flood of books on the subject—a few out- standing, most tolerable, and a few poor. Even the worst, to be sure, have some value, but their faults need to be kept constantly in mind lest error and omission take on the cloak of verity. In 1954 there was published a book of some 300 pages entitled “Scien- tific Books, Libraries and Collectors," and subtitled “A Study of Bibliog- raphy and Book Trade in Relation to Science." The authors, J. L. Thornton and R. L. J. Tully state in their preface that they have *en- deavored to record the chief writings of every prominent scientific au- thor..." They modestly add that “the professional historian of science will find little new in these pages, but the student of the subject, and the scientist searching for ‘bibliographical gaps,’ will find between 1 The Library Association London, 1956] DeWolf,—Scientific Books 333 two covers an accumulated wealth of material on the bibliography of science.” This book, it must be admitted, reads easily and pleasantly, as long as one has only a vague acquaintance with the subject. When, however, a botanist reaches page 127 and reads that the 10th edition of Linnaeus’ “Systema Naturae" (1758-59) is “of special significance because it has been taken as the basis of modern botanical and zoological binomial nomenclature,” he is inclined to be somewhat amazed—and amused. He is perhaps inclined to think that the “Hortus Cliffortianus" (1737) and *Genera Plantarum" ed. 5 & 6 (1754 & 1764), which are not men- tioned, are of somewhat more importance than the “Classes Plantarum" (1738) and “Flora Suecica" (1746 & 1755), which are mentioned. At page 165 one is surprised to be told that Bentham and Hooker's “Genera Plantarum" is a mere compilation—an assertion which, if true, would indieate that this great work was considerably less valuable than it in fact is. At page 213 one is told that the “Index Kewensis" is con- sidered to be a bibliography—one thing that it is not! One wonders why, if this, the standard nomenclator for post-Linnaean botany, is included, we do not find mention of Tournefort’s "Institutiones Rei Herbariae" or Caspar Bauhin's *Pinax Theatri Botanici,” two of the important pre-Linnaean nomenclators. For that matter, one wonders why Richter’s Codex Botanicus Linnaeanus,” which collects the botanical material from the Linnaean publications as well as providing an extensive bibliography of both the Linnaean titles and the pre-Linnaean titles, was omitted. One is distressed to find that no reference is made to J. Christian Bay’s critical “Bibliographies of Botany"—surely one of the most important bibliographies that we have. Mention is made of Pritzel’s “Thesaurus . . ." but notice is given neither to Jackson's “Guide to the Literature of Botany . . ." nor Zuchold's various botanical bibliographie titles, all of which supplement Pritzel. Some other notable, and unforgivable, omissions which may be mentioned are “The Catalogue of the Library of the Royal Botanic Gardens, Kew" and its Supplement; Merrill and Walker's * Bibliography of Eastern Asiatic Botany” and Nissen’s “Die Botanische Buch Illustra- tion” both of which are copiously annotated. Likewise, the sales catalogs of the German firm of Junk are omitted. For that matter, the catalogs of any of the larger firms dealing in antiquarian natural history books are mines of information for the bibliographer. Abstracting journals, which are certainly useful bibliographic tools, also come off badly. ‘“Botan- ical Abstracts" and its successor “Biological Abstracts" seem to have escaped the authors’ notice. Absent also is reference to the very useful United States Department of Agriculture publication “Bibliography of Agriculture.” Also among the missing is the “Journal of the Society for the Bibliography of Natural History.” Finally, we note the omis- sion of the “Union List of Serials,” a work which lists the library holdings of upwards of 120,000 titles in the United States and Canada. 334 Rhodora (Vor. 58 The next to the last chapter deals with “Scientific Publishing and Bookselling." Despite the subtitle of the book, one finds here very little about the book trade. The house of Caxton receives some discussion, and a few continental publishers of that period are mentioned. We look in vain for mention of the house of Plantin. Longmans, John Murray, and Macmillan are mentioned but we search without success for Lovell Reeve & Co., Wilhelm Engelmann, or Martinus Nijhoff. In the United States, only D. Appleton & Co. and McGraw-Hill Book Co. are considered worthy of note. No mention is made of the various university presses which here carry on so much of the scientific publica- tion today. Finally, and rather importantly, no mention is made of the several presses which are responsible for the production of technical journals. The final chapter, “Scientific Libraries of Today," vies with the pre- ceding for the title of “worst chapter in the book.” Only Great Britain and the United States are considered at all, and the latter very super- fieially. We are told of the destruction, during the last war, of the library of the Manchester (England) Literary and Philosophical Society. No mention is made of the destruction of the library of the Botanical Museum at Berlin-Dahlem or the other scientific libraries in that city. We are told of the libraries at Kew and the British Museum (N. H.), but no mention is made of the library of the old Royal Botanie Garden at St. Petersburg (now Leningrad), though its catalog receives brief mention. Nor does the splendid library at the Conservatoire Botanique in Geneva receive recognition. For this country, no mention is made of the great holdings of natural history books at the University of Cali- fornia, at Harvard, and at the United States Department of Agriculture— to mention but three. This review deals with only a small portion of the field that Thornton and Tully claim to have covered. All in all, they have not done so good a job as one might have hoped—or as other reviews would lead one to expect. True, within its limitations, the book contains a great deal of information and is, therefore, of considerable value. Most of the sins committed are those of omission—but each faet and figure needs to be checked carefully before it is quoted.—GonpoN P. DEWorr, JR., BAILEY HORTORIUM, CORNELL UNIVERSITY. Volume 58, no. 694, including pages 275—810, was issued 31 October, 1956. BAARY y REFERENCE Ut FEB 7 1957 R 47 Dodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Conducted and published for the Club, by REED CLARK ROLLINS, Editor-in-Chief ALBERT FREDERICK HILL STUART KIMBALL HARRIS l RALPH CARLETON BEAN Associate Editors RICHARD ALDEN HOWARD Í CARROLL EMORY WOOD, JR, Vol. 58 December, 1956 No. 696 CONTENTS: Reappraisal of Two Plumose Rhynchosporas of the Southeastern United States. R. B. Channell............... s 335 The Genus Peteria (Leguminosae). C. L. Porter............... 344 A New Species of Streptanthus. George J. Goodman............ 354 PPR Iso ae nn ere LUE hey eres ss 355 Inder to Volume 58.7... .. .... desee RUE ete tas sento sd 356 The New England Botanical Club, Inc. 8 and 10 West King St., Lancaster, Pa. Botanical Museum, Oxford St., Cambridge 38, Mass. RHODORA.—A monthly journal of botany, devoted primarily to the flora of the Gray’s Manual Range and regions floristically related. Price, $6.00 per year, net, postpaid, in funds payable at par in United States currency in Boston; single copies (if available) of not more than 24 pages and with 1 plate, 60 cents, numbers of more than 24 pages or with more than 1 plate mostly at higher prices (see 3rd cover-page). Back volumes can be supplied at $5.00. Some single numbers from these volumes can be supplied only at advanced prices (see 3rd cover-page). Somewhat reduced rates for complete sets can be obtained on application to Dr. Hill. Notes and short scientific papers, relating directly or indirectly to the plants of North America, will be considered for publication to the extent that the limited space of the journal permits. Illustrations can be used only if the cost of engraver’s blocks is met through the author or his institution. Forms may be closed five weeks in advance of pub- lication. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to Reed C. Rollins, Gray Herbarium, 22 Divinity Ave., Cambridge 38, Mass. Subscriptions (making all remittances payable to RHODORA) to Dr. A. F. Hill, 8 W. King St., Lancaster, Pa., or, preferably, Botanical Museum, Oxford St., Cambridge 38, Mass. Entered as second-class matter March 9, 1929, at the post office at Lancaster, Pa., under the Act of March 3, 1879. INTELLIGENCER PRINTING COMPANY Specialists in Scientific and Technical Publications EIGHT WEST KING ST., LANCASTER, PA. CARD-INDEX OF NEW GENERA, SPECIES AND VARIETIES OF AMERICAN PLANTS For all students of American Plants the Gray Herbarium Card-index of Botanical Names is indispensable. It is a work of reference essen- tial to scientific libraries and academies and all centers of botanical activity. It includes genera and species from 1885 to date. The sub- divisions of species from 1885 to date are now included and from 1753 to 1886 are in the process of being inserted. Issued quarterly, at $25.50 per thousand cards. GRAY HERBARIUM of Harvard University, Cambridge 38, Mass., U. S. A. Rbodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 58 December, 1956 No. 696 REAPPRAISAL OF TWO PLUMOSE RHYNCHOSPORAS OF THE SOUTHEASTERN UNITED STATES R. B. CHANNELL Rhynchospora, Section Eurhynchospora, in Canada, the United States and the West Indies, was treated taxonomically by Gale (Ruopora 46: 89-278.) in 1944. According to this revision, Series Plumosae consists of three species: R. plumosa ElL, R. intermedia (Chapm.) Britt. and R. oligantha Gray, all characterized by the possession of unique perianth bristles. Rhynchospora oligantha var. breviseta Gale was originally de- scribed in this work. This new variety was distinguished from R. oligantha var. oligantha (as var. typica) by the reduced condition of the perianth bristles to less than one-half the length of the achene, the ovoid shape and smaller size of the achene and the short-conic tubercle. At this time the ranges of the two varieties were considered to be mutually exclusive: var. oligantha was reported to occur along the Atlantic and Gulf Coastal Plain from New Jersey southward into Georgia and western Florida, thence westward into eastern Texas (and also in Central America); var. breviseta was reported from peninsular Florida, western Cuba, Jamaica and Hispaniola. In the treatment of the same group (Sect. Plumosae), Küken- thal (Bot. Jahrb. 74: 505-509. 1949.) recognized three species also. However, only two of these had been recognized as such by Gale. Accordingly, Rhynchospora intermedia was reduced to varietal status under R. plumosa, where, in addition, the new variety interrupta Kiikenth. appeared. The plumose- 336 Rhodora [VoL. 58 bristled R. diadon (Nees) Griseb. (R. lunata Griseb.), in effect excluded by Gale from Eurhynchospora largely on account of its “curious horned tubercle” (ef. RHopora 46: 95. 1944.), was re-instated in the Plumosae. Rhynchospora oligantha var. oligantha and var. breviseta, however, were maintained without change of status or extension of range. Material collected since the appearance of Gale’s monograph and which was not utilized by Kiikenthal reveals that Rhyn- chospora oligantha var. breviseta is not only morphologically quite distinct, but that its geographical range is sympatric with that of the typical variety over a considerable area. In some areas, in fact, the two are known to occur side by side in the field and, although apparently flowering and fruiting over the same general period, maintain their distinctive mor- phological features without any indication of intergradation. This evidence, presented in detail below, is considered to be of sufficient magnitude to require the elevation of var. breviseta to specific status. The following change of status is thus made: Rhynchospora breviseta (Gale) stat. nov. based on Rhyncho- spora oligantha A. Gray var. breviseta Gale, Ruopora 46: 129, 130. 1944. The close relationship of Rhynchospora breviseta and R. oli- gantha is not questioned. The two species have in common the rugulose achenes, plumose perianth bristles, filiform leaves and culms and the cespitose habit of Series Plumosae. In addition, they have both axillary and terminal inflorescences of 1-several pedicellate spikelets, the inflorescences being both peduneulate and sessile and the spikelets disposed singly or as decompound fascicles. Both species are perennial by means of short, firm or ligneous, somewhat bulbous, crown-like rhizomes. The lowermost leaves consist of closed, truncate sheaths which may or may not bear short, greatly reduced blades. The typical cauline leaf is comprised of a closed, slenderly cylindric sheath, truncate at the apex, and a long, linear blade about 1⁄4 or 1⁄4 the width of the sheath. The blades are more or less concave or channeled throughout the serrulate basal portion and subterete or obtusely 4-angled and entire above. The leaf tips differ according to the species: R. oligantha (Fig. 2) has entire, blunt-tipped leaves; R. breviseta (Fig. 7) 1956] Channell,—Plumose Rhynchosporas 337 RHYNCHOSPORA OLIGANTHA (Channell 3120): Fig. 1, terminal inflorescence X 4; Fig. 2, leaf tip X 12; Fig. 3, achene X 12; Fig. 4, rachis X 4. R. BREvisETA (Channell 3257): Fig. 5, prophyll X 8; Fig. 6, terminal inflorescence X 4; Fig. 7, leaf tip X 12; Fig. 8, achene X 12; Fig. 9, rachis X 4. (Drawings kindly prepared by Ding Hou.) has acute tips which are minutely serrulate on the margin. These differences have been previously overlooked. While generally representing a practical taxonomic character, it should be pointed out that the serrulations of the leaf tips of R. breviseta are so minute that they are sometimes obscured in herbarium material which has been prepared by gluing the specimens to the sheets. The glue acts to seal these minute teeth against the margin of the leaf causing the margin to appear entire or otherwise rendering the serrulations somewhat transparent and thus difficult to observe. The achenes of Rhynchospora oligantha and R. breviseta are more nearly lenticular than those of the other members of Series Plumosae. Despite this general similarity, however, 338 Rhodora [Vor. 58 the two species are remarkably distinetive on the basis of other achenial features. The achene of R. oligantha (Fig. 3) is broadly elliptic and has a conspicuous collar-like constriction at the apex, the upper flange of which surrounds the base of the conic- attenuate tubercle. The achene of R. breviseta (Fig. 8) is ovoid or obovoid in shape and is crowned by a depressed-conic tubercle giving the appearance of having been applied to the apex of the achene. Indeed, the achenial differences are so pronounced that there is little chance of confusing the two species on that basis alone. The achenes of both species have been observed occasionally to be covered with a whitish, waxy bloom, although this is by no means a general feature of herbarium material. This pruinose condition is very conspicuous in freshly collected specimens but may disappear in the process of drying the specimens, especially if artificial heat is employed. There is considerable variation in the color of the achenes, this apparently varying to a certain extent with the degree of maturation attained. In individual spikelets of R. breviseta, which are oftentimes 8-fruited, the achenes may vary in color from manila-buff in the terminal portion of the spikelet, to castaneous in the central portion and, finally, dark blue or bluish black in the basal portion. The achenes are usually promptly shed at maturity ; thus the dark-colored achenes are not often to be seen on her- barium sheets. As the epithet connotes, the plumose perianth bristles of Rhynchospora breviseta are considerably shorter than those of R. oligantha. In the latter species the bristles vary from 1% the length of the achene to slightly longer than the achene, the basal portion being heavily plumose and the tips antrorsely serrulate; in the former the bristles are less than 14 the length of the achene and usually lack the antrorsely serrulate tips so pronounced in R. oligantha and are occasionally reduced to mere plumose tufts. While there is a decided difference between Rhynchospora breviseta and R. oligantha in the size and shape of the spikelets where best developed, the considerable variation throughout the geographical ranges of the two species precludes the use of these as diagnostic characters. Nevertheless, the spikelets 1956] Channell,—Plumose Rhynchosporas 339 of R. breviseta (Fig. 6) are, in general, longer and more slenderly fusiform in shape, the surface of the spikelet being somewhat torulose due to the arrangement of the 3-8 achenes in a scarcely imbricate, spiral series along the sides of the anfractuose rachis. This spirally zigzag, somewhat winged rachis is a conspicuous feature of R. breviseta (Fig. 9) even after the achenes are shed. Varying from 5 to 8(-10) mm. in length, the rachis remains intact well after the shedding of the spikelet scales and the achenes, even though it is a weakly organized, brittle structure which breaks easily under mechanical stress. The spikelets of R. oligantha (Fig. 1) are usually only 1-2(-3)-fruited, terete and relatively smooth, with little or no tendency to be torulose. The rachis (Fig. 4) is generally short and straight, somewhat clavate, 1-2(-3) mm. in length and without evident wings. The following key summarizes the numerous differences between the two species: Achene elliptic, broadest near the middle, the apex constricted into conspicuous collar-like flange; tubercle conic-attenuate, arising out of the surrounding collar, not contracted at the base; perianth bristles exceeding one-half the length of the achene; spikelets few-flowered, 1—3-fruited, the rachis with short, congested internodes; divergent lower branch of well-developed inflorescence commonly bearing a secondary branch, each with a pulvinate bract in the axil; leaf tips abtuse, enlife. eeraa rne tne de S O A R. oligantha. Achene obovoid, broadest above the middle, the apex without a conspicuous collar-like flange; tubercle depressed-conic, abruptly contracted at the base; perianth bristles one-half the length of the achene or shorter; spikelets several- flowered, 3-8-fruited, the rachis anfractuose-elongate, the internodes produced between the achenes; divergent lower branch of well-developed inflorescence usually simple, thus with only one pulvinate bract; leaf tips acute or acuminate, minutely serrulate............... 00. r. cee eee eee ees R. breviseta. The collections of Rhynchospora breviseta cited below are in addition to those included by Gale and Kükenthal. Norra CAROLINA: Brunswick County, Wood & Clement 7063 (pUKE, GH); Craven County, Wood & Clement 6997 (GH). MuississiPPI: Jackson County, Channell 3208, 3257 & 3663 (an, duplicates to be distributed). Rhynchospora breviseta was not previously known to occur in the United States outside of Florida, and these collections extend the range of the species northward into North Carolina 340 Rhodora [Vor. 58 and westward into Mississippi, well into the range of R. oligantha in both directions. It is a significant fact that the writer’s number 3257 originally comprised a large collection of a mixture of R. breviseta and R. oligantha. Material of the latter species was segregated only after examination in the laboratory to comprise his number 3258. This information is presented as unequivocal evidence that the two species occur together in the field, at least in Jackson County, Mississippi. Even though an inconspicuous element in the flora, without doubt having generally been passed unnoticed, Rhynchospora breviseta is to be expected to occur elsewhere on the Atlantic and Gulf Coastal Plain. Future collecting is likely to result in the discovery of the species over an even wider range as well as in areas between those already reported. Differences in the structure of the inflorescences of Rhyn- chospora oligantha and R. breviseta mentioned in the preceding key apparently have been generally overlooked. Except in terms of general form, in fact, the details of inflorescence struc- ture appear nowhere in the genus to have been adequately described. The remainder of this paper is therefore devoted to the description and comparison of the inflorescences of these and related species. In terms of relating species and species groups it would appear that the inflorescence may yield characters which could serve to supplement those of the fruit upon which the taxonomy of the entire genus has been thus far so singularly dependent. Although of the same fundamental structure, both lateral and terminal inflorescences are produced by the two species in question. Both types are clearly reduced forms of a more extensive and open prototype, as suggested by a comparison with the inflorescences of other species. In addition, Rhyn- chospora oligantha and R. breviseta themselves exhibit in a variety of stages a remarkable reduction series. This series strongly suggests the course of the reduction process. In general, the inflorescences of R. breviseta (Fig. 6) have undergone reduction to a greater degree than those of R. oligantha (Fig. 1) and the two species can usually be distinguished on this basis. The lateral inflorescences (not figured) are borne on long, flattened, upwardly serrulate peduncles arising singly out of 1956] Channell,—Plumose Rhynchosporas 341 the closed leaf-sheaths of the mid and lower culm. In their simplest form these consist of a single ebracteate or bracteate spikelet. Frequently, however, the lateral inflorescences are compound and consist of fascicles of pedicellate spikelets. In such cases each fascicle is subtended by a 1-nerved, somewhat vaginate, keeled bract, serrulate at the apex. In compound inflorescences each of the component pedicellate spikelets is, in turn, subtended by an additional (ultimate) bract of distinctly scarious texture (Fig. 5). These ultimate bracts would appear to correspond to the prophylls of grasses. Ranging in length from 3-5 mm., these thin, glabrous, ultimate bracts are more or less oblong in shape and have obtusish apices. Each is longi- tudinally 2-nerved, one nerve being located on each of the two keels of the convex back. The basal portion of the blade of each of these bracts ensheathes the pedicel it subtends. Struc- turally, the lateral inflorescences thus possess only two orders of bracts: those of the inflorescence fascicles and those of the individual pedicellate spikelets. The terminal inflorescences differ from the lateral in being sessile, that is, in lacking a common peduncle. Thus each, regardless of the degree of branching, is brought into close proximity to the uppermost leaf. This leaf is regarded as the primary (lowermost) bract of the terminal inflorescence-system, although it is little modified as compared with ordinary cauline leaves. The sheath portion is open, distinctly vaginate and thin-margined; the blade portion is concave and, although not rendered in the accompanying drawings, is minutely serrulate on the margin at the base, becoming entire above and finally indistinguishable from typical leaves toward the tip. The retention by the primary bract of foliar features characteristic of ordinary leaves renders it easily distinguishable from the highly modified, distinctly keeled bracts which often subtend each fascicular unit of the inflorescence. Appearing to be a continuation of the culm, the primary bract is to each over-all terminal inflorescence-system morphologically comparable to the bract subtending each individual inflorescence-fascicle. The terminal inflorescences may thus possess bracts of three orders, as shown in Figure 1, the primary bract just described in addition to the two types described for and common to the lateral, axillary 342 Rhodora [Vor. 58 inflorescences. Bracts of secondary order may sometimes be absent altogether. In compound inflorescences the lower branch is commonly and characteristically lateral in position and bears a conspicuous tuberculate enlargement (Figs. 1 & 6) in its axil. This pulvinate structure is clearly related to the lateral position ultimately assumed by the lower branch, in growth or enlargement doubtless serving to expand the inflorescence. Of more or less callous tissue, this pouch-like process is morphologically appendicular, belonging to the ultimate bract (Fig. 5) (prophyll) of the in- florescence, that is, the bract immediately subtending the individual pedicellate spikelets. For each inflorescence, how- ever, only (0—)1-2 of these bracts bear such a process. As clarified in the accompanying key, the two species in question usually differ in this respect: Rhynchospora oligantha often bears two such bracts to each of the more well-developed in- florescences; R. breviseta usually bears only one. There is considerable variation among the inflorescences as to the degree of development, even in the same collection or individual clump. The most fully developed inflorescence is of a characteristically corymbiform type, the primary axes of which may be twice or thrice branched. In less well developed inflorescences the primary axes are simple, arranged in a sort of umbellate cluster. Due to the failure of certain spikelets to develop, there are often present more ultimate bracts (pro- phylls) per inflorescence-fascicle than spikelets, for the bract subtending each pedicel is oftentimes produced whether the spikelet develops or not. Those inflorescences in which only the lateral axis develops owe their unilateral form to the reten- tion of an axillary bract bearing a tuberculate process at the base. Occasionally the axillary inflorescences are severely reduced, represented merely by a single pedicellate, ebracteate spikelet. This condition, however, is less common than the others described and is usually associated with a more basal position of the spikelet on the culm. The inflorescences of Rhynchospora oligantha and R. breviseta are quite unlike the more congested, spiciform or glomerate inflorescences of R. plumosa (including var. interrupta) and R. intermedia, the two pairs of species usually being separable 1956] Channell,—Plumose Rhynchosporas 343 in the field on this basis. This is due to the distinctly pedicellate nature of the spikelets of the former pair as contrasted with the subsessile nature of the spikelets of the latter. Of the plumose-bristled Rhynchosporas, the inflorescences of R. oli- gantha and R. breviseta find their closest match in those of R. diadon which has only terminal inflorescences. Although more extensively branched and more floriferous, but in general more compactly corymbiform, the inflorescences of this species are nevertheless very similar to those of R. oligantha and R. breviseta in structural detail. The bracts associated with the inflorescence are of three orders, one of the ultimate, axillary bracts of each inflorescence-fascicle being dorsally spurred or pulvinate. The pattern of branching is identical and the lowermost branch of each inflorescence-fascicle assumes a lateral or divergent position. From this standpoint, therefore, the inflorescence of R. diadon suggests a logical prototype of R. oligantha and R. breviseta. The similarities in the structure of the inflorescences of these three species represent an additional correlative characteristic, not previously emphasized, which favors the retention of R. diadon in the Plumosae. In fundamental details of strueture and especially in general appearance, the inflorescences of Rhynchospora oligantha and R. breviseta are quite comparable to those of R. rarzflora (Michx.) Ell., although the latter belongs to a different series (Rariflorae) which lacks plumose perianth bristles. In the field, therefore, it is often more difficult to distinguish A. rarzflora from K. oligantha and R. breviseta than to distinguish the latter two species from R. plumosa and R. intermedia. —HARVARD UNI- VERSITY, CAMBRIDGE, MASS. 344 Rhodora [Vor. 58 THE GENUS PETERIA (LEGUMINOSAE)! C. L. PORTER HISTORY OF THE GENUS Tue North American genus Peteria was established by Asa Gray? in 1852, and named in honor of Dr. Robert Peter who had been active in research on the flora of Kentucky. The genus was based on a single species, Peteria scoparia A. Gray, typified by collections made by Wright in Trans-Pecos Texas in 1849 and in southern New Mexico in 1851, and a collection made by Wislizenus in Chihuahua in 1846. Gray referred the genus to the tribe Galegeae and indicated that it was distinct from other genera of that tribe because of the stipules consisting of a pair of divaricate prickles like those found in the genus Caragana, and because of the *'aspergilliform" style and stigma, terminated by a tuft of hairs, this latter feature reminding him of the genus Lessertia. The genus was accepted by Bentham and Hooker? in 1865, who placed it in the Galegeae between Lamprolobium, a genus found in Australia, and Barbieria, a genus of tropical South America. These authors, however, erroneously described the stipules as being “subulate,” and they also misrepresented the plants as being “glabrous” and the “stylus inflexus, apice sub stigmate cireumcirca barbatus, stigmate capitato terminali." In 1873 Sereno Watson* added another species, Peteria Thomp- sonae, based on specimens collected in southern Utah in 1872 by Mrs. Ellen P. Thompson (sometimes erroneously written as “Mrs. A. P. Thompson"), the sister of Major John Wesley Powell of Colorado River fame whose headquarters at the time were at Kanab, Utah.’ In 1876, J. G. Schaffner collected additional specimens of the genus near San Louis Potosi, Mexico, and noted that the thick, tuberous roots of the plants were eaten by the natives there ! Contributions from the Department of Botany and the Rocky Mountain Her- barium, University of Wyoming, No. 228, 2 Pl. Wright. 1: 50. 1852. 3 Gen. Pl. 1: (2): 495. 1865. ‘Am, Nat. 7: 300. 1873. 5see Ewan, J. A., Rocky Mountain Naturalists. University of Denver Press, p.321. 1950. 1956] Porter,—Genus Peteria (Leguminosae) 345 who called them “camote del monte," or mountain sweet potato; and in 1878 Parry and Palmer secured additional specimens from the same region. Based on these collections, Peteria scoparia var. glandulosa A. Gray was published by Sereno Watson® in 1882. It was noted that these plants bore a re- semblance to Peteria scoparia, but were much more viscid- glandular on the calyx and pedicels. The genus was taken up by P. Taubert? in his treatment of the family in Die natiirlichen Pflanzenfamilien where it was placed in the tribe Galegeae, subtribe Tephrosiinae. Taubert described the genus as being monotypic, and credited Peteria scoparia to New Mexico only. The species was illustrated by an excellent small figure in the text. There the matter stood until 1923, when P. A. Rydberg® published an account of the genus in which he raised var. glandu- losa to specific rank, bringing the total species to three. Ryd- berg placed the genus in the tribe Galegeae, subtribe Craccanae (the same as the subtribe Tephrosiinae), in sequence after Galega and Cracca (Tephrosia), the three genera in the North American Flora included in that subtribe. Ivar Tidestrom? published Peteria nevadensis in 1923 on the basis of plants collected by him near Las Vegas, Nevada, in 1919, which were supposed to possess characters similar to those of Peteria Thompsonae but with smaller size of plants and flowers and narrower calyx-lobes. "These differences given by Tidestrom were not convincing, being quantitative rather than qualitative, and it has been found that they fit well within the general pattern of variability of Peteria Thompsonae. Furthermore, the populations in Utah and Nevada and adjacent Arizona occupy a continuous area and very similar ecological situations. The present writer became interested in the genus while engaged in a study of Nevada legumes in the course of which it became necessary to compare the types of the above two names, and to examine all the material that could be found of these plants in the larger herbaria. It seemed worthwhile to explore the remainder of the genus, particularly when it was discovered 6 Proc. Am. Acad. 17: 342. 1882. 1 Engl. & Prantl, Nat. Pflanzenfam. III. Abt. 2, 267, 273. 1894. 8 N. Am, Fl. 24: (3): 183-184. 1923. ? Proc. Biol. Soc. Wash. 36: 183. 1923. 346 Rhodora [Vor. 58 that an apparently undescribed species has been unrecognized for a number of years. GENERAL MORPHOLOGY roots. The root system of the genus Peteria is imperfectly known, being present on herbarium specimens of only two of the species, P. Thompsonae and P. glandulosa, in both of which there is a thickened tuberous root which is deep-seated, tapering above and below. In the other two species, P. scoparia and P. pinetorum, it is presumed that a similar root type occurs, but no specimens showing this feature have been found, the plants having invariably been broken off above the level of such tubers. Kearney and Peebles!® have noted “the tuberous rootstocks of P. scoparia, known in Texas as camote-de-monte, are reported to be edible,” but that statement probably was taken from Gray’s description of P. scoparia var. glandulosa (treated hereafter as P. glandulosa), and Gray in turn derived it from the label on Schaffner 834 which became the type of Gray’s variety. The fourth species, P. pinetorum described below, is based on several duplicates of a single collection, none of which shows the roots. STEMS. The stem is that of a typical herbaceous perennial arising from a somewhat lignescent base. In P. scoparia the stems are erect and usually bushy-branched from near the base, rounded in cross section, and longitudinally striate. In P. Thompsonae the stem is erect and branched above, but usually single below. P. glandulosa is a low bushy plant with stems erect or often decumbent and sprawling, branching from near the base. The greatly elongated stems of P. pinetorum are weak and with elongated internodes, attaining a height of 1 m. or more. LEAVES. In all species the leaves are petiolate and odd- pinnate with several to many leaflets. In three of the species, P. scoparia, P. glandulosa, and P. pinetorum, the leaflets are more or less similar, small and numerous, and narrowly oblong to elliptic in outline. In P. scoparia alone the leaflets are readily deciduous from the rachises, and in late summer or during dry periods the plants take on a characteristic broom-like appear- 10 U.S. D. A. Miscel. Publ. 423: 465. 1942; Ariz. Flora 441. 1951. 1956] Porter,—Genus Peteria (Leguminosae) 347 ance which suggested the name given them by Gray. In the other species the leaflets are persistent. Peteria Thompsonae has much larger leaflets which are rounded or retuse at the apex and oval to obovate in shape, much like those of several species of Astragalus found in the same area. STIPULES. The stipules in all species consist of a pair of stiff, sharp, slender prickles which are widely divaricate, 2-10 mm. long, and stramineous or brownish in color. PUBESCENCE. ‘The plants show varying degrees of pubescence, but all have the same type of hairs which are simple, basifixed, somewhat flattened, mostly pure white in color, and appressed. The pubescence is usually more abundant on the margins of the leaflets and on their veins below. There is a tendency for the species to produce glandular hairs on the pedicels and calyx, this being best developed in P. glandulosa. INFLORESCENCE. The inflorescence is a narrow raceme of varying length, arising either terminally or opposite the leaves, never axillary. The flowers are subtended by narrow subulate bracts, and there are no bracteoles. There seems to be a general tendency, except in P. Thompsonae, for a number of the flowers to abort or drop off, leaving flowerless gaps in the inflorescence and often only a few fruits. CALYX. The calyx consists of two parts: a tubular portion which is cylindric-campanulate and slightly gibbous at the base above, and five somewhat unequal lobes which are deltoid- acuminate to narrowly lanceolate-acuminate, these lobes the same length as the tube or shorter, and the two upper ones united farther up than the three lower ones. The calyx is not much changed after anthesis. COROLLA. The corolla is papilionaceous, and the petals are long-clawed, the claws very slender and as long as the blades or a little shorter. The banner is strongly folded lengthwise and its sides are reflexed. The wings are somewhat smaller than the banner and slightly auricled. The keel petals are a little shorter than the wings, and they have a broad auricle at the base of the blade. The color of the corolla is white to ochroleucous or tinged with purple; in P. Thompsonae, at least, the keel is maculate. STAMENS. The stamens are diadelphous, nine of them with 348 Rhodora [Vor. 58 their filaments fused into a cylindric tube, the upper one entirely free. The anthers are all alike, 2-celled, and oblong. PISTIL. The pistil is unicarpellate, short-stipitate or sub- sessile, the ovary laterally flattened and without any intrusion of sutures, several-ovuled, the style bent upward and with a widened basal portion which at length becomes hardened, horny, and spirally twisted, the stigma a tuft of short hairs surrounding the apex of the style and slightly oblique. FRUIT. The fruit is a legume which is linear or narrowly oblong, nearly straight, slightly if any constricted between the seeds or at places where ovules have aborted, laterally flattened, subsessile or with a stipe up to 5 mm. long in P. pinetorum, few-seeded, dehiscent along both sutures. SEEDS. The seeds are rounded-oblong and flattened, the funiculus attached near the end. They are not strophiolate. CYTOLOGY Nothing is known concerning the cytology of this genus. It is not mentioned in the works of Chekov! nor of Senn,” An attempt to germinate a few seeds taken from herbarium specimens was unsuccessful and no root tips were obtained. In discussing relationships in the legumes, Senn (l.c.) makes the statement that the evidence of both Chekov and himself indicates that the tribes Hedysareae, Galegeae, and Genisteae are highly heterogeneous assemblages with relatively little relationship. ACKNOWLEDGMENT The following herbaria have provided material for this study, for which grateful acknowledgment is made: The Gray Herbarium, Harvard Uni- versity (GH); U. S. National Herbarium (US); U. S. National Arboretum Herbarium (NA); Missouri Botanical Garden (MO); Chicago Natural History Museum (F); University of Arizona (ARIZ); Rocky Mountain Herbarium, University of Wyoming (RM). TAXONOMY Peteria A. Gray, Pl. Wright. 1: 50. 1852; Bentham & Hooker, Gen. Pl. 1 (2): 495. 1865; Taubert in Engl. & Prantl, Nat. Pflanzenfam. III. Abt. 3, 273. 1894; Rydberg in N. Am. Fl. 24 (3): 183. 1923. Erect or sprawling herbs from a deep tuberous root. Leaves odd- 1! Bull. Soc. Nat. Mosc. Biol. 46 (4): 233-240, 1937. 12 Bibliog. Genet. 12: 175-336. 1938. 1956] Porter,—Genus Peteria (Leguminosae) 340 pinnate, 7-many-foliolate, without stipels, the stipules a pair of slender divaricate prickles. Flowers in terminal racemes or sometimes the racemes lateral and then opposite the leaves, not axillary, papilionaceous, on short pedicels, subtended by subulate bracts, ebracteolate. Calyx cylindrie-campanulate, slightly gibbous at the base above, unequally 5-lobed, the sinus between the two upper lobes shallower than the others. Corolla white or ochroleucous, sometimes tinged with purple, the petals with very slender claws; banner obovate-oblong, moderately to strongly arched, folded lengthwise and the sides reflexed; wings oblong to obovate, somewhat auricled at the base; keel petals obliquely obovate, obtuse, with a broad basal auricle. Stamens 10, diadelphous, the upper one with a free filament; anthers all alike. Ovary sessile or short-stipitate, several- ovuled, the style becoming hardened and twisted at the base and incurved, the stigma a slightly oblique tuft of short hairs on the end of the style. Pod linear or narrowly oblong, straight or nearly so, laterally flattened, little or not at all constricted where ovules have aborted, dehiscent along both sutures. Seeds few to several. Type species: Peteria scoparia A. Gray. KEY TO THE SPECIES Leaflets rounded or retuse at the apex, 5-10 mm. wide, oval to obovate; corolla 15-20 mm. long; Utah, Nevada, and northern JSV de o oss oe ye nee ee sense ee enq k Wo de 1. P. Thompsonae. Leaflets acute or at least mucronate-pointed, not more than 5 mm. wide, usually oblong or narrowly elliptical; corolla not more than 15 mm. long; eastern Arizona to western Texas and southward through Mexico. Stems about 1 m. high; leaflets 6-12 mm. long, flat; pods 6-7 em. long; plants of the pine forests of northern Mexico QS co Said t ne eR eee 2. P. pinetorum. Stems lower, seldom over 6 dm. high; leaflets mostly much smaller, flat or involute; pods shorter; plants of dry plains and hills. Calyx and pedicels strigose or slightly glandular; leaflets early deciduous; stems mostly 5-8 dm. high, with ascending branches; southeastern Arizona to western Texas and south ward to Chihuahua............ 3. P. scoparia. Calyx and pedicels densely glandular; leaflets persistent; stems mostly 1-3 dm. high, with divaricate branches; San Luis Potosi and Zacatecas to Puebla........ 4. P. glandulosa. 1. Peteria Thompsonae 5. Wats. Peteria Thompsonae 8. Wats. Am. Nat. 7: 300. 1873. Peteria nevadensis Tidestr. Proc. Biol. Soc. Wash. 36: 183. 1923. Erect and usually several-stemmed from the base, 2-5 (1-6) dm. high, the stems with a few branches near the base, glabrous to sparsely strigose. Leaves odd-pinnate, 7-16 (5-20) em. long, the rachis usually densely appressed-strigose, bearing 13-21 (11-25) oval or obovate, often slightly 350 Rhodora [Vor. 58 PETERIA THOMPSONAE A 100 PETERIA SCOPARIA e PETERIA PINETORUM + PETERIA GLANDULOSA m Fig. 1. Map showing distribution of species of Peteria. 1956] Porter,—Genus Peteria (Leguminosae) 351 emarginate leaflets 6-15 mm. long and 4-12 mm. wide, these glabrous to strigose above, strigose beneath, persistent. Stipular spines 2-5 mm. long. Racemes 1-2 dm. long, several-flowered, the pedicels and calyx slightly to moderately glandular-puberulent. Calyx tube mostly 9-10 mm. long, the teeth 7-9 mm. long. Corolla 15-23 mm. long. Pods about 5 em. long, 5 mm. wide. TYPE: Kanab, Kane Co., Utah, Mrs. E. P. Thompson (A. P. Thompson on label by error), s.n. in 1872 (GH; isotype US; phototype RM). HABITAT: desert washes and gravelly slopes of dry hills and mesas. RANGE: southern half of Utah and Nevada and adjacent northwestern Arizona. SPECIMENS ExAMINED:—Utah. Without definite locality: southern Utah, Capt. F. M. Bishop 160 in 1872 (vs), and by the same collector in 1873, s.n. (Gu, F). Emery County (or possibly GRAND County): Green River, May 9, 1890, M. E. Jones s.n. (Mo, vs); same locality May 23, 1914, M. E. Jones s.n. (RM). GRAND County: Cisco, May 9, 1890, M. E. Jones 1676 (ARIZ, RM). Juas County: desert shadscale, common in shallow washes, 5,000 ft., 11 miles south of Troutcreek, June 17, 1933, Maguire & Becraft 2664 (cn, Mo, RM, US), and by the same collectors 2665 (Mo). Kane County: Kanab, Mrs. E. P. Thompson (A. P. Thompson on label by error) s.n. in 1872 (type GH, isotype vs, phototype RM); Kanab, 5,500 ft., May 22, 1894, M. E. Jones 5286 (us) WASHINGTON County: Springdale, 4,000 ft, May 16, 1894, M. E. Jones 5230 (wo, RM, Us). Nevada. Crank County: on mesa 16 miles northeast of Las Vegas, 840 meters, April 23, 1919, T'destrom 9083 (type of P. nevadensis vs, isotypes GH, US, phototype RM). ESMERALDA County: in wash, base of Montezuma Mt., 1650 meters, June 6, 1919, T'ide- strom 9791 (NA, US). Nye County: 20 miles southwest of Currant, May 14, 1941, Eastwood & Howell 9422 (Gu, F, us); frequent, shadscale-mixed desert, gravelly-sandy soil, 34 mile north of Locke’s Ranch, May 21, 1945, Maguire & Holmgren 25127 (ARIZ, MO, NA, photo RM). 2. Peteria pinetorum sp. nov. Herba fere 1 m. alta, caulis erectus, gracilis, internodis 2-5 cm. longis; foliis 3-7 cm. longis; foliolis persistentibus, interdum 13-17, 2-3 mm. latis, 6-12 mm. longis, anguste ellipticis vel oblongis, in apice mucronatis; stipulis spinosis, 3-5 mm. longis; racemis fructu plus minusve 1-2 dm. longis; calycibus fruetu pubescentibus glandulosis; leguminibus lineato- oblanceolatis, 4-7 em. longis, 5 mm. latis, stipitibus fere 5 mm. longis; floris ignotis. Spreading, bushy herb about 1 m. high, with slender, sparingly pubescent stems bearing rather distant leaves, the internodes 2-5 em. long. Leaves 3-7 em. long, the leaflets persistent, mostly 13-17, narrowly elliptical or oblong, 2-3 mm. wide and 6-12 mm. long, acutish and mucronate at the tip, short petiolulate, with a prominent purplish-black and strigose midrib and occasional strigose hairs on the margin but otherwise glabrous, the pairs of leaflets 3-5 mm. distant from each other. Stipules a pair of divaricate setose spines 3-5 mm. long. Inflorescence a terminal raceme 1-2 dm. long or more, the upper part of the rachis bearing only distant 352 Rhodora [Vor. 58 awl-shaped bracts when in fruit, the few fruits produced only from its base, the lowest fruit from a leaf-axil. Axis and pedicels moderately glandular-pubescent. Flowers not known. Fruiting pedicels 7-14 mm. (mostly about 10 mm.) long. Fruiting calyx cylindrical, somewhat gibbous at the base, glandular-pubescent, the tube about 6 mm. long, the teeth narrowly lanceolate, about 3 mm. long. Pods linear-oblanceolate, straight, with a stipe about 5 mm. long, the body 4-7 cm. long, the maximum width about 5 mm. Ovules several, and seeds 2-5, oblong, 5-6 mm. long, 3 mm. wide. TYPE: Los Cascarones, Rio Mayo, Chihuahua, Mexico, H. S. Gentry 2658 (an; isotypes ARIZ, F, MO, Us; phototype RM). HABITAT: cool pine slopes at about 7,000 ft. elevation. RANGE: known only from the western slope of the Sierra Madre Occi- dental in the Rio Mayo region of western Chihuahua near the Sonora border. This plant had been identified by Dr. Paul Standley as Peteria glandulosa (Gray) Rydb., whieh is perhaps its nearest relative. It differs greatly from that species, however, in habitat and in general aspect, being the tallest of all species in the genus, and in having larger and less crowded leaflets than in either P. glandulosa or P. scoparia but of the same general type as in those species. It further differs from P. scoparia, a plant of dry plains and foothills, in persistence of leaflets and in lacking the char- acteristic bushy “broom-like” aspect of that species. It is known only from a single but widely distributed collection which was in fruiting condition and lacked flowers. It. is presumed to have the tuberous roots of P. glandulosa, but none was found on the specimens. 3. Peteria scoparia A. Gray Peteria scoparia A. Gray, Pl. Wright. 1: 50. 1852. Stems bushy, 5-10 (mostly about 6) dm. high, the numerous branches ascending or somewhat divaricate, glabrous or nearly so, pale green and striate, up to 5 mm. thick at the base. Leaves 5-10 em. long, with a slender, persistent rachis bearing 9-15 narrowly elliptic to oblong, mu- cronate leaflets 3-6 (2-11) mm. long and 1-2 mm. wide, the leaflets usually deciduous to give the plants a characteristic broom-like appearance during the dry season, glabrous except on margin above, strigose beneath, acutish and mucronate at the apex, the pairs of leaflets mostly 3-5 mm. distant from each other. Stipules a pair of divaricate setose spines 3—5 (2-6) mm. long. Racemes 1-2 dm. long, several-flowered, the pedicels and calyx slightly glandular-puberulent. Calyx tube about 5 mm. long, the teeth 2-3 mm. long. Corolla 13-16 mm. long. Pods 3.5-6 em. long, 4-5 mm. wide. TYPE: mountain valleys beyond the pass of the Limpia (Texas, prob- 1956] Porter,—Genus Peteria (Leguminosae) 353 ably near Alpine, Brewster County), Charles Wright 138, collected in 1849 (GH, isotype GH). HABITAT: dry hills, mesas, and rolling plains, up to about 4,600 ft. RANGE: from the Petrified Forest, Apache County, Arizona, southeast- ward through southwestern New Mexico to Trans-Pecos Texas and northeastern Chihuahua. SPECIMENS EXAMINED:—Arizona. APACHE County: Petrified Forest, July 4, 1936, Toole & Goodding s.n. (Us). New Mexico. Dona ANA County: mesa west of the Organ Mts. (Tortugas Mt.), August 19, 1906, E. O. Wooton s.n. (ARIZ, MO, RM); Tortugas Mt. southeast of Las Cruces, 4,600 ft., July and August 1906, Wooton & Standley s.n. (vs). GRANT County: Telegraph Mts., August 16, 1902, E. O. Wooton s.n. (vs). Luna County: between Frontera and Mimbres, July 1851, C. Wright 962 (au, Us). Texas. Without definite locality: Sutton Hayes 159 (au); prairie near foot of Capote Mt., October 1883, V. Havard s.n. (Us). BREwsrER County: mountain valleys beyond the pass of the Limpia, August 1849, C. Wright 138 (type Gn); rare on Sul Ross College Hill, Alpine, August 12, 1937, B. H. Warnock T429 (au, vs). CuLBERSON County: Guadalupe Mt., in 1881, V. Havard 51 (an); calcareous stony soil, banks and slopes along Pine Springs Canyon, Guadalupe Mts., July 21, 1943, U. T. Waterfall 5240 (au, mo). HvpsrETH County: light gray calcareous soil in grassy flat between low hills approaching Sierra Diablo Plateau, 1114 miles north of Allamore, July 28, 1943, U. T. Waterfall 5350 (GH, MO); igneous cliffs in the Eagle Mts., about 2,000 ft. above Quitman Valley, southeast of the old Love Ranch, about 35 miles southeast of Sierra Blanca, August 22, 1946, U. T. Waterfall 6717 (an. Mo), Jerr Davis County: Goat Canyon, Geo. Jones Ranch, Davis Mts., about 5,000 ft., July 16, 1936, L. C. Hinckley s.n. (Ariz, GH); Goat Canyon, Mt. Livermore, July 30, 1935, L. C. Hinckley 423 (F). Mexico. Chihuahua: near Lake Encinillas north of Chihuahua, August 21, 1846, A. Wislizenus 126 (mo, fragment GH). 4. Peteria glandulosa (A. Gray) Rydb. Peteria scoparia var. glandulosa A. Gray ex S. Wats., Proc. Am. Acad. 17: 342. 1882. Peteria glandulosa (A. Gray) Rydb. N. Am. FI. 24: 183. 1923. Sprawling herb from a thick tuberous root, with divaricately branched pubescent stems mostly 1-3 dm. long. Leaves 1.5-3 em. long, the leaflets mostly 9-15, narrowly elliptic, 1-3(-4) mm. wide, 2-5(-8) mm. long, very short-petiolulate, with a dark and strigose midrib and strigose margin, the pairs of leaflets 1-3 mm. distant from each other. Stipules a pair of divaricate setose spines 3-10 mm. long. Inflorescence a terminal raceme, mostly 6-16 em. long, about 3-5-flowered, the lowest flowers from leaf-axils and a pair of stipular spines, the others from slender bracts, the pedicels and calyx, and often the axis of the raceme, densely glandular-puberulent. Flowers about 14-15 mm. long, the calyx-tube 7-8 mm. long, the lanceolate teeth 4-5 mm. long. Pods linear, straight or slightly constricted between the seeds, the body 4-4.5 em. long, the maximum width 3-5 mm., and with a stipe 5-6 mm. long. Seeds about 5 per pod. TYPE: Mexico, San Luis Potosí, in arenosis circa urbem, in 1876, J. G. 354 Rhodora [Vor. 58 Schaffner 834 (GH, isotype US); paratype from the same locality in 1878, Parry & Palmer 172 (GH). HABITAT: dry, often calcareous hills and slopes. RANGE: Mexico, from southern Zacatecas and San Louis Potosi to Puebla. The geographic isolation, as well as the low aspect of the plants and their persistent leaflets, both seem to indicate specific rather than varietal rank for this plant. SPECIMENS ÉxAMINED:—Mexico. San Luis Potosi: San Luis Potosí, in arenosis circa urbem, in 1876, J. G. Schaffner 834 (type Gu, phototype RM); same locality, in 1878, Parry & Palmer 172 (paratype am). Puebla: La Canada near Tehuacan, dry calcareous slopes, August 7, 1897, C. G. Pringle 7482 (F, GH, MO, US); El Riego, July 1905, Purpus 1188 (F, GH, Mo). Queretaro: hills, San Juan del Rio, July 12, 1896, C. G. Pringle 7285 (an). Zacatecas: on mesas in dry ground, Ojo Caliente, April 30, 1892, M. E. Jones 148 (us). A NEW SPECIES OF STREPTANTHUS GEORGE J. GOODMAN Streptanthus squamiformis Goodman, sp. nov. Herba annua, saepe 0.5 m. alta in anthesi ad 1 m. alta in fructu, simplex vel interdum inflores- centia ramosa; foliis caulinis integris, ovatis, acutis, amplexicaulibus, glabratis; pedicellis ad maturate crassis ascendentibus fere rectis, usque ad 13 mm. longis, pilis late-patentibus erassis quidem prope receptaculum ; calyce purpurello, 6-8 mm. longo, sepalis acutis ad apicem intro cucullatis, saccatis, capillatis, capillis densis conspicuis lucidis, 1-2 mm. longis, cylindricis, 0.12-0.18 mm. crassis in siccatione longitudinaliter complanatis squamoideis processis; petalis 12-14 mm. longis, ungulatis, lamina amplis purpureis, 6-7 mm. latis; staminibus liberis, antheris sagittatis apicula- tisque; fructibus glabris, usque ad 14.5 em. longis, 3 mm. latis; seminibus cirea 60, oblongis, alatis, 3-3.3 mm. longis. Streptanthus squamiformis Goodman, n. sp. Annual plants, often 0.5 m. tall in flower and up to 1 m. in fruit, simple or sometimes branched in the inflorescence; stem leaves entire, ovate, acute, clasping, glabrate; pedicels, at maturity, thick, ascending, nearly straight, up to 13 mm. long, with wide-spreading thick hairs at least near the receptacle; calyx purplish, 6-8 mm. long, sepals acute and hooded near the apex within, saccate, abundantly covered with conspicuous, clear hairs, these 1-2 mm. long, cylindrie, 0.12-0.18 mm. thick, drying to longitudinally flattened scale- like processes; petals 12-14 mm. long, clawed, the broadened blade purple and 6-7 mm. wide; stamens free, anthers sagittate and apiculate; fruits glabrous, up to 14.5 em. long, 3 mm. wide; seeds about 60, oblong, winged, 3-3.3 mm. long. 1956] Errata 355 OKLAHOMA: McCurtain County: Goodman 5507, slopes of pine-oak forest, Beaver’s Bend State Park, April 27, 1952 (Type, Bebb Herbarium of University of Oklahoma); M. Hopkins and A. and R. Nelson 985, Beaver’s Bend State Park, May 20, 1944; Little and Olmstead 74, pine-oak area, north of Broken Bow, along Highway 21, June 6, 1930. ARKANSAS: Polk County: Cooley R-23, on sandstone and soft shale, Rich Mountain, east of Mena near Oklahoma line, May 13, 1944. The April collection is in flower, the three May and June collections in fruit. The three collections other than the type were distributed as S. maculatus, to which the new species is closely related. The new species belongs, then, to the section Eustreptanthus, or Streptanthus as defined in a very narrow sense by O. E. Schulz in Die Natiirlichen Pflanzenfamilien 17 b: 300. 1936. The type collection was obtained from a population of plants that extended over an area of perhaps a hundred square yards. Over seventy plants were taken and a similar number were collected by Prof. U. T. Waterfall. The plants were very uniform, the color of the vegetative parts, the leaf shape, the flower color, and the hairs on the calyces showing little or no variation. No plants of the related S. maculatus were seen on this trip. In flower the new species is quickly recognized by the conspicuous hairs on the calyces which, upon drying, resemble scales. In fruit, at least a few of these scale-like hairs may be found on the thick pedicels.—DEPARTMENT OF PLANT SCIENCES, UNIVERSITY OF OKLAHOMA, NORMAN, OKLAHOMA. Volume 58, no. 695, including pages 311—334, was issued 10 December, 1956. ERRATA ` Page 149, line 35; for Meyerii read Meyeri. Page 150, line 1; for Meyerii read Meyeri. Page 151, lines 40, and 44; for Meyerii read Meyeri. Page 208, line 4; for page read page. . Page 268, line 1; for 8a. read 8b. Cover No. 695, line 7; for Praematurus read praematurus. Cover No. 695, line 10; for Americana read americana. Page 315, line 24; for serotinum read serotina. 356 Rhodora [Vor. 58 INDEX TO VOLUME 58 New scientific names and combinations are printed in bold face type Acer saccherinum, 93 Achillea lanulosa, 247; Millifolium, 247 Acorus calamus, 322 Additional Evidence for Five as the Basic Chromosome Number of the Andropogoneae, 135 Additions and Extensions to the Flora of Manitoba, 321 Additions to the Flora of St. Paul Island, Nova Seotia, 245 Aegilops crassa, 44 Agastache anethiodora, 287; Foeni- culum, 287 Agmenellum thermale, 118 Agropyron latiglume, 281; repens var. subulatum f. Vaillantianum, 246; sericeum, 148; trachycaulum var. novae-angliae, 246 Agrostis scabra, 246; tenuis, 246 Aira cristata, 94, 96; macrantha, 94; pubescens, 96; valesiana, 95 Alaska, A New Elymus from, 144 Algae Chiefly from New Bruns- wick, A Second List of Fresh- water, 117 Allium stellatum, 282 Alnus crispa, 127, 131, 132, var. mollis, 127, 131; glutinosa, 127; rugosa, 127, 131, f. Emersoniana, 127, var. americana, 127, f. hypomalaca, 128, f. tomophylla, 128; serrulata, 128, f. novebora- censis, 128; var. subelliptiea f. emarginata, 128, f. mollescens, 128 Alopecurus ciliatus, 96 Amblystegium compactum, 248 American Botany, Correa da Serra and, 45 American Botany, Humboldt. and, 191 Amorpha microphylla, 287; nana, 287 Anabaena oscillarioides, 118; vari- abilis, 118 Anacystis dimidiata, 118; montana, 118, f. minor, 118; rupestris, 119; thermalis, 118, f. major, 118 Andromeda polifolia, 322 Andropogon Ischaemum, 44; prae- maturus, 326, 327, f. hirtivagi- natus, 327; scoparius, 326 Andropogon praematurus, a Fun- gus-induced Monstrosity, 326 Andropogoneae, Additional Evi- dence for Five as the Basic Chromosome Number of the, 135 Androsace maxima, 45 Ankistrodesmus faleatus, 118 Aphanochaete repens, 118 Aplopappus spinulosus, 282 Arabis dentata, 292; holboellii var. collinsii, 323, 324; perstellata, 202, 293, 294, var. Shortii, 292; Shortii, 292, 294 Arabis perstellata, Growth Habits of, 292 Aralia nudicaulis, 310 Arenaria stricta, 318 Arethusa pendula, 34 Aristida basiraema, 318; oligantha, 318 Arkansas, Hedyotis rosea in, 331 Arkansas, Neviusia alabamensis in, 187 Artemisia caudata, 323, 324 Asa Gray and Andrew Jackson Downing, A Bibliographic Note, 243 Asarum virginicum, 40 Asclepias syriaca, 197, var. kansana, 197, 198 Asclepias syriaca var. kansana in New York State, 197 Aster leucanthemifolius, 281; mac- rophyllus, 310 Astragalus carnosus, 281, 287, 288, 289; crassicarpus, 281, 287 Athyrium Filix-femina, 40 Atrichum undulatum, 248 Aulacomnium palustre, 248, 323 Aulosira implexa, 118 Authorship and Nomenclature of Bur Clovers (Medicago) found wild in the United States, 1 Availability of Ingredients for Plastic, 72 1956] Index to Volume 58 357 Barbilophozia barbata, 248 Bartonia decapetala, 21, 282, 285; paniculata var. sabulonensis, 247 Basiphyllaea angustifolia, 290 Batrachospermum moniliforme, 117 Bazzania trilobata, 248 Bean, R. C., Clarence Hinckley Knowlton, 157 Bean, R. C. Knowlton, C. H., and Hill, A. F., Eleventh Report of the Committee on Plant Distribution, 125 Betula alba, 128; borealis, 128, 131, 132, 247; X caerulea, 128; cae- rulea-grandis, 128, 131; glandu- losa, 128, 131, 132, 323; lenta, 128, 132, f. laciniata, 128; lutea, 128, 130, 247, var. macrolepis, 128, 131; minor, 128, 131, 132; nigra, 128, 133, 134; papyrifera, 52, 128, 130, 131, var. commu- tata, 128, 131, 247, var. cordi- folia, 128, 180, 131, var. macro- stachya, 128, 247, var. pensilis, 128, 131, X populifolia, 128; pen- dula, 128; populifolia, 128, 132, f. incisifolia, 128; pumila, 128, 133, var. renifolia, 247 Binomials of Helenium in Walter’s Flora Caroliniana, The, 311 Blake, S. F., Froelichia gracilis in Maryland, 35; The Identity of Calyptocarpus blepharolepis, 275 Blasia pusilla, 248 Blepharostoma trichophyllum, 248 Boivin, B. and Cody, W. J., The Variations of Lilium canadense Linnaeus, 14 Botrychium dissectum, 274; obli- quum, 274 Bouteloua curtipendula, 318 Brachythecium salebrosum, 248 Braun, E. L., Growth Habits of Arabis perstellata, 292; Variation in Polemonium reptans, 103 Braya alpina, 156 Broussonatia papyrifera, 129 Bryum pallens, 248 Bulbochaete minor, 118 Bur Clovers (Medicago) found wild in the United States, Au- thorship and Nomenclature of, 1 Cactus viviparus, 286 Calamagrostis canadensis, 52 Calla palustris, 322 Callirhoé digitata, 100, var. stipu- lata, 101, f. alba, 101; involu- erata, 100, 102, var. involucrata, 101, var. lineariloba, 100, 101; lineariloba, 100 Callirhoé (Malvaceae), Some Chro- mosome Numbers in the Genus, 100 Calothrix parietina, 117, 118 Caltha palustris, 322 Calypogeia Neesiana, 248; Tri- chomanis, 248 Calyptocarpus blepharolepis, 275, 277, 278; tampicanus, 275, 278; vialis, 275, 277 Calyptocarpus blepharolepis, The Identity of, 275 Camassia esculenta, 287; Quamash, 287 Camptothecium nitens, 322 Canadian Lakehead, Rubus Cha- maemorus in the, 327 Capsosiphon fulvescens, 118 Carex abdita, 322, 324; ambly- orhyncha, 322; aquatilis, 52, 64; canescens var. subloliacea, 52; cephalantha, 246; chordorrhiza, 52, 59; cusickii, 279; X fulleri, 319, 320; heleonastes, 322; heter- oneura, 279, 280; hymenocarpa, 68; illota, 280; kelloggii, 280; laeviconica, 318, 320; laevirostris, 65, 68; lasiocarpa var. americana, 322; leporinella, 279, 280; lep- talea, 246, 322; leptopoda, 278; limnophila, 279, 280; limosa, 59, 322; magellanica, 52; meadii, 318; membranacea, 55, 56, 60; microptera, 279, 280; miliaris, 279; muhlenbergii, 318; neu- rophora, 280; normalis, 318; novae-angliae, 246; oligosperma, 247, 322; palleseens var. neo- gaea, 246; paludivagans, 53, 54, 56, 57, 58, 60, 63, 64, 65, 66, 68, 69; paupercula, 322; pen- sylvanica, 318, 320; physocarpa, 56, 68, 69, 279; praeceptorum, 279; rhynchophysa, 65, 66, 68; 358 Rhodora rostrata, 51, 52, 53, 55, 56, 57, 58, 59, 60, 63, 64, 65, 66, 68, 69, 71, 322; torundata, 51, 52, 53, 55, 56, 57, 58, 59, 60, 64, 65, 66, 68, 69; salina, 246; saxatalis, 56, 68, 69; sprengelii, 318; stenoptila, 280; stipata, 279; substricta, 280; tenuiflora, 52; tolmiei, 280 Carex by Hybridization, The Ec- ology of the Natural Origin of a Species of, 51 Carex, Notes on a New Hybrid, 318 Carpinus caroliniana var. virgini- ana, 128, 132 Carphephorus corymbosus, 276 Carya cordiformis, 127, 133; glabra, 127, 132; ovalis, 127, 132, var. hirsuta, 127, var. obcordata, 127, var. obovalis, 127; ovata, 127, 132, var. Nuttallii, 127, var. pubescens, 127; tomentosa, 127, 133 Case, F. W. Jr. Some Michigan Hecords for Sarracenia purpurea forma heterophylla, 203 Castanea dentata, 128, 132; pumila, 128, 130 Celarier, R. P., Additional Evi- dence for Five as the Basic Chromosome Number of the Andropogoneae, 135 Celtis occidentalis, 129, 133, var. canina, 129, 133, var. pumila, 129, 133 Cenchrus barbatus, 44 Cephalozia bicuspidata, 248; con- nivens, 248; leucantha, 248 Cephaloziella hempeana, 248 Cerastium arvense var. viscidulum, 247 Ceratodon purpureus, 248 Ceratophyllum demersum, 77 Chaetophora incrassata, 118 Chamaedaphne calyculata var. an- gustifolia, 52 Chamaesiphon polonicus, 119 Channell, R. B., Reappraisal of Two Plumose Rhynchosporas of the Southeastern United States, 335 Chara globularis, 118 | VoL. 58 Chloris ciliata, 44 Chlorotylium cataractarum, 117 Cicuta mackenzieana, 52 Cladophora crispata, 118 Slarence Hinckley Knowlton, 157 Clarence R. Hanes, 272 Claytonia virginica, 318 Coccochloris elabens, 119; Penio- cystis, 119; stagnina, 119 Cody, W. J., and Boivin, B., The Variations of Lilium canadense Linnaeus, 14 Coelopleurum lucidum, 247 Coleochaete scutata, 119 Collinsia verna, 307, 308; violacea, 307, 308 Comptonia peregrina, 127, 132 Convolvulus pilosellaefolius, 45 Corallorrhiza innata, 34 Cornus alternifolia, 247; florida, 39; X unalaschkensis, 247 Correa da Serra and American Botany (Review), 45 Correction in the Names of Two Species of Heterokontae, A, 161 Corydalis flavula, 318 Corylus americana, 128, 132, f. missouriensis, 128; cornuta, 128, 130; heterophylla, 128 Crandall, D. L., Shortia galaci- folia in Gray's Manual Range, 38 Cranichis tenuis, 290 Croasdale, H., A Correction in the Names of Two Species of Hetero- kontae, 161 Cronquist, A., Keck, D. D., and Maguire, B., Validity of Nut- tall’s Names in Fraser's Cata- logue, 23 Cylindrocystis Brebissonii, 119; eylindrospora, 119 Cynosurus echinatus, 43 Cystopteris bulbifera, 79, 80, 81, 82, 84, 87; dickieana, 79, 321, 324; fragilis, 79, 80, 82, 87, var. fragilis, 79, 83, var. laurentiana, 79, var. mackayii, 79, 80, 82, 83, Plate 1224, 84, 85, 86, 87, var. protrusa, 79, 80, 81, 82, 83, Plate 1224, 84, 85, 86, 87, var. simulans, 79, 80, 85, var. ten- nesseensis, 79, 80; montana, 87 1956] Index to Volume 58 359 Cystopteris fragilis Complex, A Diploid Variety in the, 79 Cytotaxonomic Study of the Genus Hymenopappas (Compositae), A, 163, 208, 250, 295 Dalea enneandra, 287; laxifolia, 287, 288 Dieteria sessilifolia 282 Dentaria diphylla, 160; laciniata, 83, 160; multifida, 160 Dentaria multifida in Central Ohio, 160 Desmidium Swartzii, 119 DeWolf, G. P., Jr., Scientific Books, Libraries and Collectors (Review), 332 Dicentra cucullaria, 318 Dichanthium sericeum Plate 1226, 140 Dichothrix Orsiniana, 119 Dicliptera brachiata, 49 Dicranella heteromalla, 248; subu- lata, 248 Dicranum Bergeri, 248; Bonjeani, 248; Drummondii, 248; scopari- um, 248 Digitaria sanguinalis, 37 Diploid Variety in the Cystopteris fragilis Complex, A, 79 Dodecatheon meadia, 318 Downing: A Bibliographic Note, Asa Gray and Andrew Jackson, 243 Draba reptans, 318 Draparnaldia plumosa, 117 Drepanocladus fluitans, 249; re- volvens, 249 Drosera intermedia, 247; rotundi- folia, 59 Drouet, F. and Habeeb, H., A Second List of Freshwater Algae Chiefly from New Brunswick, 117 Drury, W. H., Jr., The Ecology of the Natural Origin of a Species of Carex by Hybridization, 51 Dryopteris intermedia, 274; nove- boracensis, 39; phegopteris, 321; spinulosa, 274 Dupree, A. H., Asa Gray and Andrew Jackson Downing: A Bibliographic Note, 243 Echinochloa muricata, 48, 331, 332; pungens, 48 Echinochloa muricata, Note on, 331 Echinochloa, Nomenclatural Change in the Grass Genus, 48 Ecology of the Natural Origin of a Species of Carex by Hybridiza- tion, The, 51 Eleocharis halophila, 246; palustris, 246; Smallii, 246 Eleventh Report of the Committee on Plant Distribution, 125 Elodea Correct without Being Con- served, 162 Elymus canadensis, 144, 145, 148; hirsutus, 144; innovatus, 279; mollis, 246; pendulosus, 144, 145, 146, 147, 148 Elymus from Alaska, A New, 144 Emery, W. H. P., Some Chromo- some Numbers in the Genus Callirhoé (Malvaceae), 100 Entophysalis Brebissonii, 119; Lem- aniae, 119; rivularis, 117, 119, f. papillosa, 119 Epidendrum cochleatum var. tri- andrum, 290; jamaicense, 290; oncidioides var. gravidum, 291; vincentinum, 291 Equisetum fluviatile, 59, 65 Eremosphaera viridis, 119 Erianthus strictus, 136 Erigonum flavum, 287 Eriophorum angustifolium, 65; Chamissonis, 65; spissum, 246 Erskine, J. S5., Additions to the Flora of St. Paul Island, Nova Scotia, 245 Eyrthronium americanum, 83, 310 Evers, R. A., Two Plants New to the Illinois Flora, 49 Ewan, J., Correa da Serra and American Botany (Review) 45; Humboldt and American Botany, 191 Fagus grandifolia, 39, 128; 130, f. pubescens, 128, var. caroliniana, 128; sylvatica, 128; var. atro- purpurea, 128 Fairbrothers, D. E., Nomencla- 360 Rhodora tural Changes in the Grass Genus Echinochloa, 48 Fell, E. W., Notes on a New Hybrid Carex, 318 Festuca cristata, 96; elatior, 246; Gerardi, 95; splendens, 95 Ficus Carica, 129, 130 Fischerella ambigua, 119 Flora Caroliniana, The Binomials of Helenium in Walter’s, 311 Flora of Winnebago County, Illinois (Review), 273 Fontinalis novae-angliae, 249 Fraser's Catalogue, Nuttall Not the Author of, 281 Fraser's Catalogue, Validity of Nuttall's names in, 23 Froelichia gracilis, 35 Froelichia gracilis in Maryland, 35 Gaillardia aestivale, 315; aristata, 288; lanceolata, 314, 315; lutea, 314, 315; serotina, 315 Gaiser, L. O., and Zavitz, C. H., Notes on Triphora trianthophora in Ontario, 31 Galax aphylla, 40 Galium Brandegei, 52; tinctorium, 52; triflorum, 310 Genus Peteria (Leguminosae), The, 344 Gerardia auriculata, 88 Gleason, H. A., On Authors’ Cita- tions in Plant Names, 88 Gloeocapsa alpicola, 118; dimidiata, 118; limnetica, 118; membranina, 118; minuta, 118; turgida, 118 Gloeocystis confluens, 121; Grevil- lei, 121 Gloeotrichia Pisum, 119 Gomphosphaeria aponina, 119; la- custris, 119 Gongrosira Debaryana, 119 Goodman, G. J., A New Species of Streptanthus, 354 Goodyera repens var. ophioides, 247 Grass Genus Echinochloa, Nomen- clatural Changes in the, 48 Graustein, J. E., Nuttall’s Quarrel with Pursh, 20 Gray, and Andrew Jackson Down- [Vor. 58 ing: A Bibliographic Note, Asa, 243 Gray’s Manual Range, Shortia galacifolia in, 38 Grimmia maritima, 249 Grindelia squarrosa, 288 Growth Habits of Arabis perstel- lata, 202 Gypsophila porrigens, 44 Habeeb, H. and Drouet, F., A Second List of Freshwater Algae Chiefly from New Brunswick, 117 Habenaria — blephariglottis, 247; Hookeri, 247; strictissima var. odontopetala, 291 Hagenah, D. J. and Wagner, W. H., Jr, A Diploid Variety in the Cystopteris fragilis Complex, 79 Hanes, Clarence R., 272 Hanes, C. R., Viability of Seed of the Black Locust, 26 Hapalosiphon fontinalis, 119; pumi- lus, 119 Haplopappus ciliatus, 50 Harpanthus scutatus, 248 Hedyotis rosea, 331 Hedyotis rosea in Arkansas, 331 Helenium aestivale, Plate 1227, 312, 313, 315, 316; autumnale, 311, 312, 315, 316, 317; brevifolium, 316; flexuosum, 315, 316; nudi- florum, 315; Nuttallii, 312, 313; pinnatifidum, 312, 313; sero- tinum, 311, Plate 1227, 312, 313, 315, 316, 317; vernale Plate 1227, 312, 313, 314, 315, 316, 317 Helenium in Walter's Flora Caro- liniana, The Binomials of, 311 Helianthemum salicifolium, 44 Helianthus doronicoides, 97, Plate 1225, 98, 99; X doronicoides, 99; giganteus, 98, 99, 100; mollis, 97, 98, 99, 100; strumosus, 97; tomentosus, 97, 98; tuberosus, 97 Helianthus doronicoides, The Hy- brid Origin of, 97 Hermann, F. J., Range Extensions in Northwestern Plants, 278 Herniaria cinerea, 44; hirsuta, 44 1956] Index to Volume 58 361 Heterokontae, A Correction in the Names of Two Species of, 161 Hevetia flexilis, 90 Hieracium aurantiacum, 247; cana- dense, 247 Hierochloe odorata, 246 Hil, A. F., Bean, R. C. and Knowlton, C. H., Eleventh Re- port of the Committee on Plant Distribution, 125 Hodgson, H. J., A New Elymus from Alaska, 144 Holly, Ilex glabra forma leuco- earpa: A White-fruited, 25 Hordeum jubatum, 246 Hormotila mucigena, 119 Humboldt and American Botany, 191 Hyalotheca dissiliens, 119; mucosa, 119 Hybrid Origin of Helianthus doro- nicoides, The, 97 Hydrangea arborescens, 39 Hydrurus foetidus, 119 Hylocomium splendens, 249 Hymenopappus anthemoides, 306; arenosus, 251, 252; artemisiae- folius, 175; 255, 259, 266, 267, 296, 302, 303, 305, 306, var. artemesiaefolius, 170, 182, 186, 210, 258, 298, 300, 302, 303, 304, 305, 306, var. riograndensis, 170, 182, 184, 186, 210, 298, 300, 303, 304, 305, 306; biennis, 180, 182, 183, 184, 186, 209, 212, 256, 257, 258, 259; canescens var. canotomentosus, 268; caro- liniensis, 295; cinereus, 228; 235, 236, 251; columbianus, 239; corymbosus, 266, 296, 297, 300, var. nuttallii, 260; douglasii, 306; engelmannianus, 297; eriopoda, 225; eriopodus, 228; filifolius, 164, 165, 166, 175, 177, 180, 184, 185, 209, 211, 214, 217, 218, 220, 230, 239, 269, 306, var. cinereus, 168, 169, 170, 174, 186, 209, 211, 215, 217, 218, 221, 222, 223, 224, 225, 228, 231, 233, 236, 251, 252, 253, 261, 262, 269, var. eriopodus, 186, 215, 218, 225, 226, 227, 228, 242, var. filifolius, 168, 170, 174, 186, 215, 216, 218, 220, 227, 239, 240, var. idahoensis, 170, 186, 215, 218, 220, 222, 242, 250, var. lugens, 165, 168, 169, 170, 172, 174, 176, 186, 215, 218, 220, 221, 222, 228, 229, 230, 231, 232, 233, 234, 238, 241, 242, 252, 255, 269, var. luteus, 169, 171, 174, 186, 215, 218, 221, 222, 223, 224, 226, 228, 252, var. megacephalus, 186, 215, 218, 223, 225, 226, 227, 228, 229, 233, 238, 239, 252, var. nanus, 168, 169, 171, 174, 186, 215, 218, 219, 220, 222, 226, 227, 233, 240, 241, 242, 250, var. nudipes, 171, 186, 214, 217, 218, 219, 220, 221, 233, var. parvulus, 168, 171, 174, 186, 214, 216, 217, 218, 219, 220, 222, 252, var. pauciflorus, 171, 186, 215, 218, 220, 224, 225, 228, 238, 239, 252, var. polycephalus, 169, 171, 172, 186, 215, 218, 231, 235, 236, 240, 261, 262, var. tomentosus, 171, 186, 215, 218, 220, 225, 228, 233, 237, 238, 239; fisheri, 264; flavescens, 182, 259, 264, 266, 267, var. cano-tomentosus, 168, 171, 174, 182, 184, 186, 209, 258, 265, 266, 268, 269, 298, var. flavescens, 169, 171, 182, 186, 209, 264, 266, 268, 269, 298, 299; flavomargina- tus, 182, 186, 209, 212, 257, 258, 259; glandulosus, 306; glaucus, 306; gloriosus, 230; integer, 254, 255; integrifolia, 306; laxiflorus, 295; ligulaeflorus, 306; lugens, 228, 230; luteus, 221, 230; macro- glottis, 230; matricarioides, 306; mexicanus, 171, 182, 185, 186, 209, 212, 216, 254, 255; nanus, 240, 241; nelsoni, 306; nevadensis, 306; newberryi, 168, 171, 173, 179, 180, 182, 183, 186, 209, 210, 211, 212, 213, 214, 257; niveus, 237, 239; nudatus, 230; nudipes, 219, var. alpestris, 219, 221; obtusifolius, 254; ochroleucous, 228, 251, 252; palmeri, 306; parvulus, 216, 217; pauciflorus 224, 228; pedatus, 306; petaloide- us, 254, 255; polycephalus, 235, 362 Rhodora 236; radiatus, 171, 180, 182, 186, 209, 211, 212, 213, 214, 255, 257; robustus, 268; scabiosaeus, 165, 186, 209, 255, 295, 296, 300, var. corymbosus, 168, 171, 173, 174, 182, 186, 210, 262, 265, 266, 296, 297, 298, 299, 300, 303, 304, 305, var. scabiosaeus, 182, 186, 210, 295, 296, 298, 299, 303; scaposus, 230; sulphureus, 297; tenuifolius, 168, 171, 174, 182, 184, 186, 209, 235, 236, 239, 258, 259, 260, 261, 262; tomentosus, 228, 237; wislizeni var. seti- formis, 307; wrightii, 307, var. viscidulus, 307 Hymenopappus (Compositae), A Cytotaxonomie Study of the Genus, 163, 208, 250, 295 Hymenothrix palmeri, 166, 178; wislizenii, 178, 179; wrightii, 178 Hyparrhenia formosa, 137, Plate 1226, 139; hirta, 136, 137, Plate 1226, 139; rufa, 137, 138, Plate 1226, 139 Hypericum perforatum, 247 Hypnum pallescens, 249 Hypogynium spathiflorum, 139, 141; spathifolius, 137, Plate 1226 Hyssopus anethiodorus, 287 Identity of Calyptocarpus blepha- rolepis, The, 275 Hex glabra, 25, f. leucocarpa, Plate 1222, 25 Ilex glabra forma leucocarpa: A White-fruited Holly, 25 Illegitimacy of Persoon’s Species of Koeleria (Graminae), 93 Illinois, Flora of Winnebago County (Review), 273 Illinois Flora, Two Plants New to the, 49 Iris cristata, 40 Jackson, R. C., The Hybrid Origin of Helianthus doronicoides, 97 Juglans cinerea, 127, 132; nigra, 127, 133, 134 Juncus alpinus, 65; arcticus, 65; biglumis, 280; castaneus, 65; compressus, 279; halli, 280; [Vor. 58 pelocarpus, 247; tracyi, 280, 281; vaseyi, 280 Juniperus communis var. depressa, 50; var. saxatilis, 50 Juniperus communis var. saxatilis from Mount Washington, 50 Kalmia latifolia, 39 Keck, D. D., Maguire, B. and Cronquist, A., Validity of Nut- talls Names in Frasers Cata- logue, 23 Knowlton, Clarence Hinckley, 157 Knowlton, C. H., Hill, A. F. and Bean, R. C., Eleventh Report of the Committee on Plant Distribution, 125 Koeleria ciliata, 96; cristata, 94, 95; Gerardi; 95; Gerrardi, 96; gracilis, 94; macrantha, 94; ni- tida, 94; pectinata, 95; phleoides 94, 95; pubescens, 96; pyra- midata, 94; splendens, 95; tube- rosa, 93, 95; valesiaca, 95; vallesiana, 94, 95; villosa, 96 Koeleria (Graminae), Illegitimacy of Persoon’s Species of, 93 Lactuca biennis, 247 Lakela, O., A New Form of Nuphar microphyllum from Minnesota, 76; Rubus Chamaemorus in the Canadian Lakehead, 327 Lappula echinata, 247 Laportea canadensis, 82 Leavenworthia aurea, 73, 74, 75, 201; exigua, 74, 75; stylosa, 73, 74, 75 Leavenworthia aurea, The Problem of, 73 Ledum groenlandicum, 323 Lepanthopsis melanantha, 291 Lepidozia reptans, 248 Leptodictyum riparium, 249 Leptoloma cognatum, 318 Leptosira Mediciana, 117, 120 Lespedeza capitata, 274, var. longi- folia, 274; hirta, 274; longifolia, 274 Lesquerella angustifolia, 199, 200, 201, 202; gracilis, 199, 200, 202; 1956] Index to Volume 58 363 longifolia, 199, 200, 201; fili- formis, 201, 202 Lesquerella angustifolia, On the Identity of, 199 Leucampyx newberryi, 210 Lilium andinum, 286; canadense, 15, 16, 19, 20, f. canadense, 16, 17, f. flavum, 17, 19, f. michiganense, 16, 19, f. peramoe- num, 16, f. rubrum, 16, 17, f. uniflorum, 16, var. canadense, 16, 17, var. editorum, 16, 17, var. flavum, 17, 19, var. luteum, 17, var. rubrum, 17, 19; ssp. canadense 16; ssp. editorum, 17; ssp. michiganense, 16, 17, f. peramoenum, 16, 18, f. uni- florum, 18; ssp. superbum, 16, 19; coccineum, 17; carolinianum, 19; catesbei, 19, 20; michauxii, 19; michiganense, 14, 15, 16, 17, 19, f. peramoenum, 18, f. umbelliferum, 17, f. uniflorum, 18, var. umbelliferum, 17; par- dalinum var. Bourgaei, 17, 18; peramoenum, 18; penduliflorum, 16; pendulum, 16; philadelphi- cum, 19, 20; superbum, 14, 15, 19, var. carolinianum, 19 Lilium eanadense Linnaeus, The Variations of, 14 Limonium leptostachyum, 45 Liparis Saundersiana, 291 Liriodendron Tulipifera, 39 Locust, Viability of Seed of the Black, 26 Long, R. W. Jr., Dentaria multi- fida in Central Ohio, 160 Lophochloa cristata, 96 Lophozia grandiretis, 248 Luzula piperi, 281 Lycopodium complanatum var. fla- belliforme, 246 Lygodium palmatum, 270 Lygodium palmatum vs. Urbaniza- tion, 270 Lyngbya Diguetii, 120; versicolor, 120 Maclura pomifera, 129 Maguire, B., Cronquist, A. and Keck, D. D., Validity of Nut- tall’s names in Fraser’s Cata- logue, 23 Maianthemum canadense, 310 Malva coccinea, 287 Mammillaria viviparus, 286 Manitoba, Additions and Exten- sions to the Flora of, 321 Maryland, Froelichia gracilis in, 35 Matricaria maritima, 247 Maxillaria rufescens, 291 Maycock, P. F., Tiarella cordifolia in Wisconsin, 309 Medeola virginiana, 40 Medicago apiculata, 5, 11, 310, var. confinis, 11; arabica, 2, 3, 6, 7; ciliaris, 2, 6, 9; cordata, 2; denticulata, 2, 3, 5, 11, var. brevispina, 11, var. ciliaris, 9, 310, var. lappacea, 8, var. mac- racantha, 8, var. vulgaris, 11, 310; falcata, 1; hirsuta, 4; hispida, 2, 3, 4, 5, 6, 7, 8, 11, 310, var. apiculata, 11, var. pentacycla, 6, 7, 8, f. longeaculeata, 7, var. polygyra, 9, f. aculeata, 9, í. inermis, 9, var. tricycla, 9, f. longispina, 9, f. microdon, 9, ssp. microcarpa, 10; laciniata, 3; lappacea, 8, var. macracantha, 8, var. pentacycla, 8, var. tricycla, 8; lupulina, 1, var. glandulosa, 2; maculata, 2; microdon, 9; minima, 1, 3, 4, 6, 7, f. mollissima, 4, f. pubescens, 4, f. viscida, 4, var. brachyodon, 4, var. com- pacta, 4, var. longiseta, 4, var. pubescens, 4, var. pulchella, 4, var. viscida, 4, var. vulgaris, 4; nigra, 8; obscura var. microdon, 9; orbicularis, 4; orbiculata, 5; pentacycla, 8; polycarpa, 5, var. tuberculata, 11, 310; polymorpha, 2, 3, 4, 5, 6, 7, 8, var. arabica, 2, var. ciliaris, 2, 6, 7, 8, 9, 10, 310, f. apiculata, 11, var. ciliaris f. ciliaris, 11, 12, f. tuberculata, 11, 12, var. coronata, 6, var. hirsuta, 4, var. laciniata, 3, var. minima, 3, 4, var. nigra, 6, 7, 8, var. orbicularis, 5, var. polygyra, 8, 9, 11, 12, var. polymorpha, 8, var. tricycla, 8, 364 Rhodora var. vulgaris, 310, f. apiculata, 310, f. tuberculata, 310; prae- cox, 12; reticulata, 9, 11, 12; sardoa, 9, 11, 12; sativa, 1; tuberculata, 12 (Medicago) found wild in the United States, Authorship and Nomenclature of Bur Clovers, 1 Megalodonta Beckii, 77 Menyanthes trifoliata, 322 Meringosphaera Nemiahi, 161; sol, 161 Merismopoedia thermalis, 118 Mesotaenium macrococcum, 120; rubescens, 120 Mexican Cruciferae, Some New Primitive, 148 Michigan Records for Sarracenia purpurea forma heterophylla, Some, 203 Microcoleus rupicola, 120 Microspora amoena, 120 Minnesota, A New Form of Nuphar microphyllum from, 76 Miscanthus floridulus, 135; japoni- cus, 135 Mitella nuda, 310 Mnium hornum, 249; punctatum, 249 Monachino, J., Yonkers Wool Mill Plant Records, 43 Monotropa Hypopitys, 247 Moore, D. M., Neuviusia ala- bamensis in Arkansas, 187 Moore, J. E., Hedyotis rosea in Arkansas, 331 Morus alba, 192; rubra, 129, 133, 134 Mougeotia laetevirens, 120 Mount Washington, Juniperus com- munis var. saxatilis from, 50 Muhlenbergia glomerata, 278 Myrica Gale, 127, 130, 247, var. subglabra, 127, 130; pensyl- vanica, 127, 132 Narenga porphyrocoma, 136, 141 Neslia paniculata, 247 Netrium digitus, 120; interruptum, 120 Neviusia alabamensis, 177, 188, 189, 190 [Vor. 58 Neviusia alabamensis in Arkansas, 187 New Brunswick, A Second List of Freshwater Algae Chiefly from, 117 New Elymus from Alaska, A, 144 New Form of Nuphar microphyllum from Minnesota, A, 76 New Species of Streptanthus, A, 354 New York State, Asclepias syriaca var. kansana in, 197 Nitella flexilis, 120 Nomenclatural Change in the Grass Genus Echinochloa, 48 Nomenclature of Bur Clovers (Medicago) found wild in the United States, Authorship and, 1 Nostoe commune, 120; microscopi- cum, 120; microscopium, 120; pruniforme, 120; sphaericum, 120; verrucosum, 120 Note on Echinochloa muricata, 331 Notes on a New Hybrid Carex, 318 Notes on Collinsia violacea, 307 Notes on Triphora trianthophora in Ontario, 31 Notes on Utricularia biflora and U. fibrosa, 41 Nova Scotia, Additions to the Flora of St. Paul Island, 245 Nuphar microphyllum, 76, f. micro- phyllum, 76, f. multisepalum, 76, 78; rubrodiscum, 77; varie- gatum, 77 Nuphar microphyllum from Min- nesota, A New Form of, 76 Nuttall Not the Author of Fraser’s Catalogue, 281 Nuttall’s names in Fraser’s Cata- logue, Validity of, 23 Nuttall’s Quarrel with Pursh, 20 Nyssa sylvatica, 39 Oenothera alata, 284; biennis, 247; caespitosa, 287, 288; macro- carpa, 282, 284, 288; missourien- sis, 284, 287, 288 Ohio, Dentaria multifida in Central, 160 On Authors’ Citations in Plant Names, 88 1956] Index to Volume 58 365 Ontario, Notes on Triphora tri- anthophora in, 31 On the Identity of Lesquerella angustifolia, 199 Oocystis Borgei, 120; crassa, 120 Ophiocytium cochleare, 120; par- vulum, 120 Oplismenus muricatus, 48, 332 Orchid Flora of Puerto Rico, Some Additions to the, 290 Oscillatoria amoena, 120; anguina, 120; articulata, 120; chlorina, 120; curviceps, 121; granulata, 121; irrigua, 121; limosa, 121; ornata, 121; princeps, 121; sim- plicissima, 121; splendida, 121; tenuis var. tergestina, 121 Ostrya virginiana, 128, 130, var. lasia, 128, 133 Oxalis violacea, 318 Oxytropis splendens, 323, 324 Pachystima Canbyi, 92 Palmella Myosurus, 119 Palmogloea protuberans, 121 Panderia pilosa, 45; turkestanica, 45 Panicum muricatum, 48, 332; pun- gens, 48; scribnerianum, 318 Penium margaritaceum, 121 Pennisetum ciliare, 44 Penstemon bradburyi, 23, 287, 288; grandiflorus, 23, 287, 288; pallidus, 318 Peteria glandulosa, 346, 347, 349, 352, 353; nevadensis, 345, 349; pinetorum, 346, 348, 349, 351; scoparia, 344, 345, 346, 349, 352, var. glandulosa, 345, 346, 353; Thompsonae, 344, 345, 346, 347, 349 Peteria (Leguminosae), The Genus, 344 Phalangium esculentum, 284, 287; Quamash, 284, 287 Phalaris pubescens, 96 Phellodendron japonicum, 43 Phormidium autumnale, 117; in- undatum, 121; purpurascens, 117, 121; Retzii, 121; Setchellianum, 117; uncinatum, 117; valderia- num, 121 Picea engelmanni, 180; glauca, 52; mariana, 52, 246, 323 Pinus palustris, 25; ponderosa, 180; serotina, 25 Plagiothecium striatellum, 249 Plastie, Availability of Ingredients for, 72 Plectonema Nostocorum, 121; To- masinianum, 121 Pleurothallis Ghiesbreghtiana, 291 Pleurotaenium trabecula, 121 Pleurozium Schreberi, 249 Poa annua, 246; cristata, 94; palustris, 246; pectinata, 94, 95; pyramidata, 94; vallesiana, 95 Podophyllum peltatum, 83 Pogonia ophioglossoides, 32; pen- dula, 33, 34; trianthophora, 31 Pohl, R. W., Andropogon prae- maturus, a Fungus-induced Mon- strosity, 326 Pohlia nutans, 249 Polyenemum arvense, 44; majus, 44 Polygonum perfoliatum, 43 Polemonium delicatum, 113, 114; occidentale, 113; pulcherrimum, 113; reptans, 103, 104, 112, 113, 114, 115, var. reptans, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, var. villosum, 103, 104, 105, 106, 107, 108, 110, 111, 112, 113, 114, 115; Van Bruntiae, 113 Polemonium reptans, Variation in, 103 Polypteris integrifolia, 276 Polystachya foliosa, 291 Polystichum acrostichoides, 39 Polytrichum gracile, 249; juni- perinum, 249 Populus alba, 125, 130; balsamifera, 51, 125, 131, var. subcordata, 125; canescens, 125; deltoides, 125, 133, 134, var. missouriensis, 125; X gileadensis, 125; grandi- dentata, 125, 130; heterophylla, 125, 133, 134; nigra var. italica, 125, 130; tremuloides, 125, 130, f. pendula, 125, f. reniformis, 125, var. magnifica, 125, 131 366 Rhodora Porter, C. L., The Genus Peteria (Leguminosae), 344 Potamogeton alpinus var. tenui- folius, 321; confervoides, 246; natans, 77; robbinsii, 321 Potentilla arguta, 323, 324; palus- tris, 59; pensylvanica var. pen- sylvanica, 323, 324 Problem of Leavenworthia aurea, The, 73 Prunus americana, 330, var. lanata, 330, var. mollis, 330; lanata, 330; mexicana, 331; mollis, 330; nigra, 330; serotina subsp. hir- suta, 115, subsp. serotina, 115; umbellata, 330; virginiana, 247, 318 Prunus americana var. lanata a Synonym of P. nigra, 330 Pseudotsuga taxifolia, 180 Puerto Rico, Some Additions to the Orchid Flora of, 290 Pursh, Nuttall's Quarrel with, 20 Pyrus americana, 247; decora, 247 Quercus alba, 128, 132, 318, f. latiloba, 128, f. repanda, 129; X Bebbiana, 129; bicolor, 129, 132; coccinea, 129, 132; faleata var. pagodaefolia, 130; X Faxoni, 129; X Fernowi, 129; ilicifolia, 129, 132; imbricaria, 129; X Leana, 129; macrocarpa, 129, 133, 318, f. olivaeformis, 129; Muehlenbergii, 129, 133, f. Alex- anderi, 129; palustris, 129, 133; X Porteri, 129; prinoides, 129, 132, var. rufescens, 129, 133; Prinus, 93, 129, 133; X Rehderi, 129; x Robbinsii, 129; Robur, 129; rubra, 93, 129, 132, var. borealis, 129, 131; X Saulii, 129; stellata, 129, 133, var. Mar- garetta, 129, 133; stelloides, 129; velutina, 129, 132, 318, f. dilaniata, 129, f. macrophylla, 129, f. pagodaeformis, 129 Radiosphaera Nemiahi, 161; sol, 161 Range Extensions in Northwestern Plants, 278 [Vor. 58 Ranunculus repens, 65 Ratibida columnaris, 287; colum- nifera, 287 Reappraisal of Two Plumose Rhyn- chosporas of the Southeastern United States, 335 Reeder, J. R., Note on Echinochloa muricata, 331 Rhizoclonium hieroglyphicum, 117, 121; riparium, 122 Rhus radicans, 82 Rhynchospora breviseta, 336, 337, 338, 339, 340, 342, 343; diadon, 336, 343; intermedia, 335, 342, 343; lunata, 336; oligantha, 335, 336, 337, 338, 339, 340, 342, 343, var. breviseta, 335, 336, var. oligantha, 335, 336; plumosa, 335, 342, 343, var. interrupta, 335, 342; rariflora, 343 Rhytidiadelphus squarrosus, 249 Ritchie, J. C., Additions and Extensions to the Flora of Mani- toba, 321 Rivularia haematites, 117, 122; minutula, 122; rufescens, 117, 122 Robinia Pseudo-Acacia, 26 Rock, H. F. L., The Binomials of Helenium in Walter’s Flora Caroliniana, 311 Rollins, R. C., Availability of Ingredients for Plastic, 72; On the Identity of Lesquerella angusti- folia, 199; The Problem of Leavenworthia aurea, 73; Some New Primitive Mexican Cruci- ferae, 148 Romanschulzia alpina, 156; apetala, 149, 155, 156; arabiformis, 154; elata, 151; Meyeri, 149, 150, 151; schistacea, 152, 154, 155; subclavata, 152, 153; turritoides, 151 Rorippa islandica, 65 Rothia artemisiaefolia, 302; caro- liniensis, 165, 295; corymbosa, 207; degenerica, 307; filifolia, 214; flavescens, 264; intermedia, 307; mexicana, 254; pinnata, 307; pusilla, 307; tenuifolius, 260 Rubus acaulis, 322; Chamaemorus, 1956] Index to Volume 58 367 323, 327, 329; florenceae, 272; hanesii, 272; kalamazoensis, 272; schooleraftianus, 272 Rubus Chamaemorus in the Cana- dian Lakehead, 327 Rudbeckia purpurea var. serotina, 286 Rumex domesticus, 247; occiden- talis, 65 Salix alba, 125, 130, var. calva, 125, var. vitellina, 125; alba X fragilis, 125; amygdaloides, 125, 133, 134; arbusculoides, 323; arctophila, 125, 131, 132; argy- rocarpa, 125, 131, 132; atha- bascensis, 323; babylonica, 126; Bebbiana, 126, 130, 247, var. capreifolia, 126; Bebbiana X dis- color, 126; Bebbiana X gracilis, 126; Bebbiana X rigida, 126; candida, 126, 133, f. denudata, 126; candida X Bebbiana, 126; candida X gracilis, 126; candida X rigida, 126; Caprea, 126; cin- erea, 126; cinerea X rigida, 126; coactilis, 126, 131; cordata, 126, var. abrasa, 126; discolor, 126, 130, var. latifolia, 126, var. Overi, 126; discolor X humilis, 126; fragilis, 126, 130; glauco- phylloides, 126, 133, f. lasioclada, 1260; gracilis, 126, 131, var. textoris, 126, 131, 323; gracilis x humilis var. microphylla, 126; gracilis X sericea, 126; gracilis X Grayi, 126; hebecarpa, 323; her- bacea, 126, 131, 132; humilis, 126, 130, f. curtifolia, 126, f. festiva, 126, f. tortifolia, 126, var. hyporhysa, 126, 133, var. keweenawensis, 126, var. micro- phylla, 126, 133; humilis X gra- cilis, 126; humilis X sericea, 126; humilis X humilis var. micro- phylla, 125; incana, 1206; interior, 126, 133, 134, f. Wheeleri, 126, var. exterior, 126, 133, 134; X Jesupi, 126; lucida, 126, 130, var. angustifolia, 126, 131, var. intonsa, 126, 131; myrtillifolia, 323; X myricoides, 1206, nigra, 126, 132; X Peasei, 126; pedicel- laris, 127, 131, var. hypoclauca, 127, 131, 322, var. tenuescens, 127; pellita, 127, 131, f. psila, 127; pentandra, 127, 130; petio- laris, 131; planifolia, 127, 131, 132; pseudocordata, 322; pur- purea, 127; pyrifolia, 127, 131; pyrifolia X discolor, 127; rigida, 127, 130, f. mollis, 127, var, angustata, 127, 131; rigida X discolor, 127; rigida X gracilis, 127; rigida X lucida, 127; rigida, X pedicellaris, 127; rigida X sub- sericea, 127; sericea, 127, 130; serissima, 127, 133; X Smithiana, 127; X subsericea, 127; Uva-ursi, 127, 131, 132, f. lasiophylla, 127; viminalis, 127 Sanicula marylandica, 247 Sanvitalia ocymoides, 275, 276, 277, 278; tragiaefolia, 275 Sarracenia heterophylla, 203; pur- purea, 203, 204, 205, f. hetero- phylla, 203, 204, 205, 206, 207 Sarracenia purpurea f. hetero- phylla, Some Michigan Records for, 203 Saururus cernuus, 125, 130, 133 Scapania irrigua, 248 Scenedesmus bijuga, 122 Schizachne purpurascens, 322, 324 Schizochlamys aurantia, 122; gela- tinosa, 122 Schizothrix arenaria, 122; calcicola, 122; Dailyi, 122; lacustris, 122; Lamyi, 122; lardacea, 122; la- teritia, 122; mellea, 122; Muel- leri, 122; purpurascens, 122 Schweinfurth, C., and Winters, H. F., Some Additions to the Orchid Flora of Puerto Rico, 290 Scientific Books, Libraries and Collectors (Review), 332 Scilla esculenta, 284, 287; Peru- viana, 2 Scirpus acutus, 44; lacustris, 44, var. Tabernaemontani, 44; micro- carpus, 65; mucronatus, 43; Tabernaemontani, 43; validus, 44 Seytonema figuratum, 122; densum, 368 Rhodora 122; mirabile, 122; tolypotrich- oides, 122 Second List of Freshwater Algae Chiefly from New Brunswick, A, 117 Setaria Faberii, 43; muricata, 48, 331 Setiscapella cleistogama, 90 Shinners, L. H., Authorship and Nomenclature of Bur Clovers (Medicago) found wild in the United States, 1; Elodea Correct without Being Conserved, 162; Illegitimacy of Persoon’s Species of Koeleria, 93; Medicago poly- morpha var. vulgaris, 310; Nut- tall Not the Author of Fraser’s Catalogue, 281; Prunus Ameri- cana var. lanata a Synonym of P. nigra, 330 Shortia galacifolia, 38 Shortia galacifolia in Gray’s Man- ual Range, 38 Silene virginica, 27 Silene virginica in the Gulf States, 27 Sisymbrium alpinum, 156; Stand- leyi, 156 Smilacina trifolia, 247 Smith, J. F., Lygodium palmatum vs. Urbanization, 270 Some Additions to the Orchid Flora of Puerto Rico, 290 Some Chromosome Numbers in the Genus Callirhoé (Malvaceae), 100 Some Michigan Records for Sar- racenia purpurea forma heter- phylla, 203 Some New Primitive Mexican Cruciferae, 148 Sonchus arvensis, 247; asper, 247 Sophora sericea, 281, 289 Sorghum brevicallosum, 136; dimi- diatum, 136; intrans, 136, Plate 1226; matarankense, 136; pur- pureo-sericeum, 136; nitidum, 136; stipoidum, 136; versicolor, 136; vulgare, 140 Sphaeralcea coccinea, 287 Sphaerella lacustris, 122 Sphagnum Angstrémii, 52, 59; [Vor. 58 balticum, 59, 63, 64; capillaceum var. tenellum, 249; centrale, 249; compactum, 249; cuspida- tum, 249; Dusenii, 59, 603; fimbriatum, 52, 249; fuscum, 59, 63, 249, 323; Girgensohnii, 249; imbricatum, 249; Lindbergii, 59, 64, 249; magellanicum, 249; papillosum, 59, 63, 249; plumu- losum, 52; pulchrum, 59; re- curvum, 52, 59, 249; riparium, 52, 59, 249; robustum, 52; squarrosum, 249; tenellum, 249; teres, 52, 59;:Warnstorfi, 249; warnstorfianum, 322, 323 Spirogyra catenaeformis, 122; com- munis, 123; condensata, 123; crassa, 123; dubia, 123; fluvia- tilis, 123; gracilis, 123; Grevil- leana, 117, 123; Hantzschii, 123; inflata, 123; Juergensii, 123; longata, 123; maxima, 123; ma- juscula, 123; neglecta, 123; ortho- spira, 123; Weberi, 123 Stanleya pinnata, 149 Staurastrum Habeebense, 117 Steele, F. L., Juniperus communis var. saxatilis from Mount Wash- ington, 50 Stelis pygmaea, 291 Steyermark, J. A. and Swink, F. A., Asclepias syriaca var. kansana in New York State, 197 Steyermark, J. A., Flora of Winne- bago County, Illinois (Review), 273 Stiburus alopecuroides, 96 Stigonema mamillosum, 123; panni- forme, 123 Streptanthus, A New Species of, 354 Streptanthus maculatus, 355; squamiformis, 354 Streptopus roseus, 310 Swink, F. A., and Steyermark, J. A., Asclepias syriaca var. kansana in New York State, 197 Symploca dubia, 123; muralis, 123 Tetmemorus granulatus, 124 Tetraedron minimum, 124 TO 1956] Index to Volume 58 369 Tetraphis pellucida, 249 Tetraspora gelatinosa, 117 Teucrium charidemi, 92 Thalictrum dasycarpum, 82 Thermopsis rhombifolia, 281 Thismia americana, 45 Thuraria herbacea, 288 Tiarella cordifolia, 309, 310 Tiarella cordifolia in Wisconsin, 309 Tolypothrix distorta, 124; lanata, 124; penicillata, 117, 124; tenius, 124 Tomanthera auriculata, 88, 89 Tribonema bombycinum, 124, f. tenue, 124 Trientalis borealis, 310 Trifolium monanthum, 281 Triglochin palustre, 65 Triphora pendula, 34; triantho- phora, 31, 33, 34 Triphora trianthophora in Ontario, Notes on, 31 Turner, B. L., A Cytotaxonomic Study of the Genus Hymeno- pappas (Compositae), 163, 208, 250, 295 Two Plants New to the Illinois Flora, 49 Ulmus americana, 91, 129, 131, f. alba, 129, f. intercedens, 129, f. laevior, 129, f. pendula, 129; glabra, 129; procera, 129; rubra, 129, 132; Thomasi, 129, 133 Ulota crispa, 249; phyllantha, 249 Ulothrix implexa, 124 United States, Authorship and Nomenclature of Bur Clovers (Medicago) found wild in the, 1 United States, Reappraisal of Two Plumose Rhynchosporas of the Southeastern, 335 Utricularia biflora, 41, 42, 43; fibrosa, 42, 43; subulata var. cleistogama, 90 Utricularia biflora and U. fibrosa, Notes on, 41 Uttal, L. J., Notes on Utricularia biflora and U. fibrosa, 41 Vaccinium Oxycoccus, 59, 322; uliginosum, 59 Validity of Nuttall’s names in Fraser’s Catalogue, 23 Vanilla barbellata, 291; Dilloniana, 291 Variation in Polemonium reptans, 103 Variations of Lilium canadense Linnaeus, The, 14 Vaucheria aversa, 124; geminata var. racemosa, 124 Veronica biloba, 44; campylopoda, 44 Vernonia illinoensis, 229 Viability of Seed of the Black Locust, 26 Vicia Benghalensis, 2; Cracca, 247 Virgilia grandiflora, 288 Vitis campestris, 286; riparia, 286; vulpina, 286 Voss, E. G. and Wilson, J. S., Clarence R. Hanes, 272 Wagner, W. H. Jr., and Hagenah, D. J., A Diploid Variety in the Cystopteris fragilis Complex, 79 Walter’s Flora Caroliniana, The Binomials of Helenium in, 311 Warea amplexifolia, 149 Wilson, J. S. and Voss, E. G., Clarence R. Hanes, 272 Winterringer, G. S., Notes on Collinsia violacea, 308 Winters, H. F. and Schweinfurth, C., Some Additions to the Orchid Flora of Puerto Rico, 290 Wisconsin, Tiarella cordifolia in, 309 Woods, F. W., Ilex glabra forma leu- cocarpa: AWhite-Fruited Holly 25 Woodsia alpina, 321, 324 Yonkers Wool Mill Plant Records, 43 Yucca angustifolia, 287; glauca, 287 Zavitz, C. H. and Gaiser, L. O., Notes on Triphora triantho- phora in Ontario, 31 Zizania aquatica, 76 Zygnema cyanosporum, 124; decus- satum, 124; insigne, 124; pec- tinatum, 124 Zygogonium ericetorum, 124